From hushujun at mail.sdu.edu.cn Thu Nov 1 03:54:44 2007 From: hushujun at mail.sdu.edu.cn (=?gb2312?B?uvrK9778?=) Date: Thu, 01 Nov 2007 10:54:44 +0800 Subject: [Pw_forum] question about the psi^2 projected on certain band. Message-ID: <393885684.16066@mail.sdu.edu.cn> Dear all, During the post-process, psi^2 projected on a certain band is necessary for us. However, pp.x with plot_num=7 gives such image on gamma point only. My question is: if a certain band is much localized (such as the 3d states in some transition metal oxide or localized donor states in the semiconductor gap), does psi^2 on gamma represent the spatial charge distribution of the whole localized band? Best wishes Shujun From sunshoutian168 at yahoo.com.cn Thu Nov 1 09:22:32 2007 From: sunshoutian168 at yahoo.com.cn (shoutian sun) Date: Thu, 1 Nov 2007 16:22:32 +0800 (CST) Subject: [Pw_forum] The pseudopotentials. Message-ID: <388514.53702.qm@web92003.mail.cnb.yahoo.com> >Dear all, >There are so many pseudopotential files available in www.pwscf.org/pseudo. >And which one is most perfect.For example Al: >Al.pbe-rrkj.UPF Al.pz-vbc.UPF Al.pbe-n-van.UPF >Al.pw91-n-van.UPF Al.blyp-n-van_ak.UPF Al.bp-n-van_ak.UPF > >The troubleshooting in the manual said that ' maybe your system does have >negative or strange phonon frequencies with the approximations you used '. > >Could anyone tell me which pseudopotential I have to choose if I encountered the >problems. Thank you so much. >Regurds --------------------------------- ?????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071101/bd13fcee/attachment.htm From degironc at sissa.it Thu Nov 1 09:53:13 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Thu, 01 Nov 2007 09:53:13 +0100 Subject: [Pw_forum] The pseudopotentials. In-Reply-To: <388514.53702.qm@web92003.mail.cnb.yahoo.com> References: <388514.53702.qm@web92003.mail.cnb.yahoo.com> Message-ID: <472993F9.20606@sissa.it> (Most of) the different pseudopotentials refer to different XC funcitonals. Which one to use depends on which approximate XC you are interested in. Stefano de Gironcoli, SISSA and DEMOCRITOS shoutian sun wrote: > >Dear all, > >There are so many pseudopotential files available in > www.pwscf.org/pseudo. > >And which one is most perfect.For example Al: > >Al.pbe-rrkj.UPF Al.pz-vbc.UPF > Al.pbe-n-van.UPF > >Al.pw91-n-van.UPF > Al.blyp-n-van_ak.UPF > Al.bp-n-van_ak.UPF > > > > >The troubleshooting in the manual said that ' maybe your system does > have >negative or strange phonon frequencies with the approximations > you used '. > > > >Could anyone tell me which pseudopotential I have to choose if I > encountered the >problems. Thank you so much. > >Regurds > > > ------------------------------------------------------------------------ > ?????????? > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From mgkrishn at ncsu.edu Thu Nov 1 11:18:05 2007 From: mgkrishn at ncsu.edu (mgkrishn at ncsu.edu) Date: Thu, 1 Nov 2007 06:18:05 -0400 (EDT) Subject: [Pw_forum] Relativistic pseudo potentials Message-ID: <1418.69.134.141.9.1193912285.squirrel@webmail.ncsu.edu> Hi I am looking for relativistic(i.e spin - angular momentum coupled) pseudo potential files for the following - Bi, Te, Sb, Se Thanks Kanna From giannozz at nest.sns.it Thu Nov 1 15:02:35 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 1 Nov 2007 15:02:35 +0100 Subject: [Pw_forum] question about the psi^2 projected on certain band. In-Reply-To: <393885684.16066@mail.sdu.edu.cn> References: <393885684.16066@mail.sdu.edu.cn> Message-ID: On Nov 1, 2007, at 3:54 , ??? wrote: > During the post-process, psi^2 projected on a certain band is > necessary for us. > However, pp.x with plot_num=7 gives such image on gamma point only. are you sure? I don't think so > My question is: if a certain band is much localized (such as the 3d > states > in some transition metal oxide or localized donor states in the > semiconductor > gap), does psi^2 on gamma represent the spatial charge distribution > of the > whole localized band? for a narrow band formed by localised states, I expect that \psi^2 doesn't change much along the band Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From degironc at sissa.it Thu Nov 1 16:03:25 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Thu, 01 Nov 2007 16:03:25 +0100 Subject: [Pw_forum] question about the psi^2 projected on certain band. In-Reply-To: References: <393885684.16066@mail.sdu.edu.cn> Message-ID: <4729EABD.5040303@sissa.it> your problem looks to me suited for a Wannier decomposition of the band structure. Have a look at www.wannier.org It is interfaced with QE Stefano de Gironcoli -SISSA and DEMOCRITOS Paolo Giannozzi wrote: > On Nov 1, 2007, at 3:54 , ??? wrote: > > >> During the post-process, psi^2 projected on a certain band is >> necessary for us. >> However, pp.x with plot_num=7 gives such image on gamma point only. >> > > are you sure? I don't think so > > >> My question is: if a certain band is much localized (such as the 3d >> states >> in some transition metal oxide or localized donor states in the >> semiconductor >> gap), does psi^2 on gamma represent the spatial charge distribution >> of the >> whole localized band? >> > > for a narrow band formed by localised states, I expect that \psi^2 > doesn't change much along the band > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From emenendez at macul.ciencias.uchile.cl Fri Nov 2 01:03:07 2007 From: emenendez at macul.ciencias.uchile.cl (Eduardo Ariel Menendez P) Date: Thu, 1 Nov 2007 21:03:07 -0300 (CLST) Subject: [Pw_forum] fixing spin in LSD Message-ID: Hi, My doubt may be trivial,as I cannot find the answer anywhere. When does one needs to fix the value of the spin in LSDA calculation? multiplicity INTEGER ( default = 0 [unspecified] ) spin multiplicity (2s+1). 1 is singlet, 2 for doublet etc. Note that this fixes the final value of the magnetization. if unspecified or a non-zero value is specified in nelup/neldw then multiplicity variable is ignored. In particular, I am interested in the case of impurities in surfaces. Should I fix the spin in this case or not? Other case is dimers and clusters. I have answered this question to senior quantum chemists and I have received answers like a) Whith Hartree-Fock equations you must specify the spin, and DFT works well when HF does, and fails when HF fails. b) The states measured by spectroscopy have well defined S_z and S^2. So you want to calculate these states. Is there any other (and more fundamental) reason? Polite request: Avoid send Word attachments. To see why, take 6 minutes to read at http://www.gnu.org/philosophy/no-word-attachments.html Eduardo A. Menendez Proupin Department of Physics Faculty of Science University of Chile Las Palmeras 3425 ?u?oa, Santiago Chile Phone: (56)(2) 978 74 11 Fax : (56)(2) 271 29 73 http://fisica.ciencias.uchile.cl/~emenendez/ From w2agz at pacbell.net Fri Nov 2 03:24:51 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Thu, 1 Nov 2007 19:24:51 -0700 Subject: [Pw_forum] negative frequencies during phonon calculations In-Reply-To: <6502ee650710250130t2523ac87ic6e2cbc759fbf5d4@mail.gmail.com> References: <6502ee650710250130t2523ac87ic6e2cbc759fbf5d4@mail.gmail.com> Message-ID: <011801c81cf7$8cf73450$a6e59cf0$@net> Dear C. H., It would be helpful if we knew a little bit more about your system. Is it a metal? How many irreducible representations? The Eliashberg formalism for e-p interactions in metals involves a "double delta" integration over the Fermi energy dispersion at k-whatever, depends on the broadening (screening) parameter chosen, and converges excruciatingly.repeat painfully.slow. As an example, see the following page from my web site (http://www.w2agz.com/snxaps07.htm) which contains the APS March Meeting talk I gave earlier this year on my (feeble) attempt to compute the gamma-point coupling in monoclinic (24 IRs !), 300 millikelvin superconductor, polysulfur nitride, (SN)x. I used a tr2_ph = 1e-11, and you'll see the lowest two values of omega are negative (open my talk). The computation took a day on a 4-year old dual Xeon box. I figure a "serious" calculation would require tr2_ph = 1e-14 at least, or a computer on the scale of a Blue Gene. Sidebar for the ph.x developers: It would be neat to be able to set a "suspension" mid-execution flag to be able to resume a long run later (maybe even on another computer). Having said all this, Quantum-Espresso is the only free package I know that even attempts to calculate alpha-squared F(omega). Viva Italia. Regards, -Paul Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford University EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of C.H. Hu Sent: Thursday, October 25, 2007 1:31 AM To: PWSCF Forum Subject: [Pw_forum] negative frequencies during phonon calculations Dear Pwscf users, I performed gamma-point phonon calculation for my system only including some light elements. Some negative frequencies appear as following. The structure has been optimized first and the nscf calculation was also performed. Does the negative frequencies indicate the structural instability? or how to eliminate these imaginary frequencies during my following calculations? ========= q = ( 0.000000000 0.000000000 0.000000000 ) ************************************************************************** omega( 1) = -3.014004 [THz] = -100.537026 [cm-1] omega( 2) = -1.833969 [THz] = -61.175042 [cm-1] omega( 3) = -1.568539 [THz] = -52.321189 [cm-1] omega( 4) = -1.362118 [THz] = -45.435661 [cm-1] omega( 5) = -0.979698 [THz] = -32.679419 [cm-1] omega( 6) = 3.631786 [THz] = 121.144148 [cm-1] omega( 7) = 4.565661 [THz] = 152.295057 [cm-1] omega( 8) = 4.870748 [THz] = 162.471753 [cm-1] omega( 9) = 5.926574 [THz] = 197.690558 [cm-1] omega(10) = 6.692205 [THz] = 223.229420 [cm-1] omega(11) = 6.751437 [THz] = 225.205202 [cm-1] omega(12) = 7.152446 [THz] = 238.581500 [cm-1] omega(13) = 7.290775 [THz] = 243.195689 [cm-1] omega(14) = 7.583548 [THz] = 252.961623 [cm-1] omega(15) = 7.848509 [THz] = 261.799816 [cm-1] omega(16) = 8.216032 [THz] = 274.059149 [cm-1] omega(17) = 8.450907 [THz] = 281.893782 [cm-1] omega(18) = 8.617843 [THz] = 287.462209 [cm-1] omega(19) = 8.874576 [THz] = 296.025963 [cm-1] omega(20) = 9.106260 [THz] = 303.754152 [cm-1] omega(21) = 9.776634 [THz] = 326.115571 [cm-1] omega(22) = 9.846084 [THz] = 328.432188 [cm-1] omega(23) = 10.441744 [THz] = 348.301414 [cm-1] omega(24) = 10.901272 [THz] = 363.629712 [cm-1] omega(25) = 11.120174 [THz] = 370.931537 [cm-1] omega(26) = 11.296950 [THz] = 376.828190 [cm-1] omega(27) = 11.744897 [THz] = 391.770187 [cm-1] omega(28) = 11.823415 [THz] = 394.389308 [cm-1] omega(29) = 12.284723 [THz] = 409.776965 [cm-1] omega(30) = 12.879001 [THz] = 429.600078 [cm-1] omega(31) = 13.104238 [THz] = 437.113223 [cm-1] omega(32) = 13.295358 [THz] = 443.488354 [cm-1] omega(33) = 13.386218 [THz] = 446.519146 [cm-1] omega(34) = 13.536055 [THz] = 451.517199 [cm-1] omega(35) = 13.571174 [THz] = 452.688634 [cm-1] omega(36) = 14.064683 [THz] = 469.150452 [cm-1] omega(37) = 14.556450 [THz] = 485.554120 [cm-1] omega(38) = 14.667300 [THz] = 489.251729 [cm-1] omega(39) = 15.110352 [THz] = 504.030430 [cm-1] omega(40) = 15.395285 [THz] = 513.534849 [cm-1] omega(41) = 16.043631 [THz] = 535.161476 [cm-1] omega(42) = 16.936124 [THz] = 564.932033 [cm-1] omega(43) = 18.733204 [THz] = 624.876561 [cm-1] omega(44) = 18.876424 [THz] = 629.653928 [cm-1] omega(45) = 19.992373 [THz] = 666.878205 [cm-1] omega(46) = 21.505666 [THz] = 717.356559 [cm-1] omega(47) = 21.617840 [THz] = 721.098322 [cm-1] omega(48) = 22.309307 [THz] = 744.163333 [cm-1] omega(49) = 23.662551 [THz] = 789.303000 [cm-1] omega(50) = 24.061062 [THz] = 802.595969 [cm-1] omega(51) = 24.064277 [THz] = 802.703198 [cm-1] omega(52) = 25.648159 [THz] = 855.536168 [cm-1] omega(53) = 26.862084 [THz] = 896.028637 [cm-1] omega(54) = 26.984089 [THz] = 900.098318 [cm-1] omega(55) = 28.413367 [THz] = 947.774196 [cm-1] omega(56) = 28.483349 [THz] = 950.108578 [cm-1] omega(57) = 28.903442 [THz] = 964.121470 [cm-1] omega(58) = 29.425892 [THz] = 981.548629 [cm-1] omega(59) = 29.797340 [THz] = 993.938893 [cm-1] omega(60) = 30.194172 [THz] = 1007.175875 [cm-1] omega(61) = 30.533468 [THz] = 1018.493632 [cm-1] omega(62) = 30.805908 [THz] = 1027.581303 [cm-1] omega(63) = 31.041973 [THz] = 1035.455634 [cm-1] omega(64) = 32.094778 [THz] = 1070.573675 [cm-1] omega(65) = 33.759643 [THz] = 1126.107965 [cm-1] omega(66) = 33.969486 [THz] = 1133.107616 [cm-1] omega(67) = 36.434092 [THz] = 1215.318565 [cm-1] omega(68) = 36.726631 [THz] = 1225.076683 [cm-1] omega(69) = 39.522021 [THz] = 1318.321478 [cm-1] omega(70) = 39.623203 [THz] = 1321.696562 [cm-1] omega(71) = 40.034020 [THz] = 1335.400026 [cm-1] omega(72) = 40.044563 [THz] = 1335.751729 [cm-1] omega(73) = 50.292137 [THz] = 1677.576257 [cm-1] omega(74) = 50.298606 [THz] = 1677.792065 [cm-1] omega(75) = 51.050051 [THz] = 1702.857719 [cm-1] omega(76) = 51.176742 [THz] = 1707.083716 [cm-1] omega(77) = 53.131701 [THz] = 1772.294552 [cm-1] omega(78) = 53.564853 [THz] = 1786.743033 [cm-1] omega(79) = 53.946325 [THz] = 1799.467683 [cm-1] omega(80) = 54.547644 [THz] = 1819.525644 [cm-1] omega(81) = 55.855114 [THz] = 1863.138433 [cm-1] omega(82) = 56.009839 [THz] = 1868.299538 [cm-1] omega(83) = 56.878920 [THz] = 1897.289172 [cm-1] omega(84) = 56.962499 [THz] = 1900.077075 [cm-1] omega(85) = 57.147227 [THz] = 1906.238980 [cm-1] omega(86) = 57.294672 [THz] = 1911.157244 [cm-1] omega(87) = 58.113524 [THz] = 1938.471399 [cm-1] omega(88) = 58.186280 [THz] = 1940.898280 [cm-1] omega(89) = 60.420046 [THz] = 2015.409202 [cm-1] omega(90) = 60.431210 [THz] = 2015.781574 [cm-1] ************************************************************************** ========= Regards, C.H. -- ============================ C.H. Hu Postdoctoral fellow Chimie et Physico-Chimie appliquees Institut Francais du Petrole (IFP) Rueil-Malmaison, France E_mail: chaohao.mse at gmail.com ============================ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071101/197f51da/attachment-0001.htm From oulihui666 at 126.com Fri Nov 2 03:26:32 2007 From: oulihui666 at 126.com (oulihui666) Date: Fri, 2 Nov 2007 10:26:32 +0800 (CST) Subject: [Pw_forum] some naive problems Message-ID: <6150849.784381193970392340.JavaMail.coremail@bj126app103.126.com> Dear pwscf users, Recently, I encounter some problems, Do pwscf only periodic hybrid density functional theory calculation, it could do any other general density functional theory calculation? Is there any principle about setting of k-point and cut-off energy? Best wishes Lihui Ou -- ====================================== Lihui Ou PH.D Candidate in Electrochemistry College of Chemistry and Molecular Science Wuhan University,430072,Hubei Province,China E-mail:oulihui666 at 126.com ====================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071102/0c11513b/attachment.htm From baroni at sissa.it Fri Nov 2 03:35:41 2007 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 2 Nov 2007 11:35:41 +0900 Subject: [Pw_forum] question about the psi^2 projected on certain band. In-Reply-To: <4729EABD.5040303@sissa.it> References: <393885684.16066@mail.sdu.edu.cn> <4729EABD.5040303@sissa.it> Message-ID: <89F0A43C-3A37-4D99-9A19-86247F23E3A7@sissa.it> Another neat way of answering your question is to compare |psi (Gamma,r)|^2 with the local density of states integrated from just below to just above the edges of your localized band. I am sure that this kind of calculation has been done many times by people using Quantum ESPRESSO. I am not sure whether or not a ready-to-use tool is included in the QE distribution, though. Others may know more about this in this forum. Cheers - Stefano Baroni On Nov 2, 2007, at 12:03 AM, Stefano de Gironcoli wrote: > your problem looks to me suited for a Wannier decomposition of the > band > structure. > Have a look at www.wannier.org > It is interfaced with QE > Stefano de Gironcoli -SISSA and DEMOCRITOS > > Paolo Giannozzi wrote: >> On Nov 1, 2007, at 3:54 , ??? wrote: >> >> >>> During the post-process, psi^2 projected on a certain band is >>> necessary for us. >>> However, pp.x with plot_num=7 gives such image on gamma point only. >>> >> >> are you sure? I don't think so >> >> >>> My question is: if a certain band is much localized (such as the 3d >>> states >>> in some transition metal oxide or localized donor states in the >>> semiconductor >>> gap), does psi^2 on gamma represent the spatial charge distribution >>> of the >>> whole localized band? >>> >> >> for a narrow band formed by localised states, I expect that \psi^2 >> doesn't change much along the band >> >> Paolo >> --- >> Paolo Giannozzi, Dept of Physics, University of Udine >> via delle Scienze 208, 33100 Udine, Italy >> Phone +39-0432-558216, fax +39-0432-558222 >> >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071102/25a0697c/attachment-0001.htm From mbaris at metu.edu.tr Fri Nov 2 09:17:31 2007 From: mbaris at metu.edu.tr (mbaris at metu.edu.tr) Date: Fri, 02 Nov 2007 10:17:31 +0200 Subject: [Pw_forum] some naive problems Message-ID: <20071102101731.subaibp7e1kcog0w@webmail2.metu.edu.tr> Dear Lihui Ou, > Do pwscf only periodic hybrid density functional theory calculation, > it could do any other general density functional theory calculation? I am not sure if I got the point of your question correctly, but Plane Wave SCF program can only solve systems thorough plane wave approximation, i.e. not through an all-electron calculation, and for simplification purposes (that is, in order to have tangible Fourier transforms etc.) periodicity/symmetries are employed. This is more than acceptable for crystals, but for macromolecules etc, you should be extra careful setting the boundaries (the most simple limit is this: a very big box surrounding the molecule you are interested in will somewhat approximate a standalone molecule, however the size of the box is limited by a combination of how the plane wave approximation is employed in pwscf and bare computational limits, you need to find the optimum value). There are other packages available such as NWChem or Gaussian for all-electron calculations. Pseudo potential wise, you can create a pseudo potential to your desire, as long as you can validify your results. Please see the documentation. > Is there any principle about setting of k-point and cut-off energy? Well, yes. The idea is: You should minimize the number of k-points and cut-off (and a number of other parameters depending on the system you are studying) in order to save computer time whilst maintaining physically meaningful results. The "physically meaningful result" part depends on your problem, keep in mind that total energy in DFT is not very well defined, due to nature of Kohn-Sham orbitals, only the differences in energy are trustworthy, or the observables, such as unit cell volume. Generally you are advised to do a number of calculations with varied parameters in order to obtain a converged set before starting the problem itself (i.e. if you are using lattice parameter as your observable, find the lattice parameter that minimizes the energy for each ecutwfc-k-point combination. You will see that after a certain limit, your lattice parameter will not change much, that value is the minimum you can use, you may later need to increase it, depending on your problem). > Best wishes > Lihui Ou Hope this was helpful, Best, Osman Baris Malcioglu Ph.D. Candidate METU, Physics Ankara, Turkey From degironc at sissa.it Fri Nov 2 11:35:57 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Fri, 02 Nov 2007 11:35:57 +0100 Subject: [Pw_forum] some naive problems In-Reply-To: <20071102101731.subaibp7e1kcog0w@webmail2.metu.edu.tr> References: <20071102101731.subaibp7e1kcog0w@webmail2.metu.edu.tr> Message-ID: <472AFD8D.1080008@sissa.it> Dear Lihui Ou and Osman Baris Malcioglu, let me re-state Osman's reply in a way I like more... Plane-wave codes (and for that matters also LMTO, FLAPW codes etc) are built so as to describe periodic systems and CAN deal with isolated molecules by the use of supercells large enough that the periodic images do not interact significantly. There are other codes (I guess NWChem is one of them) that work for isolated system only and CANNOT describe periodic systems. This may be a problem when wishing to compare isolated and periodic systems on the same footing... A plane-wave basis-set is an expensive one but is unbiased (describe with the same accuracy every point in space) and its accuracy is extremely easy to control (just increase your cutoff energy until the property you are interested in converges). A lot of effort in the past has been put in the development of pseuodopotential theory that allows to treat any element in the periodic table with good accuracy (some already using norm conserving pseudos while some others may need ultrasoft ones). Pseudopotential method is not an all-electron method but very often the results are very close to AE and always, to my experience, physically meaningful. An evolution of the US pseudopotential method is the PAW formalism that can be considered an all-electron method. We are currently implementing this formalism in PWscf. Localized basis (such as gaussian or slater-type-orbitals) are more compact but are more difficult to control and require a lot of know-how and experience (double-z ? triple-z ? triple-z+polarization ?) in order to avoid basis-set superposition error and the like. Parameters controlling plane-wave calculations: 1) kinetic energy cutoff: as said it defines the dimension of the basis set and you just need to increase it until you are satisfied. The required cutoff is a PROPERTY OF THE PSEUDOPOTENTIAL USED not of the particular system under study. It stems from the need of describing accurately the pseudopotential wavefunctions. 2) k-point sampling : it's a property of your system (insulator, metal, isolated, 2-d, 1-d) not particularly of the pseudopotential used. For isolated molecules (that is if the supercell is large enough and hence the IBZ small enough) Gamma-only sampling (for which special tricks to speed up the calculation can be used) is a good choice. If it is not the case, this means that periodic images are interacting... then increase the cell dimension and stick to Gamma sampling. As for Hybrid functionals: currently only NC pseudopotentials are implemented, sooner or later US (or most probably directly PAW) implementation will follow... any skilled volunteer would be welcome. hope this helps, stefano de Gironcoli, SISSA and DEMOCRITOS mbaris at metu.edu.tr wrote: > Dear Lihui Ou, > > >> Do pwscf only periodic hybrid density functional theory calculation, >> it could do any other general density functional theory calculation? >> > I am not sure if I got the point of your question correctly, but Plane > Wave SCF program can only solve systems thorough plane wave > approximation, i.e. not through an all-electron calculation, and for > simplification purposes (that is, in order to have tangible Fourier > transforms etc.) periodicity/symmetries are employed. This is more than > acceptable for crystals, but for macromolecules etc, you should be > extra careful setting the boundaries (the most simple limit is this: a > very big box surrounding the molecule you are interested in will > somewhat approximate a standalone molecule, however the size of the box > is limited by a combination of how the plane wave approximation is > employed in pwscf and bare computational limits, you need to find the > optimum value). There are other packages available such as NWChem or > Gaussian for all-electron calculations. Pseudo potential wise, you can > create a pseudo potential to your desire, as long as you can validify > your results. Please see the documentation. > > >> Is there any principle about setting of k-point and cut-off energy? >> > > Well, yes. The idea is: You should minimize the number of k-points and > cut-off (and a number of other parameters depending on the system you > are studying) in order to save computer time whilst maintaining > physically meaningful results. The "physically meaningful result" part > depends on your problem, keep in mind that total energy in DFT is not > very well defined, due to nature of Kohn-Sham orbitals, only the > differences in energy are trustworthy, or the observables, such as unit > cell volume. Generally you are advised to do a number of calculations > with varied parameters in order to obtain a converged set before > starting the problem itself (i.e. if you are using lattice parameter as > your observable, find the lattice parameter that minimizes the energy > for each ecutwfc-k-point combination. You will see that after a certain > limit, your lattice parameter will not change much, that value is the > minimum you can use, you may later need to increase it, depending on > your problem). > > > >> Best wishes >> Lihui Ou >> > > Hope this was helpful, > > Best, > > Osman Baris Malcioglu > Ph.D. Candidate > METU, Physics > Ankara, Turkey > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From emenendez at macul.ciencias.uchile.cl Fri Nov 2 11:57:08 2007 From: emenendez at macul.ciencias.uchile.cl (Eduardo Ariel Menendez P) Date: Fri, 2 Nov 2007 07:57:08 -0300 (CLST) Subject: [Pw_forum] fixing spin in LSD In-Reply-To: References: Message-ID: > Other case is dimers and clusters. I have answered this question to > senior quantum chemists and I have received answers like I meant that I have asked the question to senior quantum chemists. From degironc at sissa.it Fri Nov 2 13:11:41 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Fri, 02 Nov 2007 13:11:41 +0100 Subject: [Pw_forum] fixing spin in LSD In-Reply-To: References: Message-ID: <472B13FD.5010304@sissa.it> For a magnetic system the code looks for a self-consistent solution. In many cases this solution is easily found, in some other cases it may be difficult. This may be due to the fact that different solutions (with different magnetizations) are possible and the code is not able to choose. A possible strategy then is to constrain the magnetization to several values and calculate the scf solution for each. This allows to determine, a posteriori, the GS of the system and also have some idea of the low-lying alternatives that may be of some physical interest. Constraining the magnetization is one of the several tools we have in order to ask questions to the system.... answers will be meaningful, if the questions are. Stefano de Gironcoli, SISSA and DEMOCRITOS Eduardo Ariel Menendez P wrote: > Hi, > > My doubt may be trivial,as I cannot find the answer anywhere. > > When does one needs to fix the value of the spin in LSDA calculation? > > multiplicity INTEGER ( default = 0 [unspecified] ) spin > multiplicity (2s+1). 1 is singlet, 2 for doublet etc. Note > that this fixes the final value of the magnetization. if > unspecified or a non-zero value is specified in nelup/neldw > then multiplicity variable is ignored. > > In particular, I am interested in the case of impurities in surfaces. > Should I fix the spin in this case or not? > > Other case is dimers and clusters. I have answered this question to > senior quantum chemists and I have received answers like > > a) Whith Hartree-Fock equations you must specify the spin, and DFT works > well when HF does, and fails when HF fails. > > b) The states measured by spectroscopy have well defined S_z and S^2. So > you want to calculate these states. > > Is there any other (and more fundamental) reason? > > > Polite request: > Avoid send Word attachments. To see why, take 6 minutes to read at > http://www.gnu.org/philosophy/no-word-attachments.html > > Eduardo A. Menendez Proupin > Department of Physics > Faculty of Science > University of Chile > Las Palmeras 3425 > ?u?oa, Santiago > Chile > Phone: (56)(2) 978 74 11 > Fax : (56)(2) 271 29 73 > http://fisica.ciencias.uchile.cl/~emenendez/ > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From chaohao.mse at gmail.com Fri Nov 2 15:09:58 2007 From: chaohao.mse at gmail.com (C.H. Hu) Date: Fri, 2 Nov 2007 15:09:58 +0100 Subject: [Pw_forum] negative frequencies during phonon calculations In-Reply-To: <011801c81cf7$8cf73450$a6e59cf0$@net> References: <6502ee650710250130t2523ac87ic6e2cbc759fbf5d4@mail.gmail.com> <011801c81cf7$8cf73450$a6e59cf0$@net> Message-ID: <6502ee650711020709t607bebacp80fcd22c5f5d6484@mail.gmail.com> Dear P.M. Grant, Thank you for your useful presentation. No, The system I consider is an insulator. Even tr2_ph was set to 1e-12, the negative frequency still can not be eliminated. Maybe it must be further reduced as you have said. But it will spend more computational time. Regards, C.H. On 11/2/07, Paul M. Grant wrote: > > Dear C. H., > > > > It would be helpful if we knew a little bit more about your system. Is it > a metal? How many irreducible representations? > > > > The Eliashberg formalism for e-p interactions in metals involves a "double > delta" integration over the Fermi energy dispersion at k-whatever, depends > on the broadening (screening) parameter chosen, and converges > excruciatingly?repeat painfully?slow. As an example, see the following page > from my web site (http://www.w2agz.com/snxaps07.htm) which contains the > APS March Meeting talk I gave earlier this year on my (feeble) attempt to > compute the gamma-point coupling in monoclinic (24 IRs !), 300 millikelvin > superconductor, polysulfur nitride, (SN)x. I used a tr2_ph = 1e-11, and > you'll see the lowest two values of omega are negative (open my talk). The > computation took a day on a 4-year old dual Xeon box. I figure a "serious" > calculation would require tr2_ph = 1e-14 at least, or a computer on the > scale of a Blue Gene. > > > > Sidebar for the ph.x developers: It would be neat to be able to set a > "suspension" mid-execution flag to be able to resume a long run later (maybe > even on another computer). > > > > Having said all this, Quantum-Espresso is the only free package I know > that even attempts to calculate alpha-squared F(omega). Viva Italia. > > > > Regards, -Paul > > ** > -- ============================ C.H. Hu Postdoctoral fellow Chimie et Physico-Chimie appliquees Institut Francais du Petrole (IFP) Rueil-Malmaison, France E_mail: chaohao.mse at gmail.com ============================ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071102/3bb83245/attachment.htm From luotengf at msu.edu Fri Nov 2 17:19:11 2007 From: luotengf at msu.edu (Tengfei Luo) Date: Fri, 2 Nov 2007 12:19:11 -0400 Subject: [Pw_forum] ncpus exceed limit while runing pw.x using mpi Message-ID: <00a401c81d6c$1a4804a0$4e8a0923@engineering.egr.msu.edu> Dear All: I use to run pw.x with mpi with no problem. However, I recently keep getting the "ncpus exceed limit" message. Our system is a SGI Altix 3700 with mpi. I used the following pbs file to submit my job: #!/bin/sh #PBS -l ncpus=4,walltime=1:00:00,mem=8gb #PBS -M luotengf at msu.edu #PBS -m abe #PBS -N GaAs_pw #PBS -p 500 #PBS -r n cd ~/espresso-3.2.2/calculation/SAM mpirun -np 4 ~/espresso-3.2.2/bin/pw.x -npool 1 < input.in > output.out I wonder if this is a problem with the system or a problem of the grogram or a problem with my input setting. Here is my input file: &control calculation = 'scf' restart_mode='restart', wf_collect=.true. prefix='GaAs', tstress = .true. tprnfor = .true. pseudo_dir = '/mnt/home/luotengf/espresso-3.2.2/pseudo/', outdir='/mnt/home/luotengf/espresso-3.2.2/calculation/SAM/Temp' / &system ibrav=1,celldm(1)=21.3664, nat=64,ntyp=2,ecutwfc =18.0, / &electrons diagonalization='cg' mixing_mode = 'plain' mixing_beta = 0.4 conv_thr = 1.0d-8 / ATOMIC_SPECIES Ga 69.723 Ga.pz-bhs.UPF As 74.92160 As.pz-bhs.UPF ATOMIC_POSITIONS (angstrom) Ga 2.82664990425110 5.65329980850220 2.82664990425110 Ga 5.65329980850220 2.82664990425110 2.82664990425110 Ga 2.82664990425110 2.82664990425110 0.00000000000000 Ga 5.65329980850220 5.65329980850220 0.00000000000000 As 4.23997497558594 1.41332495212555 4.23997497558594 As 1.41332495212555 1.41332495212555 1.41332495212555 As 1.41332495212555 4.23997497558594 4.23997497558594 As 4.23997497558594 4.23997497558594 1.41332495212555 Ga 8.47994995117188 5.65329980850220 2.82664990425110 Ga 11.30659961700439 2.82664990425110 2.82664990425110 Ga 8.47994995117188 2.82664990425110 0.00000000000000 Ga 11.30659961700439 5.65329980850220 0.00000000000000 As 9.89327430725098 1.41332495212555 4.23997497558594 As 7.06662464141846 1.41332495212555 1.41332495212555 As 7.06662464141846 4.23997497558594 4.23997497558594 As 9.89327430725098 4.23997497558594 1.41332495212555 Ga 2.82664990425110 11.30659961700439 2.82664990425110 Ga 5.65329980850220 8.47994995117188 2.82664990425110 Ga 2.82664990425110 8.47994995117188 0.00000000000000 Ga 5.65329980850220 11.30659961700439 0.00000000000000 As 4.23997497558594 7.06662464141846 4.23997497558594 As 1.41332495212555 7.06662464141846 1.41332495212555 As 1.41332495212555 9.89327430725098 4.23997497558594 As 4.23997497558594 9.89327430725098 1.41332495212555 Ga 8.47994995117188 11.30659961700439 2.82664990425110 Ga 11.30659961700439 8.47994995117188 2.82664990425110 Ga 8.47994995117188 8.47994995117188 0.00000000000000 Ga 11.30659961700439 11.30659961700439 0.00000000000000 As 9.89327430725098 7.06662464141846 4.23997497558594 As 7.06662464141846 7.06662464141846 1.41332495212555 As 7.06662464141846 9.89327430725098 4.23997497558594 As 9.89327430725098 9.89327430725098 1.41332495212555 Ga 2.82664990425110 5.65329980850220 8.47994995117188 Ga 5.65329980850220 2.82664990425110 8.47994995117188 Ga 2.82664990425110 2.82664990425110 5.65329980850220 Ga 5.65329980850220 5.65329980850220 5.65329980850220 As 4.23997497558594 1.41332495212555 9.89327430725098 As 1.41332495212555 1.41332495212555 7.06662464141846 As 1.41332495212555 4.23997497558594 9.89327430725098 As 4.23997497558594 4.23997497558594 7.06662464141846 Ga 8.47994995117188 5.65329980850220 8.47994995117188 Ga 11.30659961700439 2.82664990425110 8.47994995117188 Ga 8.47994995117188 2.82664990425110 5.65329980850220 Ga 11.30659961700439 5.65329980850220 5.65329980850220 As 9.89327430725098 1.41332495212555 9.89327430725098 As 7.06662464141846 1.41332495212555 7.06662464141846 As 7.06662464141846 4.23997497558594 9.89327430725098 As 9.89327430725098 4.23997497558594 7.06662464141846 Ga 2.82664990425110 11.30659961700439 8.47994995117188 Ga 5.65329980850220 8.47994995117188 8.47994995117188 Ga 2.82664990425110 8.47994995117188 5.65329980850220 Ga 5.65329980850220 11.30659961700439 5.65329980850220 As 4.23997497558594 7.06662464141846 9.89327430725098 As 1.41332495212555 7.06662464141846 7.06662464141846 As 1.41332495212555 9.89327430725098 9.89327430725098 As 4.23997497558594 9.89327430725098 7.06662464141846 Ga 8.47994995117188 11.30659961700439 8.47994995117188 Ga 11.30659961700439 8.47994995117188 8.47994995117188 Ga 8.47994995117188 8.47994995117188 5.65329980850220 Ga 11.30659961700439 11.30659961700439 5.65329980850220 As 9.89327430725098 7.06662464141846 9.89327430725098 As 7.06662464141846 7.06662464141846 7.06662464141846 As 7.06662464141846 9.89327430725098 9.89327430725098 As 9.89327430725098 9.89327430725098 7.06662464141846 K_POINTS (automatic) 4 4 4 0 0 0 I would appreciate any comment! Tengfei 3547 E, Engineering Building Michigan State University East Lansing, MI, 48824 Tel: 517-355-1693 Email: luotengf at msu.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071102/904e331d/attachment-0001.htm From naromero at gmail.com Fri Nov 2 18:21:51 2007 From: naromero at gmail.com (Nichols A. Romero) Date: Fri, 2 Nov 2007 13:21:51 -0400 Subject: [Pw_forum] ncpus exceed limit while runing pw.x using mpi In-Reply-To: <00a401c81d6c$1a4804a0$4e8a0923@engineering.egr.msu.edu> References: <00a401c81d6c$1a4804a0$4e8a0923@engineering.egr.msu.edu> Message-ID: <6ac064b60711021021m4f415aafo8f16dd8ea9d7795@mail.gmail.com> It is unlikely (to say the least) that this is a problem with PWSCF. There is mostly a problem with the batch queuing system on this machine. Try looking at the man page for PBS. On 11/2/07, Tengfei Luo wrote: > > Dear All: > I use to run pw.x with mpi with no problem. However, I recently keep > getting the "ncpus exceed limit" message. Our system is a SGI Altix 3700 > with mpi. > > I used the following pbs file to submit my job: > #!/bin/sh > #PBS -l ncpus=4,walltime=1:00:00,mem=8gb > #PBS -M luotengf at msu.edu > #PBS -m abe > #PBS -N GaAs_pw > #PBS -p 500 > #PBS -r n > cd ~/espresso-3.2.2/calculation/SAM > mpirun -np 4 ~/espresso-3.2.2/bin/pw.x -npool 1 < input.in > output.out > > I wonder if this is a problem with the system or a problem of the grogram > or a problem with my input setting. > > Here is my input file: > &control > calculation = 'scf' > restart_mode='restart', > wf_collect=.true. > prefix='GaAs', > tstress = .true. > tprnfor = .true. > pseudo_dir = '/mnt/home/luotengf/espresso-3.2.2/pseudo/', > outdir='/mnt/home/luotengf/espresso-3.2.2/calculation/SAM/Temp' > / > &system > ibrav=1,celldm(1)=21.3664, > nat=64,ntyp=2,ecutwfc =18.0, > / > &electrons > diagonalization='cg' > mixing_mode = 'plain' > mixing_beta = 0.4 > conv_thr = 1.0d-8 > / > ATOMIC_SPECIES > Ga 69.723 Ga.pz-bhs.UPF > As 74.92160 As.pz-bhs.UPF > ATOMIC_POSITIONS (angstrom) > Ga 2.82664990425110 5.65329980850220 2.82664990425110 > Ga 5.65329980850220 2.82664990425110 2.82664990425110 > Ga 2.82664990425110 2.82664990425110 0.00000000000000 > Ga 5.65329980850220 5.65329980850220 0.00000000000000 > As 4.23997497558594 1.41332495212555 4.23997497558594 > As 1.41332495212555 1.41332495212555 1.41332495212555 > As 1.41332495212555 4.23997497558594 4.23997497558594 > As 4.23997497558594 4.23997497558594 1.41332495212555 > Ga 8.47994995117188 5.65329980850220 2.82664990425110 > Ga 11.30659961700439 2.82664990425110 2.82664990425110 > Ga 8.47994995117188 2.82664990425110 0.00000000000000 > Ga 11.30659961700439 5.65329980850220 0.00000000000000 > As 9.89327430725098 1.41332495212555 4.23997497558594 > As 7.06662464141846 1.41332495212555 1.41332495212555 > As 7.06662464141846 4.23997497558594 4.23997497558594 > As 9.89327430725098 4.23997497558594 1.41332495212555 > Ga 2.82664990425110 11.30659961700439 2.82664990425110 > Ga 5.65329980850220 8.47994995117188 2.82664990425110 > Ga 2.82664990425110 8.47994995117188 0.00000000000000 > Ga 5.65329980850220 11.30659961700439 0.00000000000000 > As 4.23997497558594 7.06662464141846 4.23997497558594 > As 1.41332495212555 7.06662464141846 1.41332495212555 > As 1.41332495212555 9.89327430725098 4.23997497558594 > As 4.23997497558594 9.89327430725098 1.41332495212555 > Ga 8.47994995117188 11.30659961700439 2.82664990425110 > Ga 11.30659961700439 8.47994995117188 2.82664990425110 > Ga 8.47994995117188 8.47994995117188 0.00000000000000 > Ga 11.30659961700439 11.30659961700439 0.00000000000000 > As 9.89327430725098 7.06662464141846 4.23997497558594 > As 7.06662464141846 7.06662464141846 1.41332495212555 > As 7.06662464141846 9.89327430725098 4.23997497558594 > As 9.89327430725098 9.89327430725098 1.41332495212555 > Ga 2.82664990425110 5.65329980850220 8.47994995117188 > Ga 5.65329980850220 2.82664990425110 8.47994995117188 > Ga 2.82664990425110 2.82664990425110 5.65329980850220 > Ga 5.65329980850220 5.65329980850220 5.65329980850220 > As 4.23997497558594 1.41332495212555 9.89327430725098 > As 1.41332495212555 1.41332495212555 7.06662464141846 > As 1.41332495212555 4.23997497558594 9.89327430725098 > As 4.23997497558594 4.23997497558594 7.06662464141846 > Ga 8.47994995117188 5.65329980850220 8.47994995117188 > Ga 11.30659961700439 2.82664990425110 8.47994995117188 > Ga 8.47994995117188 2.82664990425110 5.65329980850220 > Ga 11.30659961700439 5.65329980850220 5.65329980850220 > As 9.89327430725098 1.41332495212555 9.89327430725098 > As 7.06662464141846 1.41332495212555 7.06662464141846 > As 7.06662464141846 4.23997497558594 9.89327430725098 > As 9.89327430725098 4.23997497558594 7.06662464141846 > Ga 2.82664990425110 11.30659961700439 8.47994995117188 > Ga 5.65329980850220 8.47994995117188 8.47994995117188 > Ga 2.82664990425110 8.47994995117188 5.65329980850220 > Ga 5.65329980850220 11.30659961700439 5.65329980850220 > As 4.23997497558594 7.06662464141846 9.89327430725098 > As 1.41332495212555 7.06662464141846 7.06662464141846 > As 1.41332495212555 9.89327430725098 9.89327430725098 > As 4.23997497558594 9.89327430725098 7.06662464141846 > Ga 8.47994995117188 11.30659961700439 8.47994995117188 > Ga 11.30659961700439 8.47994995117188 8.47994995117188 > Ga 8.47994995117188 8.47994995117188 5.65329980850220 > Ga 11.30659961700439 11.30659961700439 5.65329980850220 > As 9.89327430725098 7.06662464141846 9.89327430725098 > As 7.06662464141846 7.06662464141846 7.06662464141846 > As 7.06662464141846 9.89327430725098 9.89327430725098 > As 9.89327430725098 9.89327430725098 7.06662464141846 > K_POINTS (automatic) > 4 4 4 0 0 0 > I would appreciate any comment! > > Tengfei > 3547 E, Engineering Building > Michigan State University > East Lansing, MI, 48824 > Tel: 517-355-1693 > Email: luotengf at msu.edu > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Nichols A. Romero, Ph.D. DoD User Productivity Enhancement and Technology Transfer (PET) Group High Performance Technologies, Inc. Reston, VA 443-567-8328 (C) 410-278-2692 (O) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071102/f779ff04/attachment.htm From naromero at gmail.com Fri Nov 2 21:08:15 2007 From: naromero at gmail.com (Nichols A. Romero) Date: Fri, 2 Nov 2007 16:08:15 -0400 Subject: [Pw_forum] Davidson algorithm Message-ID: <6ac064b60711021308y2d1f4bb6pe85700ee373373c6@mail.gmail.com> Hi, Can anyone recommend a basic reference (e.g. thesis, personal notes, computional article) that explains the Davidson algorithm used in PWSCF in more detail? P.S. I promise to add any useful information to the Developer Manual section of the QE wiki :^) -- Nichols A. Romero, Ph.D. DoD User Productivity Enhancement and Technology Transfer (PET) Group High Performance Technologies, Inc. Reston, VA 443-567-8328 (C) 410-278-2692 (O) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071102/a0e9e7ad/attachment.htm From w2agz at pacbell.net Fri Nov 2 21:40:14 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Fri, 2 Nov 2007 13:40:14 -0700 Subject: [Pw_forum] some naive problems In-Reply-To: <472AFD8D.1080008@sissa.it> References: <20071102101731.subaibp7e1kcog0w@webmail2.metu.edu.tr> <472AFD8D.1080008@sissa.it> Message-ID: <016501c81d90$92c94820$b85bd860$@net> Just to add to this general discussion...which I think is very useful...when I take up a new problem I write two scf scripts, one which increments the size of the Monkhorst-Pack (K_POINTS) net until the total energy "sort of" settles down (I say "sort of," because sometimes there are small oscillations when the net gets big. Then I write another script which takes the "best" MP size increments the cut-off upward until I'm happy. If I think I need to find new "equilibrium" lattice constants, I'll write a third script which varies these. Then I use all this to start computing observables. I don't think I've ever used the gamma option. Re psuedopotentials: I pretty much trust the ones from the PWscf library and/or from Vanderbilt. On the "na?ve" side. Most of the systems I investigate are far from cubic symmetry, and I choose the number of K_POINTS in each direction proportionately to reflect the way the dimensions of the Brillouin zone vary inversely with those of the Wigner-Seitz cell. What do others do? Also, I've never found adding an "offset" to the MP net has a significant effect...when should I expect differently? -Paul Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford University EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com ? ? -----Original Message----- From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Stefano de Gironcoli Sent: Friday, November 02, 2007 3:36 AM To: PWSCF Forum Subject: Re: [Pw_forum] some naive problems Dear Lihui Ou and Osman Baris Malcioglu, let me re-state Osman's reply in a way I like more... Plane-wave codes (and for that matters also LMTO, FLAPW codes etc) are built so as to describe periodic systems and CAN deal with isolated molecules by the use of supercells large enough that the periodic images do not interact significantly. There are other codes (I guess NWChem is one of them) that work for isolated system only and CANNOT describe periodic systems. This may be a problem when wishing to compare isolated and periodic systems on the same footing... A plane-wave basis-set is an expensive one but is unbiased (describe with the same accuracy every point in space) and its accuracy is extremely easy to control (just increase your cutoff energy until the property you are interested in converges). A lot of effort in the past has been put in the development of pseuodopotential theory that allows to treat any element in the periodic table with good accuracy (some already using norm conserving pseudos while some others may need ultrasoft ones). Pseudopotential method is not an all-electron method but very often the results are very close to AE and always, to my experience, physically meaningful. An evolution of the US pseudopotential method is the PAW formalism that can be considered an all-electron method. We are currently implementing this formalism in PWscf. Localized basis (such as gaussian or slater-type-orbitals) are more compact but are more difficult to control and require a lot of know-how and experience (double-z ? triple-z ? triple-z+polarization ?) in order to avoid basis-set superposition error and the like. Parameters controlling plane-wave calculations: 1) kinetic energy cutoff: as said it defines the dimension of the basis set and you just need to increase it until you are satisfied. The required cutoff is a PROPERTY OF THE PSEUDOPOTENTIAL USED not of the particular system under study. It stems from the need of describing accurately the pseudopotential wavefunctions. 2) k-point sampling : it's a property of your system (insulator, metal, isolated, 2-d, 1-d) not particularly of the pseudopotential used. For isolated molecules (that is if the supercell is large enough and hence the IBZ small enough) Gamma-only sampling (for which special tricks to speed up the calculation can be used) is a good choice. If it is not the case, this means that periodic images are interacting... then increase the cell dimension and stick to Gamma sampling. As for Hybrid functionals: currently only NC pseudopotentials are implemented, sooner or later US (or most probably directly PAW) implementation will follow... any skilled volunteer would be welcome. hope this helps, stefano de Gironcoli, SISSA and DEMOCRITOS mbaris at metu.edu.tr wrote: > Dear Lihui Ou, > > >> Do pwscf only periodic hybrid density functional theory calculation, >> it could do any other general density functional theory calculation? >> > I am not sure if I got the point of your question correctly, but Plane > Wave SCF program can only solve systems thorough plane wave > approximation, i.e. not through an all-electron calculation, and for > simplification purposes (that is, in order to have tangible Fourier > transforms etc.) periodicity/symmetries are employed. This is more than > acceptable for crystals, but for macromolecules etc, you should be > extra careful setting the boundaries (the most simple limit is this: a > very big box surrounding the molecule you are interested in will > somewhat approximate a standalone molecule, however the size of the box > is limited by a combination of how the plane wave approximation is > employed in pwscf and bare computational limits, you need to find the > optimum value). There are other packages available such as NWChem or > Gaussian for all-electron calculations. Pseudo potential wise, you can > create a pseudo potential to your desire, as long as you can validify > your results. Please see the documentation. > > >> Is there any principle about setting of k-point and cut-off energy? >> > > Well, yes. The idea is: You should minimize the number of k-points and > cut-off (and a number of other parameters depending on the system you > are studying) in order to save computer time whilst maintaining > physically meaningful results. The "physically meaningful result" part > depends on your problem, keep in mind that total energy in DFT is not > very well defined, due to nature of Kohn-Sham orbitals, only the > differences in energy are trustworthy, or the observables, such as unit > cell volume. Generally you are advised to do a number of calculations > with varied parameters in order to obtain a converged set before > starting the problem itself (i.e. if you are using lattice parameter as > your observable, find the lattice parameter that minimizes the energy > for each ecutwfc-k-point combination. You will see that after a certain > limit, your lattice parameter will not change much, that value is the > minimum you can use, you may later need to increase it, depending on > your problem). > > > >> Best wishes >> Lihui Ou >> > > Hope this was helpful, > > Best, > > Osman Baris Malcioglu > Ph.D. Candidate > METU, Physics > Ankara, Turkey > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From w2agz at pacbell.net Fri Nov 2 21:54:58 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Fri, 2 Nov 2007 13:54:58 -0700 Subject: [Pw_forum] Debian Platforms and Fortran Compilers Message-ID: <016601c81d92$a2532d90$e6f988b0$@net> Some info possibly useful to other PWscf users. I've found it too difficult to go through all the turmoil required to install the Intel compiler on (K)ubuntu 7.10.the install script is incredibly buggy. So I switched to gfortran, which installs effortlessly and without the additional baggage of "license verification," and found, at least for my tests (e.g., example07), there are no significant differences in cpu execution times.fractions of seconds over many hours. But, it is far less tolerant of typos in namelist strings (e.g., missing apostrophes).but, wait.this might be due to "improvements" in 3.2.3 which I also upgraded to at the same time! Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford University EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071102/5c9ab160/attachment.htm From sunshoutian168 at yahoo.com.cn Sat Nov 3 01:39:21 2007 From: sunshoutian168 at yahoo.com.cn (shoutian sun) Date: Sat, 3 Nov 2007 08:39:21 +0800 (CST) Subject: [Pw_forum] Strange error in calculation force constant with q2r.x Message-ID: <303527.70944.qm@web92002.mail.cnb.yahoo.com> >Dear all, >I have a problem in calculation force constant with q2r.x. >Here are my input files of pw.x ph.x q2r.x. I am so sad. Please tell me the reasons >Thank you so much. +++++++++++++++++++++++++++++++scf.in &CONTROL calculation = 'scf' , restart_mode = 'from_scratch' , outdir='/home/user/ecnuphyk/SUN/tmp/tmp-p1/' , pseudo_dir = '/home/user/ecnuphyk/SUN/pseudo/' , prefix='alh3' , tstress = .true. , tprnfor = .true. , etot_conv_thr = 1.0E-5 , forc_conv_thr = 1.0D-4 , / &SYSTEM ibrav = 0 , celldm(1) = 1.889725988578923203102406997093 , nat = 16 , ntyp = 2 , nspin = 1 ecutwfc = 44.0982 , / &ELECTRONS conv_thr = 1.0d-7 , mixing_beta = 0.7 , / CELL_PARAMETERS (alat) ... ATOMIC_SPECIES ... ATOMIC_POSITIONS {crystal} ... K_POINTS automatic 4 4 3 0 0 0 ++++++++++++++++++++++++++++ph.in &inputph tr2_ph=1.0d-12, prefix='alh3', ldisp=.true., nq1=4, nq2=4, nq3=3 epsil=.true. , amass(1)=26.98, amass(2)=1.0, outdir= '/home/user/ecnuphyk/SUN/tmp/tmp-p1/' , fildyn='alh3.dyn', / +++++++++++++++++++++++++++++q2r.in &input fildyn='alh3.dyn', zasr='simple', flfrc='alh3444.fc' / ++++++++++++++++++++++++++++q2r.in.out (The error ) reading grid info from file alh3.dyn0 reading force constants from file alh3.dyn1 macroscopic fields = T 4.51585 0.00000 -0.08800 0.00000 4.37104 0.00000 -0.08800 0.00000 4.43104 na= 1 2.74722 -0.27565 -0.08418 0.24590 2.66281 0.41340 -0.11507 -0.40383 2.71996 na= 2 2.74722 0.27565 -0.08418 -0.32097 2.64842 -0.40555 -0.11506 0.40383 2.71996 na= 3 2.74722 -0.27565 -0.08418 0.24590 2.66281 0.41340 -0.11507 -0.40383 2.71996 na= 4 2.74722 0.27565 -0.08418 -0.32097 2.64842 -0.40555 -0.11507 0.40383 2.71996 na= 5 -0.62424 0.00001 0.05784 -0.00008 -0.63566 -0.00071 0.06224 -0.00088 -1.42838 na= 6 -1.06619 -0.39916 0.01343 -0.40231 -0.99561 0.02342 0.01086 0.02141 -0.62960 na= 7 -1.06619 0.39915 0.01342 0.40230 -0.99561 -0.02341 0.01086 -0.02142 -0.62960 na= 8 -0.62424 -0.00002 0.05784 0.00007 -0.63566 0.00072 0.06224 0.00091 -1.42838 na= 9 -1.06613 0.39958 0.01388 0.40245 -0.99702 -0.02390 0.01136 -0.02197 -0.62981 na= 10 -1.06613 -0.39958 0.01389 -0.40245 -0.99702 0.02391 0.01136 0.02197 -0.62981 na= 11 -0.62425 0.00002 0.05784 -0.00007 -0.63567 -0.00072 0.06224 -0.00091 -1.42827 na= 12 -1.06613 0.39959 0.01389 0.40248 -0.99704 -0.02391 0.01135 -0.02197 -0.62985 na= 13 -1.06613 -0.39959 0.01388 -0.40247 -0.99704 0.02390 0.01136 0.02198 -0.62985 na= 14 -1.06619 -0.39916 0.01342 -0.40233 -0.99564 0.02341 0.01085 0.02143 -0.62963 na= 15 -1.06619 0.39916 0.01343 0.40233 -0.99563 -0.02342 0.01085 -0.02142 -0.62963 na= 16 -0.62425 -0.00001 0.05784 0.00008 -0.63567 0.00071 0.06223 0.00088 -1.42827 nqs= 1 q= 0.00000000 0.00000000 0.00000000 reading force constants from file alh3.dyn2 nqs= 2 q= 0.00095372 0.00000000 0.05148507 1 1 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from init : error # 1 nc already filled: wrong q grid or wrong nr %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... >************************************************************************ --------------------------------- ?????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071103/f4bb0e67/attachment.htm From mbaris at metu.edu.tr Sat Nov 3 08:16:02 2007 From: mbaris at metu.edu.tr (mbaris at metu.edu.tr) Date: Sat, 03 Nov 2007 09:16:02 +0200 Subject: [Pw_forum] some naive problems In-Reply-To: <016501c81d90$92c94820$b85bd860$@net> References: <20071102101731.subaibp7e1kcog0w@webmail2.metu.edu.tr> <472AFD8D.1080008@sissa.it> <016501c81d90$92c94820$b85bd860$@net> Message-ID: <20071103091602.qesj6v5arpq80soc@webmail2.metu.edu.tr> Dear Paul, If I may, I want to add some pointers > Just to add to this general discussion...which I think is very useful...when > I take up a new problem I write two scf scripts, one which increments the > size of the Monkhorst-Pack (K_POINTS) net until the total energy "sort of" > settles down (I say "sort of," because sometimes there are small > oscillations when the net gets big. Then I write another script which takes > the "best" MP size increments the cut-off upward until I'm happy. If I > think I need to find new "equilibrium" lattice constants, I'll write a third > script which varies these. Then I use all this to start computing > observables. I don't think I've ever used the gamma option. observables like the lattice parameter often settle much quicker compared to total energy, thus if you are looking for speed in your calculations, it is often better to find a minimized parameter set according to an observable. In some cases you may end up with problems other than speed using a "large" parameter set. I also have a script that minimizes the energy according to desired observable first using a golden search algorithm then spline interpolation. > > Re psuedopotentials: I pretty much trust the ones from the PWscf library > and/or from Vanderbilt. > Yes, but sometimes older pseuodos may lack full headers describing how the pseudo was created. > On the "na?ve" side. Most of the systems I investigate are far from cubic > symmetry, and I choose the number of K_POINTS in each direction > proportionately to reflect the way the dimensions of the Brillouin zone vary > inversely with those of the Wigner-Seitz cell. What do others do? Also, > I've never found adding an "offset" to the MP net has a significant > effect...when should I expect differently? Same here. The offset is used to include special points in your MP set, especially handy if you are using a rather small and even set and you want to have the gamma point explicitly. You may also increase the number of k-points in an even set to approximate gamma, or add the gamma point "by hand". Best, Baris, Osman Baris Malcioglu PhD. Candidate, METU Physics Ankara Turkey From raamesh123 at rediffmail.com Sat Nov 3 10:34:34 2007 From: raamesh123 at rediffmail.com (ramesh kumar) Date: 3 Nov 2007 09:34:34 -0000 Subject: [Pw_forum] (no subject) Message-ID: <20071103093434.4279.qmail@f5mail16.rediffmail.com> Dear friends i am new to pwgui and trying to learn linux as well as pwgui. somehow i could able to install pwgui in my pc. but when i am trying to run Si SCF in it says can't read "::pwscf::settings(CHDENS)": no such element in array can't read "::pwscf::settings(CHDENS)": no such element in array while executing "::tclu::newset run(CHDENS,$moduleObj) $::pwscf::settings(CHDENS)" (procedure "_init" line 16) invoked from within "_init $guiObj" (procedure "::pwscf::run::run" line 7) invoked from within "::pwscf::run::run ::gUI0" invoked from within ".gui0.shellchildsite.menubar.menubar.run.menu invoke active" ("uplevel" body line 1) invoked from within "uplevel #0 [list $w invoke active]" (procedure "tkMenuInvoke" line 47) invoked from within "tkMenuInvoke .gui0.shellchildsite.menubar.menubar.run.menu 1 " (command bound to event) i know that i have done some mistake in pwscf settings but i dont know how to rectify it. please help me in this regard -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071103/7b9b6296/attachment.htm From sunshoutian168 at yahoo.com.cn Sun Nov 4 08:35:39 2007 From: sunshoutian168 at yahoo.com.cn (shoutian sun) Date: Sun, 4 Nov 2007 15:35:39 +0800 (CST) Subject: [Pw_forum] a problem in calculation force constant with q2r.x. Message-ID: <541006.29975.qm@web92012.mail.cnb.yahoo.com> >Dear all, >I have a problem in calculation force constant with q2r.x. >Here are my input files of pw.x ph.x q2r.x. I am so sad. Please tell me the reasons . I need your help, anybody. Thank you. +++++++++++++++++++++++++++++++scf.in &CONTROL calculation = 'scf' , restart_mode = 'from_scratch' , outdir='/home/user/ecnuphyk/SUN/tmp/tmp-p1/' , pseudo_dir = '/home/user/ecnuphyk/SUN/pseudo/' , prefix='alh3' , tstress = .true. , tprnfor = .true. , etot_conv_thr = 1.0E-5 , forc_conv_thr = 1.0D-4 , / &SYSTEM ibrav = 0 , celldm(1) = 1.889725988578923203102406997093 , nat = 16 , ntyp = 2 , nspin = 1 ecutwfc = 44.0982 , / &ELECTRONS conv_thr = 1.0d-7 , mixing_beta = 0.7 , / CELL_PARAMETERS (alat) ... ATOMIC_SPECIES ... ATOMIC_POSITIONS {crystal} ... K_POINTS automatic 4 4 3 0 0 0 ++++++++++++++++++++++++++++ph.in &inputph tr2_ph=1.0d-12, prefix='alh3', ldisp=.true., nq1=4, nq2=4, nq3=3 epsil=.true. , amass(1)=26.98, amass(2)=1.0, outdir= '/home/user/ecnuphyk/SUN/tmp/tmp-p1/' , fildyn='alh3.dyn', / +++++++++++++++++++++++++++++q2r.in &input fildyn='alh3.dyn', zasr='simple', flfrc='alh3444.fc' / ++++++++++++++++++++++++++++q2r.in.out (The error ) reading grid info from file alh3.dyn0 reading force constants from file alh3.dyn1 macroscopic fields = T 4.51585 0.00000 -0.08800 0.00000 4.37104 0.00000 -0.08800 0.00000 4.43104 na= 1 2.74722 -0.27565 -0.08418 0.24590 2.66281 0.41340 -0.11507 -0.40383 2.71996 na= 2 2.74722 0.27565 -0.08418 -0.32097 2.64842 -0.40555 -0.11506 0.40383 2.71996 na= 3 2.74722 -0.27565 -0.08418 0.24590 2.66281 0.41340 -0.11507 -0.40383 2.71996 na= 4 2.74722 0.27565 -0.08418 -0.32097 2.64842 -0.40555 -0.11507 0.40383 2.71996 na= 5 -0.62424 0.00001 0.05784 -0.00008 -0.63566 -0.00071 0.06224 -0.00088 -1.42838 na= 6 -1.06619 -0.39916 0.01343 -0.40231 -0.99561 0.02342 0.01086 0.02141 -0.62960 na= 7 -1.06619 0.39915 0.01342 0.40230 -0.99561 -0.02341 0.01086 -0.02142 -0.62960 na= 8 -0.62424 -0.00002 0.05784 0.00007 -0.63566 0.00072 0.06224 0.00091 -1.42838 na= 9 -1.06613 0.39958 0.01388 0.40245 -0.99702 -0.02390 0.01136 -0.02197 -0.62981 na= 10 -1.06613 -0.39958 0.01389 -0.40245 -0.99702 0.02391 0.01136 0.02197 -0.62981 na= 11 -0.62425 0.00002 0.05784 -0.00007 -0.63567 -0.00072 0.06224 -0.00091 -1.42827 na= 12 -1.06613 0.39959 0.01389 0.40248 -0.99704 -0.02391 0.01135 -0.02197 -0.62985 na= 13 -1.06613 -0.39959 0.01388 -0.40247 -0.99704 0.02390 0.01136 0.02198 -0.62985 na= 14 -1.06619 -0.39916 0.01342 -0.40233 -0.99564 0.02341 0.01085 0.02143 -0.62963 na= 15 -1.06619 0.39916 0.01343 0.40233 -0.99563 -0.02342 0.01085 -0.02142 -0.62963 na= 16 -0.62425 -0.00001 0.05784 0.00008 -0.63567 0.00071 0.06223 0.00088 -1.42827 nqs= 1 q= 0.00000000 0.00000000 0.00000000 reading force constants from file alh3.dyn2 nqs= 2 q= 0.00095372 0.00000000 0.05148507 1 1 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from init : error # 1 nc already filled: wrong q grid or wrong nr %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... >************************************************************************ --------------------------------- ?????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071104/c6c54111/attachment.htm From degironc at sissa.it Sun Nov 4 08:35:51 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Sun, 04 Nov 2007 08:35:51 +0100 Subject: [Pw_forum] (no subject) In-Reply-To: <20071103093434.4279.qmail@f5mail16.rediffmail.com> References: <20071103093434.4279.qmail@f5mail16.rediffmail.com> Message-ID: <472D7657.50106@sissa.it> Dear Ramesh Kumar, I'm not familiar with the details of PWgui but chdens is a particular post-processing tool with its own GUI. Are you sure you are not mixing things ? hope this may help. Stefano de Gironcoli , SISSA and DEMOCRITOS PS: your affiliation is most welcome... commercial advertisements are not. ramesh kumar wrote: > > Dear friends > > i am new to pwgui and trying to learn linux as well as pwgui. somehow > i could able to install pwgui in my pc. but when i am trying to run Si > SCF in it says > > can't read "::pwscf::settings(CHDENS)": no such element in array > can't read "::pwscf::settings(CHDENS)": no such element in array > while executing > "::tclu::newset run(CHDENS,$moduleObj) $::pwscf::settings(CHDENS)" > (procedure "_init" line 16) > invoked from within > "_init $guiObj" > (procedure "::pwscf::run::run" line 7) > invoked from within > "::pwscf::run::run ::gUI0" > invoked from within > ".gui0.shellchildsite.menubar.menubar.run.menu invoke active" > ("uplevel" body line 1) > invoked from within > "uplevel #0 [list $w invoke active]" > (procedure "tkMenuInvoke" line 47) > invoked from within > "tkMenuInvoke .gui0.shellchildsite.menubar.menubar.run.menu 1 > " > (command bound to event) > > > i know that i have done some mistake in pwscf settings but i dont know > how to rectify it. > > please help me in this regard > > > > free petrol > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From degironc at sissa.it Sun Nov 4 08:51:41 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Sun, 04 Nov 2007 08:51:41 +0100 Subject: [Pw_forum] a problem in calculation force constant with q2r.x. In-Reply-To: <541006.29975.qm@web92012.mail.cnb.yahoo.com> References: <541006.29975.qm@web92012.mail.cnb.yahoo.com> Message-ID: <472D7A0D.4030600@sissa.it> The value of the 2nd q vector, the one that creates the problem, is very small and is assigned to the gamma-point which however is already filled... there is some problem in the definition or in the managing of the list... Two things you could do to help us helping you : 1) there used to be a stupid problem in q2r and dynmat/matdyn codes where ibrav=0 was not dealt with properly... could you please check whether your script would work with the equivalent ibrav<>0 ? 2) could you provide also the CELL_PARAMETERS card of your input ? Or even better, the complete input of your script so that one can try to reproduce your problem ? best regards, Stefano de Gironcoli - SISSA and DEMOCRITOS shoutian sun wrote: > >Dear all, > >I have a problem in calculation force constant with q2r.x. > >Here are my input files of pw.x ph.x q2r.x. I am so sad. Please tell > me the reasons . I need your help, anybody. Thank you. > +++++++++++++++++++++++++++++++scf.in > &CONTROL > calculation = 'scf' , > restart_mode = 'from_scratch' , > outdir='/home/user/ecnuphyk/SUN/tmp/tmp-p1/' , > pseudo_dir = '/home/user/ecnuphyk/SUN/pseudo/' , > prefix='alh3' , > tstress = .true. , > tprnfor = .true. , > etot_conv_thr = 1.0E-5 , > forc_conv_thr = 1.0D-4 , > / > &SYSTEM > ibrav = 0 , > celldm(1) = 1.889725988578923203102406997093 , > nat = 16 , > ntyp = 2 , > nspin = 1 > ecutwfc = 44.0982 , > / > &ELECTRONS > conv_thr = 1.0d-7 , > mixing_beta = > 0.7 , > / > CELL_PARAMETERS (alat) > ... > ATOMIC_SPECIES > ... > ATOMIC_POSITIONS {crystal} > ... > K_POINTS automatic > 4 4 3 0 0 0 > ++++++++++++++++++++++++++++ph.in > &inputph > tr2_ph=1.0d-12, > prefix='alh3', > ldisp=.true., > nq1=4, nq2=4, nq3=3 > epsil=.true. , > amass(1)=26.98, > amass(2)=1.0, > outdir= '/home/user/ecnuphyk/SUN/tmp/tmp-p1/' , > fildyn='alh3.dyn', > / > +++++++++++++++++++++++++++++q2r.in > &input > fildyn='alh3.dyn', zasr='simple', flfrc='alh3444.fc' > / > ++++++++++++++++++++++++++++q2r.in.out (The error ) > reading grid info from file alh3.dyn0 > reading force constants from file alh3.dyn1 > macroscopic fields = T > 4.51585 0.00000 -0.08800 > 0.00000 4.37104 0.00000 > -0.08800 0.00000 4.43104 > na= 1 > 2.74722 -0.27565 -0.08418 > 0.24590 2.66281 0.41340 > -0.11507 -0.40383 2.71996 > na= 2 > 2.74722 > 0.27565 -0.08418 > -0.32097 2.64842 -0.40555 > -0.11506 0.40383 2.71996 > na= 3 > 2.74722 -0.27565 -0.08418 > 0.24590 2.66281 0.41340 > -0.11507 -0.40383 2.71996 > na= 4 > 2.74722 0.27565 -0.08418 > -0.32097 2.64842 -0.40555 > -0.11507 0.40383 2.71996 > na= 5 > -0.62424 0.00001 0.05784 > -0.00008 -0.63566 -0.00071 > 0.06224 -0.00088 -1.42838 > na= 6 > -1.06619 -0.39916 0.01343 > -0.40231 -0.99561 0.02342 > 0.01086 0.02141 -0.62960 > na= 7 > -1.06619 0.39915 0.01342 > 0.40230 -0.99561 -0.02341 > 0.01086 -0.02142 -0.62960 > na= 8 > -0.62424 -0.00002 0.05784 > 0.00007 -0.63566 0.00072 > 0.06224 0.00091 -1.42838 > na= 9 > -1.06613 0.39958 0.01388 > 0.40245 -0.99702 -0.02390 > 0.01136 -0.02197 -0.62981 > na= > 10 > -1.06613 -0.39958 0.01389 > -0.40245 -0.99702 0.02391 > 0.01136 0.02197 -0.62981 > na= 11 > -0.62425 0.00002 0.05784 > -0.00007 -0.63567 -0.00072 > 0.06224 -0.00091 -1.42827 > na= 12 > -1.06613 0.39959 0.01389 > 0.40248 -0.99704 -0.02391 > 0.01135 -0.02197 -0.62985 > na= 13 > -1.06613 -0.39959 0.01388 > -0.40247 -0.99704 0.02390 > 0.01136 0.02198 -0.62985 > na= 14 > -1.06619 -0.39916 0.01342 > -0.40233 -0.99564 0.02341 > 0.01085 0.02143 -0.62963 > na= 15 > -1.06619 0.39916 0.01343 > 0.40233 -0.99563 -0.02342 > 0.01085 -0.02142 -0.62963 > na= 16 > -0.62425 -0.00001 0.05784 > 0.00008 -0.63567 0.00071 > 0.06223 0.00088 -1.42827 > nqs= 1 > q= 0.00000000 0.00000000 0.00000000 > reading force constants from file alh3.dyn2 > nqs= > 2 > q= 0.00095372 0.00000000 0.05148507 > 1 1 1 > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from init : error # 1 > nc already filled: wrong q grid or wrong nr > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > >************************************************************************ > > > ------------------------------------------------------------------------ > ?????????? > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From giannozz at nest.sns.it Sun Nov 4 19:44:03 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 4 Nov 2007 19:44:03 +0100 Subject: [Pw_forum] (no subject) In-Reply-To: <20071103093434.4279.qmail@f5mail16.rediffmail.com> References: <20071103093434.4279.qmail@f5mail16.rediffmail.com> Message-ID: <7B4B47E7-666D-4790-AADC-8A30F30426AB@nest.sns.it> On Nov 3, 2007, at 10:34 , ramesh kumar wrote: > i am new to pwgui and trying to learn linux as well as pwgui. > somehow i could able to install pwgui in my pc. but when i am > trying to run Si SCF in it says > can't read "::pwscf::settings(CHDENS)": no such element in array are you trying to install the PWGui from CVS sources? I am afraid that it doesn't work any longer since v.3.2 or so. One has to remove all references to the "chdens.x" executable and to complete the data for the "pp.x" executable with the new input format. If you discover how to do this it will be appreciated. If you are not interested in running the postprocessing codes or in the recent additions, you should still be able to use the PWGui for v.3.1 Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Sun Nov 4 19:56:17 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 4 Nov 2007 19:56:17 +0100 Subject: [Pw_forum] Debian Platforms and Fortran Compilers In-Reply-To: <016601c81d92$a2532d90$e6f988b0$@net> References: <016601c81d92$a2532d90$e6f988b0$@net> Message-ID: <5227F8BA-DE46-45B5-8667-9E8F1C36B67A@nest.sns.it> On Nov 2, 2007, at 21:54 , Paul M. Grant wrote: > I?ve found it too difficult to go through all the turmoil required > to install the Intel compiler on (K)ubuntu 7.10?the install script > is incredibly buggy. So I switched to gfortran, which installs > effortlessly and without the additional baggage of ?license > verification,? and found, at least for my tests (e.g., example07), > there are no significant differences in cpu execution times? > fractions of seconds over many hours. gfortran used to be too buggy to be usable, but recent versions seem to work fine and to produce rather fast executables. In conjunction with optimized mathematical libraries, it is becoming interesting for production runs as well. > Sidebar for the ph.x developers: It would be neat to be able to > set a ?suspension? mid-execution flag to be able to resume a long > run later (maybe even on another computer). unfortunately nobody has been working for a long time on the optimization and ease-of-use of the phonon code, and it shows... Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From krishnachaitanya.gunturu at gmail.com Mon Nov 5 04:45:41 2007 From: krishnachaitanya.gunturu at gmail.com (Krishna chaitanya) Date: Mon, 5 Nov 2007 09:15:41 +0530 Subject: [Pw_forum] Installation-ESPRESSO-3.2.3 on LINUX In-Reply-To: References: Message-ID: Hi everybody... We are trying to install ESPRESSO-3.2.3 in our AMD dual*dual core machine which is loaded with RH-AS4. We have downloaded g95 compilers directed by README file. Configuration has been done successfully but while making it is showing an error. I feel it is due to external libraries (BLAS, LAPACK and FFT). Here i have attached configure_logfile and make_logfile_error along with this e-mail. I would be greatful if anybody can suggest where i was wrong? Thank you very much With best regards G. Krishna Chaitanya C/o Dr. K. Bhanuprakash Inorganic CHemistry Division Indian Institute of Chemical Technology Hyderabad-500 007 India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071105/a9bba970/attachment-0001.htm -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: configure_logfile Url: http://www.democritos.it/pipermail/pw_forum/attachments/20071105/a9bba970/attachment-0002.txt -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: make_logfile_error Url: http://www.democritos.it/pipermail/pw_forum/attachments/20071105/a9bba970/attachment-0003.txt From akohlmey at cmm.chem.upenn.edu Mon Nov 5 05:05:02 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 5 Nov 2007 09:35:02 +0530 Subject: [Pw_forum] Installation-ESPRESSO-3.2.3 on LINUX In-Reply-To: References: Message-ID: <7b6913e90711042005p1dd42a60o5d98347f79b5f523@mail.gmail.com> On 11/5/07, Krishna chaitanya wrote: > Hi everybody... > > We are trying to install ESPRESSO-3.2.3 in our AMD dual*dual core machine > which is loaded with RH-AS4. We have downloaded g95 compilers directed by > README file. Configuration has been done successfully but while making it is > showing an error. I feel it is due to external libraries (BLAS, LAPACK > and FFT). Here i have attached configure_logfile and > make_logfile_error along with this e-mail. I would be greatful if anybody > can suggest where i was wrong? you are using BLAS/LAPACK libraries that were compiled with g77. those are not compatible with g95 (BTW: how did they get into the flib directory? did you copy them there?). if you have a copy of the matching g77 fortran runtime (libg2c.a) you may additionally link to that to work around this problem. other than that, i suggest you start by using the espresso provided libraries and use g95 also as fortran 77 compiler. for more details see the installation instructions, the QE wiki and numerous posts on this topic here in the mailing list archive. greetings from sunny bangalore, axel. > Thank you very much > > With best regards > G. Krishna Chaitanya > C/o Dr. K. Bhanuprakash > Inorganic CHemistry Division > Indian Institute of Chemical Technology > Hyderabad-500 007 > India > -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From bnrj.rudra at yahoo.com Tue Nov 6 10:53:18 2007 From: bnrj.rudra at yahoo.com (Rudra Banerjee) Date: Tue, 6 Nov 2007 01:53:18 -0800 (PST) Subject: [Pw_forum] (no subject) Message-ID: <665870.79343.qm@web62108.mail.re1.yahoo.com> how can i compile parallel version of the executable?i have mpich2 installed __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071106/4e476bd0/attachment.htm From Giovanni.Cantele at na.infn.it Tue Nov 6 11:21:06 2007 From: Giovanni.Cantele at na.infn.it (Giovanni Cantele) Date: Tue, 06 Nov 2007 11:21:06 +0100 Subject: [Pw_forum] (no subject) - PARALLEL NVIRONMENT In-Reply-To: <665870.79343.qm@web62108.mail.re1.yahoo.com> References: <665870.79343.qm@web62108.mail.re1.yahoo.com> Message-ID: <47304012.9070101@na.infn.it> Rudra Banerjee wrote: > how can i compile parallel version of the executable?i have mpich2 > installed Hi, you should first make sure that mpich executables (mpif77,mpif90) are in you PATH. Usually, just running ./configure should be enough to detect a parallel environment and let the configure script to setup compilation for the parallel version. The configure script generates a make.sys file, that should contain the -D__MPI -D__PARA flags . In the case for some reason the mpi stuff is not detected a message like "Parallel environment not detected" should be output by the configure script. There are cases in which this may happen, as some previous messages in the archive show, see for example http://www.democritos.it/pipermail/pw_forum/2006-September/004973.html http://www.democritos.it/pipermail/pw_forum/2006-September/004974.html http://www.democritos.it/pipermail/pw_forum/2004-October/001507.html You may also want to browse the "Installation" section of the Quantum-ESPRESSO Wiki: http://www.quantum-espresso.org/wiki/index.php/Installation If something goes wrong, try first to compile a sample program using the mpif90 Giovanni -- Dr. Giovanni Cantele Coherentia CNR-INFM and Dipartimento di Scienze Fisiche Universita' di Napoli "Federico II" Complesso Universitario di Monte S. Angelo - Ed. 6 Via Cintia, I-80126, Napoli, Italy Phone: +39 081 676910 Fax: +39 081 676346 E-mail: Giovanni.Cantele at na.infn.it Web: http://people.na.infn.it/~cantele From lanhaiping at gmail.com Tue Nov 6 11:23:02 2007 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 6 Nov 2007 18:23:02 +0800 Subject: [Pw_forum] (no subject) In-Reply-To: <665870.79343.qm@web62108.mail.re1.yahoo.com> References: <665870.79343.qm@web62108.mail.re1.yahoo.com> Message-ID: Dear Rudra i think you can refer to User Guide of QE package. http://www.quantum-espresso.org/wiki/index.php/Installation it describes the detailed installation of QE . Good luck !! On 11/6/07, Rudra Banerjee wrote: > > how can i compile parallel version of the executable?i have mpich2 > installed > > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071106/00be5593/attachment.htm From eyvaz_isaev at yahoo.com Tue Nov 6 11:23:10 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 6 Nov 2007 02:23:10 -0800 (PST) Subject: [Pw_forum] (no subject) In-Reply-To: <665870.79343.qm@web62108.mail.re1.yahoo.com> Message-ID: <478961.67372.qm@web60324.mail.yahoo.com> Just type ./configure and if everything (compilers and libraries) is installed correctly you will have a make.sys file. Usually "configure" says whether parallel environmet was found. Bests, Eyvaz. --- Rudra Banerjee wrote: > how can i compile parallel version of the > executable?i have mpich2 installed > > > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam > protection around > http://mail.yahoo.com > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From gaoguoying at gmail.com Tue Nov 6 13:48:17 2007 From: gaoguoying at gmail.com (Guoying Gao) Date: Tue, 6 Nov 2007 20:48:17 +0800 Subject: [Pw_forum] how to choose high-symmetry q points in the first irreducible Brillouin zone Message-ID: <6234c0c60711060448iec6477cnf2cb217cc8254473@mail.gmail.com> Dear everyone, I 'm doing the calculation of the band structure for monoclinic based-centered cell. I define the lattice vector through ibrav=0 and CELL PARAMETERS. However, when I do the calculation of the band structure, I didn't know how to choose the high-symmetry qpoints in the first irreducible Brillouin zone. Does anyone know this? Any suggestions will be appreciated. Thanks in advance. Best wishes. Gao From giannozz at nest.sns.it Tue Nov 6 14:06:27 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 6 Nov 2007 14:06:27 +0100 Subject: [Pw_forum] how to choose high-symmetry q points in the first irreducible Brillouin zone In-Reply-To: <6234c0c60711060448iec6477cnf2cb217cc8254473@mail.gmail.com> References: <6234c0c60711060448iec6477cnf2cb217cc8254473@mail.gmail.com> Message-ID: On Nov 6, 2007, at 13:48 , Guoying Gao wrote: > [...] I didn't know how to choose the high-symmetry qpoints > in the first irreducible Brillouin zone. Does anyone know this? it is explained in textbooks. One book that contains everything you may want to know about symmetry in solids is "The mathematical theory of symmetry in solids", by Bradley and Cracknell. It is a tough book, though. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Tue Nov 6 14:54:19 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 6 Nov 2007 14:54:19 +0100 Subject: [Pw_forum] Davidson algorithm In-Reply-To: <6ac064b60711021308y2d1f4bb6pe85700ee373373c6@mail.gmail.com> References: <6ac064b60711021308y2d1f4bb6pe85700ee373373c6@mail.gmail.com> Message-ID: On Nov 2, 2007, at 21:08 , Nichols A. Romero wrote: > Can anyone recommend a basic reference (e.g. thesis, personal notes, > computional article) that explains the Davidson algorithm used in > PWSCF > in more detail? The basic reference is E. R. Davidson, "The iterative calculation of a few of the lowest eigenvalues and corresponding eigenvectors of large real-symmetric matrices," Comput. Phys. 17, 87-94 (1975). The only PWscf-specific stuff I have is the following --- \section{Iterative diagonalization} Iterative diagonalization can be used whenever i) the number of states to be calculated is much smaller than the dimension of the basis set, and ii) a reasonable and economical estimate of the inverse operator $H^{-1}$ is available. Both conditions are satisfied in practical calculation in a PW basis set: the number of PW's is usually much larger than the number of bands, and the Hamiltonian matrix is dominated by the kinetic energy at large $\G$ ( the Hamiltonian is {\em diagonally dominant}). Iterative methods are based on a repeated refinement of a trial solution, which is stopped when satisfactory convergence is achieved. The number of iterative steps cannot be predicted in advance. It depends heavily on the structure of the matrix, on the type of refinement used, and on the starting point. A well-known and widely used algorithm is due to Davidson. In this method, a set of correction vectors $|\delta\psi_i\rangle$ to the $M$ trial eigenvectors $|\psi_i\rangle$ are generated as follows: \begin{equation} |\delta\psi_i\rangle = {1 \over D-\epsilon_i} (H-\epsilon_i)|\psi_i\rangle \end{equation} where the $\epsilon_i=\langle\psi_i|H|\psi_i\rangle$ are the trial eigenvalues. The $|\delta\psi_i\rangle$'s are orthogonalized and the Hamiltonian is diagonalized (with conventional techniques) in the subspace spanned by the trial and correction vectors. A new set of trial eigenvectors is obtained and the procedure is iterated until convergence is achieved. A good set of starting trial vectors is supplied by the eigenvectors at the preceding iteration of the potential. An important point is the following. The Hamiltonian matrix is never explicitly required excepted for its diagonal part. Only $H \psi_i$ products are required, which can be calculated in a very convenient way by applying the {\em dual-space technique}. In fact the kinetic energy is diagonal in \G-space, whereas the local potential term is diagonal in real space. Using FFT's (see below) one can go quickly back and forth from real to reciprocal space and perform the products where it is more convenient. There is still a nonlocal term which appears to require the storage of the matrix. The trick is to write $V_{NL}$ in a {\em separable} form: \begin{equation} V_{NL}(\k+\G,\k+\G')= \sum_{\mu=1}^{N_{at}} \sum_{j=1}^n f^\mu_j(\k+\G) g^\mu_j(\k+\G'), \end{equation} where $n$ is a small number and $N_{at}$ is the number of atoms in the unit cell. This allows us to perform the products by storing only the $f$ and $g$ vectors. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From cyrille.barreteau at cea.fr Tue Nov 6 14:05:05 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Tue, 06 Nov 2007 14:05:05 +0100 Subject: [Pw_forum] how to choose high-symmetry q points in the first irreducible Brillouin zone In-Reply-To: <6234c0c60711060448iec6477cnf2cb217cc8254473@mail.gmail.com> References: <6234c0c60711060448iec6477cnf2cb217cc8254473@mail.gmail.com> Message-ID: <47306681.9080809@cea.fr> Dear Gao, You might find this web site useful: http://www.cryst.ehu.es/cryst/get_kvec.html cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html ================================================================== Guoying Gao wrote: >Dear everyone, > > I 'm doing the calculation of the band structure for monoclinic >based-centered cell. I define the lattice vector through ibrav=0 and >CELL PARAMETERS. >However, when I do the calculation of the band structure, I didn't >know how to choose the high-symmetry qpoints in the first irreducible >Brillouin zone. Does >anyone know this? Any suggestions will be appreciated. >Thanks in advance. >Best wishes. > Gao >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From konstantin_kudin at yahoo.com Tue Nov 6 17:17:19 2007 From: konstantin_kudin at yahoo.com (Konstantin Kudin) Date: Tue, 6 Nov 2007 08:17:19 -0800 (PST) Subject: [Pw_forum] Debian Platforms and Fortran Compilers - [PH termination/restart] In-Reply-To: <5227F8BA-DE46-45B5-8667-9E8F1C36B67A@nest.sns.it> Message-ID: <672971.39769.qm@web57702.mail.re3.yahoo.com> --- Paolo Giannozzi wrote: > On Nov 2, 2007, at 21:54 , Paul M. Grant wrote: > > > Sidebar for the ph.x developers: It would be neat to be able to > > set a ?suspension? mid-execution flag to be able to resume a long > > run later (maybe even on another computer). > unfortunately nobody has been working for a long time on the > optimization and > ease-of-use of the phonon code, and it shows... Actually, there is exactly and option for this, at least, in the CVS version. Just create a file $prefix.EXIT in the working directory, and the PH code will stop at the earliest possible point. Then set 'recover=.true.' or something in the follow up input, and the run will be resumed ... Kostya __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From giannozz at nest.sns.it Tue Nov 6 18:09:23 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 6 Nov 2007 18:09:23 +0100 Subject: [Pw_forum] Davidson algorithm In-Reply-To: References: <6ac064b60711021308y2d1f4bb6pe85700ee373373c6@mail.gmail.com> Message-ID: On Nov 6, 2007, at 14:54 , Paolo Giannozzi wrote: > The basic reference is E. R. Davidson, "The iterative calculation of > a few of the lowest eigenvalues and corresponding eigenvectors of > large real-symmetric matrices," Comput. Phys. 17, 87-94 (1975). > The only PWscf-specific stuff I have is the following a correction: the following is "PW-specific", i.e. specific for the Plane-Wave basis set, but it is by no means specific to the implementation that is used in PWscf Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From oulihui666 at 126.com Wed Nov 7 04:19:12 2007 From: oulihui666 at 126.com (oulihui666) Date: Wed, 7 Nov 2007 11:19:12 +0800 (CST) Subject: [Pw_forum] A problem about relax Message-ID: <1048507.181661194405552851.JavaMail.coremail@bj126app81.126.com> Dear pwscf users: I am now doing a large job of 'relax' calculation. The unit cell contains 20 atoms and vacuum. The job has been running for seven days. While the total energy is decreasing, the amount it decreases is approximately a constant. We calculated the Pt(111) surface with pwscf, we use a five-layer slab containing 20 Pt atoms, the atoms in the bottom three layers are fixed, and the others are allowed to relax. The input file and structure can be found in below. However, calculated results indicated, the atoms in the top two layers are not inward or outward relaxation, but upward or downward relaxation, Could you tell me whether the input file is correct or not? and the final structure is correct or not please? Thanks a lot Best wishes Lihui Ou &CONTROL calculation = 'relax' , restart_mode = 'from_scratch' , outdir = '/home/olh/tmp/' , pseudo_dir = '/home/olh/pseudo/' , prefix = 'Pt111_20' , tstress = .true. , tprnfor = .true. , / &SYSTEM ibrav = 4, A = 5.549 , B = 5.549 , C = 21.062 , cosAB = -0.5 , cosAC = 0 , cosBC = 0 , nat = 20, ntyp = 1, ecutwfc = 70 , ecutrho = 0 , occupations = 'smearing' , degauss = 0.02 , smearing = 'gaussian' , / &ELECTRONS conv_thr = 1.0D-8 , mixing_beta = 0.7D0 , diagonalization = 'david' , / &IONS / ATOMIC_SPECIES Pt 195.09000 Pt.pz-rrkjus.linux.UPF ATOMIC_POSITIONS angstrom Pt -0.000000000 0.000000000 -9.062000000 0 0 0 Pt -0.000000000 2.775000000 -9.062000000 0 0 0 Pt 2.403000000 -1.387000000 -9.062000000 0 0 0 Pt 2.403000000 1.387000000 -9.062000000 0 0 0 Pt -0.000000000 -0.000000000 -2.265000000 1 1 1 Pt -0.000000000 2.775000000 -2.265000000 1 1 1 Pt 2.403000000 -1.387000000 -2.265000000 1 1 1 Pt 2.403000000 1.387000000 -2.265000000 1 1 1 Pt 1.602000000 0.000000000 -4.531000000 0 0 0 Pt 1.602000000 2.775000000 -4.531000000 0 0 0 Pt 4.005000000 -1.387000000 -4.531000000 0 0 0 Pt 4.005000000 1.387000000 -4.531000000 0 0 0 Pt 0.801000000 1.387000000 -6.796000000 0 0 0 Pt 0.801000000 4.162000000 -6.796000000 0 0 0 Pt 3.204000000 0.000000000 -6.796000000 0 0 0 Pt 3.204000000 2.775000000 -6.796000000 0 0 0 Pt 0.801000000 1.387000000 -0.000000000 1 1 1 Pt 0.801000000 4.162000000 -0.000000000 1 1 1 Pt 3.204000000 -0.000000000 -0.000000000 1 1 1 Pt 3.204000000 2.775000000 -0.000000000 1 1 1 K_POINTS automatic 3 3 1 1 1 1 -- ====================================== Lihui Ou PH.D Candidate in Electrochemistry College of Chemistry and Molecular Science Wuhan University,430072,Hubei Province,China E-mail:oulihui666 at 126.com ====================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071107/1b716e55/attachment-0001.htm From nedward at MIT.EDU Wed Nov 7 04:58:34 2007 From: nedward at MIT.EDU (Nicholas E. Singh-Miller) Date: Tue, 6 Nov 2007 22:58:34 -0500 (EST) Subject: [Pw_forum] A problem about relax In-Reply-To: <1048507.181661194405552851.JavaMail.coremail@bj126app81.126.com> References: <1048507.181661194405552851.JavaMail.coremail@bj126app81.126.com> Message-ID: Hi Lihui, From marzari at MIT.EDU Wed Nov 7 04:57:16 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Tue, 06 Nov 2007 22:57:16 -0500 Subject: [Pw_forum] A problem about relax In-Reply-To: <1048507.181661194405552851.JavaMail.coremail@bj126app81.126.com> References: <1048507.181661194405552851.JavaMail.coremail@bj126app81.126.com> Message-ID: <4731379C.2070703@mit.edu> Dear Lihui, 1) You are using a Vanderbilt ultrasoft pseudopotential - a proper cutoff is more likely in the range of 25 to 30 Ry, not 70. 2) ecutrho is by default 4 times ecutwfc (you need a cutoff radius 2 as large for the charge density than for the wfc, if the charge density if the square of wfc). The ultrasoft formulation makes the wfc smoother than that, but there is nothing that can be done for the charge density - whose natural cutoff is dictated by physics. In practice, use ecutrho 8 times ecutwfc. Ecutrho = 0 does not make sense. 3) use smaller beta than 0.7 for these systems. Of course, before doing this, you (and everyone else reading this list) should have tested extensively a 1-atom fcc bulk Pt system, to make sure everything worked properly. I strongly advice some reading (e.g. Richard Martin excellent book) - or carefully study the material at http://www.quantum-espresso.org/wiki/index.php/Tutorials_and_Examples nicola oulihui666 wrote: > Dear pwscf users: > I am now doing a large job of 'relax' calculation. The unit cell > contains 20 atoms and vacuum. The job has been running for seven days. > While the total energy is decreasing, the amount it decreases is > approximately a constant. We calculated the Pt(111) surface with pwscf, > we use a five-layer slab containing 20 Pt atoms, the atoms in the > bottom three layers are fixed, and the others are allowed to relax. The > input file and structure can be found in below. However, calculated > results indicated, the atoms in the top two layers are not inward or > outward relaxation, but upward or downward relaxation, Could you tell > me whether the input file is correct or not? and the final structure is > correct or not please? > Thanks a lot > Best wishes > Lihui Ou > > > > &CONTROL > > calculation = 'relax' , > > restart_mode = 'from_scratch' , > > outdir = '/home/olh/tmp/' , > > pseudo_dir = '/home/olh/pseudo/' , > > prefix = 'Pt111_20' , > > tstress = .true. , > > tprnfor = .true. , > > / > > &SYSTEM > > ibrav = 4, > > A = 5.549 , > > B = 5.549 , > > C = 21.062 , > > cosAB = -0.5 , > > cosAC = 0 , > > cosBC = 0 , > > nat = 20, > > ntyp = 1, > > ecutwfc = 70 , > > ecutrho = 0 , > > occupations = 'smearing' , > > degauss = 0.02 , > > smearing = 'gaussian' , > > / > > &ELECTRONS > > conv_thr = 1.0D-8 , > > mixing_beta = 0.7D0 , > > diagonalization = 'david' , > > / > > &IONS > > / > > ATOMIC_SPECIES > > Pt 195.09000 Pt.pz-rrkjus.linux.UPF > > ATOMIC_POSITIONS angstrom > > Pt -0.000000000 0.000000000 -9.062000000 0 0 0 > > Pt -0.000000000 2.775000000 -9.062000000 0 0 0 > > Pt 2.403000000 -1.387000000 -9.062000000 0 0 0 > > Pt 2.403000000 1.387000000 -9.062000000 0 0 0 > > Pt -0.000000000 -0.000000000 -2.265000000 1 1 1 > > Pt -0.000000000 2.775000000 -2.265000000 1 1 1 > > Pt 2.403000000 -1.387000000 -2.265000000 1 1 1 > > Pt 2.403000000 1.387000000 -2.265000000 1 1 1 > > Pt 1.602000000 0.000000000 -4.531000000 0 0 0 > > Pt 1.602000000 2.775000000 -4.531000000 0 0 0 > > Pt 4.005000000 -1.387000000 -4.531000000 0 0 0 > > Pt 4.005000000 1.387000000 -4.531000000 0 0 0 > > Pt 0.801000000 1.387000000 -6.796000000 0 0 0 > > Pt 0.801000000 4.162000000 -6.796000000 0 0 0 > > Pt 3.204000000 0.000000000 -6.796000000 0 0 0 > > Pt 3.204000000 2.775000000 -6.796000000 0 0 0 > > Pt 0.801000000 1.387000000 -0.000000000 1 1 1 > > Pt 0.801000000 4.162000000 -0.000000000 1 1 1 > > Pt 3.204000000 -0.000000000 -0.000000000 1 1 1 > > Pt 3.204000000 2.775000000 -0.000000000 1 1 1 > > K_POINTS automatic > > 3 3 1 1 1 1 > > > > -- > ====================================== > Lihui Ou > PH.D Candidate in Electrochemistry > College of Chemistry and Molecular Science > Wuhan University,/430072/,Hubei Province,China > /E-mail:oulihui666 at 126.com / > ====================================== -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From hushujun at mail.sdu.edu.cn Wed Nov 7 10:44:52 2007 From: hushujun at mail.sdu.edu.cn (=?gb2312?B?uvrK9778IFNodS1qdW4gSHU=?=) Date: Wed, 07 Nov 2007 17:44:52 +0800 Subject: [Pw_forum] Also the question about the psi^2 Message-ID: <394428692.27779@mail.sdu.edu.cn> Dear Stefano de Gironcoli & Stefano Baroni , The ILDOS calculation is the right way to solve my problem (psi^2 from plot_num=7 is less controllable than it). I appreciate the useful suggestions. Meanwhile i will try the want code. But another problem is that: unlike the psi^2, the ILDOS shows no sign on the states to decompose the spin up and spin down when one performs the LSDA calculation. Are there any variables like lsign for plot_num=10? Best wishes. Shujun Hu From giannozz at nest.sns.it Wed Nov 7 15:32:44 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 7 Nov 2007 15:32:44 +0100 Subject: [Pw_forum] Also the question about the psi^2 In-Reply-To: <394428692.27779@mail.sdu.edu.cn> References: <394428692.27779@mail.sdu.edu.cn> Message-ID: <8E6271F7-C292-4B66-9C29-9B2EC7F15988@nest.sns.it> On Nov 7, 2007, at 10:44 , ??? Shu-jun Hu wrote: > the ILDOS shows no sign on the states to decompose the spin up > and spin down when one performs the LSDA calculation. Are there > any variables like lsign for plot_num=10? that particular option produces the total, up+down, ILDOS. You have to modify the source code if you want something different: see PP/local_dos.f90 Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From w2agz at pacbell.net Thu Nov 8 02:50:50 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Wed, 7 Nov 2007 17:50:50 -0800 Subject: [Pw_forum] Davidson algorithm In-Reply-To: References: <6ac064b60711021308y2d1f4bb6pe85700ee373373c6@mail.gmail.com> Message-ID: <000501c821a9$cabb6ed0$60324c70$@net> Paolo, is the LaTeX script below part of a greater treatise you've written? If so, where is it for downloading (pdf preferred, but script is OK too)? -Paul Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford University EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com ? ? -----Original Message----- From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Paolo Giannozzi Sent: Tuesday, November 06, 2007 5:54 AM To: PWSCF Forum Subject: Re: [Pw_forum] Davidson algorithm On Nov 2, 2007, at 21:08 , Nichols A. Romero wrote: > Can anyone recommend a basic reference (e.g. thesis, personal notes, > computional article) that explains the Davidson algorithm used in > PWSCF > in more detail? The basic reference is E. R. Davidson, "The iterative calculation of a few of the lowest eigenvalues and corresponding eigenvectors of large real-symmetric matrices," Comput. Phys. 17, 87-94 (1975). The only PWscf-specific stuff I have is the following --- \section{Iterative diagonalization} Iterative diagonalization can be used whenever i) the number of states to be calculated is much smaller than the dimension of the basis set, and ii) a reasonable and economical estimate of the inverse operator $H^{-1}$ is available. Both conditions are satisfied in practical calculation in a PW basis set: the number of PW's is usually much larger than the number of bands, and the Hamiltonian matrix is dominated by the kinetic energy at large $\G$ ( the Hamiltonian is {\em diagonally dominant}). Iterative methods are based on a repeated refinement of a trial solution, which is stopped when satisfactory convergence is achieved. The number of iterative steps cannot be predicted in advance. It depends heavily on the structure of the matrix, on the type of refinement used, and on the starting point. A well-known and widely used algorithm is due to Davidson. In this method, a set of correction vectors $|\delta\psi_i\rangle$ to the $M$ trial eigenvectors $|\psi_i\rangle$ are generated as follows: \begin{equation} |\delta\psi_i\rangle = {1 \over D-\epsilon_i} (H-\epsilon_i)|\psi_i\rangle \end{equation} where the $\epsilon_i=\langle\psi_i|H|\psi_i\rangle$ are the trial eigenvalues. The $|\delta\psi_i\rangle$'s are orthogonalized and the Hamiltonian is diagonalized (with conventional techniques) in the subspace spanned by the trial and correction vectors. A new set of trial eigenvectors is obtained and the procedure is iterated until convergence is achieved. A good set of starting trial vectors is supplied by the eigenvectors at the preceding iteration of the potential. An important point is the following. The Hamiltonian matrix is never explicitly required excepted for its diagonal part. Only $H \psi_i$ products are required, which can be calculated in a very convenient way by applying the {\em dual-space technique}. In fact the kinetic energy is diagonal in \G-space, whereas the local potential term is diagonal in real space. Using FFT's (see below) one can go quickly back and forth from real to reciprocal space and perform the products where it is more convenient. There is still a nonlocal term which appears to require the storage of the matrix. The trick is to write $V_{NL}$ in a {\em separable} form: \begin{equation} V_{NL}(\k+\G,\k+\G')= \sum_{\mu=1}^{N_{at}} \sum_{j=1}^n f^\mu_j(\k+\G) g^\mu_j(\k+\G'), \end{equation} where $n$ is a small number and $N_{at}$ is the number of atoms in the unit cell. This allows us to perform the products by storing only the $f$ and $g$ vectors. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From w2agz at pacbell.net Thu Nov 8 03:00:52 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Wed, 7 Nov 2007 18:00:52 -0800 Subject: [Pw_forum] Debian Platforms and Fortran Compilers - [PH termination/restart] In-Reply-To: <672971.39769.qm@web57702.mail.re3.yahoo.com> References: <5227F8BA-DE46-45B5-8667-9E8F1C36B67A@nest.sns.it> <672971.39769.qm@web57702.mail.re3.yahoo.com> Message-ID: <000601c821ab$314f3d60$93edb820$@net> Kostya, I downloaded the PH directory on O-sesame...in which source file is this "escape" implemented? I had a quick look in phonon.f90 and couldn't find anything obvious therein. Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford University EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com ? ? -----Original Message----- From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Konstantin Kudin Sent: Tuesday, November 06, 2007 8:17 AM To: PWSCF Forum Subject: Re: [Pw_forum] Debian Platforms and Fortran Compilers - [PH termination/restart] --- Paolo Giannozzi wrote: > On Nov 2, 2007, at 21:54 , Paul M. Grant wrote: > > > Sidebar for the ph.x developers: It would be neat to be able to > > set a ?suspension? mid-execution flag to be able to resume a long > > run later (maybe even on another computer). > unfortunately nobody has been working for a long time on the > optimization and > ease-of-use of the phonon code, and it shows... Actually, there is exactly and option for this, at least, in the CVS version. Just create a file $prefix.EXIT in the working directory, and the PH code will stop at the earliest possible point. Then set 'recover=.true.' or something in the follow up input, and the run will be resumed ... Kostya __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From akohlmey at cmm.chem.upenn.edu Thu Nov 8 04:54:15 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 8 Nov 2007 09:24:15 +0530 (IST) Subject: [Pw_forum] Debian Platforms and Fortran Compilers - [PH termination/restart] In-Reply-To: <000601c821ab$314f3d60$93edb820$@net> References: <5227F8BA-DE46-45B5-8667-9E8F1C36B67A@nest.sns.it> <672971.39769.qm@web57702.mail.re3.yahoo.com> <000601c821ab$314f3d60$93edb820$@net> Message-ID: On Wed, 7 Nov 2007, Paul M. Grant wrote: PG> Kostya, I downloaded the PH directory on O-sesame...in which source file is PG> this "escape" implemented? I had a quick look in phonon.f90 and couldn't PG> find anything obvious therein. paul, this is a generic functionality that many QE executables use. you'll find the implementation (as well as many shared features) in the Modules subdirectory, in this case in check_stop.f90 (too obvious probably). ;-) happy diwali from bangalore, axel. PG> PG> Paul M. Grant, PhD PG> Principal, W2AGZ Technologies PG> Visiting Scholar, Applied Physics, Stanford University PG> EPRI Science Fellow (Retired) PG> IBM Research Staff Member Emeritus PG> w2agz at pacbell.net PG> http://www.w2agz.com PG> ? PG> ? PG> PG> PG> -----Original Message----- PG> From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On PG> Behalf Of Konstantin Kudin PG> Sent: Tuesday, November 06, 2007 8:17 AM PG> To: PWSCF Forum PG> Subject: Re: [Pw_forum] Debian Platforms and Fortran Compilers - [PH PG> termination/restart] PG> PG> PG> --- Paolo Giannozzi wrote: PG> PG> > On Nov 2, 2007, at 21:54 , Paul M. Grant wrote: PG> > PG> > > Sidebar for the ph.x developers: It would be neat to be able to PG> > > set a ?suspension? mid-execution flag to be able to resume a long PG> > > run later (maybe even on another computer). PG> > unfortunately nobody has been working for a long time on the PG> > optimization and PG> > ease-of-use of the phonon code, and it shows... PG> PG> Actually, there is exactly and option for this, at least, in the CVS PG> version. Just create a file $prefix.EXIT in the working directory, and PG> the PH code will stop at the earliest possible point. Then set PG> 'recover=.true.' or something in the follow up input, and the run will PG> be resumed ... PG> PG> Kostya PG> PG> __________________________________________________ PG> Do You Yahoo!? PG> Tired of spam? Yahoo! Mail has the best spam protection around PG> http://mail.yahoo.com PG> _______________________________________________ PG> Pw_forum mailing list PG> Pw_forum at pwscf.org PG> http://www.democritos.it/mailman/listinfo/pw_forum PG> PG> _______________________________________________ PG> Pw_forum mailing list PG> Pw_forum at pwscf.org PG> http://www.democritos.it/mailman/listinfo/pw_forum PG> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From hushujun at mail.sdu.edu.cn Thu Nov 8 05:15:41 2007 From: hushujun at mail.sdu.edu.cn (=?gb2312?B?uvrK9778IFNodS1qdW4gSHU=?=) Date: Thu, 08 Nov 2007 12:15:41 +0800 Subject: [Pw_forum] Also the question about the psi^2 Message-ID: <394495341.31027@mail.sdu.edu.cn> Dear all, Following is the last part of local_dos.f90: ************************************************** if (nspin == 1 .or. nspin==4) then is = 1 dos(:) = rho (:, is) else isup = 1 isdw = 2 dos(:) = rho (:, isup) + rho (:, isdw) <-------------------- end if if (lsign) then dos(:) = dos(:) * segno(:) deallocate(segno) endif if (iflag == 0) return ! ! symmetrization of the local dos ! #ifdef __PARA ! ! reduce charge density across pools ! call poolreduce (nrxx, dos) call psymrho(dos, nrx1, nrx2, nrx3, nr1, nr2, nr3, nsym, s, ftau) #else call symrho (dos, nrx1, nrx2, nrx3, nr1, nr2, nr3, nsym, s, ftau) #endif return end subroutine local_dos ************************************************** I tried to modify the line indicated by the arrow above. The decomposed spin up and spin down ILDOS seems ok. Anyone who can tell me if i was right? Best wishes. Shujun Hu From giannozz at nest.sns.it Thu Nov 8 10:20:11 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 8 Nov 2007 10:20:11 +0100 Subject: [Pw_forum] Davidson algorithm In-Reply-To: <000501c821a9$cabb6ed0$60324c70$@net> References: <6ac064b60711021308y2d1f4bb6pe85700ee373373c6@mail.gmail.com> <000501c821a9$cabb6ed0$60324c70$@net> Message-ID: On Nov 8, 2007, at 2:50 , Paul M. Grant wrote: > Paolo, is the LaTeX script below part of a greater treatise you've > written? > If so, where is it for downloading (pdf preferred, but script is OK > too)? "treatise" is an overstatement: it is part of would-be lecture notes. The most recent version is under this page: http://www.fisica.uniud.it/~giannozz/Corsi/metnum2007.html under "note sparse" (i.e. misc. notes), http://www.fisica.uniud.it/~giannozz/Corsi/metnum.pdf No panic: only the front page is in Italian. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Thu Nov 8 10:34:38 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 8 Nov 2007 10:34:38 +0100 Subject: [Pw_forum] Also the question about the psi^2 In-Reply-To: <394495341.31027@mail.sdu.edu.cn> References: <394495341.31027@mail.sdu.edu.cn> Message-ID: <96BDB3A4-BBB5-4104-B5AA-21A202EDD154@nest.sns.it> On Nov 8, 2007, at 5:15 , ??? Shu-jun Hu wrote: > I tried to modify the line indicated by the arrow above. The > decomposed > spin up and spin down ILDOS seems ok. Anyone who can tell me if i > was right? it seems correct to me Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From cyrille.barreteau at cea.fr Thu Nov 8 11:57:14 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Thu, 08 Nov 2007 11:57:14 +0100 Subject: [Pw_forum] restart problem in parallel mode Message-ID: <4732EB8A.9050202@cea.fr> dear all This is I know a question that has been asked several times but I did not find the solution to it.. I have performed a "parallel" scf calculation with the option wf_collect=.true. The calculation has ended properly and I want to restart this calculation with the option restart_mode='restart' (and the same number of nodes..) But I then get the error message IOS = 38 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from davcio : error # 10 i/o error in davcio %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... p0_30555: p4_error: : 0 p0_30555: (13.667969) net_send: could not write to fd=4, errno = 32 Let me also add that if I do a band calculation I do not get this error and it works well (but I guess the code does not need the wave function but only the electronic density). thanks in advance cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== From ihsanas at yahoo.com Thu Nov 8 15:25:37 2007 From: ihsanas at yahoo.com (ÇÍÓÇä ÚÑíÞÇÊ) Date: Thu, 8 Nov 2007 06:25:37 -0800 (PST) Subject: [Pw_forum] hybridization Message-ID: <815576.18014.qm@web52304.mail.re2.yahoo.com> Hi I plot projected density of state for CO molecule ,but how can I plot the hybridization 5 sigma ,and 1 bye . best regard Ihsan Jordan university __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From chenhanghuipwscf at gmail.com Fri Nov 9 04:02:05 2007 From: chenhanghuipwscf at gmail.com (alan chen) Date: Thu, 8 Nov 2007 22:02:05 -0500 Subject: [Pw_forum] constraints in the ionic dynamics Message-ID: <22ae3ca40711081902w2dbfa2aclad3fb0b4cb664450@mail.gmail.com> To whom it may concern, I am trying the CONSTRAINTS card for ionic dynamics. One of the 'constr_type' is 'distance'. However, if the distance of the two atoms I want to fix is larger than 'dmax' (dmax is half the shortest edge of the cell), the code crashes. I would like to ask why we can not fix two atoms whose distance is NOT larger than 'dmax'? Does the algorithm not work well if the distance of two atoms to be fixed is too large? Thank you very much. Hanghui Department of Physics, Yale University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071108/06d06045/attachment.htm From oulihui666 at 126.com Fri Nov 9 05:28:39 2007 From: oulihui666 at 126.com (oulihui666) Date: Fri, 9 Nov 2007 12:28:39 +0800 (CST) Subject: [Pw_forum] How to plot the molecular structure Message-ID: <25338427.249931194582519309.JavaMail.coremail@bj126app81.126.com> Dear pwscf users, I want to ask a simple problem about how to plot the molecular structure of input file, which software is used generally ? Thank you very much Best wishes Lihui Ou -- ====================================== Lihui Ou PH.D Candidate in Electrochemistry College of Chemistry and Molecular Science Wuhan University,430072,Hubei Province,China E-mail:oulihui666 at 126.com ====================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071109/df540a82/attachment.htm From akohlmey at cmm.chem.upenn.edu Fri Nov 9 05:46:59 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 9 Nov 2007 10:16:59 +0530 (IST) Subject: [Pw_forum] How to plot the molecular structure In-Reply-To: <25338427.249931194582519309.JavaMail.coremail@bj126app81.126.com> References: <25338427.249931194582519309.JavaMail.coremail@bj126app81.126.com> Message-ID: On Fri, 9 Nov 2007, oulihui666 wrote: LO> Dear pwscf users, LO> I want to ask a simple problem about how to plot the molecular LO> structure of input file, which software is used generally ? the use of the XCrysDen package is only mentioned in the quantum espresso webpages and the documentation about a hundred times. i guess this is not often enough... as the bible(?) says: read and thou shalt be given (wisdom)! cheers, axel. LO> LO> Thank you very much LO> Best wishes LO> Lihui Ou LO> LO> LO> LO> LO> -- LO> LO> ====================================== LO> Lihui Ou LO> PH.D Candidate in Electrochemistry LO> College of Chemistry and Molecular Science LO> Wuhan University,430072,Hubei Province,China LO> E-mail:oulihui666 at 126.com LO> ====================================== -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From baroni at sissa.it Fri Nov 9 07:07:27 2007 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 9 Nov 2007 07:07:27 +0100 Subject: [Pw_forum] constraints in the ionic dynamics In-Reply-To: <22ae3ca40711081902w2dbfa2aclad3fb0b4cb664450@mail.gmail.com> References: <22ae3ca40711081902w2dbfa2aclad3fb0b4cb664450@mail.gmail.com> Message-ID: <5D535CF6-FC5D-4EBB-80F0-551AB8621FE9@sissa.it> Dear Hanghui: I think this is just because of periodic boundary conditions. Thinking in 1D, as always, helps. Suppose you have a cell of size L. You place one atom, say A, at the origin (x=0), and another one, say B, at x=2/3 L. What will be the distance between A and B? Not 2/3 L, but L/3! Hope this helps Stefano Baroni On Nov 9, 2007, at 4:02 AM, alan chen wrote: > To whom it may concern, > I am trying the CONSTRAINTS card for ionic dynamics. One of > the 'constr_type' is 'distance'. However, if the distance of the > two atoms I want to fix is larger than 'dmax' (dmax is half the > shortest edge of the cell), the code crashes. > I would like to ask why we can not fix two atoms whose > distance is NOT larger than 'dmax'? Does the algorithm not work > well if the distance of two atoms to be fixed is too large? > Thank you very much. > > Hanghui > > Department of Physics, > > Yale University > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071109/e12379a0/attachment-0001.htm From cyrille.barreteau at cea.fr Fri Nov 9 09:28:47 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Fri, 09 Nov 2007 09:28:47 +0100 Subject: [Pw_forum] restart problem in parallel/serial mode with SOC Message-ID: <47341A3F.6090406@cea.fr> Dear all, Yesterday I posted an email about restart problems in parallel mode (version v.3.2.3) In the meanwhile I had a discussion with Conor Hogan and did some extra checks. Here are my "observations". 1) First I noticed that my problems only exist when spin-orbit coupling is present. Everything works fine if there is no SO. 2) The restart_mode='restart' (and wf_collect=.true.) does not work with SO, I get the the following message IOS = 38 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from davcio : error # 10 i/o error in davcio %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 3) the startingwfc='file' does not work either and I get the message that the wfc file was not found.. But the code continues with 'atomic' wfc. the startingpot='file' works fine. 4) I have also tried to do a dos.x calculation. For that purpose I saved the prefix.save directory on my serial machine and sent a dos.x calculation. Everything works fine without SO, but with SO I get again the error message: IOS = 38 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from davcio : error # 10 i/o error in davcio %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% I guess there is a small bug somewhere when SOC is swicthed on.. thanks for following my terrible tangle (up in blue:-) cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== From giannozz at nest.sns.it Fri Nov 9 10:14:30 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 9 Nov 2007 10:14:30 +0100 Subject: [Pw_forum] restart problem in parallel/serial mode with SOC In-Reply-To: <47341A3F.6090406@cea.fr> References: <47341A3F.6090406@cea.fr> Message-ID: <32E17EB9-17A1-465F-B2F2-861DB5E205D7@nest.sns.it> On Nov 9, 2007, at 9:28 , Cyrille Barreteau wrote: > I guess there is a small bug somewhere when SOC is switched on.. maybe this one? http://www.democritos.it:8888/O-sesame/chngview?cn=4918 also note this one: http://www.democritos.it:8888/O-sesame/chngview?cn=4926 P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From cyrille.barreteau at cea.fr Fri Nov 9 11:23:31 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Fri, 09 Nov 2007 11:23:31 +0100 Subject: [Pw_forum] restart problem in parallel/serial mode with SOC In-Reply-To: <32E17EB9-17A1-465F-B2F2-861DB5E205D7@nest.sns.it> References: <47341A3F.6090406@cea.fr> <32E17EB9-17A1-465F-B2F2-861DB5E205D7@nest.sns.it> Message-ID: <47343523.1010708@cea.fr> I thought that v3.2.3 already included these changes since it was released in october and the correction were done before. By the way I get a compilation error with the version of pw_restart.f90 from http://www.democritos.it:8888/O-sesame/chngview?cn=4926 cyrille Paolo Giannozzi wrote: >On Nov 9, 2007, at 9:28 , Cyrille Barreteau wrote: > > > >>I guess there is a small bug somewhere when SOC is switched on.. >> >> > >maybe this one? >http://www.democritos.it:8888/O-sesame/chngview?cn=4918 >also note this one: >http://www.democritos.it:8888/O-sesame/chngview?cn=4926 > >P. >--- >Paolo Giannozzi, Dept of Physics, University of Udine >via delle Scienze 208, 33100 Udine, Italy >Phone +39-0432-558216, fax +39-0432-558222 > > > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== From giannozz at nest.sns.it Fri Nov 9 11:35:43 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 9 Nov 2007 11:35:43 +0100 Subject: [Pw_forum] restart problem in parallel/serial mode with SOC In-Reply-To: <47343523.1010708@cea.fr> References: <47341A3F.6090406@cea.fr> <32E17EB9-17A1-465F-B2F2-861DB5E205D7@nest.sns.it> <47343523.1010708@cea.fr> Message-ID: On Nov 9, 2007, at 11:23 , Cyrille Barreteau wrote: > I thought that v3.2.3 already included these changes since it was > released in october and the correction were done before. only "safe" changes (those detailed in the ChangeLog-3.2.3) have been included in the 3.2.x versions. Changes that have the potential to introduce new bugs are never included in the "bugfix" versions, which could otherwise become "newbug" versions. > By the way I get a compilation error with the version of > pw_restart.f90 from > > http://www.democritos.it:8888/O-sesame/chngview?cn=4926 because there have been extensive changes and the new version of that routine is no longer compatible with the old version of the code. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From chenhanghuipwscf at gmail.com Fri Nov 9 16:33:52 2007 From: chenhanghuipwscf at gmail.com (alan chen) Date: Fri, 9 Nov 2007 10:33:52 -0500 Subject: [Pw_forum] constraints in the ionic dynamics Message-ID: <22ae3ca40711090733g4d75839fv4778cd2809c3784c@mail.gmail.com> Dear Stefano, Thank you for your comment. However my unit cell is tetragonal: the length of x and y direction is a and the length of z direction is 20*a. I want to fix the distance of two atoms in z direction which is larger than a/2. 'dmax' means half the shortest edge in the unit cell, which is a/2. So is there any better method to fix the distance of two atoms in z direction which is larger than a/2? Thank you. Hanghui Department of Physics Yale University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071109/8e2d3370/attachment.htm From oulihui666 at 126.com Sat Nov 10 07:45:49 2007 From: oulihui666 at 126.com (oulihui666) Date: Sat, 10 Nov 2007 14:45:49 +0800 (CST) Subject: [Pw_forum] How to set periodic boundary condition Message-ID: <23701527.955541194677149389.JavaMail.coremail@bj126app52.126.com> Dear pwscf users, Recently,I am now doing a large job of 'relax' calculation. The unit cell contains 20 atoms and vacuum. We calculated the Pt(111) surface with pwscf, we use a five-layer slab containing 20 Pt atoms, the atoms in the bottom three layers are fixed, and the others are allowed to relax. During calculation, the atoms in top two layers could not keep the structure of (111) plane, I guess that periodic boundary condition is not setted well, and How to set periodic boundary condition ? and how to set translational symmetry ? This is my coordinate file below, is there any problems ? Thank you very much Best wishes Lihui Ou ATOMIC_POSITIONS angstrom Pt -0.000000000 0.000000000 -9.062000000 0 0 0 Pt -0.000000000 2.775000000 -9.062000000 0 0 0 Pt 2.403000000 -1.387000000 -9.062000000 0 0 0 Pt 2.403000000 1.387000000 -9.062000000 0 0 0 Pt -0.000000000 -0.000000000 -2.265000000 1 1 1 Pt -0.000000000 2.775000000 -2.265000000 1 1 1 Pt 2.403000000 -1.387000000 -2.265000000 1 1 1 Pt 2.403000000 1.387000000 -2.265000000 1 1 1 Pt 1.602000000 0.000000000 -4.531000000 0 0 0 Pt 1.602000000 2.775000000 -4.531000000 0 0 0 Pt 4.005000000 -1.387000000 -4.531000000 0 0 0 Pt 4.005000000 1.387000000 -4.531000000 0 0 0 Pt 0.801000000 1.387000000 -6.796000000 0 0 0 Pt 0.801000000 4.162000000 -6.796000000 0 0 0 Pt 3.204000000 0.000000000 -6.796000000 0 0 0 Pt 3.204000000 2.775000000 -6.796000000 0 0 0 Pt 0.801000000 1.387000000 -0.000000000 1 1 1 Pt 0.801000000 4.162000000 -0.000000000 1 1 1 Pt 3.204000000 -0.000000000 -0.000000000 1 1 1 Pt 3.204000000 2.775000000 -0.000000000 1 1 1 -- ====================================== Lihui Ou PH.D Candidate in Electrochemistry College of Chemistry and Molecular Science Wuhan University,430072,Hubei Province,China E-mail:oulihui666 at 126.com ====================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071110/96aa0b0c/attachment.htm From degironc at sissa.it Sat Nov 10 09:27:16 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Sat, 10 Nov 2007 09:27:16 +0100 Subject: [Pw_forum] How to set periodic boundary condition In-Reply-To: <23701527.955541194677149389.JavaMail.coremail@bj126app52.126.com> References: <23701527.955541194677149389.JavaMail.coremail@bj126app52.126.com> Message-ID: <47356B64.1020309@sissa.it> The code enforce periodic boundary condition automatically so there must be something wrong in your input. how are oriented your unit cell lattice vectors in space ? are they compatible with the atomic coordinates you give ? have you visualized your cell using some visualization tool (like XCrysDen) that the cell is what you think it is ? Hope this help stefano oulihui666 wrote: > Dear pwscf users, > Recently,I am now doing a large job of 'relax' calculation. The unit > cell contains 20 atoms and vacuum. We calculated the Pt(111) surface > with pwscf, we use a five-layer slab containing 20 Pt atoms, the atoms > in the bottom three layers are fixed, and the others are allowed to > relax. During calculation, the atoms in top two layers could not keep > the structure of (111) plane, I guess that periodic boundary condition > is not setted well, and How to set periodic boundary condition ? and > how to set translational symmetry ? This is my coordinate file below, > is there any problems ? > Thank you very much > Best wishes > Lihui Ou > ATOMIC_POSITIONS angstrom > > Pt -0.000000000 0.000000000 -9.062000000 0 0 0 > > Pt -0.000000000 2.775000000 -9.062000000 0 0 0 > > Pt 2.403000000 -1.387000000 -9.062000000 0 0 0 > > Pt 2.403000000 1.387000000 -9.062000000 0 0 0 > > Pt -0.000000000 -0.000000000 -2.265000000 1 1 1 > > Pt -0.000000000 2.775000000 -2.265000000 1 1 1 > > Pt 2.403000000 -1.387000000 -2.265000000 1 1 1 > > Pt 2.403000000 1.387000000 -2.265000000 1 1 1 > > Pt 1.602000000 0.000000000 -4.531000000 0 0 0 > > Pt 1.602000000 2.775000000 -4.531000000 0 0 0 > > Pt 4.005000000 -1.387000000 -4.531000000 0 0 0 > > Pt 4.005000000 1.387000000 -4.531000000 0 0 0 > > Pt 0.801000000 1.387000000 -6.796000000 0 0 0 > > Pt 0.801000000 4.162000000 -6.796000000 0 0 0 > Pt 3.204000000 0.000000000 -6.796000000 0 0 0 > Pt 3.204000000 2.775000000 -6.796000000 0 0 0 > > Pt 0.801000000 1.387000000 -0.000000000 1 1 1 > > Pt 0.801000000 4.162000000 -0.000000000 1 1 1 > > Pt 3204000000 -0.000000000 -0.000000000 1 1 1 > > Pt 3.204000000 2.775000000 -0.000000000 1 1 1 > > > -- > ====================================== > Lihui Ou > PH.D Candidate in Electrochemistry > College of Chemistry and Molecular Science > Wuhan University,/430072/,Hubei Province,China > /E-mail:oulihui666 at 126.com / > ====================================== > > > ------------------------------------------------------------------------ > ???? ?? ?? 2007 ?? ?? ?? ?? ?? ?? ?? ???? ?? ?? 10 ?? ?? ?? ?? ?? ?? > ?? ?? ??????>> > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From dalcorso at sissa.it Sat Nov 10 09:34:34 2007 From: dalcorso at sissa.it (Andrea Dal Corso) Date: Sat, 10 Nov 2007 09:34:34 +0100 Subject: [Pw_forum] How to set periodic boundary condition Message-ID: <1194683674.47356d1ac5c5f@webmail.sissa.it> > Pt 0.801000000 1.387000000 -0.000000000 1 1 1 > > Pt 0.801000000 4.162000000 -0.000000000 1 1 1 > > Pt 3204000000 -0.000000000 -0.000000000 1 1 1 > This x coordinate seems quite strange to me. Andrea > Pt 3.204000000 2.775000000 -0.000000000 1 1 1 > > > -- > ====================================== Andrea Dal Corso Tel. 0039-040-3787428 SISSA, Via Beirut 2/4 Fax. 0039-040-3787528 34014 Trieste (Italy) e-mail: dalcorso at sissa.it ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by Horde http://www.horde.org/ From zdw2000 at gmail.com Sun Nov 11 13:43:23 2007 From: zdw2000 at gmail.com (Wei Zhou) Date: Sun, 11 Nov 2007 20:43:23 +0800 Subject: [Pw_forum] about the new espresso-3.2.3 and the variable cell BOMD Message-ID: <3e36a8fc0711110443g66eb771akfa6339a3c8db9652@mail.gmail.com> dear everyone I read the new espresso3.2.3 guide,and see the new version can do variable cell BOMD, however ,I wonder how I can apply this, for example how to set the variable cell, Is there any examle about the variable cell BOMD ? any help will be appreciated ,thank you in advace. -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071111/432d2a39/attachment.htm From njuxuyuehua at gmail.com Mon Nov 12 04:00:18 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Mon, 12 Nov 2007 11:00:18 +0800 Subject: [Pw_forum] A QUESTION about the precision of relaxation Message-ID: hi everyone , i am now do a relax job.my uint cell is a supercell . and i only want to relax a direction of lattice ,for example c . so i tried to vary the c .and compute. see the stress.but i do not know how to adjust the c on the base of the result like: entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= - 1.06 -0.00000893 -0.00000001 -0.00000001 -1.31 0.00 0.00 -0.00000001 -0.00000892 0.00000000 0.00 -1.31 0.00 -0.00000001 0.00000000 -0.00000382 0.00 0.00 -0.56 (10there is "-0.56" and the "-" stands for .i should the reduce c? (2)0.56 is a large stres? (3) is there any other way to determine the suitable c? thank you very much if you could give a advice . bests -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071112/6d03efc9/attachment.htm From baroni at sissa.it Mon Nov 12 08:05:44 2007 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 12 Nov 2007 08:05:44 +0100 Subject: [Pw_forum] A QUESTION about the precision of relaxation In-Reply-To: References: Message-ID: <8A75D215-615C-4556-B886-E647D07DFD69@sissa.it> Dear Yuehua: On Nov 12, 2007, at 4:00 AM, xu yuehua wrote: > hi everyone , > i am now do a relax job.my uint cell is a supercell . and i only > want to relax a direction of lattice ,for example c . > so i tried to vary the c .and compute. see the stress.but i do not > know how to adjust the c on the base of the result like: > entering subroutine stress ... > > total stress (Ry/bohr**3) (kbar) > P= -1.06 > -0.00000893 -0.00000001 -0.00000001 -1.31 > 0.00 0.00 > -0.00000001 -0.00000892 0.00000000 0.00 - > 1.31 0.00 > -0.00000001 0.00000000 -0.00000382 0.00 0.00 > -0.56 > > > (10there is "-0.56" and the "-" stands for .i should the reduce c? a positive pressure (or a positive component of the stress, for that matter) will make the material expand, if unconstrained. > (2)0.56 is a large stres? > the answer is "NO", but this is an answer you could have answered yourself. assume you know the elastic constants of your material (if you don't, just take the experimental values, or any reasonable guess). how much would c vary given the known values of the elastic constants and of the zz componenty of the stress? unless the material is exceedingly soft, you will find that c will vary very little. you can practice this yourself. just let c vary by, say, 5%, and see by how much would \sigma_{zz} vary ... > (3) is there any other way to determine the suitable c? > calculate and tabulate E vs. c and find the minimum. beware, though, that the curve that you will obtain should be rather rugged. this issue has been discussed several times in several places, including this forum, so I will stop here ... > thank you very much if you could give a advice . > you are most welcome Stefano Baroni --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071112/51838543/attachment.htm From njuxuyuehua at gmail.com Mon Nov 12 12:30:20 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Mon, 12 Nov 2007 19:30:20 +0800 Subject: [Pw_forum] A QUESTION about the precision of relaxation In-Reply-To: <8A75D215-615C-4556-B886-E647D07DFD69@sissa.it> References: <8A75D215-615C-4556-B886-E647D07DFD69@sissa.it> Message-ID: Dear Baroni thanks for your advice .i know how to deal with it. bests 2007/11/12, Stefano Baroni : > > Dear Yuehua: > On Nov 12, 2007, at 4:00 AM, xu yuehua wrote: > > hi everyone , > i am now do a relax job.my uint cell is a supercell . and i only want to > relax a direction of lattice ,for example c . > so i tried to vary the c .and compute. see the stress.but i do not know > how to adjust the c on the base of the result like: > > entering subroutine stress ... > > total stress (Ry/bohr**3) (kbar) P= - > 1.06 > -0.00000893 -0.00000001 -0.00000001 -1.31 0.00 0.00 > -0.00000001 -0.00000892 0.00000000 0.00 - 1.31 0.00 > -0.00000001 0.00000000 -0.00000382 0.00 0.00 -0.56 > > > (10there is "-0.56" and the "-" stands for .i should the reduce c? > > > > a positive pressure (or a positive component of the stress, for that > matter) will make the material expand, if unconstrained. > > (2)0.56 is a large stres? > > > > the answer is "NO", but this is an answer you could have answered > yourself. assume you know the elastic constants of your material (if you > don't, just take the experimental values, or any reasonable guess). how much > would c vary given the known values of the elastic constants and of the zz > componenty of the stress? unless the material is exceedingly soft, you will > find that c will vary very little. you can practice this yourself. just let > c vary by, say, 5%, and see by how much would \sigma_{zz} vary ... > > (3) is there any other way to determine the suitable c? > > calculate and tabulate E vs. c and find the minimum. beware, though, that > the curve that you will obtain should be rather rugged. this issue has been > discussed several times in several places, including this forum, so I will > stop here ... > > thank you very much if you could give a advice . > > > > you are most welcome > Stefano Baroni > > > --- > Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste > [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) > > Please, if possible, don't send me MS Word or PowerPoint attachments > Why? See: http://www.gnu.org/philosophy/no-word-attachments.html > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071112/fb8e85ed/attachment.htm From ihsanas at yahoo.com Mon Nov 12 16:20:20 2007 From: ihsanas at yahoo.com (ÇÍÓÇä ÚÑíÞÇÊ) Date: Mon, 12 Nov 2007 07:20:20 -0800 (PST) Subject: [Pw_forum] projwfc.x Message-ID: <902376.55116.qm@web52303.mail.re2.yahoo.com> Hi when I run projected wavefunction for oxygen by using this file for projwfc.x &inputpp outdir=/' prefix = 'O' Emin=-100.0, Emax=100.0, DeltaE=0.001 ngauss=1, degauss=0.01 it give me file for 1s and 2p only there is no file for 2s ..... what`s wrong ..please answer me. you don`t answer me when I asked about hybridization.... thanks Ihsan Jordan university __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From degironc at sissa.it Mon Nov 12 17:52:45 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Mon, 12 Nov 2007 17:52:45 +0100 Subject: [Pw_forum] projwfc.x In-Reply-To: <902376.55116.qm@web52303.mail.re2.yahoo.com> References: <902376.55116.qm@web52303.mail.re2.yahoo.com> Message-ID: <473884DD.4060806@sissa.it> there is no 1s in the O pseudopotential... so there cannot be any projected dos for it. stefano ????? ?????? wrote: > Hi > when I run projected wavefunction for oxygen by using > this file for projwfc.x > &inputpp > outdir=/' > prefix = 'O' > Emin=-100.0, Emax=100.0, DeltaE=0.001 > ngauss=1, degauss=0.01 > it give me file for 1s and 2p only there is no file > for 2s ..... what`s wrong ..please answer me. > you don`t answer me when I asked about > hybridization.... > thanks > Ihsan > Jordan university > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From isouza at berkeley.edu Mon Nov 12 20:59:13 2007 From: isouza at berkeley.edu (Ivo Souza) Date: Mon, 12 Nov 2007 11:59:13 -0800 Subject: [Pw_forum] Sign of magnetization Message-ID: <4738B091.1080702@berkeley.edu> Dear all, I am trying to understand whether there is a sign error in the definition of the total magnetization in pwscf. In other words, when the code tells us that, e.g., the magnetic moment of bcc iron points along +z, is it really pointing along minus z? The spin magnetic moment of a state is (n units of the Bohr magneton) <\mu_z> = - (note the minus sign) where sigma_z is the Pauli matrix. From the subroutine 'compute_magnetization' in electrons.f90: ------------------------------------------------------- IF ( lsda ) THEN ! magtot = 0.D0 absmag = 0.D0 ! DO ir = 1, nrxx ! mag = rho(ir,1) - rho(ir,2) ! magtot = magtot + mag absmag = absmag + ABS( mag ) ! END DO ! magtot = magtot * omega / ( nr1*nr2*nr3 ) absmag = absmag * omega / ( nr1*nr2*nr3 ) ------------------------------------------------------- My understanding is that rho(ir,1) and rho(ir,2) are respectively the (number) densities for spin-up ("moment-down") and spin-down ("moment-up") electrons. If so, then it seems to me that mag = rho(ir,1) - rho(ir,2) should be replaced by mag = rho(ir,2) - rho(ir,1) in order to get the correct sign for the magnetic moment per unit cell 'magtot' The above is for collinear (lsda) calculations, but the same issue with the sign is present in the definition of the total magnetization in the non-collinear case: ------------------------------------------------------- ELSE IF ( noncolin ) THEN ! magtot_nc = 0.D0 absmag = 0.D0 ! DO ir = 1,nrxx ! mag = SQRT( rho(ir,2)**2 + rho(ir,3)**2 + rho(ir,4)**2 ) ! DO i = 1, 3 ! magtot_nc(i) = magtot_nc(i) + rho(ir,i+1) ! END DO ! absmag = absmag + ABS( mag ) ! END DO ! CALL mp_sum( magtot_nc, intra_pool_comm ) CALL mp_sum( absmag, intra_pool_comm ) ! DO i = 1, 3 ! magtot_nc(i) = magtot_nc(i) * omega / ( nr1*nr2*nr3 ) ! END DO ! absmag = absmag * omega / ( nr1*nr2*nr3 ) ! ENDIF ------------------------------------------------------- Since [see subroutine 'sum_band_k'] rho(ir,2), rho(ir,3) and rho(ir,4) are respectively the x,y, and z components of the spin polarization density, proportional to psi^*(r).sigma_i.psi(r) (i=x,y,z), the line magtot_nc(i) = magtot_nc(i) + rho(ir,i+1) should be replaced by magtot_nc(i) = magtot_nc(i) - rho(ir,i+1) It is quite easy to get confused with signs, so I apologize if the confusion is on my side! Also, it could just be a matter of semantics, and what is mean by "magnetic moment" in pwscf is really "spin polarization." Any clarification would be greatly appreciated. With thanks and best regards, Ivo S. From giannozz at nest.sns.it Mon Nov 12 22:33:38 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 12 Nov 2007 22:33:38 +0100 Subject: [Pw_forum] about the new espresso-3.2.3 and the variable cell BOMD In-Reply-To: <3e36a8fc0711110443g66eb771akfa6339a3c8db9652@mail.gmail.com> References: <3e36a8fc0711110443g66eb771akfa6339a3c8db9652@mail.gmail.com> Message-ID: On Nov 11, 2007, at 13:43 , Wei Zhou wrote: > [...] how to set the variable cell, Is there any example > about the variable cell BOMD ? in examples/VCSexample there is an example for variable-cell optimization (not dynamics) Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From lyanglyan at yahoo.com.cn Tue Nov 13 13:57:59 2007 From: lyanglyan at yahoo.com.cn (=?gb2312?q?=D1=D2=20=C0=EE?=) Date: Tue, 13 Nov 2007 20:57:59 +0800 (CST) Subject: [Pw_forum] Why the calculated equilibrium lattice parameters are significantly larger than the experiment? In-Reply-To: Message-ID: <217006.37664.qm@web15805.mail.cnb.yahoo.com> Dear all, I was using a pseudopotential made by ld1.x to simulate the equation of states. But the equilibrium lattice parameters are much larger than the experiment. No improvements have been achieved although rcut is increased in the pseudopotenitals. Would you please tell me what else can affect the results in spite of the rcut? Best regards, Yan Li --------------------------------- ?????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071113/180b23eb/attachment.htm From baroni at sissa.it Tue Nov 13 14:32:32 2007 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 13 Nov 2007 13:32:32 +0000 Subject: [Pw_forum] Why the calculated equilibrium lattice parameters are significantly larger than the experiment? In-Reply-To: <217006.37664.qm@web15805.mail.cnb.yahoo.com> References: <217006.37664.qm@web15805.mail.cnb.yahoo.com> Message-ID: <7A46B40A-69F5-42FE-8B57-1192FC61A294@sissa.it> Dear Yan Li, please let us know which system are you studying, as well as more details on your calculation. The behavior you observed may depend on your pseudopotential, or on some know failure of dft, or on many other reasons. Also, please do not forget to include your affiliation in the signature. Best Stefano Baroni On Nov 13, 2007, at 12:57 PM, ? ? wrote: > Dear all, > > > I was using a pseudopotential made by ld1.x to simulate the > equation of states. But the equilibrium lattice parameters are much > larger than the experiment. No improvements have been achieved > although rcut is increased in the pseudopotenitals. Would you > please tell me what else can affect the results in spite of the rcut? > > Best regards, > Yan Li > > ?????????? > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071113/33e9035c/attachment.htm From hqzhou at nju.edu.cn Tue Nov 13 14:37:44 2007 From: hqzhou at nju.edu.cn (Huiqun Zhou) Date: Tue, 13 Nov 2007 21:37:44 +0800 Subject: [Pw_forum] Convergence test Message-ID: <010801c825fa$5fab6960$1d00a8c0@solarflare> List users, I'm testing convergence on ecutwfc for chromite using USPP. Below are results of different ecutrho with different ecutwfc. I may be wrong, I think I should watch the difference of total energy between successive runs till it's less than 1-2 mRy. But the trend is not apparent in the results below and it seems that the larger the ecutwfc, the lower the total energy. I calculated E-V relations using two sets of ecutwfc and ecutrho: for ecutwfc = 35 ry and ecutrho = 240 ry, a = 8.405, B0 = 172.8, B0' = 4.3 for ecutwfc = 55 ry and ecutrho = 250 ry, a = 8.407, B0 = 178.6, B0' = 3.6 There are no big differences in unit cell parameter and elastic properties. So, my questions are (1) usually, the ecutwfc of USPP should be around 35 ry or below, is my case an exception? (2) how small the difference of total energy of successive runs should be in judging convergence against ecutwfc? Thank you in advance. Dr. Huiqun Zhou @Earth Sciences, Nanjing University, China ecutrho = 240 ry ecutwfc total energy 35 -1461.54573299 Ry 40 -1461.64206824 Ry 45 -1461.71978835 Ry 50 -1461.76587314 Ry 55 -1461.76316597 Ry 60 -1461.76692396 Ry ecutrho = 250 ry 40 -1461.64351876 Ry 45 -1461.70130510 Ry 50 -1461.74724764 Ry 55 -1461.76462943 Ry 60 -1461.76838812 Ry ecutrho = 280 ry 50 -1461.74778158 Ry 55 -1461.76516515 Ry 60 -1461.76892327 Ry From aryjunior at gmail.com Tue Nov 13 19:15:18 2007 From: aryjunior at gmail.com (Ary Junior) Date: Tue, 13 Nov 2007 16:15:18 -0200 Subject: [Pw_forum] Crystal coordinates output Message-ID: <4c9c61c20711131015t612baef2u293c5d55f000c0e9@mail.gmail.com> Hi! I'm having a problem with PWSCF v.3.2.2... when I run a calculation with a hexagonal structure without optimization, the crystall coordinates in output file aren't the same as the input file... when I open both on XCrysden I check this... I've found the thread http://www.democritos.it/pipermail/pw_forum/2003-November/000705.html , but this can't solve my problem.... My output_tau.f90 source file has been fixed... Thanks very much!!! -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071113/47329e35/attachment-0001.htm From baroni at sissa.it Tue Nov 13 19:29:00 2007 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 13 Nov 2007 18:29:00 +0000 Subject: [Pw_forum] Crystal coordinates output In-Reply-To: <4c9c61c20711131015t612baef2u293c5d55f000c0e9@mail.gmail.com> References: <4c9c61c20711131015t612baef2u293c5d55f000c0e9@mail.gmail.com> Message-ID: Dear Mr. Unknown, Just a guess: I think that the code outputs atomic positions in cartesian coordinates. If you choose to input them as "crystal" (i.e. as linear combinations of the lattice basis vectors), I would not be surprised if the output would differ from the input. To make sure, just post the input and the relevant lines of the output files. Also, please do not forget to sign your posts to this forum. Thanks - Stefano Baroni On Nov 13, 2007, at 6:15 PM, Ary Junior wrote: > Hi! I'm having a problem with PWSCF v.3.2.2... when I run a > calculation with a hexagonal structure without optimization, the > crystall coordinates in output file aren't the same as the input > file... when I open both on XCrysden I check this... I've found > the thread http://www.democritos.it/pipermail/pw_forum/2003- > November/000705.html , but this can't solve my problem.... My > output_tau.f90 source file has been fixed... > > Thanks very much!!! > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071113/25cb7b9d/attachment.htm From bnrj.rudra at yahoo.com Tue Nov 13 20:04:19 2007 From: bnrj.rudra at yahoo.com (Rudra Banerjee) Date: Wed, 14 Nov 2007 00:34:19 +0530 (IST) Subject: [Pw_forum] ion problem Message-ID: <360668.39837.qm@web94110.mail.in2.yahoo.com> dear PWSCF users, I am new to pwscf and stuck in a question......is there any way to put anion or cation as a unit?the total cluster is charge less but have two different types of ion.......can anyone tell me what to do in this situation? Regards -- Rudra Have a Nice Time http://www.bose.res.in/~rudra Please, if possible, don't send me MS Word or PowerPoint attachments Why?See: http://www.gnu.org/philosophy/no-word-attachments.html Forgot the famous last words? Access your message archive online at http://in.messenger.yahoo.com/webmessengerpromo.php -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071114/fde1d748/attachment.htm From degironc at sissa.it Tue Nov 13 21:56:20 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Tue, 13 Nov 2007 21:56:20 +0100 Subject: [Pw_forum] ion problem In-Reply-To: <360668.39837.qm@web94110.mail.in2.yahoo.com> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> Message-ID: <473A0F74.5000905@sissa.it> Unless you put some constraint the code will distribute the charge density so as to minimize the total energy. If the atoms behaves as ions in your cluster this will be automatically found by the code. A way (one of the many arbitrary ways) to measure this atomic charge is to project the wfc on atomic orbitals and counts how many electrons are projected on each atom... this is done by projwfc.x in the PP directory. If you want to force the system to do something else w.r.t. what it would do automatically you need to add some constraint. For instance in a non collinear magnetic calculation one can constraint the size of the atomic magnetization, its direction or one of its commponents. In the cvs version magnetization size can be constrained also in the collinear case. Currently there is no constraint on the atomic charge but using the constrained_magnetization part as template should not be too difficult implement such a feature. Hope this helps, Don't forget to sign your posts including your affilation. Stefano de Gironcoli, SISSA and DEMOCRITOS Rudra Banerjee wrote: > dear PWSCF users, > I am new to pwscf and stuck in a question.......is there any way to > put anion or cation as a unit?the total cluster is charge less > but have two different types of ion.......can anyone tell me what to > do in this situation? > Regards > > -- > Rudra > Have a Nice Time > http://www.bose.res.in/~rudra > Please, if possible, don't send me MS Word or PowerPoint attachments > Why?See: http://www.gnu.org/philosophy/no-word-attachments.html > > ------------------------------------------------------------------------ > Forgot the famous last words? Access your message archive online. > Click here. > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From b.walker at irl.cri.nz Tue Nov 13 22:42:06 2007 From: b.walker at irl.cri.nz (Dr Brent Walker) Date: Wed, 14 Nov 2007 10:42:06 +1300 (NZDT) Subject: [Pw_forum] QE with ESSL on BlueGene Message-ID: Hi all, Does anyone know whether QE has been used successfully with the IBM ESSL libraries on the Blue Gene L architecture (running linux)? Google unfortunately hasn't provided me with much in relation to this. I have spent some time trying to get QE (well really PWscf) running on such a machine and am at the stage of deciding whether to persevere or just give up and use locally compiled versions of fftw and lapack/blas (following say the provided "Make.bgl" file, which seems to work fine for me). Is this worth pursuing or should I just file it in the "too hard" basket for the time being? If people think there is some hope that I can get this to work, I'll provide more details (make.sys, etc.). Thanks very much for any information/thoughts/anecdotes on this! Brent. PS. I have noted AK's comment "good luck (you'll be needing it)" regarding compilation of QE on BG/L on 31 Aug, which of course doesn't bode well! -- ############################## Dr Brent Walker Industrial Research Limited, P.O. Box 31-310, 69 Gracefield Road, Lower Hutt 5040, New Zealand. Email: b.walker at irl.cri.nz Phone: +64 (0) 4 931 3783 (extn: 4783) Fax: +64 (0) 4 931 3754 http://www.irl.cri.nz http://www.vuw.ac.nz/scps/research/compnanotech ############################## This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From chenhanghuipwscf at gmail.com Wed Nov 14 02:36:54 2007 From: chenhanghuipwscf at gmail.com (alan chen) Date: Tue, 13 Nov 2007 20:36:54 -0500 Subject: [Pw_forum] constraints in the ionic dynamics Message-ID: <22ae3ca40711131736w3ed45aacx3a3a10c5b16b3ad2@mail.gmail.com> Dear PWSCF users, I am now trying to impose some constraints in the ionic dynamics. However, after I checked the code I found that if I use 'bfgs' algorithm for 'relax' calculation, the constraints are not implemented in the code. It seems that contraints can only be imposed in the molecule dynamics using Verlet algorithm. Is that true? Any one can share his or her experience in imposing contraints in the ionic dynamics? Thank you very much. Hanghui Department of Physics Yale University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071113/0066f5a4/attachment.htm From lyanglyan at yahoo.com.cn Wed Nov 14 04:03:42 2007 From: lyanglyan at yahoo.com.cn (=?gb2312?q?=D1=D2=20=C0=EE?=) Date: Wed, 14 Nov 2007 11:03:42 +0800 (CST) Subject: [Pw_forum] Why the calculated equilibrium lattice parameters Message-ID: <854555.88476.qm@web15806.mail.cnb.yahoo.com> Dear Baroni, Thanks a lot for your advice. I want to simulate equation of states for cubic CsI. I considered 5s2 5p6 6s1 as the valence electron configurations for Cs. In the input file of the ld1.x for Cs is: &input title='Cs', zed=55., rlderiv=2.50, eminld=-4.0, emaxld=4.0, deld=0.02, nld=5, iswitch=3, config='[Kr] 4d10 5s2 5p6 6s1' dft='PZ', / &test nconf=1 pseudotype=2, configts(1)= '5s2 5p6 6s1' / &inputp lloc=0, file_pseudopw='Cs.UPF', zval=9.d0, / 3 5S 1 0 2.00 0.00 2.7 2.7 0.5 5P 2 1 6.00 0.00 2.7 2.7 0.5 6S 1 0 1.00 0.00 2.7 2.7 0.5 Best regards, Yan Li National Lab of Superhard Materials, Jilin University, Changchun 130012, P. R. China. Email: lyanglyan at yahoo.com.cn --------------------------------- ?????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071114/e721866d/attachment.htm From akohlmey at cmm.chem.upenn.edu Wed Nov 14 06:16:53 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 14 Nov 2007 10:46:53 +0530 (IST) Subject: [Pw_forum] ion problem In-Reply-To: <360668.39837.qm@web94110.mail.in2.yahoo.com> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> Message-ID: On Wed, 14 Nov 2007, Rudra Banerjee wrote: RB> dear PWSCF users, dear rudra, RB> I am new to pwscf and stuck in a question......is there any way RB> toput anion or cation as a unit?the total cluster is charge less but RB> have two different types of ion.......can anyone tell me what to do let me guess: you did classical MD before? in a first principles calculation, there is usually no need to distinguish between atoms and ions. you put in the core charges (as point charges or pseudopotentials) and then add the proper number of electrons, and the QM calculation (in DFT approximation when using PWSCF) will take care of how the electrons are distributed. so you only have to take care of the total charge, because it determines in combination with the sum of the core charges how many electrons need to be added. HTH, axel. RB> in this situation? Regards -- Rudra Have a Nice Time http://www.bose.res.in/~rudra Please, if possible, don't send me MS Word or PowerPoint attachments Why?See: http://www.gnu.org/philosophy/no-word-attachments.html Forgot the famous last words? Access your message archive online at http://in.messenger.yahoo.com/webmessengerpromo.php -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From akohlmey at cmm.chem.upenn.edu Wed Nov 14 06:19:17 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 14 Nov 2007 10:49:17 +0530 (IST) Subject: [Pw_forum] constraints in the ionic dynamics In-Reply-To: <22ae3ca40711131736w3ed45aacx3a3a10c5b16b3ad2@mail.gmail.com> References: <22ae3ca40711131736w3ed45aacx3a3a10c5b16b3ad2@mail.gmail.com> Message-ID: On Tue, 13 Nov 2007, alan chen wrote: AC> Dear PWSCF users, dear hanghui, AC> I am now trying to impose some constraints in the ionic dynamics. AC> However, after I checked the code I found that if I use 'bfgs' algorithm for AC> 'relax' calculation, the constraints are not implemented in the code. It implementing constraints in a geometry optimization (unlike in an MD) is a non-trivial task. many codes i know "fake" them by using a very, very stiff force constant. using damped dynamics is a reasonable alternative to BFGS and will allow constraints. ceers, axel. AC> seems that contraints can only be imposed in the molecule dynamics using AC> Verlet algorithm. Is that true? Any one can share his or her experience in AC> imposing contraints in the ionic dynamics? AC> Thank you very much. AC> AC> Hanghui AC> Department of Physics AC> Yale University AC> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From akohlmey at cmm.chem.upenn.edu Wed Nov 14 06:24:00 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 14 Nov 2007 10:54:00 +0530 (IST) Subject: [Pw_forum] Why the calculated equilibrium lattice parameters In-Reply-To: <854555.88476.qm@web15806.mail.cnb.yahoo.com> References: <854555.88476.qm@web15806.mail.cnb.yahoo.com> Message-ID: On Wed, 14 Nov 2007, ?? ?? wrote: YL> Dear Baroni, dear yan li, this is only a partial answer to stefano's questions. what are the input parameters where you get the bad results. what tests did you do to confirm the validity of your and the I pseudopotential? you have to explain what you did and why you chose certain paramters and why you think that your results are bad. nobody here has the time to do those tests for you or can determine from simply looking at an input for the atomic code why you are getting bad results. there are far too many ways to mess up calculations. cheers, axel. YL> Thanks a lot for your advice. I want to simulate equation of states for cubic CsI. I considered 5s2 5p6 6s1 as the valence electron configurations for Cs. In the input file of the ld1.x for Cs is: YL> YL> &input YL> title='Cs', YL> zed=55., YL> rlderiv=2.50, YL> eminld=-4.0, YL> emaxld=4.0, YL> deld=0.02, YL> nld=5, YL> iswitch=3, YL> config='[Kr] 4d10 5s2 5p6 6s1' YL> dft='PZ', YL> / YL> &test YL> nconf=1 YL> pseudotype=2, YL> configts(1)= '5s2 5p6 6s1' YL> / YL> &inputp YL> lloc=0, YL> file_pseudopw='Cs.UPF', YL> zval=9.d0, YL> / YL> 3 YL> 5S 1 0 2.00 0.00 2.7 2.7 0.5 YL> 5P 2 1 6.00 0.00 2.7 2.7 0.5 YL> 6S 1 0 1.00 0.00 2.7 2.7 0.5 YL> YL> Best regards, YL> YL> Yan Li YL> National Lab of Superhard Materials, YL> Jilin University, Changchun 130012, P. R. China. YL> Email: lyanglyan at yahoo.com.cn YL> YL> YL> YL> YL> --------------------------------- YL> ???????????????????? -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From akohlmey at cmm.chem.upenn.edu Wed Nov 14 06:31:35 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 14 Nov 2007 11:01:35 +0530 (IST) Subject: [Pw_forum] Convergence test In-Reply-To: <010801c825fa$5fab6960$1d00a8c0@solarflare> References: <010801c825fa$5fab6960$1d00a8c0@solarflare> Message-ID: On Tue, 13 Nov 2007, Huiqun Zhou wrote: HZ> List users, dear huiqun zhou, HZ> I'm testing convergence on ecutwfc for chromite using USPP. Below are HZ> results of different ecutrho with different ecutwfc. I may be wrong, I think HZ> I should watch the difference of total energy between successive runs till HZ> it's less than 1-2 mRy. But the trend is not apparent in the results below HZ> and it seems that the larger the ecutwfc, the lower the total energy. you are using an ultra-soft pseudopotential and for those actually the density cutoff is _much_ more important than the wavefunction cutoff. depending on the choice of the pseudopotential parameters, 35-40ry for the wavefunction cutoff should be a conservative choice, however a density cutoff of 250 is not. if you want to converge the energy very tightly, you may need to go up to 10x the wavefunction cutoff. you should also keep in mind that the total energy converges rather slowly, and that for most calculations you need energy differences or derivatives, which may converge faster, i.e. you may want to look at forces (use a non-symmetric input for that!!) or stress instead. what is sufficiently converged depends on your the kind of calculation you want to do. so for simple geometry optimization or MD 35/210 or less may already be sufficient. HZ> HZ> I calculated E-V relations using two sets of ecutwfc and ecutrho: HZ> for ecutwfc = 35 ry and ecutrho = 240 ry, a = 8.405, B0 = 172.8, B0' = 4.3 HZ> for ecutwfc = 55 ry and ecutrho = 250 ry, a = 8.407, B0 = 178.6, B0' = 3.6 HZ> HZ> There are no big differences in unit cell parameter and elastic properties. HZ> So, my questions are HZ> (1) usually, the ecutwfc of USPP should be around 35 ry or below, is my HZ> case an exception? no. it does not look like an exception. HZ> (2) how small the difference of total energy of successive runs should be HZ> in judging convergence against ecutwfc? total energy is not a good parameter. see above. cheers, axel. HZ> HZ> Thank you in advance. HZ> HZ> Dr. Huiqun Zhou HZ> @Earth Sciences, Nanjing University, China HZ> HZ> HZ> HZ> HZ> ecutrho = 240 ry HZ> ecutwfc total energy HZ> 35 -1461.54573299 Ry HZ> 40 -1461.64206824 Ry HZ> 45 -1461.71978835 Ry HZ> 50 -1461.76587314 Ry HZ> 55 -1461.76316597 Ry HZ> 60 -1461.76692396 Ry HZ> HZ> ecutrho = 250 ry HZ> 40 -1461.64351876 Ry HZ> 45 -1461.70130510 Ry HZ> 50 -1461.74724764 Ry HZ> 55 -1461.76462943 Ry HZ> 60 -1461.76838812 Ry HZ> HZ> ecutrho = 280 ry HZ> 50 -1461.74778158 Ry HZ> 55 -1461.76516515 Ry HZ> 60 -1461.76892327 Ry HZ> HZ> _______________________________________________ HZ> Pw_forum mailing list HZ> Pw_forum at pwscf.org HZ> http://www.democritos.it/mailman/listinfo/pw_forum HZ> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From akohlmey at cmm.chem.upenn.edu Wed Nov 14 06:44:08 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 14 Nov 2007 11:14:08 +0530 (IST) Subject: [Pw_forum] QE with ESSL on BlueGene In-Reply-To: References: Message-ID: On Wed, 14 Nov 2007, Dr Brent Walker wrote: BW> Hi all, hi brent, BW> Does anyone know whether QE has been used successfully with the IBM BW> ESSL libraries on the Blue Gene L architecture (running linux)? Google BW> unfortunately hasn't provided me with much in relation to this. yep. i've managed to compile and run QE on a BG/L. BW> I have spent some time trying to get QE (well really PWscf) running on BW> such a machine and am at the stage of deciding whether to persevere or BW> just give up and use locally compiled versions of fftw and lapack/blas BW> (following say the provided "Make.bgl" file, which seems to work fine BW> for me). due to the (lack of) features in the BG/L cpu, you may actually get reasonable performance with regular BLAS/LAPACK. you can try the "double hummer" libraries, but then you are limited to coprocessor mode. this is probably needed anyways, because the limitations to jobs on BG/L are very hard. there is no local storage, so pw.x with its default setting of storing wavefunctions as files, is not scaling well. you'll have to use the (experimental) feature of storing those in memory, but with 512MB/node there is not much memory available. on top of that, the cpus on BG/L are very slow, so you need to parallelize across a large number of cpus to get decent performance. in my view for a code like pw.x it is currently not worth the hassle. your chances with cp.x are much better, but then again, you are limited by the supported feature set of cp.x. altogether, you have to keep in mind, that BG/L is mainly a machine to get a good ranking in the top500 and thus please administrators, politicians and generally people who are not using it. from the user's perspective it is a constant struggle and a PITA. if i had the choice, i'd rather skip the top500 placement and get a machine that is usable. the majority of QE jobs are run on rather small clusters, so to run well on those machines is where most of the effort goes. BW> Is this worth pursuing or should I just file it in the "too hard" BW> basket for the time being? If people think there is some hope that I BW> can get this to work, I'll provide more details (make.sys, etc.). BW> BW> Thanks very much for any information/thoughts/anecdotes on this! well, i've been struggling a lot with finding _any_ project that runs well on a BG/L that does not run better on a cray xt3/xt4 or even a reasonably well laid out PC cluster with DDR infiniband. my best results were so far with classical MD using LAMMPS on systems that have no coulomb interactions. there i am scaling out on the BG/L at half the performance of the scaleout timing on a cray xt3. for most codes, particularly plane wave pseudopotential DFT the difference is about a factor of 10. so before putting in more effort, it might be worth to discuss what kind of calculations you intend to run and how much cpu time across how many nodes you have at your disposal. cheers, axel. BW> BW> Brent. BW> BW> PS. I have noted AK's comment "good luck (you'll be needing it)" BW> regarding compilation of QE on BG/L on 31 Aug, which of course doesn't BW> bode well! BW> BW> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From cyrille.barreteau at cea.fr Wed Nov 14 08:53:19 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Wed, 14 Nov 2007 08:53:19 +0100 Subject: [Pw_forum] ion problem In-Reply-To: <473A0F74.5000905@sissa.it> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> <473A0F74.5000905@sissa.it> Message-ID: <473AA96F.30201@cea.fr> Dear all I have a question related to this problem. I am interested in the study of charged clusters. If I understand well, due to to the periodicity of the plane wave codes one has to substract the correct "background" in order to keep the energy from going to infinity.. I would like to compare the energy of clusters with different charges -e -2e etc... Do you have some advices to do such calculation? Is it necessary to take even larger supercells than for neutral clusters? thanks in advance cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html ================================================================== Stefano de Gironcoli wrote: >Unless you put some constraint the code will distribute the charge >density so as to minimize the total energy. If the atoms behaves as ions >in your cluster this will be automatically found by the code. >A way (one of the many arbitrary ways) to measure this atomic charge is >to project the wfc on atomic orbitals and counts how many electrons are >projected on each atom... this is done by projwfc.x in the PP directory. >If you want to force the system to do something else w.r.t. what it >would do automatically you need to add some constraint. For instance in >a non collinear magnetic calculation one can constraint the size of the >atomic magnetization, its direction or one of its commponents. In the >cvs version magnetization size can be constrained also in the collinear >case. >Currently there is no constraint on the atomic charge but using the >constrained_magnetization part as template should not be too difficult >implement such a feature. > >Hope this helps, > >Don't forget to sign your posts including your affilation. > >Stefano de Gironcoli, SISSA and DEMOCRITOS > >Rudra Banerjee wrote: > > >>dear PWSCF users, >>I am new to pwscf and stuck in a question.......is there any way to >>put anion or cation as a unit?the total cluster is charge less >>but have two different types of ion.......can anyone tell me what to >>do in this situation? >>Regards >> >>-- >>Rudra >>Have a Nice Time >>http://www.bose.res.in/~rudra >>Please, if possible, don't send me MS Word or PowerPoint attachments >>Why?See: http://www.gnu.org/philosophy/no-word-attachments.html >> >>------------------------------------------------------------------------ >>Forgot the famous last words? Access your message archive online. >>Click here. >> >> >>------------------------------------------------------------------------ >> >>_______________________________________________ >>Pw_forum mailing list >>Pw_forum at pwscf.org >>http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From degironc at sissa.it Wed Nov 14 09:02:23 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 14 Nov 2007 09:02:23 +0100 Subject: [Pw_forum] ion problem In-Reply-To: <473AA96F.30201@cea.fr> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> <473A0F74.5000905@sissa.it> <473AA96F.30201@cea.fr> Message-ID: <473AAB8F.2040606@sissa.it> The Makov-Payne correction (assume_isolated flag in &electrons) corrects for the leading term in the cluster-backgound interaction . This should reduce the size of the supercell needed, which should however be large and cheched for conergence. stefano Cyrille Barreteau wrote: > Dear all > > I have a question related to this problem. > I am interested in the study of charged clusters. > If I understand well, due to to the periodicity of the plane wave codes > one has to substract the correct "background" in order to keep the > energy from going to infinity.. > > I would like to compare the energy of clusters with different > charges -e -2e etc... Do you have some advices to do such calculation? > Is it necessary to take even larger supercells than for neutral clusters? > > thanks in advance > > cyrille > > From Giovanni.Cantele at na.infn.it Wed Nov 14 09:20:17 2007 From: Giovanni.Cantele at na.infn.it (Giovanni Cantele) Date: Wed, 14 Nov 2007 09:20:17 +0100 Subject: [Pw_forum] ion problem In-Reply-To: <473AA96F.30201@cea.fr> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> <473A0F74.5000905@sissa.it> <473AA96F.30201@cea.fr> Message-ID: <473AAFC1.60002@na.infn.it> Cyrille Barreteau wrote: > Dear all > > I have a question related to this problem. > I am interested in the study of charged clusters. > If I understand well, due to to the periodicity of the plane wave codes > one has to substract the correct "background" in order to keep the > energy from going to infinity.. > This is automatically done by the code, as you may read in the INPUT_PW documentation file: nelec REAL number of electron in the unit cell (may be noninteger if you wish) Default: the same as ionic charge (neutral cell) A compensating jellium background is inserted to remove divergences if the cell is not neutral So, you set nelec in the input file, and that should be enough. > I would like to compare the energy of clusters with different > charges -e -2e etc... Do you have some advices to do such calculation? > Is it necessary to take even larger supercells than for neutral clusters? > The problem with charged clusters is that spurious interactions are introduced by the periodic boundary conditions, so as convergence would be achieved at very large supercells. Neverthelessm the total energy can be corrected, to get rid of such interactions, as described by Makov-Payne, PRB (1995). Such corrections are currently implemented in Quantum-ESPRESSO, with the only warning that, as far as I know, corrections are applied to the total energy but are not to the forces. If you are using espresso-3.2, you should - run a self-consistent calculation for you charged cluster - run post-processing code pp.x, for obtaining the total charge density (plot_num=0) in a suitable format - run dipole.x, you may read the header of PP/dipole.f90 for very short explanation You finally end up with a Makov-Payne correction that must be added to the total energy WITH its sign. You may try the same charged cluster with two different supercells, and compare the two uncorrected energies, as well as the corrected ones, finding the latter much more converged. Next releases will probably have an easier procedure. In the CVS version I understand that a new variable (assume_isolated) has been added to the pw.x input, that, if set to .TRUE., should allow to get the MP correction directly within the scf loop (so post-processing is not mandatory). Giovanni -- Dr. Giovanni Cantele Coherentia CNR-INFM and Dipartimento di Scienze Fisiche Universita' di Napoli "Federico II" Complesso Universitario di Monte S. Angelo - Ed. 6 Via Cintia, I-80126, Napoli, Italy Phone: +39 081 676910 Fax: +39 081 676346 E-mail: Giovanni.Cantele at na.infn.it Web: http://people.na.infn.it/~cantele From cyrille.barreteau at cea.fr Wed Nov 14 09:38:32 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Wed, 14 Nov 2007 09:38:32 +0100 Subject: [Pw_forum] ion problem In-Reply-To: <473AAFC1.60002@na.infn.it> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> <473A0F74.5000905@sissa.it> <473AA96F.30201@cea.fr> <473AAFC1.60002@na.infn.it> Message-ID: <473AB408.3090709@cea.fr> Thanks a lot for this detailed explanation... Hopefully I will succeed to follow this tortuous way, else I will wait for the next release :-) cyrille Giovanni Cantele wrote: >Cyrille Barreteau wrote: > > >>Dear all >> >>I have a question related to this problem. >>I am interested in the study of charged clusters. >>If I understand well, due to to the periodicity of the plane wave codes >>one has to substract the correct "background" in order to keep the >>energy from going to infinity.. >> >> >> >This is automatically done by the code, as you may read in the INPUT_PW >documentation file: >nelec REAL > number of electron in the unit cell > (may be noninteger if you wish) > Default: the same as ionic charge (neutral cell) > A compensating jellium background is inserted > to remove divergences if the cell is not neutral > >So, you set nelec in the input file, and that should be enough. > > > >>I would like to compare the energy of clusters with different >>charges -e -2e etc... Do you have some advices to do such calculation? >>Is it necessary to take even larger supercells than for neutral clusters? >> >> >> >The problem with charged clusters is that spurious interactions are >introduced by the >periodic boundary conditions, so as convergence would be achieved at >very large supercells. >Neverthelessm the total energy can be corrected, to get rid of such >interactions, as >described by Makov-Payne, PRB (1995). Such corrections are currently >implemented >in Quantum-ESPRESSO, with the only warning that, as far as I know, >corrections are >applied to the total energy but are not to the forces. > >If you are using espresso-3.2, you should >- run a self-consistent calculation for you charged cluster >- run post-processing code pp.x, for obtaining the total charge density >(plot_num=0) in >a suitable format >- run dipole.x, you may read the header of PP/dipole.f90 for very short >explanation > >You finally end up with a Makov-Payne correction that must be added to >the total energy >WITH its sign. >You may try the same charged cluster with two different supercells, and >compare the >two uncorrected energies, as well as the corrected ones, finding the >latter much more >converged. > >Next releases will probably have an easier procedure. In the CVS version >I understand that a new variable >(assume_isolated) has been added to the pw.x input, that, if set to >.TRUE., should allow to get >the MP correction directly within the scf loop (so post-processing is >not mandatory). > >Giovanni > > > > -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== From giannozz at nest.sns.it Wed Nov 14 10:14:22 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 14 Nov 2007 10:14:22 +0100 Subject: [Pw_forum] QE with ESSL on BlueGene In-Reply-To: References: Message-ID: <46DC6C03-C38A-4BBC-B753-34549A70B6C4@nest.sns.it> On Nov 14, 2007, at 6:44 , Axel Kohlmeyer wrote: > you'll have to use the (experimental) feature of storing those in > memory, but with 512MB/node there is not much memory available. in this case you may take advantage of recent improvements in memory usage: many statically dimensioned arrays have been removed and more arrays distributed across processors. In the current cvs version of PWscf, use diagonalization='david+distpara'. For CP a further level of parallelization ("task groups") has been implemented for both FFT's and iterative orthonormalization. It was done for BlueGene machines but it works in general. It can be activated with a command line option (- ntg N), but I don't remember if and where this is documented, I'll inquire. With such improvements, Carlo Cavazzoni has been able to run a realistic benchmark on up to 4000 processors of a cray XT4 (with CP). I have no data on the BlueGene and on PWscf. The last time I tried to run PWscf on a machine with a lot of CPUs and little memory per CPU it was 1992 on the Connection Machine CM2 in Pisa. It did run, but instead of "speedup", it showed a nice "speedown". I used to refer to it as a machine for "perpendicular calculation". Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Wed Nov 14 10:24:25 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 14 Nov 2007 10:24:25 +0100 Subject: [Pw_forum] Why the calculated equilibrium lattice parameters In-Reply-To: <854555.88476.qm@web15806.mail.cnb.yahoo.com> References: <854555.88476.qm@web15806.mail.cnb.yahoo.com> Message-ID: <5F617A72-4813-4617-A103-D52CB51EDDAF@nest.sns.it> On Nov 14, 2007, at 4:03 , ? ? wrote: > Thanks a lot for your advice. I want to simulate equation of states > for cubic CsI. get the pseudopotentials for Cs and I on pwscf.org and try those first. Once you have obtained a decent equation of state, you can start to experiment with different pseudopotentials, different functionals, different cutoffs/k-points/whatever, better if changing things one at the time P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From b.walker at irl.cri.nz Wed Nov 14 10:25:14 2007 From: b.walker at irl.cri.nz (Dr Brent Walker) Date: Wed, 14 Nov 2007 22:25:14 +1300 (NZDT) Subject: [Pw_forum] QE with ESSL on BlueGene In-Reply-To: Message-ID: Hi Axel, Thanks very much for your insightful comments. Just before seeing your email I actually managed to get PWscf going on our BG/L. I found it was crashing deep inside a call to the "dgetri" routine (called in invmat.f90) which is provided by essl. Without the source for the implementation of that routine in essl it was hard to see what was happening, however, I managed to get around the problem firstly by setting a significantly larger "work" array. The vendor provided documentation says that for "optimal performance" you should set the size of the work array to be >= 100 times the size of the array being inverted. I guess by "optimal performance" they meant "won't cause your code to crash". I then found that the essl implementation of that routine has an extension to the LAPACK spec whereby you can have it automatically determine and allocate the optimal size for its work array (this setting works also and is what I guess I will use now, though one could imagine scenarios where this might go wrong with the fairly severe memory limitations on the BG/L). So far, I'd say the BG/L isn't the easiest piece of equipment I've used (especially in comparison with say a p5 series machine, where things like QE tend to work "out of the box"). At the moment however that's what I've got available, so of course I'll use it as best I can! As a very crude guess (that is, on one test system containing ~250 electrons) I'd say using the vendor-provided libs amounts to a 30 % improvement over locally compiled blas/lapack/fftw, though for larger numbers of CPUs (~512) this seems to be much smaller (which is probably related to the issues you mention regarding the ill fit of PWscf to BG/L). All that being said, I won't be surprised if I can break things again .... Your comments about the perfomance of LAMMPS are encouaging however. Some of my "non-DFT" colleagues are setting themselves up to use that code, so it's good to know that they should be able to expect reasonable performance from it. And while I'm getting a bit off-topic, it seems my troubles getting PWscf/QE running on BG/L are much less severe than another of my colleagues who is attempting to run a popular commercial plane-wave DFT code. Kindest regards, Brent. On Wed, 14 Nov 2007, Axel Kohlmeyer wrote: > On Wed, 14 Nov 2007, Dr Brent Walker wrote: > > BW> Hi all, > > hi brent, > > BW> Does anyone know whether QE has been used successfully with the IBM > BW> ESSL libraries on the Blue Gene L architecture (running linux)? Google > BW> unfortunately hasn't provided me with much in relation to this. > > yep. i've managed to compile and run QE on a BG/L. > > BW> I have spent some time trying to get QE (well really PWscf) running on > BW> such a machine and am at the stage of deciding whether to persevere or > BW> just give up and use locally compiled versions of fftw and lapack/blas > BW> (following say the provided "Make.bgl" file, which seems to work fine > BW> for me). > > due to the (lack of) features in the BG/L cpu, you may actually > get reasonable performance with regular BLAS/LAPACK. you can > try the "double hummer" libraries, but then you are limited to > coprocessor mode. this is probably needed anyways, because the > limitations to jobs on BG/L are very hard. there is no local > storage, so pw.x with its default setting of storing wavefunctions > as files, is not scaling well. you'll have to use the (experimental) > feature of storing those in memory, but with 512MB/node there is > not much memory available. on top of that, the cpus on BG/L are > very slow, so you need to parallelize across a large number of > cpus to get decent performance. in my view for a code like pw.x > it is currently not worth the hassle. your chances with cp.x > are much better, but then again, you are limited by the supported > feature set of cp.x. altogether, you have to keep in mind, that > BG/L is mainly a machine to get a good ranking in the top500 > and thus please administrators, politicians and generally people > who are not using it. from the user's perspective it is a constant > struggle and a PITA. if i had the choice, i'd rather skip the > top500 placement and get a machine that is usable. the majority > of QE jobs are run on rather small clusters, so to run well on > those machines is where most of the effort goes. > > BW> Is this worth pursuing or should I just file it in the "too hard" > BW> basket for the time being? If people think there is some hope that I > BW> can get this to work, I'll provide more details (make.sys, etc.). > BW> > BW> Thanks very much for any information/thoughts/anecdotes on this! > > well, i've been struggling a lot with finding _any_ project that > runs well on a BG/L that does not run better on a cray xt3/xt4 > or even a reasonably well laid out PC cluster with DDR infiniband. > > my best results were so far with classical MD using LAMMPS on > systems that have no coulomb interactions. there i am scaling > out on the BG/L at half the performance of the scaleout timing > on a cray xt3. for most codes, particularly plane wave > pseudopotential DFT the difference is about a factor of 10. > > so before putting in more effort, it might be worth to discuss > what kind of calculations you intend to run and how much > cpu time across how many nodes you have at your disposal. > > cheers, > axel. > > BW> > BW> Brent. > BW> > BW> PS. I have noted AK's comment "good luck (you'll be needing it)" > BW> regarding compilation of QE on BG/L on 31 Aug, which of course doesn't > BW> bode well! > BW> > BW> > > -- ############################## Dr Brent Walker Industrial Research Limited, P.O. Box 31-310, 69 Gracefield Road, Lower Hutt 5040, New Zealand. Email: b.walker at irl.cri.nz Phone: +64 (0) 4 931 3783 (extn: 4783) Fax: +64 (0) 4 931 3754 http://www.irl.cri.nz http://www.vuw.ac.nz/scps/research/compnanotech ############################## This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From katalin.gaal-nagy at physik.uni-regensburg.de Wed Nov 14 13:25:05 2007 From: katalin.gaal-nagy at physik.uni-regensburg.de (Katalin Gaal-Nagy) Date: Wed, 14 Nov 2007 13:25:05 +0100 (CET) Subject: [Pw_forum] ph.x - from broyden : error ... Message-ID: Dear all, once again a question about a crashed phonon job... It happened to me as it happened to Marcel Mohr (30/10/2007), that a phonon calculation crashed and the recover did not work ... well, that's the question. In my case the original calculation ended bcause of ... Representation # 62 mode # 62 Self-consistent Calculation iter # 1 total cpu time : 61.8 secs av.it.: 3.0 thresh= 0.100E-01 alpha_mix = 0.700 |ddv_scf|^2 = 0.819E-11 iter # 2 total cpu time : 62.0 secs av.it.: 16.0 thresh= 0.286E-06 alpha_mix = 0.700 |ddv_scf|^2 = 0.134E-11 iter # 3 total cpu time : 62.1 secs av.it.: 17.0 thresh= 0.116E-06 alpha_mix = 0.700 |ddv_scf|^2 = 0.122E-12 iter # 4 total cpu time : 62.3 secs av.it.: 17.0 thresh= 0.349E-07 alpha_mix = 0.700 |ddv_scf|^2 = 0.989E-14 IOS = 38 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from davcio : error # 99 i/o error in davcio %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... I tried a restart and got ... Representation # 62 mode # 62 Self-consistent Calculation %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from broyden : error # 4 factorization %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... Since ph.x recognized correctly that the first 61 representations have already been calculated, I don't think that it is a restart problem. I don't think neither that there is a problem with the input, since I did similar calculations without any problems using the same PS and parameters for a slightly different configuration of the molecule. What I found in the archive is, that the first error is due to some in/out, maybe diskspace, network, or whatever. The second due to the matrix factorization, maybe due to reading a corrupt file. My question is now: Do I have to restart the calculation from scratch or is there some way to recover it avoiding this error? Maybe it would be enough to restart the calculation of Representation 62 from scratch by deleting/editing some files in the tmp_dir? Any hint is welcome! All the best, Katalin ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, , >>>>>>> Katalin Gaal-Nagy <<<<<<< , , Phys.Dept./INFM , , University of Milan , , Via Celoria 16 , , +39 02 50317377 , ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, , >>>>>>> katalin.gaal-nagy at physik.uni-regensburg.de <<<<<<< , ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, , >>>>>>> Don't dream your life, live your dream <<<<<<< , ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, From giannozz at nest.sns.it Wed Nov 14 14:59:19 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 14 Nov 2007 14:59:19 +0100 Subject: [Pw_forum] Sign of magnetization In-Reply-To: <4738B091.1080702@berkeley.edu> References: <4738B091.1080702@berkeley.edu> Message-ID: <658A5C85-0A39-4EB7-A956-072E7700AE26@nest.sns.it> On Nov 12, 2007, at 20:59 , Ivo Souza wrote: > I am trying to understand whether there is a sign error in the > definition > of the total magnetization in pwscf. In other words, when the code > tells > us that, e.g., the magnetic moment of bcc iron points along +z, is > it really > pointing along minus z? short answer: yes, or at least, this is what I understood after the exchange of several e-mails with several other developers ... > The spin magnetic moment of a state is (n units of the Bohr > magneton) > > <\mu_z> = - or, equivalently, <\mu_z> = -\mu_B where \mu_B = |e|\hbar/(2mc) is the Bohr magneton (a positive number). For historical reasons, going back to the first implementations of colinear and noncolinear magnetism in PWscf, the Bohr magneton is instead implicitly defined as \mu_B = e\hbar/(2mc) with e = true (negative) electron charge, so : <\mu_z> = \mu_B This convention is used in Andrea's papers and in other papers on noncolinear magnetism as well. There is nothing wrong in this convention, but its usage in PWscf is a little bit at odds with the rest of the code, where 'e', the electron charge, is actually |e| and is positive (as all the other physical constants are). Since the exchange-correlation energy depends on the modulus of the magnetization, the theory is invariant with respect to its sign. Of course this is no longer true when one introduces an external magnetic field and one has to take care to use a consistent convention. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From yukihiro_okuno at fujifilm.co.jp Wed Nov 14 15:25:52 2007 From: yukihiro_okuno at fujifilm.co.jp (yukihiro_okuno at fujifilm.co.jp) Date: Wed, 14 Nov 2007 23:25:52 +0900 Subject: [Pw_forum] =?iso-2022-jp?b?RXJyb3IgaW4gdmMtcmVsYXggGyRCISEbKEIo?= =?iso-2022-jp?b?bm90IGltcGxlbWVudGVkIHJpbm5lciBvciBucWYp?= Message-ID: Dear PWSCF users and developers. I'm now want to use virtual crystal approximation, and using virtual.x for making the virtual atom from the pseudopotentials of real atom. But frequently the errors occurs to with the message below, For example1, from Virtual : error # different rinner are not implemented (yet) stopping 2 or example2 from Virtual : error# different nqf are not implemented yet I think example2 may be the pseudo charge part of pseudo potential , but for example1 is unclear. When such errors occure, is it not available to use the pseudopotential of real atom and remake pseudo potential? (If possible I want to use the pseudopotential I have because it is reliable to use) Or is it available to correct part of virtual.f90 ( for example, in the case of different mesh interpotation) ? Sincerely. From giannozz at nest.sns.it Wed Nov 14 16:56:24 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 14 Nov 2007 16:56:24 +0100 Subject: [Pw_forum] ph.x - from broyden : error ... In-Reply-To: References: Message-ID: <67D43A28-5E94-482E-B632-DD72DCE8A806@nest.sns.it> On Nov 14, 2007, at 13:25 , Katalin Gaal-Nagy wrote: > once again a question about a crashed phonon job... > It happened to me as it happened to Marcel Mohr (30/10/2007), that a > phonon calculation crashed and the recover did not work ... well, > that's > the question. > > In my case the original calculation ended because of > from davcio : error # 99 > i/o error in davcio it looks like the disk is full or, worse, faulty... > Do I have to restart the calculation from scratch or is there some > way to > recover it avoiding this error? Maybe it would be enough to restart > the > calculation of Representation 62 from scratch by deleting/editing some > files in the tmp_dir? the way restart is currently implemented makes it very hard to deal with a case like yours. First of all, files with information on the status of the calculation are unformatted; second, the overall logic is complicated and one is never sure what has been already done and what hasn't. Until the day somebody with enough good will and knowledge will need to restart long phonon runs in a reliable way, the phonon restart procedure will stay the way it is, i.e., adventurous. In my opinion, your best chance is to modify the routine PH/phq_recover.f90 so that the code restarts from the beginning of the current mode (the one where the code crashed). I vaguely remember that there was a potential problem with some pieces of the dynamical matrix (dyn0 and the like) but I don't remember exactly what it was. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From katalin.gaal-nagy at physik.uni-regensburg.de Wed Nov 14 17:02:02 2007 From: katalin.gaal-nagy at physik.uni-regensburg.de (Katalin Gaal-Nagy) Date: Wed, 14 Nov 2007 17:02:02 +0100 (CET) Subject: [Pw_forum] ph.x - from broyden : error ... In-Reply-To: <67D43A28-5E94-482E-B632-DD72DCE8A806@nest.sns.it> References: <67D43A28-5E94-482E-B632-DD72DCE8A806@nest.sns.it> Message-ID: Dear Paolo, thanks for your answer! I will see what I can do by changing something in the routine. All the best, Katalin On Wed, 14 Nov 2007, Paolo Giannozzi wrote: > > On Nov 14, 2007, at 13:25 , Katalin Gaal-Nagy wrote: > >> once again a question about a crashed phonon job... >> It happened to me as it happened to Marcel Mohr (30/10/2007), that a >> phonon calculation crashed and the recover did not work ... well, >> that's >> the question. >> >> In my case the original calculation ended because of >> from davcio : error # 99 >> i/o error in davcio > > it looks like the disk is full or, worse, faulty... > >> Do I have to restart the calculation from scratch or is there some >> way to >> recover it avoiding this error? Maybe it would be enough to restart >> the >> calculation of Representation 62 from scratch by deleting/editing some >> files in the tmp_dir? > > the way restart is currently implemented makes it very hard to deal > with a case like yours. First of all, files with information on the > status > of the calculation are unformatted; second, the overall logic is > complicated > and one is never sure what has been already done and what hasn't. > Until the day somebody with enough good will and knowledge will > need to restart long phonon runs in a reliable way, the phonon restart > procedure will stay the way it is, i.e., adventurous. > > In my opinion, your best chance is to modify the routine > PH/phq_recover.f90 so that the code restarts from the > beginning of the current mode (the one where the code > crashed). I vaguely remember that there was a potential > problem with some pieces of the dynamical matrix (dyn0 > and the like) but I don't remember exactly what it was. > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From asafis at yahoo.com.br Wed Nov 14 19:01:27 2007 From: asafis at yahoo.com.br (=?iso-8859-1?q?=C1lvaro=20Alves?=) Date: Wed, 14 Nov 2007 15:01:27 -0300 (ART) Subject: [Pw_forum] pseudopotential Message-ID: <980542.73510.qm@web52301.mail.re2.yahoo.com> Hi. How to generate a pseudopotencial LDA for the fluorine.. ?lvaro --------------------------------- Abra sua conta no Yahoo! Mail, o ?nico sem limite de espa?o para armazenamento! -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071114/bab1d52e/attachment.htm From baroni at sissa.it Wed Nov 14 19:26:32 2007 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 14 Nov 2007 18:26:32 +0000 Subject: [Pw_forum] pseudopotential In-Reply-To: <980542.73510.qm@web52301.mail.re2.yahoo.com> References: <980542.73510.qm@web52301.mail.re2.yahoo.com> Message-ID: <9EC09245-46D8-402F-93BC-68C24B2E470A@sissa.it> the same way would for any other element ;-) S. On Nov 14, 2007, at 6:01 PM, ?lvaro Alves wrote: > Hi. How to generate a pseudopotencial LDA for the fluorine.. > > ?lvaro > > > > > > Abra sua conta no Yahoo! Mail, o ?nico sem limite de espa?o para > armazenamento! > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071114/ede15c5f/attachment-0001.htm From isouza at berkeley.edu Wed Nov 14 19:54:25 2007 From: isouza at berkeley.edu (Ivo Souza) Date: Wed, 14 Nov 2007 10:54:25 -0800 Subject: [Pw_forum] Sign of magnetization In-Reply-To: <473AE65D.9000009@ictp.it> References: <473AE65D.9000009@ictp.it> Message-ID: <473B4461.7040806@berkeley.edu> Ralph, I agree with your comment, except for the part where you imply that spin-orbit effects are of little relevance. They are very relevant to many people. I can name at least three or four :) In fact, it was by investigating a spin-orbit-induced effect (the so-called anomalous Hall effect) in iron that we concluded(*) that there was a sign problem in the definition of the magnetization in pwscf. For bcc iron with the magnetization along the positive z-direction it is known, both experimentally and from first-principles calculations [see PRL 92, 037204 ('04)], that the anomalous Hall conductivity sigma_xy is *positive*. Instead, we found a negative value with pwscf. Of course, what really happened is that pwscf was giving us a ground state with a negative magnetization, thus flippling the sign sigma_xy. Regards, and thanks to the pwscf community for the wonderful code! Ivo -- --------------------------------------------------- Ivo Souza Department of Physics, Univ. of California Berkeley tel. +1.510.642.3597 fax +1.510.643.8497 isouza at berkeley.edu http://smilodon.berkeley.edu/ --------------------------------------------------- (*) We actually made a compensating sign error in our code for computing the anomalous Hall conductivity. We just realized that recently... Ralph Gebauer wrote: > Dear Ivo, > > I think that you are right. Probably the sign convention we have adopted > corresponds to a "spin polarization" rather than to a "magnetic moment". > > Since in the usual LDA-GGA functionals the energy depends only on the > absolute value of the local magnetic moment (and not on its direction), > the energy is invariant with respect to a global rotation of the > direction of the magnetization. The direction of \vec{m} is linked to > the lattice axis only if one considers spin-orbit interactions or > externally applied fields, and the like. So the sign ambiguity you > mention is probably of little relevance, but of course, one should still > be precise. Best would probably be to put spin polarization or something > like that in the output, in order to avoid changing sings all over in > the code. > > Ralph > > > > > > >> Dear all, >> >> I am trying to understand whether there is a sign error in the >> definition of the total magnetization in pwscf. In other words, when the >> code tells us that, e.g., the magnetic moment of bcc iron points along >> +z, is it really pointing along minus z? >> >> The spin magnetic moment of a state is (n units of the Bohr magneton) >> >> <\mu_z> = - >> >> (note the minus sign) where sigma_z is the Pauli matrix. >> From the subroutine 'compute_magnetization' in electrons.f90: >> >> ------------------------------------------------------- >> IF ( lsda ) THEN >> ! >> magtot = 0.D0 >> absmag = 0.D0 >> ! >> DO ir = 1, nrxx >> ! >> mag = rho(ir,1) - rho(ir,2) >> ! >> magtot = magtot + mag >> absmag = absmag + ABS( mag ) >> ! >> END DO >> ! >> magtot = magtot * omega / ( nr1*nr2*nr3 ) >> absmag = absmag * omega / ( nr1*nr2*nr3 ) >> ------------------------------------------------------- >> >> My understanding is that rho(ir,1) and rho(ir,2) are respectively the >> (number) densities for spin-up ("moment-down") and spin-down >> ("moment-up") electrons. If so, then it seems to me that >> >> mag = rho(ir,1) - rho(ir,2) >> >> should be replaced by >> >> mag = rho(ir,2) - rho(ir,1) >> >> in order to get the correct sign for the magnetic moment per unit cell >> 'magtot' >> >> The above is for collinear (lsda) calculations, but the same issue with >> the sign is present in the definition of the total magnetization in the >> non-collinear case: >> >> ------------------------------------------------------- >> ELSE IF ( noncolin ) THEN >> ! >> magtot_nc = 0.D0 >> absmag = 0.D0 >> ! >> DO ir = 1,nrxx >> ! >> mag = SQRT( rho(ir,2)**2 + rho(ir,3)**2 + rho(ir,4)**2 ) >> ! >> DO i = 1, 3 >> ! >> magtot_nc(i) = magtot_nc(i) + rho(ir,i+1) >> ! >> END DO >> ! >> absmag = absmag + ABS( mag ) >> ! >> END DO >> ! >> CALL mp_sum( magtot_nc, intra_pool_comm ) >> CALL mp_sum( absmag, intra_pool_comm ) >> ! >> DO i = 1, 3 >> ! >> magtot_nc(i) = magtot_nc(i) * omega / ( nr1*nr2*nr3 ) >> ! >> END DO >> ! >> absmag = absmag * omega / ( nr1*nr2*nr3 ) >> ! >> ENDIF >> ------------------------------------------------------- >> >> Since [see subroutine 'sum_band_k'] rho(ir,2), rho(ir,3) and rho(ir,4) >> are respectively the x,y, and z components of the spin polarization >> density, proportional to psi^*(r).sigma_i.psi(r) (i=x,y,z), the line >> >> magtot_nc(i) = magtot_nc(i) + rho(ir,i+1) >> >> should be replaced by >> >> magtot_nc(i) = magtot_nc(i) - rho(ir,i+1) >> >> >> It is quite easy to get confused with signs, so I apologize if the >> confusion is on my side! Also, it could just be a matter of semantics, >> and what is mean by "magnetic moment" in pwscf is really "spin >> polarization." Any clarification would be greatly appreciated. >> >> With thanks and best regards, >> >> Ivo S. > > From marzari at MIT.EDU Wed Nov 14 23:05:26 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 14 Nov 2007 17:05:26 -0500 Subject: [Pw_forum] ion problem In-Reply-To: <473AA96F.30201@cea.fr> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> <473A0F74.5000905@sissa.it> <473AA96F.30201@cea.fr> Message-ID: <473B7126.8070705@mit.edu> Dear Cyrille, if you are interested in this problem, you might want to have a look at http://arxiv.org/abs/0709.4647 , and in particular at Fig 5. The published Makov-Payne correction is actually incorrect in the charge*quadrupole term. nicola Cyrille Barreteau wrote: > Dear all > > I have a question related to this problem. > I am interested in the study of charged clusters. > If I understand well, due to to the periodicity of the plane wave codes > one has to substract the correct "background" in order to keep the > energy from going to infinity.. > > I would like to compare the energy of clusters with different > charges -e -2e etc... Do you have some advices to do such calculation? > Is it necessary to take even larger supercells than for neutral clusters? > > thanks in advance > > cyrille > -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From marzari at MIT.EDU Wed Nov 14 23:11:44 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 14 Nov 2007 17:11:44 -0500 Subject: [Pw_forum] ion problem In-Reply-To: <360668.39837.qm@web94110.mail.in2.yahoo.com> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> Message-ID: <473B72A0.7070001@mit.edu> Rudra Banerjee wrote: > dear PWSCF users, > I am new to pwscf and stuck in a question.......is there any way to put > anion or cation as a unit?the total cluster is charge less but have > two different types of ion.......can anyone tell me what to do in this > situation? > Regards Dear Rudra, several have answered this question. Let me add, though, that standard approximation to the exact exchange-correlation functional will often fail in "depositing" the right amount of charge on an ion. This becomes a very relevant failure if you have in the same unit cell ions of the same type, but of different valence, e.g. Fe 2+ and Fe 3+ in an oxide, or as ferrous/ferric ions in solution. GGA+U formulations help to correct this failure in the solid state (see e.g. Cococcioni / de Gironcoli, PRB 2005); charge constraints can also be implemented successfully (see Sit, Cococcioni, Marzari, PRL 2006, and Van Voorhis PRA 72 024502 (2005)). Other references are given below. Best, nicola Van Voorhis J. Chem. Theo. Comp. 2 765 (2006); JPCA 110 9212 (2006). Akai, Blugel, Zeller and Dederichs PRL 1986 Hybertsen, Schluter and Christensen, PRB 1989 Zhang and Satpathy PRB 1991 Ujfalussy, Wang, Nicholson, Shelton, Stocks, Wang and Gyory J. App. Phys. 1999 Prezhdo, Kindt and Tully JCP 1999 Behler, Delley, Lorenz, Reuter and Scheffler PRL 2005, and Behler et al PRB 2007 --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From isouza at berkeley.edu Thu Nov 15 00:52:37 2007 From: isouza at berkeley.edu (Ivo Souza) Date: Wed, 14 Nov 2007 15:52:37 -0800 Subject: [Pw_forum] Sign of magnetization In-Reply-To: References: Message-ID: <473B8A45.1080700@berkeley.edu> Paolo, Thanks for looking into this. > > Since the exchange-correlation energy depends on the modulus of the > magnetization, the theory is invariant with respect to its sign. Of > course > this is no longer true when one introduces an external magnetic field > and one has to take care to use a consistent convention. I believe there are two related but somewhat different issues here: one, as you say, is to make sure that a consistent convention is used for the sign of the magnetization and that of the magnetic field. For example, if one does an scf calculation on a non-magnetic system and specifies in the input file that the applied magnetic field is along +z, then at the end of the calculation the induced magnetization given in the output file should also be along +z. I did this exercise and it is fine. The second issue is whether the absolute (as opposed to relative) signs of both magnetic field and magnetization are correctly reported in the output file. One way to check that is to evaluate a third quantity which flips sign upon magnetization reversal, such as the anomalous Hall conductivity. As Ralph mentioned, changing signs all over the code can be tricky, but it is probably worthwhile to clarify the sign convention in the output file. Ralph suggested writing in the output "spin polarization" instead of "magnetic moment". My preference would be to report the magnetic moment (a more "physical" quantity), and reverse the sign of the actual number. This has the additional advantage of making it easier to add the orbital magnetization contribution, when its evaluation becomes feasible. Regards, Ivo -- --------------------------------------------------- Ivo Souza Department of Physics, Univ. of California Berkeley tel. +1.510.642.3597 fax +1.510.643.8497 isouza at berkeley.edu http://smilodon.berkeley.edu/ --------------------------------------------------- From lyanglyan at yahoo.com.cn Thu Nov 15 03:26:30 2007 From: lyanglyan at yahoo.com.cn (=?gb2312?q?=D1=D2=20=C0=EE?=) Date: Thu, 15 Nov 2007 10:26:30 +0800 (CST) Subject: [Pw_forum] Why the calculated equilibrium lattice parameters Message-ID: <911492.96793.qm@web15803.mail.cnb.yahoo.com> Thanks to all of you for the helpful advices. The Cs pseudopotential on pwscf.org considers only one electron. But more electrons should be considered as valence electrons if I want study the characters of Cs under very high pressure. Therefore, I considered the configurations 5s2 5p6 6s1 for Cs. And then I test the validity of the artificial pseudopotential by simulating the equation of state under low pressure. But the calculated results are not in good agreement with the experiment. Anyhow, thank your very much. I will try the other calculation parameters to test the validity of the Cs pseudopotential. Best regards, Yan Li --------------------------------- @yahoo.cn ????????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071115/b21ff07b/attachment.htm From hqzhou at nju.edu.cn Thu Nov 15 03:43:38 2007 From: hqzhou at nju.edu.cn (Huiqun Zhou) Date: Thu, 15 Nov 2007 10:43:38 +0800 Subject: [Pw_forum] Convergence test References: <010801c825fa$5fab6960$1d00a8c0@solarflare> Message-ID: <002701c82731$56729610$1d00a8c0@solarflare> Axel, Your advices are valuable, thanks a lot!. Seeing most of my calculations (geometry optimization) are still effective, I can set my heart at rest :-). Dr. Huiqun Zhou @Nanjing University ----- Original Message ----- From: "Axel Kohlmeyer" To: "Huiqun Zhou" Cc: Sent: Wednesday, November 14, 2007 1:31 PM Subject: Re: [Pw_forum] Convergence test > On Tue, 13 Nov 2007, Huiqun Zhou wrote: > > HZ> List users, > > dear huiqun zhou, > > HZ> I'm testing convergence on ecutwfc for chromite using USPP. Below are > HZ> results of different ecutrho with different ecutwfc. I may be wrong, I > think > HZ> I should watch the difference of total energy between successive runs > till > HZ> it's less than 1-2 mRy. But the trend is not apparent in the results > below > HZ> and it seems that the larger the ecutwfc, the lower the total energy. > > you are using an ultra-soft pseudopotential and for those > actually the density cutoff is _much_ more important than > the wavefunction cutoff. depending on the choice of the > pseudopotential parameters, 35-40ry for the wavefunction > cutoff should be a conservative choice, however a density > cutoff of 250 is not. if you want to converge the energy > very tightly, you may need to go up to 10x the wavefunction > cutoff. you should also keep in mind that the total energy > converges rather slowly, and that for most calculations > you need energy differences or derivatives, which may converge > faster, i.e. you may want to look at forces (use a non-symmetric > input for that!!) or stress instead. what is sufficiently > converged depends on your the kind of calculation you want > to do. so for simple geometry optimization or MD 35/210 > or less may already be sufficient. > > HZ> > HZ> I calculated E-V relations using two sets of ecutwfc and ecutrho: > HZ> for ecutwfc = 35 ry and ecutrho = 240 ry, a = 8.405, B0 = 172.8, B0' = > 4.3 > HZ> for ecutwfc = 55 ry and ecutrho = 250 ry, a = 8.407, B0 = 178.6, B0' = > 3.6 > HZ> > HZ> There are no big differences in unit cell parameter and elastic > properties. > HZ> So, my questions are > HZ> (1) usually, the ecutwfc of USPP should be around 35 ry or below, is > my > HZ> case an exception? > > no. it does not look like an exception. > > HZ> (2) how small the difference of total energy of successive runs should > be > HZ> in judging convergence against ecutwfc? > > total energy is not a good parameter. see above. > > cheers, > axel. > > HZ> > HZ> Thank you in advance. > HZ> > HZ> Dr. Huiqun Zhou > HZ> @Earth Sciences, Nanjing University, China > HZ> > HZ> > HZ> > HZ> > HZ> ecutrho = 240 ry > HZ> ecutwfc total energy > HZ> 35 -1461.54573299 Ry > HZ> 40 -1461.64206824 Ry > HZ> 45 -1461.71978835 Ry > HZ> 50 -1461.76587314 Ry > HZ> 55 -1461.76316597 Ry > HZ> 60 -1461.76692396 Ry > HZ> > HZ> ecutrho = 250 ry > HZ> 40 -1461.64351876 Ry > HZ> 45 -1461.70130510 Ry > HZ> 50 -1461.74724764 Ry > HZ> 55 -1461.76462943 Ry > HZ> 60 -1461.76838812 Ry > HZ> > HZ> ecutrho = 280 ry > HZ> 50 -1461.74778158 Ry > HZ> 55 -1461.76516515 Ry > HZ> 60 -1461.76892327 Ry > HZ> > HZ> _______________________________________________ > HZ> Pw_forum mailing list > HZ> Pw_forum at pwscf.org > HZ> http://www.democritos.it/mailman/listinfo/pw_forum > HZ> > > -- > ======================================================================= > Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu > Center for Molecular Modeling -- University of Pennsylvania > Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 > tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 > ======================================================================= > If you make something idiot-proof, the universe creates a better idiot. > From bopengchemist at gmail.com Thu Nov 15 04:15:58 2007 From: bopengchemist at gmail.com (Bo Peng) Date: Thu, 15 Nov 2007 11:15:58 +0800 Subject: [Pw_forum] How to do a Potential energy scan in PWSCF Message-ID: <3faf0730711141915u667facabn9b0531765aaac3a@mail.gmail.com> Hi all, I, now, want to plot a contour of potential energy for a hydrogen molecule dissociated on a metal surface. So the variables are the bond length of H2 and the distance between the middle point of H2 and the surface. Can anybody give me some suggestion on how to do it in PWSCF? Thank you in advance. Bo Peng ----------------------------------------- Bo Peng, NKU, CHN Department of Chemistry -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071115/f1b1c939/attachment-0001.htm From yukihiro_okuno at fujifilm.co.jp Thu Nov 15 05:34:33 2007 From: yukihiro_okuno at fujifilm.co.jp (yukihiro_okuno at fujifilm.co.jp) Date: Thu, 15 Nov 2007 13:34:33 +0900 Subject: [Pw_forum] =?iso-2022-jp?b?RXJyb3IgaW4gdmlydHVhbC54IBskQiEhGyhC?= =?iso-2022-jp?b?KG5vdCBpbXBsZW1lbnRlZCByaW5uZXIgb3IgbnFmKQ==?= Message-ID: Dear PWSCF users. Sorry I wrote e-mail with wrong title. So, I tried again e-mail with correct title. I'm now want to use virtual crystal approximation, and using virtual.x for making the virtual atom from the pseudopotentials of real atom. But frequently the errors occurs to with the message below, For example1, from Virtual : error # different rinner are not implemented (yet) stopping 2 or example2 from Virtual : error# different nqf are not implemented yet I think example2 may be the pseudo charge part of pseudo potential , but for example1 is unclear. When such errors occure, is it not available to use the pseudopotential of real atom and remake pseudo potential? (If possible I want to use the pseudopotential I have because it is reliable to use) Or is it available to correct part of virtual.f90 ( for example, in the case of different mesh interpotation) ? Sincerely. From sunshoutian168 at yahoo.com.cn Thu Nov 15 06:36:03 2007 From: sunshoutian168 at yahoo.com.cn (shoutian sun) Date: Thu, 15 Nov 2007 13:36:03 +0800 (CST) Subject: [Pw_forum] Why ph.x cost so much time. Message-ID: <732290.39875.qm@web92011.mail.cnb.yahoo.com> > I calculate phonons with ph.x. It cost me more than a week, some even two > months. > could anyone tell me that is mormal? > Thank you. --------------------------------- @yahoo.cn ????????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071115/2cae4a78/attachment.htm From akohlmey at cmm.chem.upenn.edu Thu Nov 15 07:48:47 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 15 Nov 2007 12:18:47 +0530 Subject: [Pw_forum] Error in virtual.x (not implemented rinner or nqf) In-Reply-To: References: Message-ID: <7b6913e90711142248p6a72b782tf3bebe6ceb3808cd@mail.gmail.com> On Nov 15, 2007 10:04 AM, wrote: > > Dear PWSCF users. dear yukihiro, > I'm now want to use virtual crystal approximation, and using virtual.x > for making the virtual atom from the pseudopotentials of real atom. > > But frequently the errors occurs to with the message below, > > example1: > from Virtual : error # > different rinner are not implemented (yet) > > or example2 > > from Virtual : error# > different nqf are not implemented yet if you read the recent discussions in the mailing list archive on using virtual.x, you should see, that for the virtual crystal approach it is important that the elements, that you want to mix are as similar as possible. as a consequence the pseudopotentials themselves have to be as similar as well. rinner is the "norm-conserving" cutoff of the pseudo potential and nqf is the number of Q-functions. you will most likely have to (re-)generate a matching pair of pseudopotentials first. this problem is not as simple to solve as the previously discussed non-matching grids. it would be very helpful to know what kinds of pseudopotentials you want to mix... cheers, axel. > When such errors occure, is it not available to use > the pseudopotential of real atom and remake pseudo potential? > (If possible I want to use the pseudopotential I have because > it is reliable to use) > > Or is it available to correct part of virtual.f90 > ( for example, in the case of different mesh interpotation) ? > > Sincerely. > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From marzari at MIT.EDU Thu Nov 15 08:39:34 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 15 Nov 2007 02:39:34 -0500 Subject: [Pw_forum] ion problem In-Reply-To: <473B72A0.7070001@mit.edu> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> <473B72A0.7070001@mit.edu> Message-ID: <473BF7B6.3080902@mit.edu> > Rudra Banerjee wrote: >> dear PWSCF users, >> I am new to pwscf and stuck in a question.......is there any way to >> put anion or cation as a unit?the total cluster is charge less >> but have two different types of ion.......can anyone tell me what to >> do in this situation? >> Regards Dear Rudra, several have answered this question. Let me add, though, that standard approximation to the exact exchange-correlation functional will often fail in "depositing" the right amount of charge on an ion. This becomes a very relevant failure if you have in the same unit cell ions of the same type, but of different valence, e.g. Fe 2+ and Fe 3+ in an oxide, or as ferrous/ferric ions in solution. GGA+U formulations help to correct this failure in the solid state (see e.g. Cococcioni / de Gironcoli, PRB 2005); charge constraints can also be implemented successfully (see Sit, Cococcioni, Marzari, PRL 2006, and Van Voorhis PRA 72 024502 (2005)). Other references are given below. Best, nicola Van Voorhis J. Chem. Theo. Comp. 2 765 (2006); JPCA 110 9212 (2006). Akai, Blugel, Zeller and Dederichs PRL 1986 Hybertsen, Schluter and Christensen, PRB 1989 Zhang and Satpathy PRB 1991 Ujfalussy, Wang, Nicholson, Shelton, Stocks, Wang and Gyory J. App. Phys. 1999 Prezhdo, Kindt and Tully JCP 1999 Behler, Delley, Lorenz, Reuter and Scheffler PRL 2005, and Behler et al PRB 2007 -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From cyrille.barreteau at cea.fr Thu Nov 15 08:47:24 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Thu, 15 Nov 2007 08:47:24 +0100 Subject: [Pw_forum] ion problem In-Reply-To: <473B7126.8070705@mit.edu> References: <360668.39837.qm@web94110.mail.in2.yahoo.com> <473A0F74.5000905@sissa.it> <473AA96F.30201@cea.fr> <473B7126.8070705@mit.edu> Message-ID: <473BF98C.5060504@cea.fr> Dear Nicola Thank you for this interesting reference which is exactly the subject of my "worries".. cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html ================================================================== Nicola Marzari wrote: >Dear Cyrille, > >if you are interested in this problem, you might want to >have a look at http://arxiv.org/abs/0709.4647 , and in >particular at Fig 5. > >The published Makov-Payne correction is actually incorrect >in the charge*quadrupole term. > > nicola > > >Cyrille Barreteau wrote: > > >>Dear all >> >>I have a question related to this problem. >>I am interested in the study of charged clusters. >>If I understand well, due to to the periodicity of the plane wave codes >>one has to substract the correct "background" in order to keep the >>energy from going to infinity.. >> >>I would like to compare the energy of clusters with different >>charges -e -2e etc... Do you have some advices to do such calculation? >>Is it necessary to take even larger supercells than for neutral clusters? >> >>thanks in advance >> >> cyrille >> >> >> > > > > From eyvaz_isaev at yahoo.com Thu Nov 15 09:27:14 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 15 Nov 2007 00:27:14 -0800 (PST) Subject: [Pw_forum] Why ph.x cost so much time. In-Reply-To: <732290.39875.qm@web92011.mail.cnb.yahoo.com> Message-ID: <628046.15431.qm@web60313.mail.yahoo.com> Hi, --- shoutian sun wrote: > I calculate phonons with ph.x. It cost me more > than a week, some even two months. > could anyone tell me that is mormal? Please do not forget provide your affiliation. Well, It may depend on: cutoff energy k-points symmetry tolerance number of atoms (modes) CPU's number network problem etc., etc., If you mean the same calculation takes considerably different time (on the same computer system) for the same q-point, I guess, it is a problem due to your system. If you mean time elapsed is different for different q-points that is normal. Bests, Eyvaz. > > Thank you. > > > > > --------------------------------- > @yahoo.cn ??????????????????????????> _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs From giannozz at nest.sns.it Thu Nov 15 11:46:33 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 15 Nov 2007 11:46:33 +0100 Subject: [Pw_forum] Why the calculated equilibrium lattice parameters In-Reply-To: <911492.96793.qm@web15803.mail.cnb.yahoo.com> References: <911492.96793.qm@web15803.mail.cnb.yahoo.com> Message-ID: <2D416457-8A84-4BD6-BE73-387B27CD1785@nest.sns.it> On Nov 15, 2007, at 3:26 , ? ? wrote: > The Cs pseudopotential on pwscf.org considers only one electron. > But more electrons should be considered as valence electrons if > I want study the characters of Cs under very high pressure. > Therefore, I considered the configurations 5s2 5p6 6s1 for Cs. my first first-principle calculation was about phase transitions in CsI under pressure (PRB 35, 765 (1987)) : see the "Notes on Pseudopotential Generation" in atomic_doc/pseudo-gen.pdf . We used a 9-electron PP (in semilocal form, so unsuitable for today's usage). Most likely, if you get bad results, your pseudopotential has a ghost, as explained in the above mentioned "Notes", or you haven't followed the correct procedure in generating it. I have four 9-e PP's on my disk, generated in 2005. They are PBE and PZ (Perdew-Zunger LDA), 5s2 5p6 6s0 and 5s2 5p6 6s1, L=0 is the local part for all of them. I don't remember why they are there and whether they work or not. They come with some atomic tests, plus tests for metallic Cs that somebody did somewhere, whose results I ignore. Here they are: http://www.fisica.uniud.it/~giannozz/public/Cs.tar.gz No warranty. If some of them seem to work, please report (nobody does it, but I keep asking) Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From okunoyukihiro2000 at yahoo.co.jp Thu Nov 15 15:24:36 2007 From: okunoyukihiro2000 at yahoo.co.jp (Yukihiro Okuno) Date: Thu, 15 Nov 2007 23:24:36 +0900 Subject: [Pw_forum] Error in virtual.x (not implemented rinner or nqf) Message-ID: <200711151424.AA00386@Einstein.yahoo.co.jp> Dear PWSCF users and Dr.Kohlmeyer. >if you read the recent discussions in the mailing >list archive on using virtual.x, you should see, >that for the virtual crystal approach it is important >that the elements, that you want to mix are as >similar as possible. as a consequence the pseudopotentials >themselves have to be as similar as well. > >rinner is the "norm-conserving" cutoff of the pseudo >potential and nqf is the number of Q-functions. >you will most likely have to (re-)generate a matching >pair of pseudopotentials first. this problem is not >as simple to solve as the previously discussed >non-matching grids. > >it would be very helpful to know what kinds of >pseudopotentials you want to mix... > Thank you for reply. I want to use virtual atom for example Ti and Zr which is well used in PZT, and I want to use virtual atom among the 3d- trasitional metal, for example, Fe and Co , or Fe and Mn. I think Fe and Co, for example, their pseudo potential is very similar because it is neighebor of atomic tables. So, if we apply the virtual.x and get the message below, > from Virtual :error# > different rinner are not implemented (yet) > > or example2 > > from Virtual : error# > different nqf are not implemented yet , is it needed to make the pseudo potential again ? Sincerely. ---- Yukihiro Okuno okunoyukihiro2000 at yahoo.co.jp From mgkrishn at ncsu.edu Thu Nov 15 16:40:58 2007 From: mgkrishn at ncsu.edu (mgkrishn at ncsu.edu) Date: Thu, 15 Nov 2007 10:40:58 -0500 (EST) Subject: [Pw_forum] pseudopotential In-Reply-To: <9EC09245-46D8-402F-93BC-68C24B2E470A@sissa.it> References: <980542.73510.qm@web52301.mail.re2.yahoo.com> <9EC09245-46D8-402F-93BC-68C24B2E470A@sissa.it> Message-ID: <1926.152.1.142.204.1195141258.squirrel@webmail.ncsu.edu> Hi All I am a graduate student in Buongiorno-Nardelli, Marco group at NCSU in Raleigh NC and urgently need to generate Relativistic pseudo potentials with spin-orbit angular momentum coupling for Bi, Te, Sb and Se. Any help on how to proceed will be greatly appreciated Thanks Krishnan M G > the same way would for any other element ;-) > S. > > On Nov 14, 2007, at 6:01 PM, ?lvaro Alves wrote: > >> Hi. How to generate a pseudopotencial LDA for the fluorine.. >> >> ?lvaro >> >> >> >> >> >> Abra sua conta no Yahoo! Mail, o ?nico sem limite de espa?o para >> armazenamento! >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum > > --- > Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - > Trieste > [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) > > Please, if possible, don't send me MS Word or PowerPoint attachments > Why? See: http://www.gnu.org/philosophy/no-word-attachments.html > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From giannozz at nest.sns.it Thu Nov 15 17:18:48 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 15 Nov 2007 17:18:48 +0100 Subject: [Pw_forum] pseudopotential In-Reply-To: <1926.152.1.142.204.1195141258.squirrel@webmail.ncsu.edu> References: <980542.73510.qm@web52301.mail.re2.yahoo.com> <9EC09245-46D8-402F-93BC-68C24B2E470A@sissa.it> <1926.152.1.142.204.1195141258.squirrel@webmail.ncsu.edu> Message-ID: On Nov 15, 2007, at 16:40 , mgkrishn at ncsu.edu wrote: > I [...] urgently need to generate Relativistic pseudo potentials > with spin-orbit > angular momentum coupling for Bi, Te, Sb and Se. Any help on how to > proceed will be greatly appreciated did you have a look to atomic_doc/ ? --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Thu Nov 15 17:26:55 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 15 Nov 2007 17:26:55 +0100 Subject: [Pw_forum] Error in virtual.x (not implemented rinner or nqf) In-Reply-To: <200711151424.AA00386@Einstein.yahoo.co.jp> References: <200711151424.AA00386@Einstein.yahoo.co.jp> Message-ID: On Nov 15, 2007, at 15:24 , Yukihiro Okuno wrote: > So, if we apply the virtual.x and get the message below, > >> from Virtual :error# >> different rinner are not implemented (yet) >> >> or example2 >> >> from Virtual : error# >> different nqf are not implemented yet > > , is it needed to make the pseudo potential again ? in principle, it is not: you could generate the Q functions on the entire grid, interpolate them, re-fit them with the original form, but I am not sure it is worth the effort Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From saxteam at unimore.it Fri Nov 16 13:41:11 2007 From: saxteam at unimore.it (SaX developers Team) Date: Fri, 16 Nov 2007 13:41:11 +0100 (CET) Subject: [Pw_forum] SaX optical suite, release 1.0.0 GNU-GPL available Message-ID: <36708.155.185.13.180.1195216871.squirrel@webmail2.unimore.it> The SaX developers Team and the INFM-CNR S3 in Modena, Italy, is proud to announce that the SaX optical suite, release 1.0.0 is available at www.sax-project.org. SaX is an Open source GNU-GPL package for the calculation of electronic excitations, developed at the National Research Center on nanoStructures and bioSystems at Surfaces (S3) of the Italian INFM-CNR. SaX works in the framework of Many Body Perturbation Theory within the GW approximation. Sax calculations will provide the user with: - Full dynamical structure of the Random-Phase-Approximation Polarizability - Quasiparticle energies and wave functions - Different implementations of the GW scheme - Different degree of self-consistency in the GW scheme - Excitonic effects within the Bethe-Salpeter Equation SaX works, in principle, as postprocessing of a previous DFT calculation. In the present version, the package is interfaced with DFT codes of the Quantum-Espresso distribution (www.quantumespresso.org). SaX developers Team -- SaX developers Team saxteam at unimore.it INFM-CNR S3 via Campi 213/a 41100 Modena Italy From ierrea001 at ikasle.ehu.es Fri Nov 16 16:53:05 2007 From: ierrea001 at ikasle.ehu.es (Ion Errea) Date: Fri, 16 Nov 2007 16:53:05 +0100 Subject: [Pw_forum] matdyn.modes Message-ID: <1195228385.6791.22.camel@localhost> Dear all, I'm calculating phonons in different points using ph.x q2r.x and matdyn.x. Basically my problem is that when I get the results and want to see the normal modes, in the file matyn.modes I do not understand the meaning of all rows. There are 6 different rows for each phonon. That is for one phonon (there are 6 atoms per unit cell) at a given K point I get q = 0.5333 0.0667 0.0000 ************************************************************************** omega( 1) = 5.975199 [THz] = 199.312497 [cm-1] 0.000000 0.000000 0.000000 0.000000 0.468260 -0.006324 0.000000 0.000000 0.000000 0.000000 -0.270615 0.382197 0.000977 -0.014434 -0.003796 0.056105 -0.089053 -0.006026 -0.000977 0.014434 0.003796 -0.056105 -0.089053 -0.006026 0.000000 0.000000 0.000000 0.000000 -0.294161 0.427580 0.000000 0.000000 0.000000 0.000000 0.518994 0.000000 The rows 1,3,5 I understand that give the displacement of the atoms. However I don't know what the even ones mean. Normally I get 0.0000000 for these, but not in my lst calculations in agiven system. I wonder if anyone can help me to undesrtand the meaning of these even rows and let me know wheter this non-zero result is worrying. Thanks a lot for your attention. ==================================================================== Ion Errea Lope Department of Condensed Matter, Unibersity of the Basque Country Apdo. 644 48080 Bilbao (Spain) Tlf: +34 94 601 5322 From giannozz at nest.sns.it Fri Nov 16 17:18:35 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 16 Nov 2007 17:18:35 +0100 Subject: [Pw_forum] matdyn.modes In-Reply-To: <1195228385.6791.22.camel@localhost> References: <1195228385.6791.22.camel@localhost> Message-ID: <0E465B16-8AD0-4509-8595-9B8E19255119@nest.sns.it> On Nov 16, 2007, at 16:53 , Ion Errea wrote: > There are 6 different rows for each phonon. real and imaginary parts, x y z components, per atom P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From baroni at sissa.it Fri Nov 16 19:00:54 2007 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 16 Nov 2007 19:00:54 +0100 Subject: [Pw_forum] Why ph.x cost so much time. In-Reply-To: <732290.39875.qm@web92011.mail.cnb.yahoo.com> References: <732290.39875.qm@web92011.mail.cnb.yahoo.com> Message-ID: <6E2CF233-821F-4CA4-A107-5460EFCCBDCD@sissa.it> Dear Mr. Unknown, it depends on the system you are treating. Each individual phonon mode costs roughly as much as a single ground-state calculation. So, the total cost scale as (number of atoms) * (cost of a ground-state calc), i.e. of the orther of the fourth power of the numbr of atoms ... Also, please do not forget to sign your posts - that practice is very much appreciated by this community. Cheers - Stefano Baroni On Nov 15, 2007, at 6:36 AM, shoutian sun wrote: > > I calculate phonons with ph.x. It cost me more than a week, some > even two > > months. > > could anyone tell me that is mormal? > > Thank you. > > > > @yahoo.cn ????????????? > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071116/bc1608bd/attachment.htm From baroni at sissa.it Fri Nov 16 19:07:26 2007 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 16 Nov 2007 19:07:26 +0100 Subject: [Pw_forum] pseudopotential In-Reply-To: <1926.152.1.142.204.1195141258.squirrel@webmail.ncsu.edu> References: <980542.73510.qm@web52301.mail.re2.yahoo.com> <9EC09245-46D8-402F-93BC-68C24B2E470A@sissa.it> <1926.152.1.142.204.1195141258.squirrel@webmail.ncsu.edu> Message-ID: <5FDA57D0-6D1A-41DE-B928-8419112DF0BE@sissa.it> On Nov 15, 2007, at 4:40 PM, mgkrishn at ncsu.edu wrote: > > Hi All > > I am a graduate student in Buongiorno-Nardelli, Marco group at NCSU in > Raleigh NC and urgently need .... Hi All, I am the former supervisor of Marco, and I do not remember to have ever asked him to do anything "urgently". I will call him right away and ask him how does he manage to be such an effective supervisor. One never stops to learn ;-) Just kidding ... Have a nice day and say hello to Marco from me. Stefano Baroni > to generate Relativistic pseudo potentials > with spin-orbit angular momentum coupling for Bi, Te, Sb and Se. > Any help > on how to proceed will be greatly appreciated > > Thanks > > Krishnan M G > > >> the same way would for any other element ;-) >> S. >> >> On Nov 14, 2007, at 6:01 PM, ?lvaro Alves wrote: >> >>> Hi. How to generate a pseudopotencial LDA for the fluorine.. >>> >>> ?lvaro >>> >>> >>> >>> >>> >>> Abra sua conta no Yahoo! Mail, o ?nico sem limite de espa?o para >>> armazenamento! >>> _______________________________________________ >>> Pw_forum mailing list >>> Pw_forum at pwscf.org >>> http://www.democritos.it/mailman/listinfo/pw_forum >> >> --- >> Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - >> Trieste >> [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) >> >> Please, if possible, don't send me MS Word or PowerPoint attachments >> Why? See: http://www.gnu.org/philosophy/no-word-attachments.html >> >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071116/9bdcd368/attachment-0001.htm From baroni at sissa.it Fri Nov 16 19:09:01 2007 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 16 Nov 2007 19:09:01 +0100 Subject: [Pw_forum] SaX optical suite, release 1.0.0 GNU-GPL available In-Reply-To: <36708.155.185.13.180.1195216871.squirrel@webmail2.unimore.it> References: <36708.155.185.13.180.1195216871.squirrel@webmail2.unimore.it> Message-ID: Welcome on bord SaX guys (and dolls!) Stefano On Nov 16, 2007, at 1:41 PM, SaX developers Team wrote: > The SaX developers Team and the INFM-CNR S3 in Modena, Italy, is > proud to > announce that the SaX optical suite, release 1.0.0 is available at > www.sax-project.org. > > SaX is an Open source GNU-GPL package for the calculation of > electronic > excitations, developed at the National Research Center on > nanoStructures > and bioSystems at Surfaces (S3) of the Italian INFM-CNR. > > SaX works in the framework of Many Body Perturbation Theory within > the GW > approximation. Sax calculations will provide the user with: > > - Full dynamical structure of the Random-Phase-Approximation > Polarizability > - Quasiparticle energies and wave functions > - Different implementations of the GW scheme > - Different degree of self-consistency in the GW scheme > - Excitonic effects within the Bethe-Salpeter Equation > > SaX works, in principle, as postprocessing of a previous DFT > calculation. > In the present version, the package is interfaced with DFT codes of > the > Quantum-Espresso distribution (www.quantumespresso.org). > > SaX developers Team > > -- > SaX developers Team > saxteam at unimore.it > INFM-CNR S3 > via Campi 213/a > 41100 Modena > Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071116/4eef7665/attachment.htm From baroni at sissa.it Sat Nov 17 09:28:42 2007 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 17 Nov 2007 09:28:42 +0100 Subject: [Pw_forum] Why the calculated equilibrium lattice parameters In-Reply-To: <2D416457-8A84-4BD6-BE73-387B27CD1785@nest.sns.it> References: <911492.96793.qm@web15803.mail.cnb.yahoo.com> <2D416457-8A84-4BD6-BE73-387B27CD1785@nest.sns.it> Message-ID: <07EFCE39-BC6A-4A01-868E-0AC17CDD8A80@sissa.it> Hi! I understand from Paolo that the PP that he refers to is not the one originally used in our paper. It may be boring to copy by hands, but the form of the PP used therein is reported in the text of the paper. Cheers - Stefano Baroni On Nov 15, 2007, at 11:46 AM, Paolo Giannozzi wrote: > > On Nov 15, 2007, at 3:26 , ? ? wrote: > >> The Cs pseudopotential on pwscf.org considers only one electron. >> But more electrons should be considered as valence electrons if >> I want study the characters of Cs under very high pressure. >> Therefore, I considered the configurations 5s2 5p6 6s1 for Cs. > > my first first-principle calculation was about phase transitions in > CsI under pressure (PRB 35, 765 (1987)) : see the "Notes on > Pseudopotential Generation" in atomic_doc/pseudo-gen.pdf . > We used a 9-electron PP (in semilocal form, so unsuitable > for today's usage). Most likely, if you get bad results, your > pseudopotential has a ghost, as explained in the above > mentioned "Notes", or you haven't followed the correct > procedure in generating it. > > I have four 9-e PP's on my disk, generated in 2005. > They are PBE and PZ (Perdew-Zunger LDA), 5s2 5p6 6s0 > and 5s2 5p6 6s1, L=0 is the local part for all of them. > I don't remember why they are there and whether they > work or not. They come with some atomic tests, plus > tests for metallic Cs that somebody did somewhere, > whose results I ignore. Here they are: > http://www.fisica.uniud.it/~giannozz/public/Cs.tar.gz > No warranty. If some of them seem to work, please > report (nobody does it, but I keep asking) > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071117/f6c3013a/attachment.htm From ezadshojaee at hotmail.com Sun Nov 18 13:45:15 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Sun, 18 Nov 2007 12:45:15 +0000 Subject: [Pw_forum] Gamma phonons in Si lattice Message-ID: Hi if we define the Si lattice as an FCC with 2-atoms per basis, we have the Gamma phonons as obtained in the example#2. but it seems that if we define the lattice as a SC with 8-atoms per basis, we have a lot of frequencies which seems to have nothing in common with the first one. if there is no problem, what physics is behind it? did i miss something? and also the code seems to have problem with converging some of the modes in the 8-atoms per basis one(there are some negative freq's) Thanx for any comment _________________________________________________________________ News, entertainment and everything you care about at Live.com. Get it now! http://www.live.com/getstarted.aspx -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071118/6b7edf20/attachment.htm From oulihui666 at 126.com Sun Nov 18 14:36:38 2007 From: oulihui666 at 126.com (oulihui666) Date: Sun, 18 Nov 2007 21:36:38 +0800 (CST) Subject: [Pw_forum] A problem about electrons Message-ID: <26840889.885151195392998773.JavaMail.coremail@bj126app104.126.com> Dear pwscf users, Recently, I want to ask a problem about electrons, how to assign an excess or deficit number of electrons ne to the metal and a corresponding electronic charge q=?ne to the system with respect to neutrality in the process of calculation with pwscf. Thanks in advance Best wishes Lihui Ou -- ====================================== Lihui Ou PH.D Candidate in Electrochemistry College of Chemistry and Molecular Science Wuhan University,430072,Hubei Province,China E-mail:oulihui666 at 126.com ====================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071118/5fd606e5/attachment.htm From paulatto at sissa.it Sun Nov 18 15:07:17 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Sun, 18 Nov 2007 15:07:17 +0100 (CET) Subject: [Pw_forum] A problem about electrons In-Reply-To: <26840889.885151195392998773.JavaMail.coremail@bj126app104.126.com> References: <26840889.885151195392998773.JavaMail.coremail@bj126app104.126.com> Message-ID: <15518.82.54.232.48.1195394837.squirrel@webmail.sissa.it> On Sun, November 18, 2007 14:36, oulihui666 wrote: > Dear pwscf users, > Recently, I want to ask a problem about electrons, how to assign an > excess or deficit > number of electrons ne to the metal and a corresponding electronic charge > q=???ne to the system with respect to neutrality in the process of > calculation with pwscf. Dear Lihui Ou, a similar question has been asked a few days ago. If you had looked in Doc/INPUT_PW you would have found: nelec REAL number of electron in the unit cell (may be noninteger if you wish) Default: the same as ionic charge (neutral cell) A compensating jellium background is inserted to remove divergences if the cell is not neutral Goodbye, LP -- Lorenzo Paulatto +39 040 3787 511 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From degironc at sissa.it Sun Nov 18 17:18:17 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Sun, 18 Nov 2007 17:18:17 +0100 Subject: [Pw_forum] Gamma phonons in Si lattice In-Reply-To: References: Message-ID: <474065C9.4070903@sissa.it> The SC contains 8 atoms since it is four time larger than the 2-atom fcc cell. Therefore you MUST have 24 frequency in each point in the supercell BZ... What they are ? where they come from ? In order to answer this question take into consideration the fact that the supercell BZ is four time smaller that the original one because some q-vectors that would be inside the BZ of the 2-atom cell have become G-vectors (i.e. equivalent to Gamma point!) in the supercell. Check carefully your k-point sampling and other convergence parameters in your calculation... there should not be any imaginary frequencies in Si at equilibrium lattice constant. Don't forget to indicate your affiliation. Stefano de Gironcoli - SISSA and DEMOCRITOS Ezad Shojaee wrote: > Hi > if we define the Si lattice as an FCC with 2-atoms per basis, we have > the Gamma phonons as obtained in the example#2. but it seems that if > we define the lattice as a SC with 8-atoms per basis, we have a lot of > frequencies which seems to have nothing in common with the first one. > if there is no problem, what physics is behind it? did i miss > something? and also the code seems to have problem with converging > some of the modes in the 8-atoms per basis one(there are some negative > freq's) > Thanx for any comment > > > ------------------------------------------------------------------------ > Get news, entertainment and everything you care about at Live.com. > Check it out! > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From sks.jnc at gmail.com Sun Nov 18 17:46:02 2007 From: sks.jnc at gmail.com (S. K. S.) Date: Sun, 18 Nov 2007 22:16:02 +0530 Subject: [Pw_forum] Doc/INPUT_PW: A compensating jellium background is inserted Message-ID: Dear all, It's written in the documentation that a compensating jellium background is inserted. I'll be highly obliged if someone kindly tell me the name of the subroutine where it has been inserted. Yours sincerely, S. K. S R&D Assistant. Jncasr India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071118/6897937c/attachment.htm From paulatto at sissa.it Sun Nov 18 17:59:42 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Sun, 18 Nov 2007 17:59:42 +0100 (CET) Subject: [Pw_forum] Doc/INPUT_PW: A compensating jellium background is inserted In-Reply-To: References: Message-ID: <21693.82.54.232.48.1195405182.squirrel@webmail.sissa.it> On Sun, November 18, 2007 17:46, S. K. S. wrote: > Dear all, > > It's written in the documentation that > a compensating jellium background is inserted. > > I'll be highly obliged if someone kindly tell me the name > of the subroutine where it has been inserted. > Actually the jellium is not really added: the operation is equivalent to throwing away the G=0 component of the hartree potential. You can see how it works in subroutine v_h, inside PW/v_of_rho.f90 keep in mind that gstart is the index of the first non-zero G vector. Goodbye -- Lorenzo Paulatto +39 040 3787 511 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From ezadshojaee at hotmail.com Sun Nov 18 20:35:54 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Sun, 18 Nov 2007 19:35:54 +0000 Subject: [Pw_forum] Gamma phonons in Si lattice Message-ID: we have obtained exactly 4 times grater in energy and 4 times smaller in cell volume compared to the original FCC one. when the structure seems to be right in the energy calculation, what may cause the problem? and beyond the converging prob, what do you expect of the frequencies to be, compared to the FCC one? like for example you expect the total energy to be 4 times grater? should there be any relation between the freq's in the SC and FCC one? which physical quantity should be the same? Thanx Ezad Shojaee - Tehran University _________________________________________________________________ Connect to the next generation of MSN Messenger? http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&source=wlmailtagline -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071118/d69ab82b/attachment.htm From degironc at sissa.it Sun Nov 18 20:58:14 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Sun, 18 Nov 2007 20:58:14 +0100 Subject: [Pw_forum] Gamma phonons in Si lattice In-Reply-To: References: Message-ID: <47409956.7040608@sissa.it> Ezad Shojaee wrote: > we have obtained exactly 4 times grater in energy and 4 times smaller > in cell volume compared to the original FCC one. when the structure > seems to be right in the energy calculation, what may cause the problem? > and beyond the converging prob, what do you expect of the frequencies > to be, compared to the FCC one? like for example you expect the total > energy to be 4 times grater? should there be any relation between the > freq's in the SC and FCC one? which physical quantity should be the same? what are the q-points in the 2-atom BZ that map into the Gamma-point of the supercell ? Stefano de Gironcoli - SISSA and DEMOCRITOS > Thanx > Ezad Shojaee - Tehran University > > ------------------------------------------------------------------------ > Connect to the next generation of MSN Messenger Get it now! > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From njuxuyuehua at gmail.com Mon Nov 19 08:36:19 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Mon, 19 Nov 2007 15:36:19 +0800 Subject: [Pw_forum] about the influence of lo-to split on the vibrational model Message-ID: hi everyone : i am now computing ir-intensity.and in the input file, i add lo-to split .and the result is that some vibration frequency is changed corresponding to the *.dynG. and my question is once adding lo-to split,does the vibrational model including vibrational vector changed ?if it does ,how to get the changed vector ,because i need to plot the graph about the mode . maybe some of you could give me a piece of advise. i need your help thank you -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071119/a055efbd/attachment.htm From lyanglyan at yahoo.com.cn Mon Nov 19 08:45:30 2007 From: lyanglyan at yahoo.com.cn (=?gb2312?q?=D1=D2=20=C0=EE?=) Date: Mon, 19 Nov 2007 15:45:30 +0800 (CST) Subject: [Pw_forum] =?gb2312?b?UmWjuiBXaHkgdGhlIGNhbGN1bGF0ZWQgZXF1aWxp?= =?gb2312?b?YnJpdW0gbGF0dGljZSBwYXJhbWV0ZXJz?= In-Reply-To: Message-ID: <682668.90811.qm@web15806.mail.cnb.yahoo.com> Thanks to all of you for the helpful suggestions. I will test the feasibility of the Cs pp and report the results. Yours sincerely, Yan Li PhD National Lab of Superhard Materials, Jilin University, Changchun 130012, P. R. China. Email: liyanpcl at yahoo.com.cn pw_forum-request at pwscf.org ??? Send Pw_forum mailing list submissions to pw_forum at pwscf.org To subscribe or unsubscribe via the World Wide Web, visit http://www.democritos.it/mailman/listinfo/pw_forum or, via email, send a message with subject or body 'help' to pw_forum-request at pwscf.org You can reach the person managing the list at pw_forum-owner at pwscf.org When replying, please edit your Subject line so it is more specific than "Re: Contents of Pw_forum digest..." Today's Topics: 1. Re: Why the calculated equilibrium lattice parameters (Stefano Baroni) ---------------------------------------------------------------------- Message: 1 Date: Sat, 17 Nov 2007 09:28:42 +0100 From: Stefano Baroni Subject: Re: [Pw_forum] Why the calculated equilibrium lattice parameters To: PWSCF Forum Message-ID: <07EFCE39-BC6A-4A01-868E-0AC17CDD8A80 at sissa.it> Content-Type: text/plain; charset="utf-8" Hi! I understand from Paolo that the PP that he refers to is not the one originally used in our paper. It may be boring to copy by hands, but the form of the PP used therein is reported in the text of the paper. Cheers - Stefano Baroni On Nov 15, 2007, at 11:46 AM, Paolo Giannozzi wrote: > > On Nov 15, 2007, at 3:26 , ? ? wrote: > >> The Cs pseudopotential on pwscf.org considers only one electron. >> But more electrons should be considered as valence electrons if >> I want study the characters of Cs under very high pressure. >> Therefore, I considered the configurations 5s2 5p6 6s1 for Cs. > > my first first-principle calculation was about phase transitions in > CsI under pressure (PRB 35, 765 (1987)) : see the "Notes on > Pseudopotential Generation" in atomic_doc/pseudo-gen.pdf . > We used a 9-electron PP (in semilocal form, so unsuitable > for today's usage). Most likely, if you get bad results, your > pseudopotential has a ghost, as explained in the above > mentioned "Notes", or you haven't followed the correct > procedure in generating it. > > I have four 9-e PP's on my disk, generated in 2005. > They are PBE and PZ (Perdew-Zunger LDA), 5s2 5p6 6s0 > and 5s2 5p6 6s1, L=0 is the local part for all of them. > I don't remember why they are there and whether they > work or not. They come with some atomic tests, plus > tests for metallic Cs that somebody did somewhere, > whose results I ignore. Here they are: > http://www.fisica.uniud.it/~giannozz/public/Cs.tar.gz > No warranty. If some of them seem to work, please > report (nobody does it, but I keep asking) > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071117/f6c3013a/attachment.html ------------------------------ _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum End of Pw_forum Digest, Vol 5, Issue 28 *************************************** --------------------------------- ?????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071119/b989f876/attachment.htm From ierrea001 at ikasle.ehu.es Mon Nov 19 12:31:03 2007 From: ierrea001 at ikasle.ehu.es (Ion Errea) Date: Mon, 19 Nov 2007 12:31:03 +0100 Subject: [Pw_forum] Imaginary displacements Message-ID: <1195471863.31296.31.camel@localhost> Dear all, Does anyonone which are the reasons why you can get imaginary displacements for phonon eigenmodes? As far as I remember from algebra, the dynamical matrix is real and symmetric, thus hermitian, and hermitian matrices have real eigenvectors and eigenvalues. Therefore I don understand why I get this results. Problems with convergence maybe? Thanks for your answers, Ion ==================================================================== Ion Errea Lope Department of Condensed Matter, Unibersity of the Basque Country Apdo. 644 48080 Bilbao (Spain) Tlf: +34 94 601 5322 From degironc at sissa.it Mon Nov 19 12:46:23 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Mon, 19 Nov 2007 12:46:23 +0100 Subject: [Pw_forum] Imaginary displacements In-Reply-To: <1195471863.31296.31.camel@localhost> References: <1195471863.31296.31.camel@localhost> Message-ID: <4741778F.80508@sissa.it> check your algebra text-book. An hermitian matrix has real eigenvalues and in general complex eigenvectors. If you are surprised think of simple example where you KNOW the answer... For instance free electrons: the hamiltonian is real, the eigenvalues are real... the eigenvectors are plane-waves (i.e. complex funcions) Stefano de Gironcoli Ion Errea wrote: > Dear all, > > Does anyonone which are the reasons why you can get imaginary > displacements for phonon eigenmodes? As far as I remember from algebra, > the dynamical matrix is real and symmetric, thus hermitian, and > hermitian matrices have real eigenvectors and eigenvalues. Therefore I > don understand why I get this results. Problems with convergence maybe? > > Thanks for your answers, > > Ion > > ==================================================================== > Ion Errea Lope > Department of Condensed Matter, Unibersity of the Basque Country > Apdo. 644 > 48080 Bilbao (Spain) > Tlf: +34 94 601 5322 > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From giannozz at nest.sns.it Mon Nov 19 13:03:06 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 19 Nov 2007 13:03:06 +0100 Subject: [Pw_forum] Imaginary displacements In-Reply-To: <1195471863.31296.31.camel@localhost> References: <1195471863.31296.31.camel@localhost> Message-ID: <534DC6E4-9164-4948-9DB0-5E055A3E0E0F@nest.sns.it> On Nov 19, 2007, at 12:31 , Ion Errea wrote: > Does anyonone which are the reasons why you can get imaginary > displacements for phonon eigenmodes? at q=0 both the dynamical matrix and its eigenvectors are real, but for a general q, the dynamical matrix is complex hermitian and its eigenvectors are in general complex Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Mon Nov 19 13:17:12 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 19 Nov 2007 13:17:12 +0100 Subject: [Pw_forum] about the influence of lo-to split on the vibrational model In-Reply-To: References: Message-ID: On Nov 19, 2007, at 8:36 , xu yuehua wrote: > i am now computing ir-intensity.and in the input file, i add lo- > to split . > and the result is that some vibration frequency is changed > corresponding > to the *.dynG. and my question is once adding lo-to split, does the > vibrational > model including vibrational vector changed ? if it does, how to get > the changed > vector, because i need to plot the graph about the mode . program "dynmat.x" recalculates the normal modes: just calculate them with lo-to split, save them to file for plotting Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From njuxuyuehua at gmail.com Mon Nov 19 13:45:03 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Mon, 19 Nov 2007 20:45:03 +0800 Subject: [Pw_forum] about the influence of lo-to split on the vibrational model In-Reply-To: References: Message-ID: hi Paolo: thank your reply firstly ,yes, the mode has been recalculate,but there is only new frequency in outfile,no new vibrational vector present which is my wanted. how can i get it ? look forward to your further advise . thank you 2007/11/19, Paolo Giannozzi : > > > On Nov 19, 2007, at 8:36 , xu yuehua wrote: > > > i am now computing ir-intensity.and in the input file, i add lo- > > to split . > > and the result is that some vibration frequency is changed > > corresponding > > to the *.dynG. and my question is once adding lo-to split, does the > > vibrational > > model including vibrational vector changed ? if it does, how to get > > the changed > > vector, because i need to plot the graph about the mode . > > program "dynmat.x" recalculates the normal modes: just calculate them > with > lo-to split, save them to file for plotting > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071119/e111574b/attachment.htm From baroni at sissa.it Mon Nov 19 13:59:29 2007 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 19 Nov 2007 12:59:29 +0000 Subject: [Pw_forum] Imaginary displacements In-Reply-To: <534DC6E4-9164-4948-9DB0-5E055A3E0E0F@nest.sns.it> References: <1195471863.31296.31.camel@localhost> <534DC6E4-9164-4948-9DB0-5E055A3E0E0F@nest.sns.it> Message-ID: <979AA826-BAD3-4E9F-BDA7-44AAE4802A66@sissa.it> Just to add a (hopefully, clarifying) comment. Hermitean matrix: eignevalues are real, eigenvectors are in in general complex Real symmetric matrix: eigenvalues are real (because the matrix is Hermitean as well). The eigenvectors *CAN* be chose to be real. This does not imply that they *MUST* be chosen so. You may want to rethink to the previous replies in the light of this remark. Also, I am almost certain that this very question has been already discussed in this forum. You may want to check the archive. Take care Stefano B. On Nov 19, 2007, at 12:03 PM, Paolo Giannozzi wrote: > > On Nov 19, 2007, at 12:31 , Ion Errea wrote: > >> Does anyonone which are the reasons why you can get imaginary >> displacements for phonon eigenmodes? > > at q=0 both the dynamical matrix and its eigenvectors are real, > but for a general q, the dynamical matrix is complex hermitian > and its eigenvectors are in general complex > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071119/600792c9/attachment-0001.htm From giannozz at nest.sns.it Mon Nov 19 14:17:36 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 19 Nov 2007 14:17:36 +0100 Subject: [Pw_forum] about the influence of lo-to split on the vibrational model In-Reply-To: References: Message-ID: <93C314E0-A00E-49BD-9339-4A06EDEF0EA1@nest.sns.it> On Nov 19, 2007, at 13:45 , xu yuehua wrote: > the mode has been recalculate,but there is only new frequency in > outfile, > no new vibrational vector present which is my wanted. how can i get > it ? see the header of file dynmat.f90 -- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From yaoxinxin at mail.sdu.edu.cn Mon Nov 19 14:51:04 2007 From: yaoxinxin at mail.sdu.edu.cn (Xinxin Yao) Date: Mon, 19 Nov 2007 21:51:04 +0800 Subject: [Pw_forum] questions about the calculation time Message-ID: <395480264.13371@mail.sdu.edu.cn> Dear all, Recently I calculated a large supercell with 96 atoms and 400 valence electrons. I found some strange phenomenons of calculation time. This is a part of the output file: # Self-consistent Calculation # # iteration # 1 ecut= 30.00 Ry beta=0.20 # Davidson diagonalization with overlap (about 5min) # ethr = 1.00E-05, avg # of iterations = 12.2 (longer than 10min) # # total cpu time spent up to now is 440.62 secs # # total energy = -3750.38793850 Ry # Harris-Foulkes estimate = -3754.78120012 Ry # estimated scf accuracy < 14.37337387 Ry The diagonalization process takes a few time about 5 minutes, but after the line "ethr =" appeares, until the next display of output file, it takes much long time than the former diagonalization process, larger than 10 minutes. What cost so much time in this process? And the cpu time is less than the real time. Does it mean the cpu efficiency is low? How could I improve the cpu efficiency? Thank you! Xinxin Yao From bnrj.rudra at yahoo.com Mon Nov 19 15:07:36 2007 From: bnrj.rudra at yahoo.com (Rudra Banerjee) Date: Mon, 19 Nov 2007 19:37:36 +0530 (IST) Subject: [Pw_forum] stoping the pw.x Message-ID: <750630.77166.qm@web94111.mail.in2.yahoo.com> dear colleagues, is it possible to stop a running pw job inbetween and restart it with "restart" mode? if we explicitly kill a job is it possible to resume it from where we left using "restart" tag? -- Rudra Have a Nice Time http://www.bose.res.in/~rudra Please, if possible, don't send me MS Word or PowerPoint attachments Why?See: http://www.gnu.org/philosophy/no-word-attachments.html Unlimited freedom, unlimited storage. Get it now, on http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071119/853e6336/attachment.htm From marcel at physik.tu-berlin.de Mon Nov 19 15:25:10 2007 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Mon, 19 Nov 2007 15:25:10 +0100 (CET) Subject: [Pw_forum] Plot phonon dispersion Message-ID: <20071119151946.V44136@rosa.physik-pool.tu-berlin.de> Hi, what do you usually do to convert the *.freq file, that comes from matdyn.x ? I found an old post from Eyvaz Isaev from (Feb 4 2007), but couldn't fetch the attachments from the archive. Cheers Marcel ________________________________________________________________________ Marcel Mohr Institut f?r Festk?rperphysik, TU Berlin marcel(at)physik.tu-berlin.de Sekr. EW 5-4 TEL: +49-30-314 24442 Hardenbergstr. 36 FAX: +49-30-314 27705 10623 Berlin From eyvaz_isaev at yahoo.com Mon Nov 19 16:24:32 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 19 Nov 2007 07:24:32 -0800 (PST) Subject: [Pw_forum] Plot phonon dispersion In-Reply-To: <20071119151946.V44136@rosa.physik-pool.tu-berlin.de> Message-ID: <1218.39769.qm@web60322.mail.yahoo.com> Dear Marcel, I will send you all files and script asap. Bests, Eyvaz --- Marcel Mohr wrote: > Hi, > > what do you usually do to convert the *.freq file, > that comes from > matdyn.x ? > > I found an old post from Eyvaz Isaev from (Feb 4 > 2007), but > couldn't fetch the attachments from the archive. > > Cheers > Marcel > > > ________________________________________________________________________ > Marcel Mohr Institut f?r Festk?rperphysik, TU > Berlin > marcel(at)physik.tu-berlin.de Sekr. EW 5-4 > TEL: +49-30-314 24442 Hardenbergstr. 36 > FAX: +49-30-314 27705 10623 Berlin > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs From giannozz at nest.sns.it Mon Nov 19 16:53:14 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 19 Nov 2007 16:53:14 +0100 Subject: [Pw_forum] questions about the calculation time In-Reply-To: <395480264.13371@mail.sdu.edu.cn> References: <395480264.13371@mail.sdu.edu.cn> Message-ID: <15FD60DC-39ED-4FC4-8914-656A147AF68C@nest.sns.it> On Nov 19, 2007, at 14:51 , Xinxin Yao wrote: > The diagonalization process takes a few time about 5 minutes, > but after the line "ethr =" appeares, until the next display of output > file, it takes much long time than the former diagonalization process, > larger than 10 minutes. What cost so much time in this process? look at the time report, printed at the end of the calculation. You can perform just one scf step if you do not want to wait too much. > And the cpu time is less than the real time. I would be very surprised of the opposite. How much less? > How could I improve the cpu efficiency? you didn't specify the most basic pieces of information, i.e. what the time reports says, whether you are running serially or in parallel on how many processors on which communication hardware. How can anybody answer? Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From chenhanghuipwscf at gmail.com Mon Nov 19 19:14:59 2007 From: chenhanghuipwscf at gmail.com (alan chen) Date: Mon, 19 Nov 2007 13:14:59 -0500 Subject: [Pw_forum] ultrasoft pseudopotential Message-ID: <22ae3ca40711191014o5673a2d2n43dcf2a8d8675cf1@mail.gmail.com> Dear PWSCF users, I have just begun to use ultrasoft pseudopotentials (uspp). There are two energy cutoffs in PWSCF code. One is ecutwfc and the other is ecutrho. By default, ecutrho = 4*ecutwfc. However, it is said that for uspp, ecutrho should be larger than default value. So how large an ecutrho shall I set with respect to ecutwfc? Five times or eight times? The elements I am now calculating are O, Al, Sr, Ti and La. Thank you very much for your advice. Hanghui Department of Physics Yale University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071119/a815a557/attachment.htm From eyvaz_isaev at yahoo.com Mon Nov 19 19:23:51 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 19 Nov 2007 10:23:51 -0800 (PST) Subject: [Pw_forum] ultrasoft pseudopotential In-Reply-To: <22ae3ca40711191014o5673a2d2n43dcf2a8d8675cf1@mail.gmail.com> Message-ID: <496919.48397.qm@web60312.mail.yahoo.com> Dear Hanghui, This question has been discussed many times, and please search the forum archive. Bests, Eyvaz. --- alan chen wrote: > Dear PWSCF users, > I have just begun to use ultrasoft > pseudopotentials (uspp). There are > two energy cutoffs in PWSCF code. One is ecutwfc and > the other is ecutrho. > By default, ecutrho = 4*ecutwfc. However, it is said > that for uspp, ecutrho > should be larger than default value. So how large an > ecutrho shall I set > with respect to ecutwfc? Five times or eight times? > The elements I am now calculating are O, Al, > Sr, Ti and La. > Thank you very much for your advice. > > > Hanghui > Department of Physics > Yale University > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Be a better sports nut! Let your teams follow you with Yahoo Mobile. Try it now. http://mobile.yahoo.com/sports;_ylt=At9_qDKvtAbMuh1G1SQtBI7ntAcJ From Giovanni.Cantele at na.infn.it Mon Nov 19 19:31:22 2007 From: Giovanni.Cantele at na.infn.it (Giovanni Cantele) Date: Mon, 19 Nov 2007 19:31:22 +0100 Subject: [Pw_forum] ultrasoft pseudopotential In-Reply-To: <22ae3ca40711191014o5673a2d2n43dcf2a8d8675cf1@mail.gmail.com> References: <22ae3ca40711191014o5673a2d2n43dcf2a8d8675cf1@mail.gmail.com> Message-ID: <4741D67A.50103@na.infn.it> alan chen wrote: > Dear PWSCF users, > I have just begun to use ultrasoft pseudopotentials (uspp). > There are two energy cutoffs in PWSCF code. One is ecutwfc and the > other is ecutrho. By default, ecutrho = 4*ecutwfc. However, it is said > that for uspp, ecutrho should be larger than default value. So how > large an ecutrho shall I set with respect to ecutwfc? Five times or > eight times? > The elements I am now calculating are O, Al, Sr, Ti and La. > Thank you very much for your advice. > > > Hanghui > Department of Physics > Yale University > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > This issue has been discussed several times in the forum. You'll probably find the answer in one of these links: http://www.democritos.it/pipermail/pw_forum/2005-December/003396.html http://www.democritos.it/pipermail/pw_forum/2005-October/003128.html http://www.democritos.it/pipermail/pw_forum/2007-November/007621.html ..... that are easily found searching for ecutrho in the forum. Giovanni -- Dr. Giovanni Cantele Coherentia CNR-INFM and Dipartimento di Scienze Fisiche Universita' di Napoli "Federico II" Complesso Universitario di Monte S. Angelo - Ed. 6 Via Cintia, I-80126, Napoli, Italy Phone: +39 081 676910 Fax: +39 081 676346 E-mail: Giovanni.Cantele at na.infn.it Web: http://people.na.infn.it/~cantele From lyanglyan at yahoo.com.cn Tue Nov 20 02:13:48 2007 From: lyanglyan at yahoo.com.cn (=?gb2312?q?=D1=D2=20=C0=EE?=) Date: Tue, 20 Nov 2007 09:13:48 +0800 (CST) Subject: [Pw_forum] The calculated results using the pp file of Cs from Mr. Paolo. Message-ID: <702854.31445.qm@web15814.mail.cnb.yahoo.com> Dear Paolo and all users, I got a good areeement between the theory and experiment by using the Cs pp file that you sent to me. The simulated results even reproduce the experiment data under very high pressure. But I just tested the one with PBE and 9 valence electrons. Thanks a lot! Best regards, Yan Li --------------------------------- ?????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071120/cd6c7d0e/attachment-0001.htm From w2agz at pacbell.net Tue Nov 20 04:07:27 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Mon, 19 Nov 2007 19:07:27 -0800 Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation Message-ID: <007a01c82b22$7bc79920$7356cb60$@net> To All Forum Members: I'm planning on building a new Linux box (or boxes) to explore highly correlated systems (e-p coupling plus LDA+U), and am seeking the collective experience and advice of the PWscf community on a suitable, inexpensive (< 2000 USD, MB+CPUs+RAM, exclusive of power supplies, enclosures, and accessories) hardware platform. I emphasize that the principal purpose of this new box would be exploratory, or for development, not production. I've built several past platforms, both Windows and Linux based, using server boards manufactured by Supermicro and have had generally good experience and service (the factory is only 15 miles from where I live). Currently, I use two machines with dual Xeon processors (single core, 32 bits), one with 1 GHz cpus, 1 GB RAM, the other 2.4 GHz and 3.25 GB RAM, both with bus speeds of 133 MHz, the newest 3 years old. However, occasionally I run PWscf exercises on my little Thinkpad X41 tablet (single processor, 1.5 GHz, 1.5 GB), and the scf computation will run 3-5 times faster than on the other machines! I suspect this rather surprising result is because the Thinkpad has a 400 MHz bus clock speed. One option I'm considering is using a "gaming" or server class motherboard with dual 2.33 GHz quad-core 64-bit processors, a 1333 MHz FSB, and 16 GB RAM. Having said this, I'm not sure PWscf (and the Fortran compilers available) can handle all this parallelism efficiently on a single motherboard. I've noticed when running pw.x, the CPU activity "flips" between processors every several seconds, instead of sharing each at 90-100% full time. On the other hand, one could consider building a small MPI-connected cluster for about the same amount of money. When IBM announced a couple of years ago the incredible performance details about the Cell processor that would go into Playstation 3, I thought, "Wow, maybe the future of computational physics rests with gamers." I'm sure most of you know this is actually beginning to happen, spurred on by the fact that the PS3 is "open architecture" and can run a Linux distro. Moreover, there apparently are "open software" numerical analysis tools available from IBM. At least four US universities are experimenting with off-the-shelf PS3 clusters, perhaps one of the more interesting is at UMass, http://gravity.phy.umassd.edu/ps3.html. In the last week or so, Sony lowered the entry level price of the PS3 to 400 USD. So, a cluster of four with a cheap switch could be purchased for about the same price at the single motherboard configuration I mentioned above. My teenager, a gamer, tells me the PS3 has problems. He says it's unreliable and overheats and only has 256 MB RAM on board (he owns a Wii, which outsells the PS3 in the US by a factor of three). Has anybody tried porting PWscf to a PS3? Any and all advice is welcome. Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford University EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071119/16ba6dd0/attachment.htm From xu11249249 at gmail.com Tue Nov 20 04:16:37 2007 From: xu11249249 at gmail.com (xu yuehua) Date: Tue, 20 Nov 2007 11:16:37 +0800 Subject: [Pw_forum] about the influence of lo-to split on the vibrational model In-Reply-To: <93C314E0-A00E-49BD-9339-4A06EDEF0EA1@nest.sns.it> References: <93C314E0-A00E-49BD-9339-4A06EDEF0EA1@nest.sns.it> Message-ID: <426b166e0711191916g3763855fo858e525af0c96fd2@mail.gmail.com> hi Paolo: oh .there is a file named dynmatrix.f90 .and i got confused . i list my in file and my out file here : &input fildyn='/raid/xyh/pwscf/*/*.dynG',asr='one-dim',q(1)=0,q(2)=0,q(3)=12/ AND THE OUTFILE HERE: Thu Oct 25 11:06:00 CST 2007 Reading Dynamical Matrix from file /raid/xyh/pwscf/*/*.dynG asr rotation axis in 1D system= 3 Acoustic Sum Rule: || Z*(ASR) - Z*(orig)|| = 0.815659E-01 Acoustic Sum Rule: ||dyn(ASR) - dyn(orig)||= 0.697493E-01 Polarizability (A^3 units) multiply by 0.971817 for Clausius-Mossotti correction 5.625414 0.000658 0.000133 0.000658 5.624128 0.000114 0.000133 0.000114 8.860593 IR cross sections are in (D/A)^2/amu units # mode [cm-1] [THz] IR 1 0.00 0.0000 0.0000 2 0.00 0.0000 0.0000 3 0.00 0.0000 0.0000 4 0.00 0.0000 0.0000 5 49.16 1.4738 0.0064 6 63.88 1.9151 0.0043 7 68.56 2.0554 0.0016 8 106.29 3.1865 0.0002 9 156.91 4.7040 0.0002 10 252.93 7.5826 0.2436 11 254.66 7.6344 0.4708 and there is not the vector at all!! and the job script: #!/bin/bash # #$ -cwd #$ -j y #$ -S /bin/bash # MPI_DIR=/home/software/mpich/intel/bin date $MPI_DIR/mpirun -np $NSLOTS -machinefile $TMPDIR/machines -nolocal /home/xyh/intel_espresso-3.2/bin/dynmat.x: > > > On Nov 19, 2007, at 13:45 , xu yuehua wrote: > > > the mode has been recalculate,but there is only new frequency in > > outfile, > > no new vibrational vector present which is my wanted. how can i get > > it ? > > see the header of file dynmat.f90 > -- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- lalala lalala -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071120/204595f0/attachment.htm From njuxuyuehua at gmail.com Tue Nov 20 08:19:16 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Tue, 20 Nov 2007 15:19:16 +0800 Subject: [Pw_forum] about the influence of lo-to split on the vibrational model Message-ID: hi Paolo: oh .there is a file named dynmatrix.f90 .and i got confused . i list my in file and my out file here : &input fildyn='/raid/xyh/pwscf/*/*.dynG',asr='one-dim',q(1)=0,q(2)=0,q(3)=12/ AND THE OUTFILE HERE: Thu Oct 25 11:06:00 CST 2007 Reading Dynamical Matrix from file /raid/xyh/pwscf/*/*.dynG asr rotation axis in 1D system= 3 Acoustic Sum Rule: || Z*(ASR) - Z*(orig)|| = 0.815659E-01 Acoustic Sum Rule: ||dyn(ASR) - dyn(orig)||= 0.697493E-01 Polarizability (A^3 units) multiply by 0.971817 for Clausius-Mossotti correction 5.625414 0.000658 0.000133 0.000658 5.624128 0.000114 0.000133 0.000114 8.860593 IR cross sections are in (D/A)^2/amu units # mode [cm-1] [THz] IR 1 0.00 0.0000 0.0000 2 0.00 0.0000 0.0000 3 0.00 0.0000 0.0000 4 0.00 0.0000 0.0000 5 49.16 1.4738 0.0064 6 63.88 1.9151 0.0043 7 68.56 2.0554 0.0016 8 106.29 3.1865 0.0002 9 156.91 4.7040 0.0002 10 252.93 7.5826 0.2436 11 254.66 7.6344 0.4708 and there is not the vector at all!! and the job script: #!/bin/bash # #$ -cwd #$ -j y #$ -S /bin/bash # MPI_DIR=/home/software/mpich/intel/bin date $MPI_DIR/mpirun -np $NSLOTS -machinefile $TMPDIR/machines -nolocal /home/xyh/intel_espresso-3.2/bin/dynmat.x References: <93C314E0-A00E-49BD-9339-4A06EDEF0EA1@nest.sns.it> <426b166e0711191916g3763855fo858e525af0c96fd2@mail.gmail.com> Message-ID: <41DEF3AB-6204-4DA3-902F-D7FE74698F8E@nest.sns.it> On Nov 20, 2007, at 4:16 , xu yuehua wrote: > and there is not the vector at all!! > [...] > and what is wrong ? i do not know > please give me a detailed advise your attitude is wrong. Had you followed my previous advice (not "advise") you would have found this: ! filout character output file containing frequencies and modes ! (default: filout='dynmat.out') ! filmol character as above, in a format suitable for 'molden' ! (default: filmol='dynmat.mold') ! filxsf character as above, in axsf format suitable for xcrysden ! (default: filmol='dynmat.axsf') --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From njuxuyuehua at gmail.com Tue Nov 20 09:33:09 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Tue, 20 Nov 2007 16:33:09 +0800 Subject: [Pw_forum] about the influence of lo-to split on the vibrational model In-Reply-To: <41DEF3AB-6204-4DA3-902F-D7FE74698F8E@nest.sns.it> References: <93C314E0-A00E-49BD-9339-4A06EDEF0EA1@nest.sns.it> <426b166e0711191916g3763855fo858e525af0c96fd2@mail.gmail.com> <41DEF3AB-6204-4DA3-902F-D7FE74698F8E@nest.sns.it> Message-ID: i am very sorry .just now ,i did not find the dynmat.f90. now .i am reading . thank you for your kind criticism and your suggestion 2007/11/20, Paolo Giannozzi : > > > On Nov 20, 2007, at 4:16 , xu yuehua wrote: > > > and there is not the vector at all!! > > [...] > > and what is wrong ? i do not know > > please give me a detailed advise > > your attitude is wrong. Had you followed my previous > advice (not "advise") you would have found this: > > ! filout character output file containing frequencies and modes > ! (default: filout='dynmat.out') > ! filmol character as above, in a format suitable for 'molden' > ! (default: filmol='dynmat.mold') > ! filxsf character as above, in axsf format suitable for xcrysden > ! (default: filmol='dynmat.axsf') > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071120/d4676105/attachment.htm From giannozz at nest.sns.it Tue Nov 20 09:33:27 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 20 Nov 2007 09:33:27 +0100 Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation In-Reply-To: <007a01c82b22$7bc79920$7356cb60$@net> References: <007a01c82b22$7bc79920$7356cb60$@net> Message-ID: On Nov 20, 2007, at 4:07 , Paul M. Grant wrote: > Currently, I use two machines with dual Xeon processors dual Xeon? very bad choice. They have a single memory bus, so if you run two processes on the same motherboard, they spend most of the time waiting for access to memory. P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Tue Nov 20 09:41:07 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 20 Nov 2007 09:41:07 +0100 Subject: [Pw_forum] stoping the pw.x In-Reply-To: <750630.77166.qm@web94111.mail.in2.yahoo.com> References: <750630.77166.qm@web94111.mail.in2.yahoo.com> Message-ID: <00E63025-950D-4CA6-BB49-6B1D9CC63022@nest.sns.it> On Nov 19, 2007, at 15:07 , Rudra Banerjee wrote: > is it possible to stop a running pw job in between > and restart it with "restart" mode? yes, but it doesn't always work. You can either stop after a predefined time (see input variable "max_seconds") or when you deem it appropriate (the best way is to create a file "prefix".EXIT in the temporary directory where the other "prefix".* files are). How the code will restart depends on how much information is stored to file. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From njuxuyuehua at gmail.com Tue Nov 20 10:18:33 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Tue, 20 Nov 2007 17:18:33 +0800 Subject: [Pw_forum] still can not get the modes with the dynmat.x Message-ID: i use dynmat.x to add lo-to split ,and want to get the new frequencies and modes, and only the frequencies is present . my infile now: &input fildyn='/raid/xyh/pwscf/fourringwater/pho_scf_relax_fromlqf_via_vaspz/*.dynG', asr='one-dim', q(1)=0,q(2)=0,q(3)=12, filout= 'dynmat.out'/ but there is not dynmat.out .only the file like pwscf.sh.o8645 and in pwscf.sh.o8645 there is : ue Nov 20 17:29:58 CST 2007 Reading Dynamical Matrix from file /raid/xyh/pwscf/fourringwater/pho_scf_relax_fromlqf_via_vaspz/fourringwater.dynG asr rotation axis in 1D system= 3 Acoustic Sum Rule: || Z*(ASR) - Z*(orig)|| = 0.815659E-01 Acoustic Sum Rule: ||dyn(ASR) - dyn(orig)||= 0.697493E-01 Polarizability (A^3 units) multiply by 0.971817 for Clausius-Mossotti correction 5.625414 0.000658 0.000133 0.000658 5.624128 0.000114 0.000133 0.000114 8.860593 IR cross sections are in (D/A)^2/amu units # mode [cm-1] [THz] IR 1 0.00 0.0000 0.0000 2 0.00 0.0000 0.0000 3 0.00 0.0000 0.0000 4 0.00 0.0000 0.0000 5 49.16 1.4738 0.0064 6 63.88 1.9151 0.0043 7 68.56 2.0554 0.0016 8 106.29 3.1865 0.0002 9 156.91 4.7040 0.0002 10 252.93 7.5826 0.2436 11 254.66 7.6344 0.4708 12 257.93 7.7324 0.6958 13 564.27 16.9163 0.4540 14 631.40 18.9287 0.1605 15 632.64 18.9659 1.7223 16 633.03 18.9777 1.8109 17 705.34 21.1455 1.3172 18 750.19 22.4900 3.3584 19 750.80 22.5084 3.3466 20 793.99 23.8030 0.0001 21 854.32 25.6117 3.6181 ........... no mode ! .i am puzzled .and where is my error ? -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071120/20d45094/attachment.htm From shaposh at isp.nsc.ru Tue Nov 20 10:01:34 2007 From: shaposh at isp.nsc.ru (Alexander Shaposhnikov) Date: Tue, 20 Nov 2007 15:01:34 +0600 Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation In-Reply-To: References: <007a01c82b22$7bc79920$7356cb60$@net> Message-ID: <200711201501.37502.shaposh@isp.nsc.ru> Actually newer Woodcrest/Clovertown Xeon platform have dual independent FSBs, but this does not help much. Memory bandwidth available on each FSB is very limited - between 1/2 and 1/3 of what is available to modern desktop CPUs. There is no good "server" computational platform today, unfortunately . AMD's CPUs have weak FPU and Intel's have weak memory subsystem. Intel- based desktop single-socket platform is the best choice for now. On Tuesday 20 November 2007 14:33, Paolo Giannozzi wrote: > On Nov 20, 2007, at 4:07 , Paul M. Grant wrote: > > Currently, I use two machines with dual Xeon processors > > dual Xeon? very bad choice. They have a single memory > bus, so if you run two processes on the same motherboard, > they spend most of the time waiting for access to memory. > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From marcel at physik.tu-berlin.de Tue Nov 20 11:14:12 2007 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Tue, 20 Nov 2007 11:14:12 +0100 (CET) Subject: [Pw_forum] still can not get the modes with the dynmat.x In-Reply-To: References: Message-ID: <20071120111017.G75082@rosa.physik-pool.tu-berlin.de> Hi Yuehua Try to run dynmat interactively without using the queue manager ! There should be a files called dynmat.out, dynmat.mold, dynmat.x... See in example09. Nobody can guess, where your computer puts them ;) Cheers Marcel ________________________________________________________________________ Marcel Mohr Institut f?r Festk?rperphysik, TU Berlin marcel(at)physik.tu-berlin.de Sekr. EW 5-4 TEL: +49-30-314 24442 Hardenbergstr. 36 FAX: +49-30-314 27705 10623 Berlin On Tue, 20 Nov 2007, xu yuehua wrote: > i use dynmat.x to add lo-to split ,and want to get the new frequencies and > modes, and only the frequencies is present . > my infile now: > > &input > fildyn='/raid/xyh/pwscf/fourringwater/pho_scf_relax_fromlqf_via_vaspz/*.dynG', > asr='one-dim', > q(1)=0,q(2)=0,q(3)=12, > filout= 'dynmat.out'/ > > > > but there is not dynmat.out .only the file like pwscf.sh.o8645 > and in pwscf.sh.o8645 > > there is : > > ue Nov 20 17:29:58 CST 2007 > > Reading Dynamical Matrix from file > /raid/xyh/pwscf/fourringwater/pho_scf_relax_fromlqf_via_vaspz/fourringwater.dynG > asr rotation axis in 1D system= 3 > Acoustic Sum Rule: || Z*(ASR) - Z*(orig)|| = 0.815659E-01 > Acoustic Sum Rule: ||dyn(ASR) - dyn(orig)||= 0.697493E-01 > > Polarizability (A^3 units) > multiply by 0.971817 for Clausius-Mossotti correction > 5.625414 0.000658 0.000133 > 0.000658 5.624128 0.000114 > 0.000133 0.000114 8.860593 > > IR cross sections are in (D/A)^2/amu units > # mode [cm-1] [THz] IR > 1 0.00 0.0000 0.0000 > 2 0.00 0.0000 0.0000 > 3 0.00 0.0000 0.0000 > 4 0.00 0.0000 0.0000 > 5 49.16 1.4738 0.0064 > 6 63.88 1.9151 0.0043 > 7 68.56 2.0554 0.0016 > 8 106.29 3.1865 0.0002 > 9 156.91 4.7040 0.0002 > 10 252.93 7.5826 0.2436 > 11 254.66 7.6344 0.4708 > 12 257.93 7.7324 0.6958 > 13 564.27 16.9163 0.4540 > 14 631.40 18.9287 0.1605 > 15 632.64 18.9659 1.7223 > 16 633.03 18.9777 1.8109 > 17 705.34 21.1455 1.3172 > 18 750.19 22.4900 3.3584 > 19 750.80 22.5084 3.3466 > 20 793.99 23.8030 0.0001 > 21 854.32 25.6117 3.6181 > > ........... > > > > > no mode ! .i am puzzled .and where is my error ? > > > -- > Xu Yuehua > physics Department of Nanjing university > China > From njuxuyuehua at gmail.com Tue Nov 20 13:11:48 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Tue, 20 Nov 2007 20:11:48 +0800 Subject: [Pw_forum] still can not get the modes with the dynmat.x In-Reply-To: <20071120111017.G75082@rosa.physik-pool.tu-berlin.de> References: <20071120111017.G75082@rosa.physik-pool.tu-berlin.de> Message-ID: hi Marcel: i am just looking for the file .and at last ,i found it .it was put in my home directory.hehe! and my another question is whether i can specify the destination of the file ,like in the currently directory? and i have a third question is the interactively process,just called serial process, just now ,i tried it .i can not work,the load of cpu is zero so ,i do not know why ! i need your help thank you 2007/11/20, Marcel Mohr : > > Hi Yuehua > > Try to run dynmat interactively without using the queue manager ! > > There should be a files called dynmat.out, dynmat.mold, dynmat.x... > See in example09. > > Nobody can guess, where your computer puts them ;) > > Cheers Marcel > > ________________________________________________________________________ > Marcel Mohr Institut f?r Festk?rperphysik, TU Berlin > marcel(at)physik.tu-berlin.de Sekr. EW 5-4 > TEL: +49-30-314 24442 Hardenbergstr. 36 > FAX: +49-30-314 27705 10623 Berlin > > > On Tue, 20 Nov 2007, xu yuehua wrote: > > > i use dynmat.x to add lo-to split ,and want to get the new frequencies > and > > modes, and only the frequencies is present . > > my infile now: > > > > &input > > > fildyn='/raid/xyh/pwscf/fourringwater/pho_scf_relax_fromlqf_via_vaspz/*.dynG', > > asr='one-dim', > > q(1)=0,q(2)=0,q(3)=12, > > filout= 'dynmat.out'/ > > > > > > > > but there is not dynmat.out .only the file like pwscf.sh.o8645 > > and in pwscf.sh.o8645 > > > > there is : > > > > ue Nov 20 17:29:58 CST 2007 > > > > Reading Dynamical Matrix from file > > > /raid/xyh/pwscf/fourringwater/pho_scf_relax_fromlqf_via_vaspz/fourringwater.dynG > > asr rotation axis in 1D system= 3 > > Acoustic Sum Rule: || Z*(ASR) - Z*(orig)|| = 0.815659E-01 > > Acoustic Sum Rule: ||dyn(ASR) - dyn(orig)||= 0.697493E-01 > > > > Polarizability (A^3 units) > > multiply by 0.971817 for Clausius-Mossotti correction > > 5.625414 0.000658 0.000133 > > 0.000658 5.624128 0.000114 > > 0.000133 0.000114 8.860593 > > > > IR cross sections are in (D/A)^2/amu units > > # mode [cm-1] [THz] IR > > 1 0.00 0.0000 0.0000 > > 2 0.00 0.0000 0.0000 > > 3 0.00 0.0000 0.0000 > > 4 0.00 0.0000 0.0000 > > 5 49.16 1.4738 0.0064 > > 6 63.88 1.9151 0.0043 > > 7 68.56 2.0554 0.0016 > > 8 106.29 3.1865 0.0002 > > 9 156.91 4.7040 0.0002 > > 10 252.93 7.5826 0.2436 > > 11 254.66 7.6344 0.4708 > > 12 257.93 7.7324 0.6958 > > 13 564.27 16.9163 0.4540 > > 14 631.40 18.9287 0.1605 > > 15 632.64 18.9659 1.7223 > > 16 633.03 18.9777 1.8109 > > 17 705.34 21.1455 1.3172 > > 18 750.19 22.4900 3.3584 > > 19 750.80 22.5084 3.3466 > > 20 793.99 23.8030 0.0001 > > 21 854.32 25.6117 3.6181 > > > > ........... > > > > > > > > > > no mode ! .i am puzzled .and where is my error ? > > > > > > -- > > Xu Yuehua > > physics Department of Nanjing university > > China > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071120/5c67212b/attachment.htm From nazari at iasbs.ac.ir Tue Nov 20 18:25:47 2007 From: nazari at iasbs.ac.ir (nazari at iasbs.ac.ir) Date: Tue, 20 Nov 2007 17:25:47 -0000 (UTC) Subject: [Pw_forum] Pseudo_Generation/unknown pseudo error Message-ID: <8236.213.176.122.17.1195579547.squirrel@mail.iasbs.ac.ir> Dear All, I wish to generate a pseudo potential. I have used the instruction in Metodi Numerici in struttura Elettronica. But when I applied the pseudo potential I face with unknown pseudo potential error. Also I did search in archive and find one answer but it doesn't work. Would you please let me know How I can use the result of step 1 and 2 of the mentioned text for producing pseudo potential. Because I wrote my input file for pseudo generation by comparing with oxygen input in pseudo-gen folder and run f.in in this folder. Regards Fariba Nazari IASBS From cyrille.barreteau at cea.fr Tue Nov 20 14:59:49 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Tue, 20 Nov 2007 14:59:49 +0100 Subject: [Pw_forum] Problem with nbnd in clusters Message-ID: <4742E855.2030909@cea.fr> Dear pwscf_ers I am doing a very simple calculation on a transition metal dimer and I have a encoutered a strange problem (I love strange problems:-) I have done a first calculation with the default value of nbnd, ie nbnd=nelec/2*20% (=12 in my case) But since it is known that in clusters it is often good to increase nbnd I have performed another calculation with larger nbnd (=20). The result is quite different from the one at nbnd=12. I have then increased again nbnd up to 25 and then I recover the result obtained for nbnd=12 More problematic is the local density of states (I know it is not really a dos but a bunch of dirac peaks). In a dimer with the z axis along the direction connecting the two atoms, the xy and x^2-y^2 dos should be degenerate (if the supercell box is large enough). In fact the xy and x^2-y^2 are degenerate if nbnd=12 or 25 but there is a rather large splitting if nbnd=20. I am quite sure the result for nbnd=20 is not correct but what is the origin of this problem? Maybe I could try to use another diagonalization scheme.. thanks for reading my strange problems cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== From paulatto at sissa.it Tue Nov 20 15:12:55 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Tue, 20 Nov 2007 15:12:55 +0100 (CET) Subject: [Pw_forum] Pseudo_Generation/unknown pseudo error In-Reply-To: <8236.213.176.122.17.1195579547.squirrel@mail.iasbs.ac.ir> References: <8236.213.176.122.17.1195579547.squirrel@mail.iasbs.ac.ir> Message-ID: <4823.147.122.5.182.1195567975.squirrel@webmail.sissa.it> On Tue, November 20, 2007 18:25, nazari at iasbs.ac.ir wrote: > I face with unknown pseudo potential error. There are many things that can go wrong; conflicting input directive may result in quite obscure error messages. Anyway if you provide the error message, or the input file, it is more likely than someone will be able to solve you problem. Another hint is to have a look in atomic_doc/pseudo_library/ where you can find many examples of working input files. goodbye -- Lorenzo Paulatto +39 040 3787 511 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From nazari at iasbs.ac.ir Tue Nov 20 19:27:51 2007 From: nazari at iasbs.ac.ir (nazari at iasbs.ac.ir) Date: Tue, 20 Nov 2007 18:27:51 -0000 (UTC) Subject: [Pw_forum] Pseudo_Generation/unknown pseudo error In-Reply-To: <4823.147.122.5.182.1195567975.squirrel@webmail.sissa.it> References: <8236.213.176.122.17.1195579547.squirrel@mail.iasbs.ac.ir> <4823.147.122.5.182.1195567975.squirrel@webmail.sissa.it> Message-ID: <8946.213.176.122.17.1195583271.squirrel@mail.iasbs.ac.ir> Dear All, Here is the input / &input title='F', zed=9.0, rel=0, beta=0.5, rlderiv=2.0, eminld=-4.0, emaxld=4.0, deld=0.02d0, nld=4, iswitch=3, dft='BLYP', / 4 1S 1 0 2.0 1 2S 2 0 2.0 1 2P 2 1 5.0 1 3D 3 2 -1.0 1 &inputp pseudotype=3, lloc=2, file_pseudopw='FBLYP.VAN_AK', / 5 2S 1 0 2.00 0.00 1.10 1.70 2S 1 0 0.00 0.05 1.10 1.70 2P 2 1 5.00 0.00 1.10 1.70 2P 2 1 0.00 0.05 1.10 1.70 3D 3 2 -2.00 0.15 1.50 1.50 regards fariba nazari IASBS > On Tue, November 20, 2007 18:25, nazari at iasbs.ac.ir wrote: >> I face with unknown pseudo potential error. > > There are many things that can go wrong; conflicting input directive may > result in quite obscure error messages. Anyway if you provide the error > message, or the input file, it is more likely than someone will be able to > solve you problem. > > Another hint is to have a look in atomic_doc/pseudo_library/ where you can > find many examples of working input files. > > goodbye > > -- > Lorenzo Paulatto > +39 040 3787 511 > http://people.sissa.it/~paulatto/ > > > ---------------------------------------------------------------- > SISSA Webmail https://webmail.sissa.it/ > Powered by SquirrelMail http://www.squirrelmail.org/ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From cyrille.barreteau at cea.fr Tue Nov 20 15:55:17 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Tue, 20 Nov 2007 15:55:17 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <4742E855.2030909@cea.fr> References: <4742E855.2030909@cea.fr> Message-ID: <4742F555.9070201@cea.fr> I just wanted to add that changing the diagonalization scheme to conjugate gradient 'cg' instead of the default 'david' seems to solve the problem in my case. I now find the same result with nbnd=12,20,25.. Does anyone has experience about the behaviour of the various diagonalization schemes? Which one is better, in which case? cyrille Cyrille Barreteau wrote: >Dear pwscf_ers > >I am doing a very simple calculation on a transition metal >dimer and I have a encoutered a strange problem (I love >strange problems:-) > >I have done a first calculation with the default value >of nbnd, ie nbnd=nelec/2*20% (=12 in my case) > >But since it is known that in clusters it is often good >to increase nbnd I have performed another calculation with >larger nbnd (=20). The result is quite different from >the one at nbnd=12. > >I have then increased again nbnd up to 25 and then I recover >the result obtained for nbnd=12 > >More problematic is the local density of states (I know >it is not really a dos but a bunch of dirac peaks). >In a dimer with the z axis along the direction connecting >the two atoms, the xy and x^2-y^2 dos should be degenerate >(if the supercell box is large enough). >In fact the xy and x^2-y^2 are degenerate if nbnd=12 or 25 >but there is a rather large splitting if nbnd=20. > >I am quite sure the result for nbnd=20 is not correct but >what is the origin of this problem? > >Maybe I could try to use another diagonalization scheme.. > > thanks for reading my strange problems > > cyrille > > > > -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== From marzari at MIT.EDU Tue Nov 20 16:48:21 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Tue, 20 Nov 2007 10:48:21 -0500 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <4742E855.2030909@cea.fr> References: <4742E855.2030909@cea.fr> Message-ID: <474301C5.1050502@mit.edu> Hi Cyrille, your tm dimer will likely require spin polarization, and will have several scf solutions that correspond to different S_z (that you can fix in several ways, either specifying the number of spin up and spin down, or using a fermi_up and fermi_down energy, to specify n_up - n_down but still treating the system with a smearing and fractional occupations). Even for a given S_z, there will be different scf solutions that have different spatial symmetry. So, in your case I would definitely switch spin-polarization, and maybe play around with small and large smearings, to see what happens, and look at the states, and the occupations. Anyhow, if you get to a scf solution, that is a "good" solution for the GGA problem, although might not be very physical. Also, have a look at Kulik et al PRL 2006. nicola Cyrille Barreteau wrote: > Dear pwscf_ers > > I am doing a very simple calculation on a transition metal > dimer and I have a encoutered a strange problem (I love > strange problems:-) > > I have done a first calculation with the default value > of nbnd, ie nbnd=nelec/2*20% (=12 in my case) > > But since it is known that in clusters it is often good > to increase nbnd I have performed another calculation with > larger nbnd (=20). The result is quite different from > the one at nbnd=12. > > I have then increased again nbnd up to 25 and then I recover > the result obtained for nbnd=12 > > More problematic is the local density of states (I know > it is not really a dos but a bunch of dirac peaks). > In a dimer with the z axis along the direction connecting > the two atoms, the xy and x^2-y^2 dos should be degenerate > (if the supercell box is large enough). > In fact the xy and x^2-y^2 are degenerate if nbnd=12 or 25 > but there is a rather large splitting if nbnd=20. > > I am quite sure the result for nbnd=20 is not correct but > what is the origin of this problem? > > Maybe I could try to use another diagonalization scheme.. > > thanks for reading my strange problems > > cyrille > > -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From matteo at umn.edu Tue Nov 20 16:58:29 2007 From: matteo at umn.edu (Matteo Cococcioni) Date: Tue, 20 Nov 2007 09:58:29 -0600 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <474301C5.1050502@mit.edu> References: <4742E855.2030909@cea.fr> <474301C5.1050502@mit.edu> Message-ID: <47430425.8020205@umn.edu> Dear Cyrille, Nicola Marzari's is a good advice for these systems. Another thing that you could check is starting from random wave functions (instead of atomic). I am not expert of diagonalization algorithms but maybe what happens is that davidson with 20 bands (who knows why...) gets trapped in some local minimum just for numerical reasons. So maybe some noise could help. How do the energies compare? Regards, Matteo Nicola Marzari wrote: > Hi Cyrille, > > your tm dimer will likely require spin polarization, and > will have several scf solutions that correspond to different > S_z (that you can fix in several ways, either specifying the > number of spin up and spin down, or using a fermi_up and > fermi_down energy, to specify n_up - n_down but still treating > the system with a smearing and fractional occupations). > > Even for a given S_z, there will be different scf solutions > that have different spatial symmetry. > > So, in your case I would definitely switch spin-polarization, > and maybe play around with small and large smearings, to see what > happens, and look at the states, and the occupations. Anyhow, if > you get to a scf solution, that is a "good" solution for the GGA > problem, although might not be very physical. > > Also, have a look at Kulik et al PRL 2006. > > nicola > > > Cyrille Barreteau wrote: > >> Dear pwscf_ers >> >> I am doing a very simple calculation on a transition metal >> dimer and I have a encoutered a strange problem (I love >> strange problems:-) >> >> I have done a first calculation with the default value >> of nbnd, ie nbnd=nelec/2*20% (=12 in my case) >> >> But since it is known that in clusters it is often good >> to increase nbnd I have performed another calculation with >> larger nbnd (=20). The result is quite different from >> the one at nbnd=12. >> >> I have then increased again nbnd up to 25 and then I recover >> the result obtained for nbnd=12 >> >> More problematic is the local density of states (I know >> it is not really a dos but a bunch of dirac peaks). >> In a dimer with the z axis along the direction connecting >> the two atoms, the xy and x^2-y^2 dos should be degenerate >> (if the supercell box is large enough). >> In fact the xy and x^2-y^2 are degenerate if nbnd=12 or 25 >> but there is a rather large splitting if nbnd=20. >> >> I am quite sure the result for nbnd=20 is not correct but >> what is the origin of this problem? >> >> Maybe I could try to use another diagonalization scheme.. >> >> thanks for reading my strange problems >> >> cyrille >> >> >> > > > -------------- next part -------------- A non-text attachment was scrubbed... Name: matteo.vcf Type: text/x-vcard Size: 294 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20071120/29749b8b/attachment.vcf From paulatto at sissa.it Tue Nov 20 16:59:19 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Tue, 20 Nov 2007 16:59:19 +0100 (CET) Subject: [Pw_forum] Pseudo_Generation/unknown pseudo error In-Reply-To: <8946.213.176.122.17.1195583271.squirrel@mail.iasbs.ac.ir> References: <8236.213.176.122.17.1195579547.squirrel@mail.iasbs.ac.ir> <4823.147.122.5.182.1195567975.squirrel@webmail.sissa.it> <8946.213.176.122.17.1195583271.squirrel@mail.iasbs.ac.ir> Message-ID: <18136.147.122.5.182.1195574359.squirrel@webmail.sissa.it> I regret to say that your input file works for me (QE-3.2). Also the output PP looks quite ok (as much as you can say without testing it). What version of QE are you using? P.S. I would just change this line: file_pseudopw='FBLYP.VAN_AK', to file_pseudopw='FBLYP.VAN_AK.UPF', in order to use "new" file format. -- Lorenzo Paulatto +39 040 3787 511 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From marzari at MIT.EDU Tue Nov 20 17:01:53 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Tue, 20 Nov 2007 11:01:53 -0500 Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation In-Reply-To: <007a01c82b22$7bc79920$7356cb60$@net> References: <007a01c82b22$7bc79920$7356cb60$@net> Message-ID: <474304F1.4030203@mit.edu> Hi Paul, let us know if you get PWSCF on a PS3 ! I generally echo the other suggestions - the speed of the memory bus is critical, having fast memory (matched with the bus) is good, having the fastest clock on current Intel chip is good, having two cores is a 25-50% improvement, and having 4 cores is a negligible improvement on 2. If you plan to link blades in a cluster, it's vital to have fast, low latency communication - if you use gigabit, you might get good parallelism with as many as 8 blades, or as little as 2, depending on the quality of your switch, using a very recent mpi implementation, and the quality of your ethernet controllers on board. We have some tests on different platforms on http://quasiamore.mit.edu/pmwiki/index.php?n=Main.CP90Timings nicola Paul M. Grant wrote: > To All Forum Members: > > I?m planning on building a new Linux box (or boxes) to explore highly > correlated systems (e-p coupling plus LDA+U), and am seeking the > collective experience and advice of the PWscf community on a suitable, > inexpensive (< 2000 USD, MB+CPUs+RAM, exclusive of power supplies, > enclosures, and accessories) hardware platform. I emphasize that the > principal purpose of this new box would be exploratory, or for > development, not production. > > I?ve built several past platforms, both Windows and Linux based, using > server boards manufactured by Supermicro and have had generally good > experience and service (the factory is only 15 miles from where I > live). Currently, I use two machines with dual Xeon processors (single > core, 32 bits), one with 1 GHz cpus, 1 GB RAM, the other 2.4 GHz and > 3.25 GB RAM, both with bus speeds of 133 MHz, the newest 3 years old. > However, occasionally I run PWscf exercises on my little Thinkpad X41 > tablet (single processor, 1.5 GHz, 1.5 GB), and the scf computation will > run 3-5 times faster than on the other machines! I suspect this rather > surprising result is because the Thinkpad has a 400 MHz bus clock speed. > > One option I?m considering is using a ?gaming? or server class > motherboard with dual 2.33 GHz quad-core 64-bit processors, a 1333 MHz > FSB, and 16 GB RAM. Having said this, I?m not sure PWscf (and the > Fortran compilers available) can handle all this parallelism efficiently > on a single motherboard. I?ve noticed when running pw.x, the CPU > activity ?flips? between processors every several seconds, instead of > sharing each at 90-100% full time. > > On the other hand, one could consider building a small MPI-connected > cluster for about the same amount of money. > > When IBM announced a couple of years ago the incredible performance > details about the Cell processor that would go into Playstation 3, I > thought, ?Wow, maybe the future of computational physics rests with > gamers.? I?m sure most of you know this is actually beginning to > happen, spurred on by the fact that the PS3 is ?open architecture? and > can run a Linux distro. Moreover, there apparently are ?open software? > numerical analysis tools available from IBM. At least four US > universities are experimenting with off-the-shelf PS3 clusters, perhaps > one of the more interesting is at UMass, > _http://gravity.phy.umassd.edu/ps3.html_. In the last week or so, Sony > lowered the entry level price of the PS3 to 400 USD. So, a cluster of > four with a cheap switch could be purchased for about the same price at > the single motherboard configuration I mentioned above. > > My teenager, a gamer, tells me the PS3 has problems. He says it?s > unreliable and overheats and only has 256 MB RAM on board (he owns a > Wii, which outsells the PS3 in the US by a factor of three). > > Has anybody tried porting PWscf to a PS3? > > Any and all advice is welcome. > -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From cyrille.barreteau at cea.fr Tue Nov 20 17:14:28 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Tue, 20 Nov 2007 17:14:28 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <474301C5.1050502@mit.edu> References: <4742E855.2030909@cea.fr> <474301C5.1050502@mit.edu> Message-ID: <474307E4.8060906@cea.fr> Hi Nicola Of course my (iron) cluster is magnetic and my calculation is spin polarized. I also did play with the various smearings (fd,mp,mv) and the broadening but the result is quite insensitive to this parameter. I really have the feeling that for some reasons as Matteo mentionned davidson with 20 bands (who knows why...) gets trapped in some local minimum just for numerical reasons... But this trap seems quite stable since it remains even when changing the smearing. It was just bad (or good) luck. cyrille Nicola Marzari wrote: >Hi Cyrille, > >your tm dimer will likely require spin polarization, and >will have several scf solutions that correspond to different >S_z (that you can fix in several ways, either specifying the >number of spin up and spin down, or using a fermi_up and >fermi_down energy, to specify n_up - n_down but still treating >the system with a smearing and fractional occupations). > >Even for a given S_z, there will be different scf solutions >that have different spatial symmetry. > >So, in your case I would definitely switch spin-polarization, >and maybe play around with small and large smearings, to see what >happens, and look at the states, and the occupations. Anyhow, if >you get to a scf solution, that is a "good" solution for the GGA >problem, although might not be very physical. > >Also, have a look at Kulik et al PRL 2006. > > nicola > > >Cyrille Barreteau wrote: > > >>Dear pwscf_ers >> >>I am doing a very simple calculation on a transition metal >>dimer and I have a encoutered a strange problem (I love >>strange problems:-) >> >>I have done a first calculation with the default value >>of nbnd, ie nbnd=nelec/2*20% (=12 in my case) >> >>But since it is known that in clusters it is often good >>to increase nbnd I have performed another calculation with >>larger nbnd (=20). The result is quite different from >>the one at nbnd=12. >> >>I have then increased again nbnd up to 25 and then I recover >>the result obtained for nbnd=12 >> >>More problematic is the local density of states (I know >>it is not really a dos but a bunch of dirac peaks). >>In a dimer with the z axis along the direction connecting >>the two atoms, the xy and x^2-y^2 dos should be degenerate >>(if the supercell box is large enough). >>In fact the xy and x^2-y^2 are degenerate if nbnd=12 or 25 >>but there is a rather large splitting if nbnd=20. >> >>I am quite sure the result for nbnd=20 is not correct but >>what is the origin of this problem? >> >>Maybe I could try to use another diagonalization scheme.. >> >> thanks for reading my strange problems >> >> cyrille >> >> >> >> > > > > -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== From giannozz at nest.sns.it Tue Nov 20 17:37:20 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 20 Nov 2007 17:37:20 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <4742F555.9070201@cea.fr> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> Message-ID: <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> On Nov 20, 2007, at 15:55 , Cyrille Barreteau wrote: > Does anyone has experience about the behaviour > of the various diagonalization schemes? > Which one is better, in which case? the so-called conjugate-gradient diagonalization basically never fails; it uses less memory but it is much slower than the block Davidson algorithm. The latter is also quite robust but there is a tiny probability to end up in the wrong ground state. The only such case I know happens in some high-symmetry perovskites, if you start from superposition of atomic states, with occupied states only, and if the phase of the moon is close to 42. The problem disappears if you add a few more states, or if you start from random wavefunctions. I am quite surprised to hear that there is a case in which you get good results with 12 states and bad with 20. I wouldn't be surprised by the opposite. Could you please post one jobs showing the problem? Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From w2agz at pacbell.net Tue Nov 20 18:02:47 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Tue, 20 Nov 2007 09:02:47 -0800 Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation In-Reply-To: <200711201501.37502.shaposh@isp.nsc.ru> References: <007a01c82b22$7bc79920$7356cb60$@net> <200711201501.37502.shaposh@isp.nsc.ru> Message-ID: <00df01c82b97$2ddc5690$899503b0$@net> Thanks for the feedback. The two dual Xeon boards I have were originally chosen to host my website. I long suspected that there was "thrashing" between the two processors when performing numerically intensive calculations. I assume the same would occur within the cores of a single processor as well? However, my understanding was this approach was supposed to better implement matrix manipulation for gaming applications. If so, one would assume improved performance for the fft operations at the "core" of pw.x. Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford University EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com ? ? -----Original Message----- From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Alexander Shaposhnikov Sent: Tuesday, November 20, 2007 1:02 AM To: PWSCF Forum Subject: Re: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation Actually newer Woodcrest/Clovertown Xeon platform have dual independent FSBs, but this does not help much. Memory bandwidth available on each FSB is very limited - between 1/2 and 1/3 of what is available to modern desktop CPUs. There is no good "server" computational platform today, unfortunately . AMD's CPUs have weak FPU and Intel's have weak memory subsystem. Intel- based desktop single-socket platform is the best choice for now. On Tuesday 20 November 2007 14:33, Paolo Giannozzi wrote: > On Nov 20, 2007, at 4:07 , Paul M. Grant wrote: > > Currently, I use two machines with dual Xeon processors > > dual Xeon? very bad choice. They have a single memory > bus, so if you run two processes on the same motherboard, > they spend most of the time waiting for access to memory. > > P. > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From w2agz at pacbell.net Wed Nov 21 04:24:26 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Tue, 20 Nov 2007 19:24:26 -0800 Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation In-Reply-To: <474304F1.4030203@mit.edu> References: <007a01c82b22$7bc79920$7356cb60$@net> <474304F1.4030203@mit.edu> Message-ID: <011d01c82bee$0522b2b0$0f681810$@net> Hi Nicola, Thanks for your feedback. I was aware of your page on Quasiamore and actually have visited it in the past, and should have had a look before I sent in my forum posting. Re the PS3, I told my teenage gamer today I'm going to get a unit to experiment with (I've already begun downloading Yellow Dog Linux). He was very happy, because if I fail porting PWscf, he gets it! Over the last couple of days, my research reflects exactly what you mentioned on the detailed performance of the switch. Also, there may be some issues regarding Fortran compilers, but I've got some friends at IBM that could help with this. What concerns me most is the 256 MB on board memory limit, and I've found no hints online on how to hack this, soldering iron or otherwise. Ciao, -Paul Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford University EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com ? ? -----Original Message----- From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Nicola Marzari Sent: Tuesday, November 20, 2007 8:02 AM To: PWSCF Forum Subject: Re: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation Hi Paul, let us know if you get PWSCF on a PS3 ! I generally echo the other suggestions - the speed of the memory bus is critical, having fast memory (matched with the bus) is good, having the fastest clock on current Intel chip is good, having two cores is a 25-50% improvement, and having 4 cores is a negligible improvement on 2. If you plan to link blades in a cluster, it's vital to have fast, low latency communication - if you use gigabit, you might get good parallelism with as many as 8 blades, or as little as 2, depending on the quality of your switch, using a very recent mpi implementation, and the quality of your ethernet controllers on board. We have some tests on different platforms on http://quasiamore.mit.edu/pmwiki/index.php?n=Main.CP90Timings nicola Paul M. Grant wrote: > To All Forum Members: > > I?m planning on building a new Linux box (or boxes) to explore highly > correlated systems (e-p coupling plus LDA+U), and am seeking the > collective experience and advice of the PWscf community on a suitable, > inexpensive (< 2000 USD, MB+CPUs+RAM, exclusive of power supplies, > enclosures, and accessories) hardware platform. I emphasize that the > principal purpose of this new box would be exploratory, or for > development, not production. > > I?ve built several past platforms, both Windows and Linux based, using > server boards manufactured by Supermicro and have had generally good > experience and service (the factory is only 15 miles from where I > live). Currently, I use two machines with dual Xeon processors (single > core, 32 bits), one with 1 GHz cpus, 1 GB RAM, the other 2.4 GHz and > 3.25 GB RAM, both with bus speeds of 133 MHz, the newest 3 years old. > However, occasionally I run PWscf exercises on my little Thinkpad X41 > tablet (single processor, 1.5 GHz, 1.5 GB), and the scf computation will > run 3-5 times faster than on the other machines! I suspect this rather > surprising result is because the Thinkpad has a 400 MHz bus clock speed. > > One option I?m considering is using a ?gaming? or server class > motherboard with dual 2.33 GHz quad-core 64-bit processors, a 1333 MHz > FSB, and 16 GB RAM. Having said this, I?m not sure PWscf (and the > Fortran compilers available) can handle all this parallelism efficiently > on a single motherboard. I?ve noticed when running pw.x, the CPU > activity ?flips? between processors every several seconds, instead of > sharing each at 90-100% full time. > > On the other hand, one could consider building a small MPI-connected > cluster for about the same amount of money. > > When IBM announced a couple of years ago the incredible performance > details about the Cell processor that would go into Playstation 3, I > thought, ?Wow, maybe the future of computational physics rests with > gamers.? I?m sure most of you know this is actually beginning to > happen, spurred on by the fact that the PS3 is ?open architecture? and > can run a Linux distro. Moreover, there apparently are ?open software? > numerical analysis tools available from IBM. At least four US > universities are experimenting with off-the-shelf PS3 clusters, perhaps > one of the more interesting is at UMass, > _http://gravity.phy.umassd.edu/ps3.html_. In the last week or so, Sony > lowered the entry level price of the PS3 to 400 USD. So, a cluster of > four with a cheap switch could be purchased for about the same price at > the single motherboard configuration I mentioned above. > > My teenager, a gamer, tells me the PS3 has problems. He says it?s > unreliable and overheats and only has 256 MB RAM on board (he owns a > Wii, which outsells the PS3 in the US by a factor of three). > > Has anybody tried porting PWscf to a PS3? > > Any and all advice is welcome. > -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From thonhauser at MIT.EDU Wed Nov 21 05:04:09 2007 From: thonhauser at MIT.EDU (Timo Thonhauser) Date: Tue, 20 Nov 2007 23:04:09 -0500 Subject: [Pw_forum] time-reversal symmetry broken Message-ID: Dear All, I have a question concerning the breaking of time-reversal symmetry within PWscf. For my proposes I have modified the code and included a (periodic) vector potential into the Hamiltonian. The code gives very good results for the quantity that I am interested in for molecules, i.e. gamma-point calculations. However, when I try to do calculations for periodic solids, the results worsen. I think the code (unless you do spin-orbit coupling) assumes time- reversal symmetry. In principle, including a vector potential breaks time-reversal symmetry. But, in the cases of my gamma-point calculations that did not cause any problems ( psi_(-k)=psi_(k)* for k=gamma ) and hence, I get good results for molecules. However, maybe this is the problem with my calculations for extended systems: including the vector potential breaks time-reversal symmetry, and in general psi_(-k) is not psi_(k)*, but the code is not "aware" that the symmetry has been broken. Thus, my question is: is there a simple way to tell the code that time-reversal symmetry is broken without getting into spin-orbit stuff? Thanks, Timo Thonhauser, MIT ______________________________________________________________________ Dr. TIMO THONHAUSER Department of Materials Science and Engineering (617) 253 6026 MIT, 13-4084, 77 Massachusetts Avenue thonhauser at mit.edu Cambridge, MA 02139-4307, USA www.mit.edu/~thonhaus ______________________________________________________________________ From nss at isp.nsc.ru Wed Nov 21 05:38:05 2007 From: nss at isp.nsc.ru (Nekrashevich Sergey) Date: Wed, 21 Nov 2007 10:38:05 +0600 Subject: [Pw_forum] Problem with postprocessing of the vc-cp Message-ID: <4743B62D.5060007@isp.nsc.ru> Dear all, I've found a bit strange behavior of cppp.x tool while dealing with vc-cp dynamics of 167-atomic supercell of silicon nitride. I perfomed the 500 steps of variable-cell Car-Parrinello molecular dynamics and then ran it through cppp.x to have an AXSF-file as output. After opening the animation in the XCRYSDEN I can see, that during some number of steps atoms are moving properly from their initial positions (which were achieved after ionic relaxation). But then at some certain time step atoms revert to their original positions specified in the input file for cp.x. After that atoms are starting to move continuously again, but after some steps the leap is reapeted. It's obvious that such movement does not correspond to real behavior of the system, at least because in the cp.x output file there are no leaps in atomic positions (I've checked it by putting 'iprint = 1'). Any help is greatly appreciated. Best regards, Nekrashevich Sergey. From nazari at iasbs.ac.ir Wed Nov 21 10:12:47 2007 From: nazari at iasbs.ac.ir (nazari at iasbs.ac.ir) Date: Wed, 21 Nov 2007 09:12:47 -0000 (UTC) Subject: [Pw_forum] Pseudo_Generation/unknown pseudo error In-Reply-To: <18136.147.122.5.182.1195574359.squirrel@webmail.sissa.it> References: <8236.213.176.122.17.1195579547.squirrel@mail.iasbs.ac.ir> <4823.147.122.5.182.1195567975.squirrel@webmail.sissa.it> <8946.213.176.122.17.1195583271.squirrel@mail.iasbs.ac.ir> <18136.147.122.5.182.1195574359.squirrel@webmail.sissa.it> Message-ID: <10846.213.176.122.17.1195636367.squirrel@mail.iasbs.ac.ir> Dear Lorenzo Paulatto, You are wright. I have also produced the potential before sending a question. But when I wnat to do optimization by using this potential I face with the following error: from read_pseudo_header : error # 1 unknown pseudo type Regards Fariba Nazari IASBS From mbaris at metu.edu.tr Wed Nov 21 07:40:20 2007 From: mbaris at metu.edu.tr (mbaris at metu.edu.tr) Date: Wed, 21 Nov 2007 08:40:20 +0200 Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation Message-ID: <20071121084020.24548yyiy25ccwko@webmail2.metu.edu.tr> Hi Paul, Some pointers about the PS3: instead of yellow dog, opensuse 10.3ppc seems to be better implemented (at least I am more used to it), and has a set of cell development tools coming right out of the package. yes the amount of ram is a big problem, I have tried implementing a swap on a very fast flash memory card, at least to lighten the problem, but once you format the card, the performance drops significantly. I will be playing with it as soon as I have time. Since this is a console, it is -very- hard to tinker with the ram, it is a bga chip soldered directly on the second level of mainboard pcb, you need a wave soldering technique to remove&replace it. In fact, there is a doable method if you are feeling lucky, by using a heat gun you can remove the chip, and solder back again with applying flux & solder to the pads beforehand, assuming you are -very- good with the heat gun and do not fry the chips right away, there are still three problems: 1)since this is a console, the communication between the cpu and the ram will most probably be carried on a trusted platform, to prevent copied games from running, thus changing the chip will tend to brick the system even if you are good with soldering. 2) where do you plan to buy the xddr3 chip? Vandalize from a graphics card perhaps? But then again you need to find the exact amount PS3 uses... 3) How do you plan the architecture to recognize the ram you have installed? XDDR3 is a multi chip platform, and most probably you should replace also the controller. as with the programming, you should be aware that cell programming is not equal to usual parallel programming, in fact the idea is in some sense reversed, in the cell, the inter-communication is cheap but the ram available to each node is very low (256 kB) and the processors themselves are rather low performance. So you should divide your program to small&and easy blocks as much as possible, otherwise you get a terrible performance. I nowadays plan to experiment with the new fftw3.2alpha with cell support to see if there is a difference. well, if you are looking for some alternatives, what about nvidia tesla? It is basically a high end fpu& generic development tools added to your pc via a pci-x slot. I am very optimistic about this technology, looking at the amount of work a gpu can pull of nowadays. The FPU is not very good at generic calculations, but luckily FFT is one of the strong points. As soon as I have the money&time I will try implementing it. Best, O. Baris Malcioglu METU Physics. PhD. Candidate Ankara From paulatto at sissa.it Wed Nov 21 08:52:31 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Wed, 21 Nov 2007 08:52:31 +0100 (CET) Subject: [Pw_forum] Pseudo_Generation/unknown pseudo error In-Reply-To: <10846.213.176.122.17.1195636367.squirrel@mail.iasbs.ac.ir> References: <8236.213.176.122.17.1195579547.squirrel@mail.iasbs.ac.ir> <4823.147.122.5.182.1195567975.squirrel@webmail.sissa.it> <8946.213.176.122.17.1195583271.squirrel@mail.iasbs.ac.ir> <18136.147.122.5.182.1195574359.squirrel@webmail.sissa.it> <10846.213.176.122.17.1195636367.squirrel@mail.iasbs.ac.ir> Message-ID: <21423.82.48.222.219.1195631551.squirrel@webmail.sissa.it> On Wed, November 21, 2007 10:12, nazari at iasbs.ac.ir wrote: > Dear Lorenzo Paulatto, > > You are wright. I have also produced the potential before sending a > question. But when I wnat to do optimization by using this potential > I face with the following error: > > from read_pseudo_header : error # 1 > unknown pseudo type Ok, I will look at this. Anyway add .UPF at the end of the file_pseudo ps parameters, as have stated in my previous message; this will force ld1 to use a different format for the file that should work around the problem. -- Lorenzo Paulatto +39 040 3787 511 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From cyrille.barreteau at cea.fr Wed Nov 21 08:54:40 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Wed, 21 Nov 2007 08:54:40 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> Message-ID: <4743E440.1080508@cea.fr> Dear Paolo, Maybe I found the second exception due to another very special phase of the moon:-) What do I win in this case:-) You will find at the following links: the input file, the pseudo file and the three output corresponding to nbnd=12,20,25 respectively. http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer.scf.in http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe.pw91.vdb.UPF http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd12.scf.out http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd20.scf.out http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd25.scf.out thanks for having a look at this case. cyrille - ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== Paolo Giannozzi wrote: >On Nov 20, 2007, at 15:55 , Cyrille Barreteau wrote: > > > >>Does anyone has experience about the behaviour >>of the various diagonalization schemes? >>Which one is better, in which case? >> >> > >the so-called conjugate-gradient diagonalization basically >never fails; it uses less memory but it is much slower than >the block Davidson algorithm. The latter is also quite >robust but there is a tiny probability to end up in the wrong >ground state. The only such case I know happens in some >high-symmetry perovskites, if you start from superposition >of atomic states, with occupied states only, and if the phase >of the moon is close to 42. The problem disappears if you >add a few more states, or if you start from random wavefunctions. >I am quite surprised to hear that there is a case in which >you get good results with 12 states and bad with 20. >I wouldn't be surprised by the opposite. Could you please >post one jobs showing the problem? > >Paolo >--- >Paolo Giannozzi, Dept of Physics, University of Udine >via delle Scienze 208, 33100 Udine, Italy >Phone +39-0432-558216, fax +39-0432-558222 > > > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > - From degironc at sissa.it Wed Nov 21 10:39:30 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 21 Nov 2007 10:39:30 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <4743E440.1080508@cea.fr> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> Message-ID: <4743FCD2.604@sissa.it> methfessel-paxton smearing is dangerous (because it is not positive-defnite) when k-point sampling is not sufficient. When dealing with a "fake metal" as your molecule where the Fermi energy should fall in gap of the spectrum I think it is more advisable to use simple gaussian or marzari-vanderbilt smearing. The spurious negative weights of m-p smearing the occupation of levels just above the Fermi energy can lock some spourios state that shouldn't be there ... If you have many kpoints this (in my experience... Nicola has a different opinion) is never a problem. In any case the smearing contribution for a molecule should be zero ... which is not the case in your output ... particularly so for the problematic cases with 12 or 25 bands.. TEST> grep -e TS -e ! Fe*.out Fe_dimer_gamma_nbnd12.scf.out:! total energy = -121.76376536 Ry Fe_dimer_gamma_nbnd12.scf.out: smearing contrib. (-TS) = -0.00157785 Ry Fe_dimer_gamma_nbnd20.scf.out:! total energy = -121.76907297 Ry Fe_dimer_gamma_nbnd20.scf.out: smearing contrib. (-TS) = 0.00048313 Ry Fe_dimer_gamma_nbnd25.scf.out:! total energy = -121.76376546 Ry Fe_dimer_gamma_nbnd25.scf.out: smearing contrib. (-TS) = -0.00157856 Ry Does the problem (at least the one with 25 bands) disappears if you use "m-v" of "gaussian" smearing ? I would expect so... otherwise there is still something else. hope it helps, stefano Cyrille Barreteau wrote: > Dear Paolo, > > Maybe I found the second exception due to another very special > phase of the moon:-) What do I win in this case:-) > > You will find at the following links: > the input file, the pseudo file and the three output > corresponding to nbnd=12,20,25 respectively. > > http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer.scf.in > http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe.pw91.vdb.UPF > http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd12.scf.out > http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd20.scf.out > http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd25.scf.out > > thanks for having a look at this case. > > cyrille > > - > ================================================================== > Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 > CEA Saclay | fax : +33 (0)1 69 08 84 46 > DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr > Batiment 462 | > 91191 Gif sur Yvette Cedex FRANCE > ~~~~~~~~~~~~~~~~~~~~~~~~ > http://www-drecam.cea.fr/spcsi/index.php > http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html > http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php > ================================================================== > > > > > Paolo Giannozzi wrote: > > >> On Nov 20, 2007, at 15:55 , Cyrille Barreteau wrote: >> >> >> >> >>> Does anyone has experience about the behaviour >>> of the various diagonalization schemes? >>> Which one is better, in which case? >>> >>> >>> >> the so-called conjugate-gradient diagonalization basically >> never fails; it uses less memory but it is much slower than >> the block Davidson algorithm. The latter is also quite >> robust but there is a tiny probability to end up in the wrong >> ground state. The only such case I know happens in some >> high-symmetry perovskites, if you start from superposition >> of atomic states, with occupied states only, and if the phase >> of the moon is close to 42. The problem disappears if you >> add a few more states, or if you start from random wavefunctions. >> I am quite surprised to hear that there is a case in which >> you get good results with 12 states and bad with 20. >> I wouldn't be surprised by the opposite. Could you please >> post one jobs showing the problem? >> >> Paolo >> --- >> Paolo Giannozzi, Dept of Physics, University of Udine >> via delle Scienze 208, 33100 Udine, Italy >> Phone +39-0432-558216, fax +39-0432-558222 >> >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> > > > - > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From cyrille.barreteau at cea.fr Wed Nov 21 10:55:43 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Wed, 21 Nov 2007 10:55:43 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <4743FCD2.604@sissa.it> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> Message-ID: <4744009F.7030904@cea.fr> Stefano I did try with simple fermi-dirac (and very small broadening) and got the same result...The problematic case is nbnd=20 and not ndnd=12 or 25 cyrille Stefano de Gironcoli wrote: >methfessel-paxton smearing is dangerous (because it is not >positive-defnite) when k-point sampling is not sufficient. > >When dealing with a "fake metal" as your molecule where the Fermi energy >should fall in gap of the spectrum I think it is more advisable to use >simple gaussian or marzari-vanderbilt smearing. The spurious negative >weights of m-p smearing the occupation of levels just above the Fermi >energy can lock some spourios state that shouldn't be there ... If you >have many kpoints this (in my experience... Nicola has a different >opinion) is never a problem. > >In any case the smearing contribution for a molecule should be zero >... which is not the case in your output ... particularly so for the >problematic cases with 12 or 25 bands.. > >TEST> grep -e TS -e ! Fe*.out >Fe_dimer_gamma_nbnd12.scf.out:! total energy = >-121.76376536 Ry >Fe_dimer_gamma_nbnd12.scf.out: smearing contrib. (-TS) = >-0.00157785 Ry >Fe_dimer_gamma_nbnd20.scf.out:! total energy = >-121.76907297 Ry >Fe_dimer_gamma_nbnd20.scf.out: smearing contrib. (-TS) = >0.00048313 Ry >Fe_dimer_gamma_nbnd25.scf.out:! total energy = >-121.76376546 Ry >Fe_dimer_gamma_nbnd25.scf.out: smearing contrib. (-TS) = >-0.00157856 Ry > >Does the problem (at least the one with 25 bands) disappears if you use >"m-v" of "gaussian" smearing ? >I would expect so... otherwise there is still something else. >hope it helps, >stefano > >Cyrille Barreteau wrote: > > >>Dear Paolo, >> >>Maybe I found the second exception due to another very special >>phase of the moon:-) What do I win in this case:-) >> >>You will find at the following links: >>the input file, the pseudo file and the three output >>corresponding to nbnd=12,20,25 respectively. >> >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer.scf.in >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe.pw91.vdb.UPF >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd12.scf.out >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd20.scf.out >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd25.scf.out >> >>thanks for having a look at this case. >> >> cyrille >> >>- >>================================================================== >>Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 >>CEA Saclay | fax : +33 (0)1 69 08 84 46 >>DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr >>Batiment 462 | >>91191 Gif sur Yvette Cedex FRANCE >> ~~~~~~~~~~~~~~~~~~~~~~~~ >>http://www-drecam.cea.fr/spcsi/index.php >>http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html >>http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php >>================================================================== >> >> >> >> >>Paolo Giannozzi wrote: >> >> >> >> >>>On Nov 20, 2007, at 15:55 , Cyrille Barreteau wrote: >>> >>> >>> >>> >>> >>> >>>>Does anyone has experience about the behaviour >>>>of the various diagonalization schemes? >>>>Which one is better, in which case? >>>> >>>> >>>> >>>> >>>> >>>the so-called conjugate-gradient diagonalization basically >>>never fails; it uses less memory but it is much slower than >>>the block Davidson algorithm. The latter is also quite >>>robust but there is a tiny probability to end up in the wrong >>>ground state. The only such case I know happens in some >>>high-symmetry perovskites, if you start from superposition >>>of atomic states, with occupied states only, and if the phase >>>of the moon is close to 42. The problem disappears if you >>>add a few more states, or if you start from random wavefunctions. >>>I am quite surprised to hear that there is a case in which >>>you get good results with 12 states and bad with 20. >>>I wouldn't be surprised by the opposite. Could you please >>>post one jobs showing the problem? >>> >>>Paolo >>>--- >>>Paolo Giannozzi, Dept of Physics, University of Udine >>>via delle Scienze 208, 33100 Udine, Italy >>>Phone +39-0432-558216, fax +39-0432-558222 >>> >>> >>> >>>_______________________________________________ >>>Pw_forum mailing list >>>Pw_forum at pwscf.org >>>http://www.democritos.it/mailman/listinfo/pw_forum >>> >>> >>> >>> >>> >>- >> >>_______________________________________________ >>Pw_forum mailing list >>Pw_forum at pwscf.org >>http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== From baroni at sissa.it Wed Nov 21 11:30:22 2007 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 21 Nov 2007 10:30:22 +0000 Subject: [Pw_forum] time-reversal symmetry broken In-Reply-To: References: Message-ID: <9D6E1F6F-C235-4D3F-B708-DC27F5A74682@sissa.it> Timo: I do not have a solution handy, and others may have deeper ideas. I think the problem is that when simmetrizing the charge density the code always assumes e(k)=e(-k) and psi(-k)=psi(k)*, and so it never uses more than 1/2 the BZ. One thing you may try is to pass explicitly all the k-points, and wish that the code is not so smart as to remove 1/2 of them ... Otherwise, I do not see any other options than hacking the code (but, again, others may have smarter ideas). SB On Nov 21, 2007, at 4:04 AM, Timo Thonhauser wrote: > Dear All, > > I have a question concerning the breaking of time-reversal symmetry > within PWscf. For my proposes I have modified the code and included a > (periodic) vector potential into the Hamiltonian. The code gives very > good results for the quantity that I am interested in for molecules, > i.e. gamma-point calculations. However, when I try to do calculations > for periodic solids, the results worsen. > > I think the code (unless you do spin-orbit coupling) assumes time- > reversal symmetry. In principle, including a vector potential breaks > time-reversal symmetry. But, in the cases of my gamma-point > calculations > that did not cause any problems ( psi_(-k)=psi_(k)* for k=gamma ) and > hence, I get good results for molecules. However, maybe this is the > problem with my calculations for extended systems: including the > vector potential breaks time-reversal symmetry, and > in general psi_(-k) is not psi_(k)*, but the code is not "aware" > that the symmetry has been broken. > > Thus, my question is: is there a simple way to tell the code that > time-reversal symmetry is broken without getting into spin-orbit > stuff? > > Thanks, > Timo Thonhauser, MIT > > ______________________________________________________________________ > > Dr. TIMO THONHAUSER > Department of Materials Science and Engineering (617) 253 6026 > MIT, 13-4084, 77 Massachusetts Avenue thonhauser at mit.edu > Cambridge, MA 02139-4307, USA www.mit.edu/~thonhaus > ______________________________________________________________________ > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071121/b566c291/attachment-0001.htm From nazari at iasbs.ac.ir Wed Nov 21 16:28:07 2007 From: nazari at iasbs.ac.ir (nazari at iasbs.ac.ir) Date: Wed, 21 Nov 2007 15:28:07 -0000 (UTC) Subject: [Pw_forum] Pseudo_Generation/unknown pseudo error In-Reply-To: <21423.82.48.222.219.1195631551.squirrel@webmail.sissa.it> References: <8236.213.176.122.17.1195579547.squirrel@mail.iasbs.ac.ir> <4823.147.122.5.182.1195567975.squirrel@webmail.sissa.it> <8946.213.176.122.17.1195583271.squirrel@mail.iasbs.ac.ir> <18136.147.122.5.182.1195574359.squirrel@webmail.sissa.it> <10846.213.176.122.17.1195636367.squirrel@mail.iasbs.ac.ir> <21423.82.48.222.219.1195631551.squirrel@webmail.sissa.it> Message-ID: <17546.213.176.122.17.1195658887.squirrel@mail.iasbs.ac.ir> > > On Wed, November 21, 2007 10:12, nazari at iasbs.ac.ir wrote: >> Dear Lorenzo Paulatto, >> >> You are wright. I have also produced the potential before sending a >> question. But when I wnat to do optimization by using this potential >> I face with the following error: >> >> from read_pseudo_header : error # 1 >> unknown pseudo type > > Ok, I will look at this. Anyway add .UPF at the end of the file_pseudo ps > parameters, as have stated in my previous message; this will force ld1 to > use a different format for the file that should work around the problem. Thank you I add the .UPF and the program works. Regards Fariba Nazari IASBS > > -- > Lorenzo Paulatto > +39 040 3787 511 > http://people.sissa.it/~paulatto/ > > > ---------------------------------------------------------------- > SISSA Webmail https://webmail.sissa.it/ > Powered by SquirrelMail http://www.squirrelmail.org/ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From marzari at MIT.EDU Wed Nov 21 14:18:31 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 21 Nov 2007 08:18:31 -0500 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <4743FCD2.604@sissa.it> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> Message-ID: <47443027.20901@mit.edu> Stefano de Gironcoli wrote: > If you > have many kpoints this (in my experience... Nicola has a different > opinion) is never a problem. :-) . A proton falling on a metal surface. Antimatter, right away... nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From cyrille.barreteau at cea.fr Wed Nov 21 14:28:18 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Wed, 21 Nov 2007 14:28:18 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <4743FCD2.604@sissa.it> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> Message-ID: <47443272.7090008@cea.fr> Stefano, I did various checks on the smearing (mv,mp,fd..) and the broadening. For nbnd=20 I systematically find the same 'odd' result if diagonalization='david' and I recover a 'reasonable' result if diagonalization='cg' (similar to the one obtained with nbnd=12,25). Let me also mention that the "odd" result gives a lower energy than the "reasonable" one. cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html ================================================================== Stefano de Gironcoli wrote: >methfessel-paxton smearing is dangerous (because it is not >positive-defnite) when k-point sampling is not sufficient. > >When dealing with a "fake metal" as your molecule where the Fermi energy >should fall in gap of the spectrum I think it is more advisable to use >simple gaussian or marzari-vanderbilt smearing. The spurious negative >weights of m-p smearing the occupation of levels just above the Fermi >energy can lock some spourios state that shouldn't be there ... If you >have many kpoints this (in my experience... Nicola has a different >opinion) is never a problem. > >In any case the smearing contribution for a molecule should be zero >... which is not the case in your output ... particularly so for the >problematic cases with 12 or 25 bands.. > >TEST> grep -e TS -e ! Fe*.out >Fe_dimer_gamma_nbnd12.scf.out:! total energy = >-121.76376536 Ry >Fe_dimer_gamma_nbnd12.scf.out: smearing contrib. (-TS) = >-0.00157785 Ry >Fe_dimer_gamma_nbnd20.scf.out:! total energy = >-121.76907297 Ry >Fe_dimer_gamma_nbnd20.scf.out: smearing contrib. (-TS) = >0.00048313 Ry >Fe_dimer_gamma_nbnd25.scf.out:! total energy = >-121.76376546 Ry >Fe_dimer_gamma_nbnd25.scf.out: smearing contrib. (-TS) = >-0.00157856 Ry > >Does the problem (at least the one with 25 bands) disappears if you use >"m-v" of "gaussian" smearing ? >I would expect so... otherwise there is still something else. >hope it helps, >stefano > >Cyrille Barreteau wrote: > > >>Dear Paolo, >> >>Maybe I found the second exception due to another very special >>phase of the moon:-) What do I win in this case:-) >> >>You will find at the following links: >>the input file, the pseudo file and the three output >>corresponding to nbnd=12,20,25 respectively. >> >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer.scf.in >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe.pw91.vdb.UPF >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd12.scf.out >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd20.scf.out >>http://www-drecam.cea.fr/spcsi/cbarreteau/tempo/Fe_dimer_gamma_nbnd25.scf.out >> >>thanks for having a look at this case. >> >> cyrille >> >>- >>================================================================== >>Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 >>CEA Saclay | fax : +33 (0)1 69 08 84 46 >>DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr >>Batiment 462 | >>91191 Gif sur Yvette Cedex FRANCE >> ~~~~~~~~~~~~~~~~~~~~~~~~ >>http://www-drecam.cea.fr/spcsi/index.php >>http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html >>http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php >>================================================================== >> >> >> >> >>Paolo Giannozzi wrote: >> >> >> >> >>>On Nov 20, 2007, at 15:55 , Cyrille Barreteau wrote: >>> >>> >>> >>> >>> >>> >>>>Does anyone has experience about the behaviour >>>>of the various diagonalization schemes? >>>>Which one is better, in which case? >>>> >>>> >>>> >>>> >>>> >>>the so-called conjugate-gradient diagonalization basically >>>never fails; it uses less memory but it is much slower than >>>the block Davidson algorithm. The latter is also quite >>>robust but there is a tiny probability to end up in the wrong >>>ground state. The only such case I know happens in some >>>high-symmetry perovskites, if you start from superposition >>>of atomic states, with occupied states only, and if the phase >>>of the moon is close to 42. The problem disappears if you >>>add a few more states, or if you start from random wavefunctions. >>>I am quite surprised to hear that there is a case in which >>>you get good results with 12 states and bad with 20. >>>I wouldn't be surprised by the opposite. Could you please >>>post one jobs showing the problem? >>> >>>Paolo >>>--- >>>Paolo Giannozzi, Dept of Physics, University of Udine >>>via delle Scienze 208, 33100 Udine, Italy >>>Phone +39-0432-558216, fax +39-0432-558222 >>> >>> >>> >>>_______________________________________________ >>>Pw_forum mailing list >>>Pw_forum at pwscf.org >>>http://www.democritos.it/mailman/listinfo/pw_forum >>> >>> >>> >>> >>> >>- >> >>_______________________________________________ >>Pw_forum mailing list >>Pw_forum at pwscf.org >>http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From marzari at MIT.EDU Wed Nov 21 14:29:34 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 21 Nov 2007 08:29:34 -0500 Subject: [Pw_forum] time-reversal symmetry broken In-Reply-To: <9D6E1F6F-C235-4D3F-B708-DC27F5A74682@sissa.it> References: <9D6E1F6F-C235-4D3F-B708-DC27F5A74682@sissa.it> Message-ID: <474432BE.5010600@mit.edu> Stefano Baroni wrote: > Timo: I do not have a solution handy, and others may have deeper ideas. > I think the problem is that when simmetrizing the charge density the > code always assumes e(k)=e(-k) and psi(-k)=psi(k)*, and so it never uses > more than 1/2 the BZ. One thing you may try is to pass explicitly all > the k-points, and wish that the code is not so smart as to remove 1/2 of > them ... Otherwise, I do not see any other options than hacking the code > (but, again, others may have smarter ideas). SB Timo - what about the no_sym flag ? I thought that that would allow to force pwscf not to try to symmetrize the points (e.g. if one wanted to use a Baldereschi point, such as 1/4 1/4 1/4) - since it does not attempt to symmetrize anything, it might not use time reversal (rambling hypothesis - I'm sure SdG or PG know better, as always the case :-) ). As a side note - once k and -k are treated separately, even if the Hamiltonian is invariant for time reversal, in principle it's not psi(-k)=psi(k)* , but you might find more generally that psi(-k) = exp (i alpha) psi(k)* , where alpha is random. nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From degironc at sissa.it Wed Nov 21 14:57:52 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 21 Nov 2007 14:57:52 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <47443272.7090008@cea.fr> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> Message-ID: <47443960.7050201@sissa.it> Cyrille Barreteau wrote: > Stefano, > > Let me also mention that the "odd" result gives a > lower energy than the "reasonable" one. > this is why I thought that nbnd=20 was the correct one... however I fail to get it in a scalar run .... it is strange. is this strange behavior robust w.r.t. number of processors in your case ? stefano > cyrille > > From marzari at MIT.EDU Wed Nov 21 15:00:10 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 21 Nov 2007 09:00:10 -0500 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <47443272.7090008@cea.fr> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> Message-ID: <474439EA.5080009@mit.edu> Cyrille Barreteau wrote: > Stefano, > > I did various checks on the smearing (mv,mp,fd..) > and the broadening. For nbnd=20 I systematically > find the same 'odd' result if diagonalization='david' > and I recover a 'reasonable' result if diagonalization='cg' > (similar to the one obtained with nbnd=12,25). > Let me also mention that the "odd" result gives a > lower energy than the "reasonable" one. > > cyrille > Dear Cyrille, 1) I wouldn't focus on cg vs. david . They both converge to a scf state - the fact that it's different it's just related to the fact that both might be scf state, and one recipe takes a path that lead to one, and another to the other. 2) it looks like the difference might be driven by the fact that in the 12 or 25 bands calculation, you get 3 spin down states around -3.75 to -3.72 eV , close enough that they will have fractional occupations. In fact, -TS is more significant in the 12/25 bands calculation, than the 20 bands calculation 3) tightening the scf accuracy, and decreasing the smearing, in my opinion would ultimately lead the system to occupy fully the lower of the 3 states, and empty the others 4) note that the energy that pwscf writes is what I'd call E-TS, and for these systems you want to subtract -TS (i.e. add TS) to obtain E. Otherwise, if the system at zero T has fractional occupations (say because two states are exactly degenerate) and you increase T (and any other empty states are much higher) the energy that PWSCF writes as total energy does change, while in principle nothing is happening (i.e. only T is changeing in E-TS, while E and S do not change). 5) the case still remains - for the finite broadening, do you have really two very similar scf states ? My wilde guess that in reality the system wants to break symmetry, and occupy fully one of the three states, but is living on a very shallow potential energy surface, that become a tight less shallow if you decrease the smearing 6) this open up my favourite question - should B be spherical or cylindrical in DFT ? I.e. should we look like for occupation 1 0 0 for the 2p levels, or 1/3 1/3 1/3 ? 7) Last, but important - even using Gamma sampling, PWSCF finds for your geometry 16 symmetry operations, and thus heavily symmetrizes the system. That means that in 6) even 1 0 0 would become spherical. I.e. the symmetrization has a similar effect to temperature, in creating a charge density that is more symmetrical /degenerate occupations than it could be. I would be *Very curious* to see what happens if you were to put the dimer in a very random position, so the code misses most of the symmetry operations, and also you specify no_sym, so no symmetrization is performed. nicola -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From thonhauser at MIT.EDU Wed Nov 21 15:05:42 2007 From: thonhauser at MIT.EDU (Timo Thonhauser) Date: Wed, 21 Nov 2007 09:05:42 -0500 Subject: [Pw_forum] time-reversal symmetry broken In-Reply-To: <474432BE.5010600@mit.edu> References: <9D6E1F6F-C235-4D3F-B708-DC27F5A74682@sissa.it> <474432BE.5010600@mit.edu> Message-ID: <47443B36.1040808@mit.edu> Thanks Stefano, Nicola, I already use the no_sym flag, which prevents the charge density from being symmetrized with respect to all symmetry operations. However, even with the no_sym flag, only half the k-points are generated (e.g. for a 4x4x4 mesh the code generates 32 points and all the ones with negative k_x are missing). Here Stefano's idea of supplying a k-point list myself, would help. But, it is not that simple, is it? If inversion symmetry and time reversal are present, the wave functions can be chosen real (by real I mean that the wave functions only have a constant arbitrary phase factor). So, if I break time reversal and give the code my own, complete k-point list, does the code anywhere still assume that the wave functions are "real"? Maybe breaking inversion symmetry by hand could help? Timo Nicola Marzari wrote: > Stefano Baroni wrote: > >> Timo: I do not have a solution handy, and others may have deeper ideas. >> I think the problem is that when simmetrizing the charge density the >> code always assumes e(k)=e(-k) and psi(-k)=psi(k)*, and so it never uses >> more than 1/2 the BZ. One thing you may try is to pass explicitly all >> the k-points, and wish that the code is not so smart as to remove 1/2 of >> them ... Otherwise, I do not see any other options than hacking the code >> (but, again, others may have smarter ideas). SB >> > > Timo - what about the no_sym flag ? I thought that that would allow > to force pwscf not to try to symmetrize the points (e.g. if > one wanted to use a Baldereschi point, such as 1/4 1/4 1/4) - since > it does not attempt to symmetrize anything, it might not use > time reversal (rambling hypothesis - I'm sure SdG or PG know better, > as always the case :-) ). > > As a side note - once k and -k are treated separately, even > if the Hamiltonian is invariant for time reversal, > in principle it's not psi(-k)=psi(k)* , but you might > find more generally that psi(-k) = exp (i alpha) psi(k)* , > where alpha is random. > > > nicola > > > --------------------------------------------------------------------- > Prof Nicola Marzari Department of Materials Science and Engineering > 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA > tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- ______________________________________________________________________ Dr. TIMO THONHAUSER Department of Materials Science and Engineering (617) 253 6026 MIT, 13-4084, 77 Massachusetts Avenue thonhauser at mit.edu Cambridge, MA 02139-4307, USA www.mit.edu/~thonhaus ______________________________________________________________________ From degironc at sissa.it Wed Nov 21 15:09:59 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 21 Nov 2007 15:09:59 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <474439EA.5080009@mit.edu> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> <474439EA.5080009@mit.edu> Message-ID: <47443C37.6070700@sissa.it> Nicola Marzari wrote: > 6) this open up my favourite question - should B be spherical or > cylindrical in DFT ? I.e. should we look like for occupation 1 0 0 > for the 2p levels, or 1/3 1/3 1/3 ? > > 7) Last, but important - even using Gamma sampling, PWSCF finds > for your geometry 16 symmetry operations, and thus heavily > symmetrizes the system. That means that in 6) even 1 0 0 would > become spherical. I.e. the symmetrization has a similar effect > to temperature, in creating a charge density that is more symmetrical > /degenerate occupations than it could be. > > I would be *Very curious* to see what happens if you were to > put the dimer in a very random position, so the code misses most > of the symmetry operations, and also you specify no_sym, so no > symmetrization is performed. > nicola\ > > I agree with Nicola that this is very relevant and that symmetrization induces a degeneracy that could not be there in the best "broken symmetry" solution. This one can be achieved by 1) occupations='from_input' 2) nosym=.true. no smearing is required then. stefano From giannozz at nest.sns.it Wed Nov 21 15:14:40 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 21 Nov 2007 15:14:40 +0100 Subject: [Pw_forum] time-reversal symmetry broken In-Reply-To: <47443B36.1040808@mit.edu> References: <9D6E1F6F-C235-4D3F-B708-DC27F5A74682@sissa.it> <474432BE.5010600@mit.edu> <47443B36.1040808@mit.edu> Message-ID: <3C38B67F-213E-4CB8-8794-26BB56AA4864@nest.sns.it> On Nov 21, 2007, at 15:05 , Timo Thonhauser wrote: > If inversion symmetry and time reversal are present, the wave > functions > can be chosen real (by real I mean that the wave functions only have a > constant arbitrary phase factor). So, if I break time reversal and > give the > code my own, complete k-point list, does the code anywhere still > assume > that the wave functions are "real"? no, inversion symmetry is not used to force the wavefunctions to be real (by real I mean real). It should, it used to be once upon a time in some specialized and archaic version of PWscf, it would be easy to implement, but it isn't implemented right now. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From cyrille.barreteau at cea.fr Wed Nov 21 16:07:13 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Wed, 21 Nov 2007 16:07:13 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <47443C37.6070700@sissa.it> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> <474439EA.5080009@mit.edu> <47443C37.6070700@sissa.it> Message-ID: <474449A1.2010402@cea.fr> Stefano de Gironcoli wrote: >Nicola Marzari wrote: > > >>6) this open up my favourite question - should B be spherical or >>cylindrical in DFT ? I.e. should we look like for occupation 1 0 0 >>for the 2p levels, or 1/3 1/3 1/3 ? >> >>7) Last, but important - even using Gamma sampling, PWSCF finds >>for your geometry 16 symmetry operations, and thus heavily >>symmetrizes the system. That means that in 6) even 1 0 0 would >>become spherical. I.e. the symmetrization has a similar effect >>to temperature, in creating a charge density that is more symmetrical >>/degenerate occupations than it could be. >> >>I would be *Very curious* to see what happens if you were to >>put the dimer in a very random position, so the code misses most >>of the symmetry operations, and also you specify no_sym, so no >>symmetrization is performed. >> nicola\ >> >> >> >> > I agree with Nicola that this is very relevant and that symmetrization > induces > a degeneracy that could not be there in the best "broken symmetry" > solution. > This one can be achieved by 1) occupations='from_input' 2) nosym=.true. > no smearing is required then. > I agree that symmetrization (due to the periodization) induces a degeneracy but the "real" system that I want to simulate (a dimer alone in space) has even more symmetries. In particular it has a rotational symmetry that will never be there in a periodic code!! life is hard:-) Anyway I will try to put the dimer in a random position to see what happens. Systems with discrete levels are always a bit tricky and often present multiple solutions.. cyrille From marzari at MIT.EDU Wed Nov 21 16:17:46 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 21 Nov 2007 10:17:46 -0500 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <474449A1.2010402@cea.fr> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> <474439EA.5080009@mit.edu> <47443C37.6070700@sissa.it> <474449A1.2010402@cea.fr> Message-ID: <47444C1A.3070405@mit.edu> > I agree that symmetrization (due to the periodization) induces a > degeneracy but the > "real" system that I want to simulate (a dimer alone in space) has even more > symmetries. In particular it has a rotational symmetry that will never > be there Well, this is the part I've never understood - that's why I was mentioning the boron atom. I wouldn't necessarily agree that your dimer charge density should have the same symmetry of the ionic potential - as much as a dissociated H2+ molecular ion doesn't have mirror symmetry on the plan bisecting the bond, as you dissociate the molecule, or antiferromagnetic MnO has a lower electronic-structure symmetry, since the Mn are inequivalent from the point of view of spin density. Also, the potential in atoms is spherical, but I'm sure there are solutions lower in energy that are cylindrical. Think at a transition metal with only one d electron: do you fill up 1 orbital 1.0, 2 orbitals 0.5, 3 orbitals 0.33333, 5 orbitals 0.2 ? Only the last is spherical, all others are cylindrical, although of course an ensemble of measurements on atoms will always converge on the spherical limit. Am I the only one worrying about this ? It seems a key problem that I've never seen addressed - Englisch and Englisch in 1983 had a paper on the fact that fractional occupations are somewhat not compatible with v-representability, but I don't recall it very well. nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From cyrille.barreteau at cea.fr Wed Nov 21 16:52:01 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Wed, 21 Nov 2007 16:52:01 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <47444C1A.3070405@mit.edu> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> <474439EA.5080009@mit.edu> <47443C37.6070700@sissa.it> <474449A1.2010402@cea.fr> <47444C1A.3070405@mit.edu> Message-ID: <47445421.4000103@cea.fr> Well this is an interesting discussion which deviates from the original "technical" problem. If you want to obtain the "right" configuration of an atom standard LSDA is surely not the appropriate method and the strong degeneracy that wou will obtain will lead to "existential" questions such as "how to fill degenerate states". First you must include spin-orbit coupling which will induce a some splitting (and give you second Hund's rule). But if you want to get the third Hund's rule you also have to include the so-called orbital polarization (with B Racah parameter), that you can obtain in a full LSDA+U method including all the matrix elements of the Coulomb interaction At the very end I think you never have any degeneracy (ouf):-) Let us also not that this degeneracy problem is of course very relevant for cluster physics and often linked to Jahn-Teller distorsion (but maybe atoms would also like to distort and become elliptic:-) cyrille Nicola Marzari wrote: >>I agree that symmetrization (due to the periodization) induces a >>degeneracy but the >>"real" system that I want to simulate (a dimer alone in space) has even more >>symmetries. In particular it has a rotational symmetry that will never >>be there >> >> > > >Well, this is the part I've never understood - that's why I was >mentioning the boron atom. I wouldn't necessarily agree that your >dimer charge density should have the same symmetry of the ionic >potential - as much as a dissociated H2+ molecular ion doesn't have >mirror symmetry on the plan bisecting the bond, as you dissociate the >molecule, or antiferromagnetic MnO has a lower electronic-structure >symmetry, since the Mn are inequivalent from the point of view of spin >density. > >Also, the potential in atoms is spherical, but I'm sure there are >solutions lower in energy that are cylindrical. Think at a transition >metal with only one d electron: do you fill up 1 orbital 1.0, 2 orbitals >0.5, 3 orbitals 0.33333, 5 orbitals 0.2 ? Only the last is spherical, >all others are cylindrical, although of course an ensemble of >measurements on atoms will always converge on the spherical limit. > >Am I the only one worrying about this ? It seems a key problem >that I've never seen addressed - Englisch and Englisch in 1983 >had a paper on the fact that fractional occupations are somewhat >not compatible with v-representability, but I don't recall it very well. > > > nicola > > >--------------------------------------------------------------------- >Prof Nicola Marzari Department of Materials Science and Engineering >13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA >tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/spcsi/index.php http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html http://www-drecam.cea.fr/Phocea/Membres/Cours/index.php ================================================================== From degironc at sissa.it Wed Nov 21 16:58:56 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 21 Nov 2007 16:58:56 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <47443C37.6070700@sissa.it> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> <474439EA.5080009@mit.edu> <47443C37.6070700@sissa.it> Message-ID: <474455C0.7050107@sissa.it> Stefano de Gironcoli wrote: > Nicola Marzari wrote: > >> 6) this open up my favourite question - should B be spherical or >> cylindrical in DFT ? I.e. should we look like for occupation 1 0 0 >> for the 2p levels, or 1/3 1/3 1/3 ? >> >> 7) Last, but important - even using Gamma sampling, PWSCF finds >> for your geometry 16 symmetry operations, and thus heavily >> symmetrizes the system. That means that in 6) even 1 0 0 would >> become spherical. I.e. the symmetrization has a similar effect >> to temperature, in creating a charge density that is more symmetrical >> /degenerate occupations than it could be. >> >> I would be *Very curious* to see what happens if you were to >> put the dimer in a very random position, so the code misses most >> of the symmetry operations, and also you specify no_sym, so no >> symmetrization is performed. >> nicola\ >> >> >> > I agree with Nicola that this is very relevant and that symmetrization > induces > a degeneracy that could not be there in the best "broken symmetry" solution. > This one can be achieved by 1) occupations='from_input' 2) nosym=.true. > no smearing is required then. > > indeed with this input I get ! total energy = -121.76929092 Ry Harris-Foulkes estimate = -121.76929090 Ry estimated scf accuracy < 1.9E-09 Ry which I think is the correct result... stefano > stefano > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From degironc at sissa.it Wed Nov 21 17:08:24 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 21 Nov 2007 17:08:24 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <47445421.4000103@cea.fr> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> <474439EA.5080009@mit.edu> <47443C37.6070700@sissa.it> <474449A1.2010402@cea.fr> <47444C1A.3070405@mit.edu> <47445421.4000103@cea.fr> Message-ID: <474457F8.3080603@sissa.it> I think that given a DFT functional form the GS is, by definition, the solution that gives the lowest energy and that in those cases where the "true" functional would lead to degeneracy (open shell atoms to name a typical system) the approximate DFT's we all use will break the symmetry and produce a cylindrical (or even lower) symmetrical solution (for Jahn-Teller related reasons). This is the case for C and O atoms for sure. stefano Cyrille Barreteau wrote: > Well this is an interesting discussion which deviates from the original > "technical" problem. > > If you want to obtain the "right" configuration of an atom standard > LSDA is surely not the appropriate method and the strong degeneracy > that wou will obtain will lead to "existential" questions such as > "how to fill degenerate states". > > First you must include spin-orbit coupling which will induce a > some splitting (and give you second Hund's rule). > But if you want to get the third Hund's rule you also have > to include the so-called orbital polarization (with B Racah > parameter), that you can obtain in a full LSDA+U method > including all the matrix elements of the Coulomb interaction > > At the very end I think you never have any degeneracy (ouf):-) > > Let us also not that this degeneracy problem is of course very > relevant for cluster physics and often linked to Jahn-Teller > distorsion (but maybe atoms would also like to distort and > become elliptic:-) > > cyrille > > > > > Nicola Marzari wrote: > > >>> I agree that symmetrization (due to the periodization) induces a >>> degeneracy but the >>> "real" system that I want to simulate (a dimer alone in space) has even more >>> symmetries. In particular it has a rotational symmetry that will never >>> be there >>> >>> >>> >> Well, this is the part I've never understood - that's why I was >> mentioning the boron atom. I wouldn't necessarily agree that your >> dimer charge density should have the same symmetry of the ionic >> potential - as much as a dissociated H2+ molecular ion doesn't have >> mirror symmetry on the plan bisecting the bond, as you dissociate the >> molecule, or antiferromagnetic MnO has a lower electronic-structure >> symmetry, since the Mn are inequivalent from the point of view of spin >> density. >> >> Also, the potential in atoms is spherical, but I'm sure there are >> solutions lower in energy that are cylindrical. Think at a transition >> metal with only one d electron: do you fill up 1 orbital 1.0, 2 orbitals >> 0.5, 3 orbitals 0.33333, 5 orbitals 0.2 ? Only the last is spherical, >> all others are cylindrical, although of course an ensemble of >> measurements on atoms will always converge on the spherical limit. >> >> Am I the only one worrying about this ? It seems a key problem >> that I've never seen addressed - Englisch and Englisch in 1983 >> had a paper on the fact that fractional occupations are somewhat >> not compatible with v-representability, but I don't recall it very well. >> >> >> nicola >> >> >> --------------------------------------------------------------------- >> Prof Nicola Marzari Department of Materials Science and Engineering >> 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA >> tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> > > > From matteo at umn.edu Wed Nov 21 17:13:43 2007 From: matteo at umn.edu (Matteo Cococcioni) Date: Wed, 21 Nov 2007 10:13:43 -0600 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <47444C1A.3070405@mit.edu> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> <474439EA.5080009@mit.edu> <47443C37.6070700@sissa.it> <474449A1.2010402@cea.fr> <47444C1A.3070405@mit.edu> Message-ID: <47445937.8010601@umn.edu> Hi Nicola, I would like to express my opinion on this very interesting topic that you mentioned. These are just my very arguable ideas or what I think I have understood from a little thinking and studying of this problem. Let's start from B and H2+ or 1 e- TM atoms with only 1 d state occupied (e.g. Sc or Y): I think that the "real" ground state is cylindrical. So for example in the H2+ molecule at large interionic distance one should get the localization of the e- on one of the protons. The symmetry of the problem (and of the potential) is re-established on a statistical level because the physical situation can be described by a statistical mixture of states with 1 e- on one proton or the other one that have the same probability. So in the case of B it would be a misture of px, py, pz with weights of 0.333 each (BTW this case is a bit like the Scroedinger cat because any cartesian coordinate system you fix is arbitrary; you would probably need a filed to break this equivalence but in that case you would perturb the system....). Now problem is: would should we expect from DFT? I think this is at the core of the old problem that DFT "prefers" the solution where the charge density has the same symmetry of the potential (a "spherical" one for B) instead one of the possible and equivalent materializations of the statistical mixture for which symmetry is broken. Why? I think because a) you would need to break the symmetry "by hand"; b) there is a finite, negative contribution to the energy (coming from an approximate Hxc potential) even for one electron systems that shouldn't be there and gives lower energy for the symmetrical solution (e.g. the electron delocalized over the H2+ molecule at any distance). In fact corrections as "simple" as SIC can solve these problems (I guess/hope). Situation is more involved with multi-electron systems like, for example H2. If you pull the two nuclei apart one gets (in real life) 1 e- localized on each side. What is the spin of these two e- (of course they were in a singlet state at short distances)? Any relative orientation is good. So again I think the infinite distance solution is a statistical mixture of 4 states with equal probability. How this solution connects to the one at finite distances? I think it continuously connects to a multi-reference ground state described by a combination of slater determinants. What to expect from DFT? I think that to get the best out of DFT one has to impose a broken symmetry solution (e.g. an antiferromagnetic one for H2); this corresponds in fact to one of the states in the statistical mixture of the infinite distance case and to one of the slater determinants in the real ground state wavefunction. How about the charge density? If we sum over the spin it still has the symmetry of the problem. If we had the perfect DFT I think that, with a broken symmetry solution (that has the advantage also that the two body density matrix rho(x,x')-->0 for large interatomic distances), we would get the right energy and density (not wfc that is a fake one) because that is what DFT is meant to do. if we have a symmetrical solution with each electron split between both sites I think this won't come out right even with a perfect xc functional (which nobody has). For multi-electron, multi-states systems of course the situation is more complicated than this. But I would prefer a broken symmetry solution in any case (and hopefully a better xc functional). hope this helps, Matteo Nicola Marzari wrote: >> I agree that symmetrization (due to the periodization) induces a >> degeneracy but the >> "real" system that I want to simulate (a dimer alone in space) has even more >> symmetries. In particular it has a rotational symmetry that will never >> be there >> > > > Well, this is the part I've never understood - that's why I was > mentioning the boron atom. I wouldn't necessarily agree that your > dimer charge density should have the same symmetry of the ionic > potential - as much as a dissociated H2+ molecular ion doesn't have > mirror symmetry on the plan bisecting the bond, as you dissociate the > molecule, or antiferromagnetic MnO has a lower electronic-structure > symmetry, since the Mn are inequivalent from the point of view of spin > density. > > Also, the potential in atoms is spherical, but I'm sure there are > solutions lower in energy that are cylindrical. Think at a transition > metal with only one d electron: do you fill up 1 orbital 1.0, 2 orbitals > 0.5, 3 orbitals 0.33333, 5 orbitals 0.2 ? Only the last is spherical, > all others are cylindrical, although of course an ensemble of > measurements on atoms will always converge on the spherical limit. > > Am I the only one worrying about this ? It seems a key problem > that I've never seen addressed - Englisch and Englisch in 1983 > had a paper on the fact that fractional occupations are somewhat > not compatible with v-representability, but I don't recall it very well. > > > nicola > > > --------------------------------------------------------------------- > Prof Nicola Marzari Department of Materials Science and Engineering > 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA > tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- A non-text attachment was scrubbed... Name: matteo.vcf Type: text/x-vcard Size: 294 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20071121/21a419b5/attachment-0001.vcf From degironc at sissa.it Wed Nov 21 18:06:17 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 21 Nov 2007 18:06:17 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <47445937.8010601@umn.edu> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <4743FCD2.604@sissa.it> <47443272.7090008@cea.fr> <474439EA.5080009@mit.edu> <47443C37.6070700@sissa.it> <474449A1.2010402@cea.fr> <47444C1A.3070405@mit.edu> <47445937.8010601@umn.edu> Message-ID: <47446589.30803@sissa.it> let me add one reference I find (not completely ir-)relevant in this context: Volker Bach , Elliott H. Lieb , Michael Loss , and Jan Philip Solovej , "There are no unfilled shells in unrestricted Hartree-Fock theory " Phys. Rev. Lett. 72, 2981 - 2983 (1994) [ http://link.aps.org/abstract/PRL/v72/p2981] where they demonstrate (in 1994 not in 1930) that in the case of unrestricted HF (which means allowing symmetry breaking wfc's) one never gets problems of anomalous or partial filling of shells... DFT is not HF but I would expect something qualitatively similar. stefano From giannozz at nest.sns.it Wed Nov 21 19:15:28 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 21 Nov 2007 19:15:28 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <4743E440.1080508@cea.fr> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> Message-ID: <724F150B-C9FC-4BA8-855C-AF275B167DF1@nest.sns.it> On Nov 21, 2007, at 8:54 , Cyrille Barreteau wrote: > Maybe I found the second exception due to another very special > phase of the moon:-) What do I win in this case:-) you won a nice discussion on DFT. Now you just have to write the paper :-) Actually your case is different from the one I was mentioning. In that particular case, the final state was "self-inconsistent", i.e. diagonalization of the final potential yielded different Kohn-Sham states, so it was a 100% guaranteed bug. In your case, it looks like there are two different, and very close in energy, self-consistent states P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From dkorotin at gmail.com Thu Nov 22 05:53:38 2007 From: dkorotin at gmail.com (Dmitry Korotin) Date: Thu, 22 Nov 2007 09:53:38 +0500 Subject: [Pw_forum] time-reversal symmetry broken In-Reply-To: References: Message-ID: <2fd252650711212053p5f7f392et57b37fe3266d0063@mail.gmail.com> Dear Timo, I had the same problem with ESPRESSO. And I solved with the following hack (file kpoint_grid.f90 from PW directory is attached). You should add only three new strings to original file (marked with "! *** MODIFICATION"). Please note, that my file is from ESPRESSO 3.0 code (not the latest). And you forced to use smearing technique. Best regards, Dmitry Korotin PhD student, Institute of Metal Physics, Ural branch of RAS, Russia P.S.: Yes, I know that use of "goto" constructions is bad :) -------------- next part -------------- A non-text attachment was scrubbed... Name: kpoint_grid.f90 Type: text/x-fortran Size: 10631 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20071122/c463f3ad/attachment.bin From akohlmey at cmm.chem.upenn.edu Thu Nov 22 06:13:09 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Nov 2007 00:13:09 -0500 (EST) Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation In-Reply-To: <007a01c82b22$7bc79920$7356cb60$@net> References: <007a01c82b22$7bc79920$7356cb60$@net> Message-ID: On Mon, 19 Nov 2007, Paul M. Grant wrote: PG> To All Forum Members: hi paul, due to (longish) travel i'm a bit late in this discussion, but i still want to put in my two rupees (i got plenty of those these days ;-) ). PG> I'm planning on building a new Linux box (or boxes) to explore highly PG> correlated systems (e-p coupling plus LDA+U), and am seeking the collective PG> experience and advice of the PWscf community on a suitable, inexpensive (< PG> 2000 USD, MB+CPUs+RAM, exclusive of power supplies, enclosures, and PG> accessories) hardware platform. I emphasize that the principal purpose of PG> this new box would be exploratory, or for development, not production. with $2k you have a pretty reduced budget and in the business of computer hardware, great bargains are extremely rare. you mostly get what you pay for. however, you can pay in two currencies: money and your time. so first you have to decide about how much time you want to spend on putting together the machine and adapting the codes. assuming, that you don't want to rewrite large parts of QE, i'd suggest to stick with a PC (multi-core) platform. hacking consoles with cell and GPUs is out of the question, because of two reasons: a) they are single precision and b) they work best on (very) small compute kernels. there is a recent paper by the VMD/NAMD crowd on GPU computing with NVIDIA cards and also a paper on adapting a classical MD code for cell. in both cases a large speedup was only achieved by using a restricted compute kernel, that was specifically (re-)written for the hardware in question and the speedup for more generic problems was less. i can dig out the references if people are interested in details. as in most scientific papers these days, one has to carefully read between the lines to see the (IMNSHO major) problems that make those platforms of limited use for plane wave pseudopotential codes and general multi-feature general purpose codes. for small analysis tools or utilities with a small compute kernel they are a good way to improve the performance, as in most cases, you'll have to adapt/rewrite those codes anyways. having to use single precision in the vector/GPU units is even worse, as this limits your accuracy severely and would at best allow you to offload some scf cycles to the vector/GPU module, but the final convergence has to be achieved with double precision. PG> I've built several past platforms, both Windows and Linux based, using PG> server boards manufactured by Supermicro and have had generally good PG> experience and service (the factory is only 15 miles from where I live). PG> Currently, I use two machines with dual Xeon processors (single core, 32 PG> bits), one with 1 GHz cpus, 1 GB RAM, the other 2.4 GHz and 3.25 GB RAM, PG> both with bus speeds of 133 MHz, the newest 3 years old. However, PG> occasionally I run PWscf exercises on my little Thinkpad X41 tablet (single PG> processor, 1.5 GHz, 1.5 GB), and the scf computation will run 3-5 times PG> faster than on the other machines! I suspect this rather surprising result PG> is because the Thinkpad has a 400 MHz bus clock speed. for current machines, memory transfer rates _and_ latencies are decisive for speed of codes operating on large data sets (as QE). PG> One option I'm considering is using a "gaming" or server class motherboard PG> with dual 2.33 GHz quad-core 64-bit processors, a 1333 MHz FSB, and 16 GB PG> RAM. Having said this, I'm not sure PWscf (and the Fortran compilers PG> available) can handle all this parallelism efficiently on a single PG> motherboard. I've noticed when running pw.x, the CPU activity "flips" PG> between processors every several seconds, instead of sharing each at 90-100% PG> full time. if you have a new enough kernel, there is a tool called "numactl" where you can adjust the placement of the cpu. some MPI packages, like OpenMPI have flags to do this automatically (although they are still not able to handle the intel quad core properly, when using only half of the cores). the overhead of having the jobs "jump around" is rather small (~5%). since memory bandwidth is important, going for the fastest frontside bus is imperative and a "gaming" board may be the ticket (if you are willing to experiment, you can slightly overclock and gain a few % extra). intel quad cores are very competitively priced, so even if you don't use all cores (only two instead of), you can a significant speed increase, since you double the cpu cache per MPI task (n.b., the intel quad is actually a dual-dual cpu with a shared cache for each dual core part). clock rate is not that important and with your budget, the amount of memory is going to be the major decider. i.e. i would look for how much memory you need, add %20 and then buy a good board with an intel quad, that you can afford. PG> On the other hand, one could consider building a small MPI-connected cluster PG> for about the same amount of money. i don't think it would be worth it. you can only afford gigabit ethernet and i'd rather go for a dual cpu "server board" than for two desktop boards. the dual-quad will give you more flexibility (you can run a big serial and up to 8 mpi threads). particularly, for development, this is a desirable setup. in production, things are different... PG> When IBM announced a couple of years ago the incredible performance PG> details about the Cell processor that would go into Playstation 3, I PG> thought, "Wow, maybe the future of computational physics rests with PG> gamers." I'm sure most of you know this is actually beginning to PG> happen, spurred on by the fact that the PS3 is "open architecture" PG> and can run a Linux distro. Moreover, there apparently are "open PG> software" numerical analysis tools available from IBM. At least PG> four US universities are experimenting with off-the-shelf PS3 PG> clusters, perhaps one of the more interesting is at UMass, PG> http://gravity.phy.umassd.edu/ps3.html. In the last week or so, PG> Sony lowered the entry level price of the PS3 to 400 USD. So, a PG> cluster of four with a cheap switch could be purchased for about the PG> same price at the single motherboard configuration I mentioned PG> above. this is primarily (like IBMs blue gene) an exercise is excellent marketing. they really learned from p.t. barnum ;-). not that this is bad hardware, but codes have to be redesigned to work well on those. this is straightforward for a (simple) matrix multiplication, but a PITA for a general purpose code, that is written mostly by people that know very little about clock cycles, TLBs, cache lines, pipelines, speculative execution, etc. etc. and care mostly about getting the physics right (which is a hard enough job already). PG> My teenager, a gamer, tells me the PS3 has problems. He says it's PG> unreliable and overheats and only has 256 MB RAM on board (he owns a Wii, PG> which outsells the PS3 in the US by a factor of three). PG> PG> Has anybody tried porting PWscf to a PS3? you can (probably) port it using only the power pc "frontend" cpu core, but using the rest (and the PS3 does not activate all special purpose cores to boot) will be hard. i understand that people want to tap into this, but they don't factor in the amount ot time needed to get it _and_ the applications working. this has been tried before and so far, the machine (and codes) were usable by the time they were obsolete (same is true for special purpose MD cpus). so, now i've probably destroyed some peoples hopes (or offended them), but one has to be realistic about what people can do, and reality is that we are already struggling to inflict enough programming knowledge on researches, to get their work done with ancient languages like fortran, where there is tons of experience and teaching material around. expeciting low-level understanding of machine architecture is something for a very limited number of geeks, and they don't usually understand the physics well enough. just look into the problems most people running into when compiling/running simple MPI parallel executables (which is quite trivial, all things considered. best regards and good luck, axel. PG> PG> Any and all advice is welcome. PG> PG> Paul M. Grant, PhD PG> Principal, W2AGZ Technologies PG> Visiting Scholar, Applied Physics, Stanford University PG> EPRI Science Fellow (Retired) PG> IBM Research Staff Member Emeritus PG> w2agz at pacbell.net PG> http://www.w2agz.com PG> PG> PG> PG> PG> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From akohlmey at cmm.chem.upenn.edu Thu Nov 22 06:25:44 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Nov 2007 00:25:44 -0500 (EST) Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation In-Reply-To: <00df01c82b97$2ddc5690$899503b0$@net> References: <007a01c82b22$7bc79920$7356cb60$@net> <200711201501.37502.shaposh@isp.nsc.ru> <00df01c82b97$2ddc5690$899503b0$@net> Message-ID: On Tue, 20 Nov 2007, Paul M. Grant wrote: PG> Thanks for the feedback. The two dual Xeon boards I have were originally PG> chosen to host my website. I long suspected that there was "thrashing" PG> between the two processors when performing numerically intensive PG> calculations. I assume the same would occur within the cores of a single PG> processor as well? However, my understanding was this approach was supposed PG> to better implement matrix manipulation for gaming applications. If so, one PG> would assume improved performance for the fft operations at the "core" of PG> pw.x. paul, you have to be careful to not compare apples and oranges. older xeons (netburst arch), particularly the 2.4GHz ones were notorius for over saturating memory bandwidth with even a single core (hence the significant step up of cache for the xeon variety over desktop version and double cache increase for the 4-way versions). those actually get slower with using both cpus for memory intensive applications. i have some benchmark numbers with CPMD (same as QE it uses plane waves and psps) on my (mostly obsolete) homepage in bochum. current, "core" xeons are _much_ better and also the memory bus has improved a lot since. the "trashing" is of minor importance. it the actively used memory subset that counts. if it fits into the cache the cpu is very fast, if not, it is horribly slow (in two way config). cheers, axel. PG> PG> Paul M. Grant, PhD PG> Principal, W2AGZ Technologies PG> Visiting Scholar, Applied Physics, Stanford University PG> EPRI Science Fellow (Retired) PG> IBM Research Staff Member Emeritus PG> w2agz at pacbell.net PG> http://www.w2agz.com PG> ? PG> ? PG> PG> -----Original Message----- PG> From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On PG> Behalf Of Alexander Shaposhnikov PG> Sent: Tuesday, November 20, 2007 1:02 AM PG> To: PWSCF Forum PG> Subject: Re: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf PG> Implementation PG> PG> Actually newer Woodcrest/Clovertown Xeon platform have dual independent PG> FSBs, but this does not help much. Memory bandwidth available on each FSB is PG> PG> very limited - between 1/2 and 1/3 of what is available to modern desktop PG> CPUs. PG> PG> There is no good "server" computational platform today, unfortunately . PG> AMD's CPUs have weak FPU and Intel's have weak memory subsystem. Intel- PG> based desktop single-socket platform is the best choice for now. PG> PG> On Tuesday 20 November 2007 14:33, Paolo Giannozzi wrote: PG> > On Nov 20, 2007, at 4:07 , Paul M. Grant wrote: PG> > > Currently, I use two machines with dual Xeon processors PG> > PG> > dual Xeon? very bad choice. They have a single memory PG> > bus, so if you run two processes on the same motherboard, PG> > they spend most of the time waiting for access to memory. PG> > PG> > P. PG> > --- PG> > Paolo Giannozzi, Dept of Physics, University of Udine PG> > via delle Scienze 208, 33100 Udine, Italy PG> > Phone +39-0432-558216, fax +39-0432-558222 PG> > PG> > PG> > PG> > _______________________________________________ PG> > Pw_forum mailing list PG> > Pw_forum at pwscf.org PG> > http://www.democritos.it/mailman/listinfo/pw_forum PG> _______________________________________________ PG> Pw_forum mailing list PG> Pw_forum at pwscf.org PG> http://www.democritos.it/mailman/listinfo/pw_forum PG> PG> _______________________________________________ PG> Pw_forum mailing list PG> Pw_forum at pwscf.org PG> http://www.democritos.it/mailman/listinfo/pw_forum PG> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From shaposh at isp.nsc.ru Thu Nov 22 06:44:49 2007 From: shaposh at isp.nsc.ru (Alexander Shaposhnikov) Date: Thu, 22 Nov 2007 11:44:49 +0600 Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation In-Reply-To: References: <007a01c82b22$7bc79920$7356cb60$@net> Message-ID: <200711221144.50125.shaposh@isp.nsc.ru> On Thursday 22 November 2007 11:13, Axel Kohlmeyer wrote: > PG> On the other hand, one could consider building a small MPI-connected > cluster PG> for about the same amount of money. > > i don't think it would be worth it. you can only afford gigabit > ethernet and i'd rather go for a dual cpu "server board" than for > two desktop boards. the dual-quad will give you more flexibility > (you can run a big serial and up to 8 mpi threads). particularly, > for development, this is a desirable setup. in production, things > are different... > I want to comment on this one, for the small budget dual Xeon paltform is just a waste of money as performance will be inadequate. Going for the small 2-node quad-core cluster instead can give upto 100% faster exec times in pw.x and upto 150% faster for cp.x/cpmd.x. From my experience with dual quad-core Xeon machine, cp/cpmd codes scale very poorly, mostly less than 3x on 8 cores, and pw.x code scales 4x-4.5x on average. My home desktop machine with Intel C2Q at 3.6GHz (4 cores) is as fast as dual Xeon X5355 2.66Ghz (8 cores) in pw.x and 10-20% faster in cp.x/cpmd.x. Just my 2 cents. From akohlmey at cmm.chem.upenn.edu Thu Nov 22 07:01:52 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Nov 2007 01:01:52 -0500 (EST) Subject: [Pw_forum] Seeking Advice on Small Hardware Platforms for PWscf Implementation In-Reply-To: <200711221144.50125.shaposh@isp.nsc.ru> References: <007a01c82b22$7bc79920$7356cb60$@net> <200711221144.50125.shaposh@isp.nsc.ru> Message-ID: On Thu, 22 Nov 2007, Alexander Shaposhnikov wrote: AS> I want to comment on this one, AS> for the small budget dual Xeon paltform is just a waste of money as AS> performance will be inadequate. performance, on what kind of system. for large scale inputs, you are very likely correct, but for development and tests, with small inputs, flexibility and latency and last but not least convenience matters more. AS> Going for the small 2-node quad-core cluster instead can give upto 100% faster AS> exec times in pw.x and upto 150% faster for cp.x/cpmd.x. AS> >From my experience with dual quad-core Xeon machine, cp/cpmd codes scale very AS> poorly, mostly less than 3x on 8 cores, and pw.x code scales 4x-4.5x on please also consider testing using only half the cores. i found that to be the fastest option on several dual quad-core machines for a number of different codes and inputs. also, as was noted by others, memory bus bandwidth is the primary performance decider. ... and with a budget of $2k, i doubt that there will be enough money to buy either two nodes or a dual processor board, anyways. cheers, axel. AS> average. My home desktop machine with Intel C2Q at 3.6GHz (4 cores) is as AS> fast as dual Xeon X5355 2.66Ghz (8 cores) in pw.x and 10-20% faster in AS> cp.x/cpmd.x. AS> AS> Just my 2 cents. AS> _______________________________________________ AS> Pw_forum mailing list AS> Pw_forum at pwscf.org AS> http://www.democritos.it/mailman/listinfo/pw_forum AS> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From hyello25 at gmail.com Thu Nov 22 12:21:45 2007 From: hyello25 at gmail.com (hyello25 Qgrui) Date: Thu, 22 Nov 2007 19:21:45 +0800 Subject: [Pw_forum] ld: cannot find -lmpi_f90 Message-ID: Hi every pwscf user, When I tried to compile pwscf-3.2.3 with openmpi-1.2.4, the error "ld: cannot find -lmpi_f90" came out. I had tried many way to solve it but failed to find the answer in openmpi forum.Can anyone give me some advices about it. My system is emt64 architecture, ifort 10, mkl 9 and icc 10 compiler, RedHat 4.3 os. I have succeeded to compiler the pwscf code using mpich2. But the network data exchange is too large so that the efficiency is low when parallel calculation performed. Bests. G.R. Qian From eyvaz_isaev at yahoo.com Thu Nov 22 12:42:26 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 22 Nov 2007 03:42:26 -0800 (PST) Subject: [Pw_forum] ld: cannot find -lmpi_f90 In-Reply-To: Message-ID: <580869.20928.qm@web60321.mail.yahoo.com> Hi, --- hyello25 Qgrui wrote: > When I tried to compile pwscf-3.2.3 with > openmpi-1.2.4, the error > "ld: cannot find -lmpi_f90" came out. this means that the linker could not find libmpi_f90.a file. Try find the path (using "locate libmpi_f90" command) and then add -L/path/to/library/file -lmpi_f90 Hopefully you have compiled openmpi using the same comiler. Hope it helps. Bests, Eyvaz. >I had tried many way to solve it > but failed to find the answer in openmpi forum.Can > anyone give me some > advices about it. > My system is emt64 architecture, ifort 10, mkl 9 > and icc 10 > compiler, RedHat 4.3 os. > I have succeeded to compiler the pwscf code using > mpich2. But the > network data exchange is too large so that the > efficiency is low when > parallel calculation performed. > > > Bests. > > G.R. Qian > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Be a better pen pal. Text or chat with friends inside Yahoo! Mail. See how. http://overview.mail.yahoo.com/ From akohlmey at cmm.chem.upenn.edu Thu Nov 22 13:35:18 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Nov 2007 07:35:18 -0500 (EST) Subject: [Pw_forum] ld: cannot find -lmpi_f90 In-Reply-To: References: Message-ID: On Thu, 22 Nov 2007, hyello25 Qgrui wrote: GQ> Hi every pwscf user, GQ> GQ> When I tried to compile pwscf-3.2.3 with openmpi-1.2.4, the error GQ> "ld: cannot find -lmpi_f90" came out. I had tried many way to solve it GQ> but failed to find the answer in openmpi forum.Can anyone give me some GQ> advices about it. that means, that either you messed up the OpenMPI installation or are not using the right mpif90 wrapper as linker. since you don't provide any information about either, there is little we can do for you. can you compiler/run a simple "hello world" mpi program with openmpi? GQ> My system is emt64 architecture, ifort 10, mkl 9 and icc 10 GQ> compiler, RedHat 4.3 os. GQ> I have succeeded to compiler the pwscf code using mpich2. But the GQ> network data exchange is too large so that the efficiency is low when GQ> parallel calculation performed. switching to OpenMPI is not going to change that! i assume you are using gigabit ethernet interconnect. but (again) since you don't provide any specifics of the network, job type and system size, or how you run pwscf, there is little room for giving any advice. do you parallelize over k-points (using -npools)? cheers, axel. GQ> GQ> GQ> Bests. GQ> GQ> G.R. Qian GQ> _______________________________________________ GQ> Pw_forum mailing list GQ> Pw_forum at pwscf.org GQ> http://www.democritos.it/mailman/listinfo/pw_forum GQ> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From cyrille.barreteau at cea.fr Thu Nov 22 14:30:39 2007 From: cyrille.barreteau at cea.fr (Cyrille Barreteau) Date: Thu, 22 Nov 2007 14:30:39 +0100 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <724F150B-C9FC-4BA8-855C-AF275B167DF1@nest.sns.it> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <724F150B-C9FC-4BA8-855C-AF275B167DF1@nest.sns.it> Message-ID: <4745847F.4080505@cea.fr> A final comment concerning this tricky problem. I think I got the reason of my troubles. I have two scf solutions: 1) one that (almost) satisfies the symmetry of the real dimer and for which xy and x^2-y^2 eigenfunctions are degenerate. However this solution has the bad idea of having an unfilled homo state. 2) an other that do not satisfy the symmetry of the real dimer (but of course satisfy the symmetry of the crystal) since a rather large spitting of xy and x^2-y^2 is observed. The second solution is the most stable for reasons related to the so called Jahn-Teller effect. There is no true Jahn Teller distorsion but the dissymetrization of xy and x^2-y^2 leads to a splitting of levels that decreases energy.... bad luck! I guess this problem will disappear if I add spin-orbit coupling since it will induce splittings and the (bad) chance to have an unfilled homo state is smaller. that's all folks. cyrille -- ================================================================== Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 CEA Saclay | fax : +33 (0)1 69 08 84 46 DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr Batiment 462 | 91191 Gif sur Yvette Cedex FRANCE ~~~~~~~~~~~~~~~~~~~~~~~~ http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html ================================================================== Paolo Giannozzi wrote: >On Nov 21, 2007, at 8:54 , Cyrille Barreteau wrote: > > > >>Maybe I found the second exception due to another very special >>phase of the moon:-) What do I win in this case:-) >> >> > >you won a nice discussion on DFT. Now you just have to write >the paper :-) > >Actually your case is different from the one I was mentioning. >In that particular case, the final state was "self-inconsistent", >i.e. diagonalization of the final potential yielded different >Kohn-Sham states, so it was a 100% guaranteed bug. >In your case, it looks like there are two different, and very >close in energy, self-consistent states > >P. >--- >Paolo Giannozzi, Dept of Physics, University of Udine >via delle Scienze 208, 33100 Udine, Italy >Phone +39-0432-558216, fax +39-0432-558222 > > > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From ihsanas at yahoo.com Thu Nov 22 16:16:53 2007 From: ihsanas at yahoo.com (ÇÍÓÇä ÚÑíÞÇÊ) Date: Thu, 22 Nov 2007 07:16:53 -0800 (PST) Subject: [Pw_forum] neb Message-ID: <893938.83865.qm@web52310.mail.re2.yahoo.com> Hi below is neb output file ,it`s still running ,my question can this neb job converged ? If no,how could I make it better best regard Ihsan Jordan university -Amman calculation = neb restart_mode = restart opt_scheme = broyden num_of_images = 8 nstep = 50 CI_scheme = no-CI first_last_opt = F coarse-grained phase-space = F use_freezing = F ds = 1.0000 a.u. k_max = 0.3000 a.u. k_min = 0.2000 a.u. suggested k_max = 0.6169 a.u. suggested k_min = 0.4112 a.u. path_thr = 0.1000 eV / A ------------------------------ iteration 5 ------------------------------ tcpu = 0.1 self-consistency for image 2 tcpu = 233610.5 self-consistency for image 3 tcpu = 495616.2 self-consistency for image 4 tcpu = 525905.7 self-consistency for image 5 tcpu = 557074.3 self-consistency for image 6 tcpu = 586500.5 self-consistency for image 7 activation energy (->) = 5.301332 eV activation energy (<-) = 1.314140 eV image energy (eV) error (eV/A) frozen 1 -24438.8519036 1.707793 T 2 -24436.1036226 3.559105 F 3 -24436.1776232 1.594526 F 4 -24434.4137114 3.410501 F 5 -24433.5505712 3.090315 F 6 -24434.3100502 2.919468 F 7 -24434.8021457 3.902409 F 8 -24434.8647109 3.415599 T path length = 11.684 bohr inter-image distance = 1.669 bohr ____________________________________________________________________________________ Be a better pen pal. Text or chat with friends inside Yahoo! Mail. See how. http://overview.mail.yahoo.com/ From akohlmey at cmm.chem.upenn.edu Thu Nov 22 17:28:03 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Nov 2007 11:28:03 -0500 (EST) Subject: [Pw_forum] neb In-Reply-To: <893938.83865.qm@web52310.mail.re2.yahoo.com> References: <893938.83865.qm@web52310.mail.re2.yahoo.com> Message-ID: On Thu, 22 Nov 2007, ????? ?????? wrote: IA> Hi IA> below is neb output file ,it`s still running ,my IA> question can this neb job converged ? If no,how could hi! how should anybody be able to tell it from the little information you provide??? your question is like asking "will the brown horse win the race?" without telling anybody the horses' names are, who is riding them, and what the race is. if you want help, you have give people a chance. nobody here has psychic abilities and can read your computer's mind. even if it may look like it on occasion... cheers, axel. IA> I make it better IA> best regard IA> Ihsan IA> Jordan university -Amman IA> IA> calculation = neb IA> restart_mode = restart IA> opt_scheme = broyden IA> num_of_images = 8 IA> nstep = 50 IA> CI_scheme = no-CI IA> first_last_opt = F IA> coarse-grained phase-space = F IA> use_freezing = F IA> ds = 1.0000 a.u. IA> k_max = 0.3000 a.u. IA> k_min = 0.2000 a.u. IA> suggested k_max = 0.6169 a.u. IA> suggested k_min = 0.4112 a.u. IA> path_thr = 0.1000 eV / A IA> IA> ------------------------------ iteration 5 IA> ------------------------------ IA> IA> tcpu = 0.1 self-consistency for image 2 IA> tcpu = 233610.5 self-consistency for image 3 IA> tcpu = 495616.2 self-consistency for image 4 IA> tcpu = 525905.7 self-consistency for image 5 IA> tcpu = 557074.3 self-consistency for image 6 IA> tcpu = 586500.5 self-consistency for image 7 IA> IA> activation energy (->) = 5.301332 eV IA> activation energy (<-) = 1.314140 eV IA> IA> image energy (eV) error (eV/A) IA> frozen IA> IA> 1 -24438.8519036 1.707793 IA> T IA> 2 -24436.1036226 3.559105 IA> F IA> 3 -24436.1776232 1.594526 IA> F IA> 4 -24434.4137114 3.410501 IA> F IA> 5 -24433.5505712 3.090315 IA> F IA> 6 -24434.3100502 2.919468 IA> F IA> 7 -24434.8021457 3.902409 IA> F IA> 8 -24434.8647109 3.415599 IA> T IA> IA> path length = 11.684 bohr IA> inter-image distance = 1.669 bohr IA> IA> IA> IA> ____________________________________________________________________________________ IA> Be a better pen pal. IA> Text or chat with friends inside Yahoo! Mail. See how. http://overview.mail.yahoo.com/ IA> _______________________________________________ IA> Pw_forum mailing list IA> Pw_forum at pwscf.org IA> http://www.democritos.it/mailman/listinfo/pw_forum IA> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From tone.kokalj at ijs.si Thu Nov 22 18:08:36 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Thu, 22 Nov 2007 18:08:36 +0100 Subject: [Pw_forum] neb In-Reply-To: References: <893938.83865.qm@web52310.mail.re2.yahoo.com> Message-ID: <1195751316.2598.58.camel@localhost.localdomain> On Thu, 2007-11-22 at 11:28 -0500, Axel Kohlmeyer wrote: > On Thu, 22 Nov 2007, ????? ?????? wrote: > > IA> Hi > IA> below is neb output file ,it`s still running ,my > IA> question can this neb job converged ? If no,how could > > hi! > > how should anybody be able to tell it from the little > information you provide??? I agree with Axel, but at least something can be said from the Ihsan's provided data. Usually in NEB calculation, one connects the two local minima (i.e. 1st and last image). But this is not the case in Ihsan' calculation as 1st and last image have large total force: 1.707793 and 3.415599 eV/Angs, respectively. It is possible to optimize them on the fly by setting the first_last_opt = .true. , but my suggestion is to optimize these first by the bfgs 'relax' run. Otherwise the errors on the order of 1 eV/Angs in 5th iteration is nothing awkward. The NEB calculation may easily take as long as 100 iteration (yes, the NEB is a CPU time monster). Regards, Tone -- J. Stefan Institute, Ljubljana, Slovenia From matteo at umn.edu Thu Nov 22 21:22:47 2007 From: matteo at umn.edu (Matteo Cococcioni) Date: Thu, 22 Nov 2007 14:22:47 -0600 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: <4745847F.4080505@cea.fr> References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <724F150B-C9FC-4BA8-855C-AF275B167DF1@nest.sns.it> <4745847F.4080505@cea.fr> Message-ID: Dear Cyrille let me add last comments too and ask few questions. On Nov 22, 2007, at 7:30 AM, Cyrille Barreteau wrote: > > A final comment concerning this tricky problem. > I think I got the reason of my troubles. > I have two scf solutions: > > 1) one that (almost) satisfies the symmetry of the real dimer > and for which xy and x^2-y^2 eigenfunctions are degenerate. > However this solution has the bad idea of having an unfilled homo > state. I think this is the same bad idea of e.g. transition metal oxides that turn out to be metals (instead of insulators) in DFT. The difference is that in your case you don't have a particular reason to suppose that the electron should instead be localized on one of the degenerate states, while in a crystal you may appeal to crystal field, symmetry etc. Are the degenerate xy and x^2-y^2 orbitals part of the unfilled homo? if so, do you get one of them above the "Fermi energy" and one below when you get the other, lower energy solution? If this is true I think that you would obtain the same energy with the two states inverted (the one that was above ef goes below and vice versa). Can you check this? I agree with Stefano in general and prefer the broken symmetry solution. I think this gives you the "right" (of course in a DFT sense) energy. However I don't feel completely confident about the fact that you can obtain it just with nosym = true and placing the molecule in a generic orientation w r t the cell. Your ground will still have the symmetry of the fake crystal that you construct (even f you don't impose it) unless a broken symmetry solution exists. I guess my question is: is the numerical noise produced in the diagonalization (and of course not washed out by the imposing of the symmetry) always enough for the code to recognize a broken symmetry solution and to choose it? I'm not sure. In your case I would feel more confident simulating a "supercell" with few molecules randomly oriented wrt each other. This would be a more faithful simulation of the gas phase you are probably thinking of than the fake crystal you are using. Of course this will be more expensive though... Maybe the answer to this doubt of mine is trivial. Does anybody have more clear ideas than mine on how to avoid symmetry? Regards, Matteo > > 2) an other that do not satisfy the symmetry of the real dimer > (but of course satisfy the symmetry of the crystal) since a > rather large spitting of xy and x^2-y^2 is observed. > > The second solution is the most stable for reasons related > to the so called Jahn-Teller effect. There is no true Jahn > Teller distorsion but the dissymetrization of xy and x^2-y^2 > leads to a splitting of levels that decreases energy.... > bad luck! > > I guess this problem will disappear if I add spin-orbit coupling > since it will induce splittings and the (bad) chance to have > an unfilled homo state is smaller. > > that's all folks. > > cyrille > > -- > ================================================================== > Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 > CEA Saclay | fax : +33 (0)1 69 08 84 46 > DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr > Batiment 462 | > 91191 Gif sur Yvette Cedex FRANCE > ~~~~~~~~~~~~~~~~~~~~~~~~ > http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html > ================================================================== > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > Matteo Cococcioni > Department of Chemical Engineering and Materials Science > University of Minnesota > 151 Amundson Hall > 421 Washington Av. SE > Minneapolis, MN 55455 > Tel. +1 612 624 9056 Fax +1 612 626 7246 > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071122/af69c595/attachment-0001.htm From matteo at umn.edu Thu Nov 22 21:39:54 2007 From: matteo at umn.edu (Matteo Cococcioni) Date: Thu, 22 Nov 2007 14:39:54 -0600 Subject: [Pw_forum] Problem with nbnd in clusters In-Reply-To: References: <4742E855.2030909@cea.fr> <4742F555.9070201@cea.fr> <74A5928C-2495-4967-8315-879BFC54DD6B@nest.sns.it> <4743E440.1080508@cea.fr> <724F150B-C9FC-4BA8-855C-AF275B167DF1@nest.sns.it> <4745847F.4080505@cea.fr> Message-ID: <71EC25E1-AAD6-4C05-86DC-A2135433C847@umn.edu> Well my doubt was trivial indeed: placing a molecule in a generic orientation wrt the cell corresponds to effectively destroy the symmetry of the "basis" of the fake crystal. so there is no (or little) symmetry that the code can see even with nosym = false. at least I won't waste anybody's time.... Matteo On Nov 22, 2007, at 2:22 PM, Matteo Cococcioni wrote: > > Dear Cyrille > > let me add last comments too and ask few questions. > > > On Nov 22, 2007, at 7:30 AM, Cyrille Barreteau wrote: > >> >> A final comment concerning this tricky problem. >> I think I got the reason of my troubles. >> I have two scf solutions: >> >> 1) one that (almost) satisfies the symmetry of the real dimer >> and for which xy and x^2-y^2 eigenfunctions are degenerate. >> However this solution has the bad idea of having an unfilled homo >> state. > > I think this is the same bad idea of e.g. transition metal oxides > that turn out to be metals (instead of insulators) in DFT. > The difference is that in your case you don't have a particular > reason to suppose that the electron should instead > be localized on one of the degenerate states, while in a crystal > you may appeal to crystal field, symmetry etc. > Are the degenerate xy and x^2-y^2 orbitals part of the unfilled > homo? if so, do you get one of them above the "Fermi energy" and > one below > when you get the other, lower energy solution? If this is true I > think that you would obtain the same energy with the two states > inverted (the one that was above ef goes below and vice versa). Can > you check this? > > I agree with Stefano in general and prefer the broken symmetry > solution. I think this gives you the "right" (of course in a DFT > sense) energy. > > However I don't feel completely confident about the fact that you > can obtain it just with nosym = true and placing the molecule in a > generic > orientation w r t the cell. Your ground will still have the > symmetry of the fake crystal that you construct (even f you don't > impose it) > unless a broken symmetry solution exists. > I guess my question is: is the numerical noise produced in the > diagonalization (and of course not washed out by the imposing of > the symmetry) > always enough for the code to recognize a broken symmetry solution > and to choose it? I'm not sure. > In your case I would feel more confident simulating a "supercell" > with few molecules randomly oriented wrt each other. This would be > a more faithful simulation of the gas phase you are probably > thinking of than the fake crystal you are using. Of course this > will be more expensive though... > Maybe the answer to this doubt of mine is trivial. Does anybody > have more clear ideas than mine on how to avoid symmetry? > > Regards, > > Matteo > > >> >> 2) an other that do not satisfy the symmetry of the real dimer >> (but of course satisfy the symmetry of the crystal) since a >> rather large spitting of xy and x^2-y^2 is observed. >> >> The second solution is the most stable for reasons related >> to the so called Jahn-Teller effect. There is no true Jahn >> Teller distorsion but the dissymetrization of xy and x^2-y^2 >> leads to a splitting of levels that decreases energy.... >> bad luck! >> >> I guess this problem will disappear if I add spin-orbit coupling >> since it will induce splittings and the (bad) chance to have >> an unfilled homo state is smaller. >> >> that's all folks. >> >> cyrille >> >> -- >> ================================================================== >> Cyrille Barreteau | phone : +33 (0)1 69 08 29 51 >> CEA Saclay | fax : +33 (0)1 69 08 84 46 >> DSM/DRECAM/SPCSI | email cyrille.barreteau at cea.fr >> Batiment 462 | >> 91191 Gif sur Yvette Cedex FRANCE >> ~~~~~~~~~~~~~~~~~~~~~~~~ >> http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html >> ================================================================== >> >> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% >> Matteo Cococcioni >> Department of Chemical Engineering and Materials Science >> University of Minnesota >> 151 Amundson Hall >> 421 Washington Av. SE >> Minneapolis, MN 55455 >> Tel. +1 612 624 9056 Fax +1 612 626 7246 > >> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071122/b353547a/attachment.htm From r95222066 at ntu.edu.tw Fri Nov 23 03:56:51 2007 From: r95222066 at ntu.edu.tw (r95222066 at ntu.edu.tw) Date: Fri, 23 Nov 2007 10:56:51 +0800 Subject: [Pw_forum] hi, I want to ask how to construct R-3c symmetry structure? Message-ID: <20071123105651.tae0wa1pmowcwk4g@wmail1.cc.ntu.edu.tw> Hello, I want to ask how to construct R-3c symmetry ? I have used wien2k to construct hexegonal and rhombohedral structure on Ca3Co2O6,and I write the position to scf file. But the scf.out file appeal "no symmetry" message!! Here is my input data: ---------------------------------------------------------------------- (a) Ca3Co2O6 hexagonal structure scf file ibrav = 4, celldm(1) =17.1228,celldm(3)=1.1441342, nat= 66, ntyp= 3,nspin=2, ATOMIC_POSITIONS Ca 0.369000 0.000000 0.250000 Ca 0.702333 0.666666 0.916666 Ca 0.035666 0.333333 0.583333 Ca 0.000000 0.369000 0.250000 Ca 0.333333 0.035666 0.916666 Ca 0.666666 0.702333 0.583333 Ca -0.369000 -0.369000 0.250000 Ca -0.035666 0.297666 0.916666 Ca 0.297666 -0.035666 0.583333 Ca -0.369000 0.000000 0.750000 Ca -0.035666 0.666666 0.416666 Ca 0.297666 0.333333 0.083333 Ca 0.000000 -0.369000 0.750000 Ca 0.333333 0.297666 0.416666 Ca 0.666666 -0.035666 0.083333 Ca 0.369000 0.369000 0.750000 Ca 0.702333 0.035666 0.416666 Ca 0.035666 0.702333 0.083333 Co 0.000000 0.000000 0.250000 Co 0.333333 0.666666 0.916666 Co 0.666666 0.333333 0.583333 Co 0.000000 0.000000 0.750000 Co 0.333333 0.666666 0.416666 Co 0.666666 0.333333 0.083333 Co 0.000000 0.000000 0.000000 Co 0.333333 0.666666 0.666666 Co 0.666666 0.333333 0.333333 Co 0.000000 0.000000 0.500000 Co 0.333333 0.666666 0.166666 Co 0.666666 0.333333 0.833333 O 0.176400 0.022700 0.114000 O 0.509733 0.689366 0.780666 O 0.843066 0.356033 0.447333 O -0.176400 -0.022700 -0.114000 O 0.156933 0.643966 0.552666 O 0.490266 0.310633 0.219333 O -0.022700 -0.176400 0.614000 O 0.310633 0.490266 0.280666 O 0.643966 0.156933 0.947333 O 0.022700 0.176400 0.386000 O 0.356033 0.843066 0.052666 O 0.689366 0.509733 0.719333 O -0.022700 0.153700 0.114000 O 0.310633 0.820366 0.780666 O 0.643966 0.487033 0.447333 O 0.022700 -0.153700 -0.114000 O 0.356033 0.512966 0.552666 O 0.689366 0.179633 0.219333 O 0.176400 0.153700 0.614000 O 0.509733 0.820366 0.280666 O 0.843066 0.487033 0.947333 O -0.153700 -0.176400 0.114000 O 0.179633 0.490266 0.780666 O 0.512966 0.156933 0.447333 O 0.153700 0.176400 -0.114000 O 0.487033 0.509733 0.552666 O 0.820366 0.843066 0.219333 O -0.153700 0.022700 0.614000 O 0.179633 0.689366 0.280666 O 0.512966 0.356033 0.447333 O 0.153700 -0.022700 0.386000 O 0.487033 0.643966 0.052666 O 0.820366 0.310633 0.719333 ----------------------------------------------------------------------- (b) Ca3Co2O6 rhombohedral structure file ibrav = 5, celldm(1) =11.86937,celldm(4)=-0.04466, nat= 22, ntyp= 3,nspin=2, ATOMIC_POSITIONS Ca 0.619000 0.881000 0.250000 Ca 0.250000 0.619000 0.881000 Ca 0.881000 0.250000 0.619000 Ca 0.381000 0.118999 0.750000 Ca 0.750000 0.381000 0.118999 Ca 0.118999 0.750000 0.381000 Co 0.500000 0.499999 0.500000 Co 0.000000 0.000000 0.000000 Co 0.250000 0.250000 0.250000 Co 0.750000 0.749999 0.750000 O 0.539100 0.209500 0.410000 O 0.410000 0.539100 0.209500 O 0.209500 0.410000 0.539100 O 0.290499 0.960900 0.090000 O 0.960900 0.090000 0.290499 O 0.090000 0.290499 0.960900 O 0.460900 0.790499 0.590000 O 0.590000 0.460900 0.790499 O 0.790499 0.590000 0.460900 O 0.709500 0.039100 0.910000 O 0.039100 0.910000 0.709500 O 0.910000 0.709500 0.039100 --------------------------------------------------------------------- Here is the scf.out file message Starting magnetic structure atomic species magnetization Ca 0.100 Co 1.000 O 0.100 No symmetry! Cartesian axes site n. atom positions (a_0 units) 1 Ca tau( 1) = ( 0.6190000 0.8810000 0.2500000 ) ----------------------------------------------------------------------------- Thanks for your help Bing-Hong Chen From r95222066 at ntu.edu.tw Fri Nov 23 03:58:05 2007 From: r95222066 at ntu.edu.tw (r95222066 at ntu.edu.tw) Date: Fri, 23 Nov 2007 10:58:05 +0800 Subject: [Pw_forum] Hello,I have a question about SOC picture? Message-ID: <20071123105805.4yty0gur0gwcgoo4@wmail1.cc.ntu.edu.tw> hello, I have make a pratice for spin-orbit coupling about Fe like example 22 in PWSCF. when I finished the process of plotband.x . The xmgr.* file will generated. But I have little confused the picture has some straight line in bandstructure picture at the nearby special kpoints. the attached file is iron-soc bandstructure!! Thanks for your help! Bing-Hong Chen From r95222066 at ntu.edu.tw Fri Nov 23 04:00:14 2007 From: r95222066 at ntu.edu.tw (r95222066 at ntu.edu.tw) Date: Fri, 23 Nov 2007 11:00:14 +0800 Subject: [Pw_forum] Hi, I would like to know how to use the restart_mode='restart'? Message-ID: <20071123110014.nxt2xmw9gk40ksc0@wmail1.cc.ntu.edu.tw> Hello, I have read the user guide and the INPUT_PW doc file.I wolud like to ask how to use the restart_mode='restart'. Do I need to save the wfc file? what is the process if I have not scf yet,and I want to scf again by used the restart_mode="restart"? From akohlmey at cmm.chem.upenn.edu Fri Nov 23 05:43:24 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Nov 2007 23:43:24 -0500 Subject: [Pw_forum] hi, I want to ask how to construct R-3c symmetry structure? In-Reply-To: <20071123105651.tae0wa1pmowcwk4g@wmail1.cc.ntu.edu.tw> References: <20071123105651.tae0wa1pmowcwk4g@wmail1.cc.ntu.edu.tw> Message-ID: <7b6913e90711222043m7ef45d30j80374e70433a4683@mail.gmail.com> On Nov 22, 2007 9:56 PM, wrote: > Hello, > I want to ask how to construct R-3c symmetry ? I have used wien2k > to construct > hexegonal and rhombohedral structure on Ca3Co2O6,and I write the > position to scf file. > But the scf.out file appeal "no symmetry" message!! hi! this is no surprise. please read the INPUT_PW file more carefully and understand the difference between alat and cystal units of the ATOMIC_POSITIONS section of the input. cheers, axel. > Here is my input data: > ---------------------------------------------------------------------- > (a) Ca3Co2O6 hexagonal structure scf file > ibrav = 4, celldm(1) =17.1228,celldm(3)=1.1441342, > nat= 66, ntyp= 3,nspin=2, > ATOMIC_POSITIONS > Ca 0.369000 0.000000 0.250000 Cartesian axes [...] > site n. atom positions (a_0 units) > 1 Ca tau( 1) = ( 0.6190000 0.8810000 0.2500000 ) > ----------------------------------------------------------------------------- > > Thanks for your help > Bing-Hong Chen > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From giannozz at nest.sns.it Fri Nov 23 11:10:29 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 23 Nov 2007 11:10:29 +0100 Subject: [Pw_forum] Hi, I would like to know how to use the restart_mode='restart'? In-Reply-To: <20071123110014.nxt2xmw9gk40ksc0@wmail1.cc.ntu.edu.tw> References: <20071123110014.nxt2xmw9gk40ksc0@wmail1.cc.ntu.edu.tw> Message-ID: On Nov 23, 2007, at 4:00 , r95222066 at ntu.edu.tw wrote: > I have read the user guide and the INPUT_PW doc file.I would like > to ask how to use the restart_mode='restart'. first learn to use the code without 'restart' ... Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From hyello25 at gmail.com Fri Nov 23 13:49:51 2007 From: hyello25 at gmail.com (hyello25 Qgrui) Date: Fri, 23 Nov 2007 20:49:51 +0800 Subject: [Pw_forum] ld: cannot find -lmpi_f90 Message-ID: Hi all, >this means that the linker could not find libmpi_f90.a >file. Try find the path (using "locate libmpi_f90" >command) and then add -L/path/to/library/file >-lmpi_f90 Thanks to Eyvaz's advice. I have succeeded to compile the pwscf code. >switching to OpenMPI is not going to change that! >i assume you are using gigabit ethernet interconnect. >but (again) since you don't provide any specifics of >the network, job type and system size, Sorry for my stupid question, till now I did not konw clear about and how to describe the the network. And you are right "OpenMPI is not going to change that!", although the openmpi performance better than mpich2. I think the pwscf with openmpi have reached my expectancy. Thanks again. Bests. From mmarques at staffmail.ed.ac.uk Fri Nov 23 13:35:31 2007 From: mmarques at staffmail.ed.ac.uk (Miriam Marques) Date: Fri, 23 Nov 2007 12:35:31 +0000 Subject: [Pw_forum] Na pseudopotential Message-ID: <20071123123531.zjtpvq15kc0cc880@www.staffmail.ed.ac.uk> Dear ESPRESSO users, Does anybody have a norm-conserving Na PP (PBE or PW91 exchange correlation functional) and including s and p semicore states in valence? Thank you very much, Miriam. Dr. Miriam Marques Centre for Science at Extreme Conditions. The University of Edinburgh. e-mail: mmarques at staffmail.ed.ac.uk -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. From calderin at physics.queensu.ca Fri Nov 23 14:20:06 2007 From: calderin at physics.queensu.ca (Lazaro Calderin) Date: Fri, 23 Nov 2007 14:20:06 +0100 Subject: [Pw_forum] Kubo-Greenwood Message-ID: <1195824017.3101.170.camel@wolfe> Hello everyone, Could some one please tell me if the Kubo-Greenwood formula for the optical conductivity has been implemented in pwscf ? Thanks, Lazaro -- ============================================================= Dr. Lazaro Calderin Dpt. of Theoretical, Atomic and Optical Physics Univ. of Valladolid, 47005 Valladolid. Spain Phone: +34 983 423891 Fax: +34 983 423013 E-mails: calderin at fta.uva.es or calderin at physics.queensu.ca From akohlmey at cmm.chem.upenn.edu Fri Nov 23 17:05:13 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 23 Nov 2007 11:05:13 -0500 (EST) Subject: [Pw_forum] Na pseudopotential In-Reply-To: <20071123123531.zjtpvq15kc0cc880@www.staffmail.ed.ac.uk> References: <20071123123531.zjtpvq15kc0cc880@www.staffmail.ed.ac.uk> Message-ID: On Fri, 23 Nov 2007, Miriam Marques wrote: MM> Dear ESPRESSO users, MM> dear miriam, MM> Does anybody have a norm-conserving Na PP MM> (PBE or PW91 exchange correlation MM> functional) and including s and p semicore states in valence? there is a cpmd format TM sodium semicore psp with PBE in the (extended) psp library maintained by mauro boero. you can find the whole library at http://www.pci.unizh.ch/gruppe.hutter/e/information.html and try to convert the pseuodpotential to UPF. since the conversion is always a bit risky and the CPMD pseudopotentials notoriously inconsistent, it may be safer to re-create the pseudopotential from the parameters using the atomic code in QE. here are the parameters. 2s rc=1.000 eref=-2.36868 2p rc=1.000 eref=-1.33583 using the s-channel as local potential and a (minimum) plane wave cutoff of 80ry. those parameters should be pretty transferable to other GGA functionals... cheers, axel. MM> MM> Thank you very much, MM> MM> Miriam. MM> MM> MM> Dr. Miriam Marques MM> Centre for Science at Extreme Conditions. MM> The University of Edinburgh. MM> e-mail: mmarques at staffmail.ed.ac.uk MM> MM> MM> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From mmarques at staffmail.ed.ac.uk Fri Nov 23 18:59:33 2007 From: mmarques at staffmail.ed.ac.uk (Miriam Marques) Date: Fri, 23 Nov 2007 17:59:33 +0000 Subject: [Pw_forum] PBE Na pseudopotential Message-ID: <20071123175933.3hbgv6yg0kko844s@www.staffmail.ed.ac.uk> Dear Axel, First of all, thank you very much for your answer. Unfortunately, I cannot find the PBE semicore pseudopotential in the CPMD extended library. Instead, I find: Na_GO_BLYP.psp Na_MT_HCTH_SEMI.psp Na_MT_PW91_SEMI.psp Na_SEMI_BLYP.psp Na_GO_LDA.psp Na_MT_NLCC.psp Na_MT_XLYP_SEMI.psp Na_SGS_LDA.psp Among them, a PW91 semi PP. Sorry for bothering you a little bit more, but could you please send me the PBE PP you found? Thank you very much in advance, Miriam. Dr. Miriam Marques Centre for Science at Extreme Conditions. The University of Edinburgh. e-mail: mmarques at staffmail.ed.ac.uk -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. From smeza at uas.uasnet.mx Fri Nov 23 19:27:29 2007 From: smeza at uas.uasnet.mx (Salvador Meza Aguilar) Date: Fri, 23 Nov 2007 11:27:29 -0700 Subject: [Pw_forum] Quad-processor Message-ID: <47471B91.4090001@uas.uasnet.mx> PW forum Somebody knows if the PWscf works on Quad-Processor? I think it is possible to run on double processor quad-core, like a 8 processors. Could you informe me about this. Thank you Salvador Meza Aguilar From akohlmey at cmm.chem.upenn.edu Fri Nov 23 19:34:24 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 23 Nov 2007 13:34:24 -0500 (EST) Subject: [Pw_forum] Quad-processor In-Reply-To: <47471B91.4090001@uas.uasnet.mx> References: <47471B91.4090001@uas.uasnet.mx> Message-ID: On Fri, 23 Nov 2007, Salvador Meza Aguilar wrote: SMA> PW forum SMA> SMA> Somebody knows if the PWscf works on Quad-Processor? yes. why not? you have to use MPI and you can run as many MPI threads as you want per node. SMA> I think it is possible to run on double processor quad-core, like a 8 SMA> processors. yes. also this is possible. but you have to expect that the total performance, may be severly limited by the memory bandwidth unless you are running _very_ small calculations. it may be faster to only use half of the cores, though and benefit from the larger cache per core. please see the recent discussion on the topic in the mailing list archives. cheers, axel. SMA> SMA> Could you informe me about this. SMA> SMA> Thank you SMA> SMA> Salvador Meza Aguilar SMA> _______________________________________________ SMA> Pw_forum mailing list SMA> Pw_forum at pwscf.org SMA> http://www.democritos.it/mailman/listinfo/pw_forum SMA> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From baroni at sissa.it Sat Nov 24 10:02:47 2007 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 24 Nov 2007 09:02:47 +0000 Subject: [Pw_forum] POSTDOCTORAL POSITION AVAILABLE @SISSA References: <34003D61-3EB1-4CEF-8758-3A9C6CBA0476@sissa.it> Message-ID: <94DAFE84-8269-4172-8508-748876ABF014@sissa.it> Postdoctoral position in Computational spectroscopy at the nanoscale, SISSA, Trieste Italy A postdoctoral position in Computational spectroscopy at the nanoscale is available at SISSA, to work with Stefano Baroni. This position is part of a theoretical effort, coordinated by the DEMOCRITOS National Simulation Center in Trieste, to develop and implement new methods that will extend the scope of ab initio calculations, based on time-dependent density-functional theory and many-body perturbation theory, to systems of hundreds of atoms. The postdoctoral fellow will concentrate on TDDFT and will have the task to implement our newly developed Liouville-Lanczos scheme in a PAW environment, and/or to develop new methods to simulate optical dichroism in extended systems, and/or electron energy-loss spectra, to name a few possible applications. The successful candidate will be part of a strong theoretical group at SISSA and DEMOCRITOS and will have the chance to collaborate with experimentalists at the ELETTRA synchrotron radiation laboratory in Trieste. We are seeking highly qualified candidates with a strong inclination towards methodological innovation, but equally versed in applications and able to speak to and to understand experimentalists. Trieste is a beautifully located medium-sized city on the side of the Adriatic sea. The quality of life in the Trieste area is consistently praised in the reports of past and present PhD students and postdoctoral fellows. Local social life is lively, with cultural and natural attractions (Venice, the Alps, and beautiful seaside resorts) less than two hours away from the center of the city. Although the knowledge of the Italian language will of course facilitate social life, this is not strictly required for professional purposes, as English is the language currently spoken at SISSA. The salary being offered is compatible with a quite decent living standard. The deadline for submitting an application is end of January 2008. Perspective candidates are invited to ask more information and to send their CV to Stefano Baroni, baroni at sissa.it. The formal announcement of the vacancy, with instructions on how to apply can be found at: http://www.sissa.it/main/?p=A3_B4&what=tpd. --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071124/575a8c4c/attachment.htm From njuxuyuehua at gmail.com Sat Nov 24 14:01:40 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Sat, 24 Nov 2007 21:01:40 +0800 Subject: [Pw_forum] a question about how to choose the best relaxed structure Message-ID: hi everyone here : i have a puzzled question about the relaxed structure : i relaxed a structure : ,i set c direction of crystal lattice is 2.7950 and the final result is entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= - 1.06 -0.00000893 -0.00000001 -0.00000001 -1.31 0.00 0.00 -0.00000001 -0.00000892 0.00000000 0.00 -1.31 0.00 -0.00000001 0.00000000 -0.00000382 0.00 0.00 -0.56 bfgs converged in 25 scf cycles and 22 bfgs steps End of BFGS Geometry Optimization Final energy = -137.7926586085 Ry so , i reduced the length of c ,set to 2.760, and the result is : entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= - 0.81 -0.00000882 -0.00000004 0.00000001 -1.30 -0.01 0.00 -0.00000004 -0.00000881 0.00000000 -0.01 -1.30 0.00 0.00000001 0.00000000 0.00000106 0.00 0.00 0.16 bfgs converged in 35 scf cycles and 32 bfgs steps End of BFGS Geometry Optimization Final energy = -137.7921913584 Ry and my problem is which the structure is better ? according to stress ,seems latter,but accrording the energy ,seems the former ? i do not know so please give me some help thank you -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071124/bcdab82d/attachment-0001.htm From scandolo at ictp.it Sat Nov 24 15:18:49 2007 From: scandolo at ictp.it (Sandro Scandolo) Date: Sat, 24 Nov 2007 15:18:49 +0100 Subject: [Pw_forum] a question about how to choose the best relaxed structure In-Reply-To: References: Message-ID: <474832C9.1080705@ictp.it> Dear Xu Yuehua, the calculated stress tensor is not exactly equal to the derivative of the energy if the basis set is not complete. For a detailed discussion see: http://www.quantum-espresso.org/wiki/index.php/Stress_calculations Regards, Sandro xu yuehua wrote: > hi everyone here : > i have a puzzled question about the relaxed structure : > i relaxed a structure : ,i set c direction of crystal lattice is > 2.7950 and the final result is > > entering subroutine stress ... > > total stress (Ry/bohr**3) (kbar) > P= -1.06 > -0.00000893 -0.00000001 -0.00000001 -1.31 0.00 0.00 > -0.00000001 -0.00000892 0.00000000 0.00 - 1.31 0.00 > -0.00000001 0.00000000 -0.00000382 0.00 0.00 -0.56 > > > bfgs converged in 25 scf cycles and 22 bfgs steps > > End of BFGS Geometry Optimization > > Final energy = -137.7926586085 Ry > > so , i reduced the length of c ,set to 2.760, and the result is : > > entering subroutine stress ... > > total stress (Ry/bohr**3) (kbar) > P= -0.81 > -0.00000882 -0.00000004 0.00000001 -1.30 -0.01 0.00 > -0.00000004 -0.00000881 0.00000000 -0.01 - 1.30 0.00 > 0.00000001 0.00000000 0.00000106 0.00 0.00 0.16 > > > bfgs converged in 35 scf cycles and 32 bfgs steps > > End of BFGS Geometry Optimization > > Final energy = -137.7921913584 Ry > > and my problem is which the structure is better ? > according to stress ,seems latter,but accrording the energy ,seems > the former ? > i do not know > so please give me some help > thank you > > > > -- > Xu Yuehua > physics Department of Nanjing university > China > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From sunshoutian168 at yahoo.com.cn Sun Nov 25 01:47:47 2007 From: sunshoutian168 at yahoo.com.cn (shoutian sun) Date: Sun, 25 Nov 2007 08:47:47 +0800 (CST) Subject: [Pw_forum] How to calculate phonon partial density of states with PWscf? Message-ID: <566498.36864.qm@web92001.mail.cnb.yahoo.com> >Dear all, > >Could anyone here tell me how to calculate phonon partial density of states with PWscf? Thank you so much. > ___________________________________________________________ ??1G??????????? http://cn.mail.yahoo.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071125/0a2d30d4/attachment.htm From eyvaz_isaev at yahoo.com Mon Nov 26 10:20:11 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 26 Nov 2007 01:20:11 -0800 (PST) Subject: [Pw_forum] How to calculate phonon partial density of states with PWscf? In-Reply-To: <566498.36864.qm@web92001.mail.cnb.yahoo.com> Message-ID: <725492.31750.qm@web60319.mail.yahoo.com> Dear all, As the request for such a kind problems is inceasing let me resubmit my own code again. And I would like to ask Paolo to consider an opportunity to distribute it together with the main code. Best regards, Eyvaz. --- shoutian sun wrote: > >Dear all, > > > >Could anyone here tell me how to calculate phonon > partial density of states with PWscf? Thank you so > much. > > > > > > > > > > > ___________________________________________________________ > > ?? ???1G? ??????????????????????????????? > http://cn.mail.yahoo.com/> _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Be a better sports nut! Let your teams follow you with Yahoo Mobile. Try it now. http://mobile.yahoo.com/sports;_ylt=At9_qDKvtAbMuh1G1SQtBI7ntAcJ From calderin at physics.queensu.ca Mon Nov 26 14:58:49 2007 From: calderin at physics.queensu.ca (Lazaro Calderin) Date: Mon, 26 Nov 2007 08:58:49 -0500 (EST) Subject: [Pw_forum] Kubo-Greenwood In-Reply-To: <1195824017.3101.170.camel@wolfe> References: <1195824017.3101.170.camel@wolfe> Message-ID: <59841.157.88.38.103.1196085529.squirrel@physics.queensu.ca> Ok guys .. I just programmed the Kubo-Greenwood conductivity formula for any symmetry and any number of k-points. But it is limited to the no-spin and no-uspp calculations. Also the correction for the non-localility of the pseudopotential is not included. If it is of interest to include this code in the espresso package just let me know. Now, I am interested in calculating the correction due to the non-local part of the pseudopotential asap but I need some info on the 'technical' details on how that is implemented in pwscf. I am digging into the code already but it would be faster and better if I can see some mathematical expressions. Thanks, Lazaro > Hello everyone, Could some one please tell me if the Kubo-Greenwood > formula for the optical conductivity has been implemented in pwscf ? > Thanks, Lazaro > > > -- > ============================================================= > Dr. Lazaro Calderin > Dpt. of Theoretical, Atomic and Optical Physics > Univ. of Valladolid, 47005 Valladolid. Spain > Phone: +34 983 423891 Fax: +34 983 423013 > E-mails: calderin at fta.uva.es or calderin at physics.queensu.ca > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From riccardo at fcs.it Mon Nov 26 15:01:26 2007 From: riccardo at fcs.it (riccardo at fcs.it) Date: Mon, 26 Nov 2007 15:01:26 +0100 (CET) Subject: [Pw_forum] Installation problem (ranlib) on Mac Intel -- HELP !!! Message-ID: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> Hello, i have a problem installing Quantum Espresso on my Mac Intel (MacBook Pro). I'm running Inter Fortran Compiler 10.1 and i installed the new gcc GNU compiler 4.3 from http://hpc.sourceforge.net/. For the last i just updated the Developer tools to the last version. The automatic configuration works grat (./configure) and the i run male all. Every is working fine for a while but then the installation stops. The problem is about ranlib. After compiling the flib the log says ..... ranlib: file: libpw.a(vhpsi.o) has no symbols ranlib: file: libpw.a(vhpsi_nc.o) has no symbols ranlib: file: libpw.a(vloc_of_g.o) has no symbols ranlib: file: libpw.a(vloc_psi.o) has no symbols ranlib: file: libpw.a(vpack.o) has no symbols ranlib: file: libpw.a(w0gauss.o) has no symbols ranlib: file: libpw.a(w1gauss.o) has no symbols ranlib: file: libpw.a(wfcinit.o) has no symbols ranlib: file: libpw.a(wgauss.o) has no symbols ranlib: file: libpw.a(write_ns.o) has no symbols ranlib: file: libpw.a(wsweight.o) has no symbols ...... and then compiling pw.x ....... ifort -o pw.x \ pwscf.o ../Modules/atom.o ../Modules/{CUTTED} ../Modules/xml_io_base.o libpw.a ../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a -llapack -latlas ipo: warning #11023: file format not recognized for ../flib/flib.a ipo: warning #11023: file format not recognized for ../iotk/src/libiotk.a ld: warning empty table of contents: ../flib/flib.a (can't load from it) ld: warning empty table of contents: ../iotk/src/libiotk.a (can't load from it) ld: Undefined symbols: _MAIN__ make[1]: *** [pw.x] Error 1 make: *** [pw] Error 2 ......... I'm stuck here and i can't go on. I tried setting ranlib=echo in the make.sys but no fortune.... If any of you can help me i would be very very happy, i'm getting mad !!! Thanks in advance, Riccardo From ferretti.andrea at unimore.it Mon Nov 26 16:58:02 2007 From: ferretti.andrea at unimore.it (Andrea Ferretti) Date: Mon, 26 Nov 2007 16:58:02 +0100 (CET) Subject: [Pw_forum] Installation problem (ranlib) on Mac Intel -- HELP !!! In-Reply-To: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> References: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> Message-ID: Hi Riccardo, it seems your problem is related to the building of .a libraries, more problably due to "ar", instead of "ranlib"... I sort of remeber I got a similar problem on intel Mac, and the solution was to set AR and ARFLAGS to some standard values (in my case they where set in a different, mac-oriented, way), like AR = ar ARFLAGS = ruv actually, don't know if this is your case... anyway, if you didn't already, see also the section related to Mac at: http://www.quantum-espresso.org/wiki/index.php/Installation cheers andrea > > i have a problem installing Quantum Espresso on my Mac Intel (MacBook > Pro). I'm running Inter Fortran Compiler 10.1 and i installed the new > gcc GNU compiler 4.3 from http://hpc.sourceforge.net/. For the last i > just updated the Developer tools to the last version. > > The automatic configuration works grat (./configure) and the i run > male all. Every is working fine for a while but then the installation > stops. The problem is about ranlib. After compiling the flib the log > says > > ..... > ranlib: file: libpw.a(vhpsi.o) has no symbols > ranlib: file: libpw.a(vhpsi_nc.o) has no symbols > ranlib: file: libpw.a(vloc_of_g.o) has no symbols > ranlib: file: libpw.a(vloc_psi.o) has no symbols > ranlib: file: libpw.a(vpack.o) has no symbols > ranlib: file: libpw.a(w0gauss.o) has no symbols > ranlib: file: libpw.a(w1gauss.o) has no symbols > ranlib: file: libpw.a(wfcinit.o) has no symbols > ranlib: file: libpw.a(wgauss.o) has no symbols > ranlib: file: libpw.a(write_ns.o) has no symbols > ranlib: file: libpw.a(wsweight.o) has no symbols > ...... > > and then > > compiling pw.x > > ....... > ifort -o pw.x \ > pwscf.o ../Modules/atom.o ../Modules/{CUTTED} > ../Modules/xml_io_base.o libpw.a ../flib/ptools.a ../flib/flib.a > ../clib/clib.a ../iotk/src/libiotk.a -llapack -latlas > ipo: warning #11023: file format not recognized for ../flib/flib.a > ipo: warning #11023: file format not recognized for ../iotk/src/libiotk.a > ld: warning empty table of contents: ../flib/flib.a (can't load from it) > ld: warning empty table of contents: ../iotk/src/libiotk.a (can't load > from it) > ld: Undefined symbols: > _MAIN__ > make[1]: *** [pw.x] Error 1 > make: *** [pw] Error 2 > ......... > > I'm stuck here and i can't go on. I tried setting ranlib=echo in the > make.sys but no fortune.... > If any of you can help me i would be very very happy, i'm getting mad !!! > > Thanks in advance, > > Riccardo > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From giannozz at nest.sns.it Mon Nov 26 20:15:09 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 26 Nov 2007 20:15:09 +0100 Subject: [Pw_forum] Installation problem (ranlib) on Mac Intel -- HELP !!! In-Reply-To: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> References: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> Message-ID: <0E0675AB-B151-4D71-ADDD-1B3F4752283F@nest.sns.it> On Nov 26, 2007, at 15:01 , riccardo at fcs.it wrote: > ipo: warning #11023: file format not recognized for ../flib/flib.a > ipo: warning #11023: file format not recognized for ../iotk/src/ > libiotk.a you have "leftover" libraries from a previous compilation. "make clean" and re-compile Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From zalahmed at hotmail.com Mon Nov 26 21:47:12 2007 From: zalahmed at hotmail.com (ZEYAD ALAHMED) Date: Mon, 26 Nov 2007 20:47:12 +0000 Subject: [Pw_forum] Installation problem on Intel i32 ubuntu In-Reply-To: <0E0675AB-B151-4D71-ADDD-1B3F4752283F@nest.sns.it> References: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> <0E0675AB-B151-4D71-ADDD-1B3F4752283F@nest.sns.it> Message-ID: Dear pwscf users, I have been trying to install the code "espresso-3.2.3" on my computer (Intel i32 - Ubuntu). I have Intel compilers (ifort 10.1.008), (icc 10.1.008), and mkl lib ( 10.0.011 ). When I issue ./configure , every things are OK. (configure: success). However, when I run (make all) I allways get the same problem : . . . es/wave_base.o ../Modules/xml_io_base.o libpw.a ../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a -lmkl_lapack -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 -lguide -lpthread -lfftw /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined reference to `mkl_serv_load_fun' /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined reference to `mkl_serv_load_dll' make[1]: *** [pw.x] Error 1 make[1]: Leaving directory `/home/A/espresso-3.2.3/PW' make: *** [pw] Error 2 What is wrong and any suggestions! Thanks for help in advance Zeyad Alahmed, PhD College of Science, King Saud University zalahmed at hotmail.com Here are make.sys and config.long files: ------------------------- ############ make.sys ############ # make.sys. Generated from make.sys.in by configure. # compilation rules .SUFFIXES : .SUFFIXES : .o .c .f .f90 # most fortran compilers can directly preprocess c-like directives: use # $(MPIF90) $(F90FLAGS) -c $< # if explicit preprocessing by the C preprocessor is needed, use: # $(CPP) $(CPPFLAGS) $< -o $*.F90 # $(MPIF90) $(F90FLAGS) -c $*.F90 -o $*.o # remember the tabulator in the first column !!! .f90.o: $(MPIF90) $(F90FLAGS) -c $< # .f.o and .c.o: do not modify .f.o: $(F77) $(FFLAGS) -c $< .c.o: $(CC) $(CFLAGS) -c $< # DFLAGS = precompilation options (possible arguments to -D and -U) # used by the C compiler and preprocessor # FDFLAGS = as DFLAGS, for the f90 compiler # See include/defs.h.README for a list of options and their meaning # With the exception of IBM xlf, FDFLAGS = $(DFLAGS) # For IBM xlf, FDFLAGS is the same as DFLAGS with separating commas DFLAGS = -D__INTEL -D__FFTW FDFLAGS = $(DFLAGS) # IFLAGS = how to locate directories where files to be included are # In most cases, IFLAGS = -I../include # If loading an external FFTW library, add the location of FFTW include files IFLAGS = -I../include -lfftw # MODFLAGS = flag used by f90 compiler to locate modules # You need to search for modules in ./, in ../iotk/src, in ../Modules # Some applications also neeed modules in ../PW, ../PH, ../CPV MODFLAGS = -I./ -I../Modules -I../iotk/src \ -I../PW -I../PH -I../CPV # Compilers: fortran-90, fortran-77, C # If a parallel compilation is desired, MPIF90 should be a fortran-90 # compiler that produces executables for parallel execution using MPI # (such as for instance mpif90, mpf90, mpxlf90,...); # otherwise, an ordinary fortran-90 compiler (f90, g95, xlf90, ifort,...) # If you have a parallel machine but no suitable candidate for MPIF90, # try to specify the directory containing "mpif.h" in IFLAGS # and to specify the location of MPI libraries in MPI_LIBS MPIF90 = ifort #F90 = ifort CC = icc F77 = ifort # C preprocessor and preprocessing flags - for explicit preprocessing, # if needed (see the compilation rules above) # preprocessing flags must include DFLAGS and IFLAGS CPP = cpp CPPFLAGS = -P -traditional $(DFLAGS) $(IFLAGS) # compiler flags: C, F90, F77 # C flags must include DFLAGS and IFLAGS # F90 flags must include MODFLAGS, IFLAGS, and FDFLAGS with appropriate syntax CFLAGS = -O3 $(DFLAGS) $(IFLAGS) F90FLAGS = $(FFLAGS) -nomodule -fpp $(FDFLAGS) $(IFLAGS) $(MODFLAGS) FFLAGS = -O2 -assume byterecl # compiler flags without optimization for fortran-77 # the latter is NEEDED to properly compile dlamch.f, used by lapack FFLAGS_NOOPT = -O0 -assume byterecl # Linker and linker-specific flags (if any) # Typically LD coincides with F90 or MPIF90 LD = ifort LDFLAGS = # External Libraries (if any) : blas, lapack, fft, MPI # If you have nothing better, use the local copy : ../flib/blas.a BLAS_LIBS = -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 -lguide -lpthread # The following lapack libraries will be available in flib/ : # ../flib/lapack.a : contains all needed routines # ../flib/lapack_atlas.a: only routines not present in the Atlas library # For IBM machines with essl: load essl BEFORE lapack ! # remember that LAPACK_LIBS precedes BLAS_LIBS in loading order LAPACK_LIBS = -lmkl_lapack # nothing needed here if the the internal copy of FFTW is compiled # (needs -D__FFTW -D__USE_INTERNAL_FFTW in DFLAGS) FFT_LIBS = -lfftw # For parallel execution, the correct path to MPI libraries must # be specified in MPI_LIBS (except for IBM if you use mpxlf) MPI_LIBS = # IBM-specific: MASS libraries, if available and if -D__MASS is defined in FDFLAGS MASS_LIBS = # pgplot libraries (used by some post-processing tools) PGPLOT_LIBS = # ar command and flags - for most architectures: AR = ar, ARFLAGS = ruv # ARFLAGS_DYNAMICS is used in iotk to produce a dynamical library, # for Mac OS-X with PowerPC and xlf compiler. In all other cases # ARFLAGS_DYNAMICS = $(ARFLAGS) AR = ar ARFLAGS = ruv ARFLAGS_DYNAMIC= ruv # ranlib command. If ranlib is not needed (it isn't in most cases) use # RANLIB = echo RANLIB = ranlib # all internal and external libraries - do not modify LIBOBJS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a LIBS = $(LAPACK_LIBS) $(BLAS_LIBS) $(FFT_LIBS) $(MPI_LIBS) $(MASS_LIBS) $(PGPLOT_LIBS) ----------------------------- ############## config.log: ############## This file contains any messages produced by compilers while running configure, to aid debugging if configure makes a mistake. It was created by ESPRESSO configure 3.2, which was generated by GNU Autoconf 2.59. Invocation command line was $ ./configure ## --------- ## ## Platform. ## ## --------- ## hostname = ubuntu-laptop uname -m = i686 uname -r = 2.6.15-29-386 uname -s = Linux uname -v = #1 PREEMPT Mon Sep 24 17:18:25 UTC 2007 /usr/bin/uname -p = unknown /bin/uname -X = unknown /bin/arch = i686 /usr/bin/arch -k = unknown /usr/convex/getsysinfo = unknown hostinfo = unknown /bin/machine = unknown /usr/bin/oslevel = unknown /bin/universe = unknown PATH: /opt/intel/idb/10.1.008/bin PATH: /opt/intel/cc/10.1.008/bin PATH: /opt/intel/fc/10.1.008/bin PATH: /usr/local/sbin PATH: /usr/local/bin PATH: /usr/sbin PATH: /usr/bin PATH: /sbin PATH: /bin PATH: /usr/bin/X11 PATH: /usr/games PATH: /home/zeyad/.local/bin ## ----------- ## ## Core tests. ## ## ----------- ## configure:1395: checking build system type configure:1413: result: i686-pc-linux-gnu configure:1598: checking for ifort configure:1614: found /opt/intel/fc/10.1.008/bin/ifort configure:1624: result: ifort configure:1639: checking for Fortran 77 compiler version configure:1642: ifort --version &5 ifort (IFORT) 10.1 20070913 Copyright (C) 1985-2007 Intel Corporation. All rights reserved. configure:1645: $? = 0 configure:1647: ifort -v &5 Version 10.1 configure:1650: $? = 0 configure:1652: ifort -V &5 Intel(R) Fortran Compiler for applications running on IA-32, Version 10.1 Build 20070913 Package ID: l_fc_p_10.1.008 Copyright (C) 1985-2007 Intel Corporation. All rights reserved. FOR NON-COMMERCIAL USE ONLY configure:1655: $? = 0 configure:1669: checking for Fortran 77 compiler default output file name configure:1672: ifort conftest.f >&5 configure:1675: $? = 0 configure:1721: result: a.out configure:1726: checking whether the Fortran 77 compiler works configure:1752: result: yes configure:1759: checking whether we are cross compiling configure:1761: result: yes configure:1764: checking for suffix of executables configure:1766: ifort -o conftest conftest.f >&5 configure:1769: $? = 0 configure:1794: result: configure:1800: checking for suffix of object files configure:1811: ifort -c conftest.f>&5 configure:1814: $? = 0 configure:1836: result: o configure:1844: checking whether we are using the GNU Fortran 77 compiler configure:1858: ifort -c conftest.F>&5 fortcom: Error: conftest.F, line 3: Syntax error, found END-OF-STATEMENT when expecting one of: => = . ( : % choke me ---------------^ fortcom: Error: conftest.F, line 3: This statement is positioned incorrectly and/or has syntax errors. choke me ---------------^ compilation aborted for conftest.F (code 1) configure:1864: $? = 1 configure: failed program was: | program main | #ifndef __GNUC__ | choke me | #endif | | end .... ... .. . _________________________________________________________________ Put your friends on the big screen with Windows Vista? + Windows Live?. http://www.microsoft.com/windows/shop/specialoffers.mspx?ocid=TXT_TAGLM_CPC_MediaCtr_bigscreen_102007 From cornycoww at gmail.com Tue Nov 27 04:04:57 2007 From: cornycoww at gmail.com (harman dev) Date: Tue, 27 Nov 2007 11:04:57 +0800 Subject: [Pw_forum] honeycomb lattices and shift in gamma point? Message-ID: <5bdab34e0711261904h2b0b3016p122b3c946452176e@mail.gmail.com> Dear all, This is short extension with reference to my previous message. I was wondering if this shift in the origin of the Brillouin Zone for graphene is associated with the honeycomb symmetry. As mentioned, a hcp cell such as that of Mg has the gamma point as the origin of its Brillouin Zone whilst the point K seems to be the origin of the Brillouin Zone for honeycomb lattices such as graphene and boron nitride. Any comments on this would be great appreciated. Thank you. Cheers, Dev Dear all, I am a little perplexed by the definition of some high symmetry k-points in the hexagonal Brillouin Zone (BZ) of say something like graphene and Magnesium (Mg). I've carried out a number of nscf type calculations (to obtain the band structure) for graphene and magnesium. Typically, one defines the gamma-point as (0,0,0) and the K point as say something like (0,0.577,0), given a set of b-vectors, both in units of 2pi/a0. However, I found that whilst this does apply to bulk Mg, which has a hcp unit cell, the same isn't true for graphene. In the case of graphene, it appears that in order to reproduce the bandstructure that we nominally see in literature, the high-symmetry K point has the coordinates (0,0,0) and the gamma point as (0,0.577,0)! It "appears" as if the origin has shifted in that regard and I'm really quite puzzled by this. If I do follow this and be careful with my selection of k-points in the BZ (using the b-vectors to define its boundaries and so on and so forth), I do in fact get the usual band structure of graphene. Whilst this is ok, I'm not quite contented since I'm not sure as to the actual mechanics of how the program works. Is there any explanation as to why this is the case? Note that I am not using a Monkhorst-Pack grid but am actually specifying the exact k-vectors to be used in the nscf type calculation. I would really appreciate any feedback on this and any suggestions.The following is the list of k-points that I used for graphene, which I should say again, actually generates the correct band structure: 0.000 0.000 0.000 1.0 !K 0.000 0.192 0.000 1.0 0.000 0.385 0.000 1.0 0.000 0.577 0.000 1.0 !GAMMA 0.083 0.433 0.000 1.0 0.167 0.288 0.000 1.0 0.250 0.144 0.000 1.0 !M 0.167 0.096 0.000 1.0 0.083 0.048 0.000 1.0 0.000 0.000 0.000 1.0 !K My apologies for the extremely lengthy description of the problem. Thank you very much for your time and happy computing. Cheers, Dev From njuxuyuehua at gmail.com Tue Nov 27 07:39:05 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Tue, 27 Nov 2007 14:39:05 +0800 Subject: [Pw_forum] how to get the *.xyz OR *XSF file which can be used in XCrySDen Message-ID: hi all: i relaxed a structure ,and want to see the details,and the XCrySDen can read the *xyz AND *XSF .SO HOW TO get the above file .my out file does not including the above . thanks for your help. bests -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/07143545/attachment.htm From lanhaiping at gmail.com Tue Nov 27 08:27:05 2007 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 27 Nov 2007 15:27:05 +0800 Subject: [Pw_forum] how to get the *.xyz OR *XSF file which can be used in XCrySDen In-Reply-To: References: Message-ID: Dear Yuehua, xcrysden can read input/output of pwscf . you can achieve just by 'xcrysden --pwi (or -pwo) pwscf.in(pwscf.out)' . good luck Regards. On Nov 27, 2007 2:39 PM, xu yuehua wrote: > > hi all: > i relaxed a structure ,and want to see the details,and the XCrySDen can > read the *xyz AND *XSF .SO HOW TO get the above file .my out file does not > including the above . > thanks for your help. > bests > -- > Xu Yuehua > physics Department of Nanjing university > China > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/90f3aa2f/attachment.htm From njuxuyuehua at gmail.com Tue Nov 27 09:43:12 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Tue, 27 Nov 2007 16:43:12 +0800 Subject: [Pw_forum] how to get the *.xyz OR *XSF file which can be used in XCrySDen In-Reply-To: References: Message-ID: hi : thank for lan haiping , and there is another problem present when i put *.o8888 (a out file produced by machine)to xcrysden ,the software ask me which to see,i choose the initial structure ,and it is OK ,but when i choose to havea look at the optimized and latest structure ,it gives error message :an error occurred while trying to update a structure;error reading /home/xyh/xcry_tmp/xc_8016/xc_structure8016 so i do not know the reason and how to resolve it .if you could give me a piece of advice .i will be thankful. bests 2007/11/27, lan haiping : > > Dear Yuehua, > > xcrysden can read input/output of pwscf . you can achieve just by > 'xcrysden --pwi (or -pwo) pwscf.in(pwscf.out)' . > good luck > > Regards. > > > On Nov 27, 2007 2:39 PM, xu yuehua < njuxuyuehua at gmail.com> wrote: > > > > > hi all: > > i relaxed a structure ,and want to see the details,and the XCrySDen can > > read the *xyz AND *XSF .SO HOW TO get the above file .my out file does not > > including the above . > > thanks for your help. > > bests > > -- > > Xu Yuehua > > physics Department of Nanjing university > > China > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > -- > Hai-Ping Lan > Department of Electronics , > Peking University , Bejing, 100871 > lanhaiping at gmail.com, hplan at pku.edu.cn > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/0a8887ca/attachment.htm From njuxuyuehua at gmail.com Tue Nov 27 10:00:07 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Tue, 27 Nov 2007 17:00:07 +0800 Subject: [Pw_forum] xcrysden error:***.rx.in is not pwscf input file Message-ID: hi everyone: I have complete the relaxed work ,and want to see the details ,but when i put my input file **.rx.in into crysden ,and the error message says:**, rx.in is not pwscf infut file ! and i want to ask what is input file in pwscf? why **.rx.in is not ? i need your help bests -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/1cc483c1/attachment.htm From tone.kokalj at ijs.si Tue Nov 27 10:19:16 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Tue, 27 Nov 2007 10:19:16 +0100 Subject: [Pw_forum] xcrysden error:***.rx.in is not pwscf input file In-Reply-To: References: Message-ID: <1196155156.2678.3.camel@localhost.localdomain> On Tue, 2007-11-27 at 17:00 +0800, xu yuehua wrote: > hi everyone: > I have complete the relaxed work ,and want to see the details ,but > when i put my input file **.rx.in into crysden ,and the error message > says:**,rx.in is not pwscf infut file ! How can anybody say? It is impossible to say anything without the input file that produces the error ! Regards, Tone From giannozz at nest.sns.it Tue Nov 27 11:09:40 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 Nov 2007 11:09:40 +0100 Subject: [Pw_forum] Installation problem on Intel i32 ubuntu In-Reply-To: References: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> <0E0675AB-B151-4D71-ADDD-1B3F4752283F@nest.sns.it> Message-ID: <44F87DF5-10E8-4AEA-9EFC-9768473CAB47@nest.sns.it> On Nov 26, 2007, at 21:47 , ZEYAD ALAHMED wrote: > [...] -lmkl_lapack -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 - > lguide -lpthread -lfftw > /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined > reference to `mkl_serv_load_fun' > /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined > reference to `mkl_serv_load_dll' like states in the Lanczos chain, each new version of Intel software is orthogonal to all preceding versions. You might try Axel's advice for newer MKL: LDFLAGS=-i-static -openmp BLAS_LIBS= -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 (without -lguide -lpthread) If this doesn't work, please read the instructions on how to use your specific version of MKL (they should be included in the MKL library) and let us know what must be done so that we may try to update "configure" Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From nazari at iasbs.ac.ir Tue Nov 27 15:38:14 2007 From: nazari at iasbs.ac.ir (nazari at iasbs.ac.ir) Date: Tue, 27 Nov 2007 14:38:14 -0000 (UTC) Subject: [Pw_forum] Installation problem on Intel i32 ubuntu In-Reply-To: <44F87DF5-10E8-4AEA-9EFC-9768473CAB47@nest.sns.it> References: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> <0E0675AB-B151-4D71-ADDD-1B3F4752283F@nest.sns.it> <44F87DF5-10E8-4AEA-9EFC-9768473CAB47@nest.sns.it> Message-ID: <7343.213.176.122.17.1196174294.squirrel@mail.iasbs.ac.ir> > On Nov 26, 2007, at 21:47 , ZEYAD ALAHMED wrote: > >> [...] -lmkl_lapack -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 - >> lguide -lpthread -lfftw >> /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined >> reference to `mkl_serv_load_fun' >> /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined >> reference to `mkl_serv_load_dll' > > like states in the Lanczos chain, each new version of Intel software > is orthogonal > to all preceding versions. You might try Axel's advice for newer MKL: > > LDFLAGS=-i-static -openmp > BLAS_LIBS= -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 > (without -lguide -lpthread) > > If this doesn't work, please read the instructions on how to use your > specific > version of MKL (they should be included in the MKL library) and let > us know > what must be done so that we may try to update "configure" > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > I just add the follwing lines in .bashrc and the problem lik this is solved. source /opt/intel/cc/10.0.026/bin/iccvars.sh source /opt/intel/fc/10.0.026/bin/ifortvars.sh source /opt/intel/idb/10.0.026/bin/idbvars.sh LD_LIBRARY_PATH=/opt/intel/mkl/9.1.023/lib/32 Regards Fariba Nazari > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From meisam_a84 at yahoo.com Tue Nov 27 12:01:28 2007 From: meisam_a84 at yahoo.com (meisam aghtar) Date: Tue, 27 Nov 2007 03:01:28 -0800 (PST) Subject: [Pw_forum] zigzag nanotube input error Message-ID: <111966.22190.qm@web30602.mail.mud.yahoo.com> hi I had a problem during runing the pw input of (5,0) nanotube . the input and the error is in following. another problem is that I dont know wich parameters should be in a pw input for a nanotube. &CONTROL calculation = 'scf' , restart_mode = 'restart', outdir = '/home/meisam/projhe/5-0/band/ scratch/' pseudo_dir = '/home/meisam/espresso-3.1.1/pseudo/', prefix = 'cnt-scf' , / &SYSTEM ibrav = 4, celldm(1) = 19, celldm(3) = 0.282460, nat = 20, ntyp = 1, ecutwfc = 30, / &ELECTRONS / ATOMIC_SPECIES C 12.00000 C.pz-vbc.UPF ATOMIC_POSITIONS alat C 0.194660000 0.000000000 0.000000000 C 0.157490000 0.114420000 0.070620000 C 0.060150000 0.185130000 0.000000000 C -0.060150000 0.185130000 0.070620000 C -0.157490000 0.114420000 0.000000000 C -0.194660000 0.000000000 0.070620000 C -0.157490000 -0.114420000 0.000000000 C -0.060150000 -0.185130000 0.070620000 C 0.060150000 -0.185130000 0.000000000 C 0.157490000 -0.114420000 0.070620000 C 0.157490000 0.114420000 0.211850000 C 0.060150000 0.185130000 0.282460000 C -0.060150000 0.185130000 0.211850000 C -0.157490000 0.114420000 0.282460000 C -0.194660000 0.000000000 0.211850000 C -0.157490000 -0.114420000 0.282460000 C -0.060150000 -0.185130000 0.211850000 C 0.060150000 -0.185130000 0.282460000 C 0.157490000 -0.114420000 0.211850000 C 0.194660000 0.000000000 0.282460000 K_POINTS automatic 2 2 14 1 1 1 erro : Starting configuration read from file cnt-scf.save Failed to open file cnt-scf.save Using input configuration %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from setup : error # 1 Wrong atomic coordinates %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... ____________________________________________________________________________________ Be a better pen pal. Text or chat with friends inside Yahoo! Mail. See how. http://overview.mail.yahoo.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/6d4b7f43/attachment-0001.htm From marcel at physik.tu-berlin.de Tue Nov 27 12:19:20 2007 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Tue, 27 Nov 2007 12:19:20 +0100 (CET) Subject: [Pw_forum] zigzag nanotube input error In-Reply-To: <111966.22190.qm@web30602.mail.mud.yahoo.com> References: <111966.22190.qm@web30602.mail.mud.yahoo.com> Message-ID: <20071127121536.K56501@rosa.physik-pool.tu-berlin.de> Hi I think your z coordinates are wrong, they are in units of lattice constants. Try to use angstrom as input coordinates - you will see errors much easier. BTW, you only need 1 k-point in the non-repeating directions, as long yuo are not interested in bundling effects. The ectuwfc looks also very small for C, did you check convergence ? Cheers Marcel ________________________________________________________________________ Marcel Mohr Institut f?r Festk?rperphysik, TU Berlin marcel(at)physik.tu-berlin.de Sekr. EW 5-4 TEL: +49-30-314 24442 Hardenbergstr. 36 FAX: +49-30-314 27705 10623 Berlin On Tue, 27 Nov 2007, meisam aghtar wrote: > hi > I had a problem during runing the pw input of (5,0) nanotube . the input and the error is in following. another problem is that I dont know wich parameters should be in a pw input for a nanotube. > &CONTROL > calculation = 'scf' , > restart_mode = 'restart', > outdir = '/home/meisam/projhe/5-0/band/ scratch/' > pseudo_dir = '/home/meisam/espresso-3.1.1/pseudo/', > prefix = 'cnt-scf' , > / > &SYSTEM > ibrav = 4, > celldm(1) = 19, > celldm(3) = 0.282460, > nat = 20, > ntyp = 1, > ecutwfc = 30, > / > > &ELECTRONS > / > ATOMIC_SPECIES > C 12.00000 C.pz-vbc.UPF > ATOMIC_POSITIONS alat > C 0.194660000 0.000000000 0.000000000 > > C 0.157490000 0.114420000 0.070620000 > C 0.060150000 0.185130000 0.000000000 > C -0.060150000 0.185130000 0.070620000 > C -0.157490000 0.114420000 0.000000000 > C -0.194660000 0.000000000 0.070620000 > C -0.157490000 -0.114420000 0.000000000 > C -0.060150000 -0.185130000 0.070620000 > C 0.060150000 -0.185130000 0.000000000 > C 0.157490000 -0.114420000 0.070620000 > C 0.157490000 0.114420000 0.211850000 > C 0.060150000 0.185130000 0.282460000 > C -0.060150000 0.185130000 0.211850000 > C -0.157490000 0.114420000 0.282460000 > C -0.194660000 0.000000000 0.211850000 > C -0.157490000 -0.114420000 0.282460000 > C -0.060150000 -0.185130000 0.211850000 > C 0.060150000 -0.185130000 0.282460000 > C 0.157490000 -0.114420000 0.211850000 > C 0.194660000 0.000000000 0.282460000 > K_POINTS automatic > 2 2 14 1 1 1 > erro : > Starting configuration read from file cnt-scf.save > Failed to open file > cnt-scf.save > Using input configuration > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from setup : error # 1 > Wrong atomic coordinates > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > > > ____________________________________________________________________________________ > Be a better pen pal. > Text or chat with friends inside Yahoo! Mail. See how. http://overview.mail.yahoo.com/ From hande at newton.physics.metu.edu.tr Tue Nov 27 12:21:22 2007 From: hande at newton.physics.metu.edu.tr (Hande Ustunel) Date: Tue, 27 Nov 2007 13:21:22 +0200 (EET) Subject: [Pw_forum] zigzag nanotube input error In-Reply-To: <111966.22190.qm@web30602.mail.mud.yahoo.com> Message-ID: Hi, If you really have the balnk line between the first and the second atomic coordinates that could be your problem. A more severe problem is that the code failed to read the configuration you meant to restart from. This could be due to the space in the name of your outdir, again of course if you really have it there in your input. As for the parameters, first off, there's a single k point should be enough in the plane perp to the NT axis. To determine the no of kpoints along the axis, you should do a convergence study, increasing the number until you are satisfied with the difference of energy values (or whichever value you are interested in). Same goes for Ecut. You are lucky in that a (5,0) has a substantial gap so you don't have to worry about smearing. Your vacuum value seems a little small to me. I usually leave about 10 A between the closest atoms belonging to neigboring images. But then again you should increase this value a little bit and see if it has any effects. These are all the parameters I can think of. Best wishes, Hande Ustunel On Tue, 27 Nov 2007, meisam aghtar wrote: > hi > I had a problem during runing the pw input of (5,0) nanotube . the input and the error is in following. another problem is that I dont know wich parameters should be in a pw input for a nanotube. > &CONTROL > calculation = 'scf' , > restart_mode = 'restart', > outdir = '/home/meisam/projhe/5-0/band/ scratch/' > pseudo_dir = '/home/meisam/espresso-3.1.1/pseudo/', > prefix = 'cnt-scf' , > / > &SYSTEM > ibrav = 4, > celldm(1) = 19, > celldm(3) = 0.282460, > nat = 20, > ntyp = 1, > ecutwfc = 30, > / > > &ELECTRONS > / > ATOMIC_SPECIES > C 12.00000 C.pz-vbc.UPF > ATOMIC_POSITIONS alat > C 0.194660000 0.000000000 0.000000000 > > C 0.157490000 0.114420000 0.070620000 > C 0.060150000 0.185130000 0.000000000 > C -0.060150000 0.185130000 0.070620000 > C -0.157490000 0.114420000 0.000000000 > C -0.194660000 0.000000000 0.070620000 > C -0.157490000 -0.114420000 0.000000000 > C -0.060150000 -0.185130000 0.070620000 > C 0.060150000 -0.185130000 0.000000000 > C 0.157490000 -0.114420000 0.070620000 > C 0.157490000 0.114420000 0.211850000 > C 0.060150000 0.185130000 0.282460000 > C -0.060150000 0.185130000 0.211850000 > C -0.157490000 0.114420000 0.282460000 > C -0.194660000 0.000000000 0.211850000 > C -0.157490000 -0.114420000 0.282460000 > C -0.060150000 -0.185130000 0.211850000 > C 0.060150000 -0.185130000 0.282460000 > C 0.157490000 -0.114420000 0.211850000 > C 0.194660000 0.000000000 0.282460000 > K_POINTS automatic > 2 2 14 1 1 1 > erro : > Starting configuration read from file cnt-scf.save > Failed to open file > cnt-scf.save > Using input configuration > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from setup : error # 1 > Wrong atomic coordinates > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > > > ____________________________________________________________________________________ > Be a better pen pal. > Text or chat with friends inside Yahoo! Mail. See how. http://overview.mail.yahoo.com/ -- Hande Ustunel Department of Physics Office 439 Middle East Technical University Ankara 06531, Turkey Tel : +90 312 210 3264 http://www.physics.metu.edu.tr/~hande From ezadshojaee at hotmail.com Tue Nov 27 13:43:43 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Tue, 27 Nov 2007 12:43:43 +0000 Subject: [Pw_forum] symmetry problem with Si lattice Message-ID: Hi i have a silly problem about Si lattice. I defined it as an SC(simple cubic) with 8 atoms per basis: ATOMIC_POSITIONS Si 0.00 0.00 0.00 Si 0.50 0.50 0.00 Si 0.00 0.50 0.50 Si 0.50 0.00 0.50 Si 0.25 0.25 0.25 Si 0.75 0.75 0.25 Si 0.25 0.75 0.75 Si 0.75 0.25 0.75K_POINTS 4 0.1250000 0.1250000 0.1250000 1.00 0.3750000 0.3750000 0.3750000 1.00 0.3750000 0.1250000 0.1250000 3.00 0.3750000 0.3750000 0.1250000 3.00 the Total Energy is 4-times as great as the Fcc structure total energy (till the 5-th digit), but the code has detected 24 symmetries, with no inversion for the SC one. What's the problem(if there is any)? Should'nt be 48 symmetries? _________________________________________________________________ News, entertainment and everything you care about at Live.com. Get it now! http://www.live.com/getstarted.aspx -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/b9bde83a/attachment.htm From njuxuyuehua at gmail.com Tue Nov 27 14:19:45 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Tue, 27 Nov 2007 21:19:45 +0800 Subject: [Pw_forum] xcrysden error:***.rx.in is not pwscf input file In-Reply-To: <1196155156.2678.3.camel@localhost.localdomain> References: <1196155156.2678.3.camel@localhost.localdomain> Message-ID: ok .i will put my ***.rx.in here :i put it into XCRYSDEN.BUT THE ERROR SAYS::**,rx.in is not pwscf input file ! &CONTROL calculation = "relax", prefix = "h2o", pseudo_dir = "/home/xyh/intel_espresso-3.2/pseudo", outdir = "/home/xyh/intel_espresso-3.2/tmp", / &SYSTEM ibrav = 0, nat = 15, ntyp = 2, ecutwfc = 29.98744D0, ecutrho =200 occupations = "smearing", smearing = "gauss", degauss = 0.003D0, / &ELECTRONS conv_thr = 1.D-9, mixing_beta = 0.2D0, / &IONS pot_extrapolation = "second_order", wfc_extrapolation = "second_order", / CELL_PARAMETERS 33.0694,19.0926 0 0.0 38.1852 0.0 0.0 0.0 5.4820 ATOMIC_SPECIES H 1.0 H.pbe-van_bm.UPF O 15.999 O.pbe-van_bm.UPF ATOMIC_POSITIONS { crystal } H 0.1102739014 -0.0774290299 0.6115517241 H 0.1389927901 -0.0435013455 1.0285862069 H -0.1054664601 0.0215451112 0.6018620690 H -0.1078330245 -0.0413062600 0.2584827586 H -0.1324189999 0.1241055395 1.0127586207 H -0.0719001157 0.1095391668 0.5982068966 H 0.0005944993 -0.0947576457 0.7058965517 H 0.0681044664 -0.1481264906 1.0700000000 H 0.0163544467 0.1248178261 0.2592758621 H 0.0639829858 0.0507411804 0.5938965517 O 0.1321617646 -0.0439521694 0.6873103448 O -0.0925246679 -0.0329455868 0.5714137931 O -0.1200031109 0.1096352188 0.6811034483 O 0.0571462440 -0.1256424289 0.7530000000 O 0.0142279685 0.0991335405 0.5491724138 K_POINTS {automatic} 1 1 6 0 0 0 2007/11/27, Tone Kokalj : > > > On Tue, 2007-11-27 at 17:00 +0800, xu yuehua wrote: > > hi everyone: > > I have complete the relaxed work ,and want to see the details ,but > > when i put my input file **.rx.in into crysden ,and the error message > > says:**,rx.in is not pwscf infut file ! > > How can anybody say? > It is impossible to say anything without the input file that produces > the error ! > > Regards, Tone > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/2b6a3429/attachment.htm From lanhaiping at gmail.com Tue Nov 27 14:34:08 2007 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 27 Nov 2007 21:34:08 +0800 Subject: [Pw_forum] xcrysden error:***.rx.in is not pwscf input file In-Reply-To: References: <1196155156.2678.3.camel@localhost.localdomain> Message-ID: this part is so strange to me !! CELL_PARAMETERS 33.0694,19.0926 0 0.0 38.1852 0.0 0.0 0.0 5.4820 On Nov 27, 2007 9:19 PM, xu yuehua wrote: > ok .i will put my ***.rx.in here :i put it into XCRYSDEN.BUT > THE ERROR SAYS::**,rx.in is not pwscf input file ! > &CONTROL > calculation = "relax", > prefix = "h2o", > pseudo_dir = "/home/xyh/intel_espresso-3.2/pseudo", > outdir = "/home/xyh/intel_espresso-3.2/tmp", > / > &SYSTEM > ibrav = 0, > nat = 15, > ntyp = 2, > ecutwfc = 29.98744D0, > ecutrho =200 > occupations = "smearing", > smearing = "gauss", > degauss = 0.003D0, > / > &ELECTRONS > conv_thr = 1.D-9, > mixing_beta = 0.2D0, > / > &IONS > pot_extrapolation = "second_order", > wfc_extrapolation = "second_order", > / > CELL_PARAMETERS > 33.0694,19.0926 0 > 0.0 38.1852 0.0 > 0.0 0.0 5.4820 > ATOMIC_SPECIES > H 1.0 H.pbe-van_bm.UPF > O 15.999 O.pbe-van_bm.UPF > ATOMIC_POSITIONS { crystal } > H 0.1102739014 -0.0774290299 0.6115517241 > H 0.1389927901 -0.0435013455 1.0285862069 > H -0.1054664601 0.0215451112 0.6018620690 > H -0.1078330245 -0.0413062600 0.2584827586 > H -0.1324189999 0.1241055395 1.0127586207 > H - 0.0719001157 0.1095391668 0.5982068966 > H 0.0005944993 -0.0947576457 0.7058965517 > H 0.0681044664 -0.1481264906 1.0700000000 > H 0.0163544467 0.1248178261 0.2592758621 > H 0.0639829858 0.0507411804 0.5938965517 > O 0.1321617646 -0.0439521694 0.6873103448 > O -0.0925246679 -0.0329455868 0.5714137931 > O -0.1200031109 0.1096352188 0.6811034483 > O 0.0571462440 -0.1256424289 0.7530000000 > O 0.0142279685 0.0991335405 0.5491724138 > K_POINTS {automatic} > 1 1 6 0 0 0 > > > > 2007/11/27, Tone Kokalj : > > > > > On Tue, 2007-11-27 at 17:00 +0800, xu yuehua wrote: > > > hi everyone: > > > I have complete the relaxed work ,and want to see the details ,but > > > when i put my input file **.rx.in into crysden ,and the error message > > > says:**,rx.in is not pwscf infut file ! > > > > How can anybody say? > > It is impossible to say anything without the input file that produces > > the error ! > > > > Regards, Tone > > > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > -- > Xu Yuehua > physics Department of Nanjing university > China > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/877326e4/attachment-0001.htm From giannozz at nest.sns.it Tue Nov 27 14:42:18 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 Nov 2007 14:42:18 +0100 Subject: [Pw_forum] symmetry problem with Si lattice In-Reply-To: References: Message-ID: On Nov 27, 2007, at 13:43 , Ezad Shojaee wrote: > i have a silly problem about Si lattice. I defined it as an SC > (simple cubic) > with 8 atoms per basis: > > ATOMIC_POSITIONS > Si 0.00 0.00 0.00 > Si 0.50 0.50 0.00 > Si 0.00 0.50 0.50 > Si 0.50 0.00 0.50 > Si 0.25 0.25 0.25 > Si 0.75 0.75 0.25 > Si 0.25 0.75 0.75 > Si 0.75 0.25 0.75 > > K_POINTS > 4 > 0.1250000 0.1250000 0.1250000 1.00 > 0.3750000 0.3750000 0.3750000 1.00 > 0.3750000 0.1250000 0.1250000 3.00 > 0.3750000 0.3750000 0.1250000 3.00 > > the Total Energy is 4-times as great as the Fcc structure > total energy (till the 5-th digit), but the code has detected > 24 symmetries, with no inversion for the SC one. What's > the problem(if there is any)? Should'nt be 48 symmetries? there isn't any problem, but there is a subtlety. The 24 symmetries that have no associated fractional translation are found; the remaining 24 have an associated fractional translation (1/4 1/4 1/4) and are also found, but not accepted. The reason is that in a supercell the choice of fractional translations is not unique and you have no (simple) way to be sure that the set of all symmetry operations is actually a group, in the mathematical sense. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Tue Nov 27 14:59:43 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 Nov 2007 14:59:43 +0100 Subject: [Pw_forum] xcrysden error:***.rx.in is not pwscf input file In-Reply-To: References: <1196155156.2678.3.camel@localhost.localdomain> Message-ID: On Nov 27, 2007, at 14:34 , lan haiping wrote: > this part is so strange to me !! > CELL_PARAMETERS > 33.0694,19.0926 0 > 0.0 38.1852 0.0 > 0.0 0.0 5.4820 it is a perfectly acceptable syntax --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From tone.kokalj at ijs.si Tue Nov 27 15:04:05 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Tue, 27 Nov 2007 15:04:05 +0100 Subject: [Pw_forum] xcrysden error:***.rx.in is not pwscf input file In-Reply-To: References: <1196155156.2678.3.camel@localhost.localdomain> Message-ID: <1196172245.2678.26.camel@localhost.localdomain> On Tue, 2007-11-27 at 21:19 +0800, xu yuehua wrote: > ok .i will put my ***.rx.in here :i put it into XCRYSDEN.BUT > THE ERROR SAYS::**,rx.in is not pwscf input file ! On my computer xcrysden says: invalid number: incomprehensible list input apparent state: internal I/O lately reading sequential formatted internal IO Luckily xcrysden is picky enough to complain against your input file, because there is typo therein. In particular: > ATOMIC_POSITIONS { crystal } > H 0.1102739014 -0.0774290299 0.6115517241 > H 0.1389927901 -0.0435013455 1.0285862069 > H -0.1054664601 0.0215451112 0.6018620690 > H -0.1078330245 -0.0413062600 0.2584827586 > H -0.1324189999 0.1241055395 1.0127586207 > H - 0.0719001157 0.1095391668 0.5982068966 Notice the whitespace after minus sign for last x-coordinate: - 0.0719001157 Regards, Tone From ezadshojaee at hotmail.com Tue Nov 27 15:11:41 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Tue, 27 Nov 2007 14:11:41 +0000 Subject: [Pw_forum] symmetry problem with Si lattice Message-ID: Hi in order to have the Gamma-point phonon frequencies, if we use the SC lattice schemma for Si, it seems that we have frequencies rather diffrerent from the FCC schemma. What do you expect of the irreps and the symmetries of the modes(and also their number) of SC schemma with 8-atom basis comparing to the FCC one(with 2 atoms per basis)? what should be the same in two shemmas? _________________________________________________________________ Invite your mail contacts to join your friends list with Windows Live Spaces. It's easy! http://spaces.live.com/spacesapi.aspx?wx_action=create&wx_url=/friends.aspx&mkt=en-us -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/68fa1c66/attachment.htm From njuxuyuehua at gmail.com Tue Nov 27 15:11:55 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Tue, 27 Nov 2007 22:11:55 +0800 Subject: [Pw_forum] xcrysden error:***.rx.in is not pwscf input file In-Reply-To: <1196172245.2678.26.camel@localhost.localdomain> References: <1196155156.2678.3.camel@localhost.localdomain> <1196172245.2678.26.camel@localhost.localdomain> Message-ID: about the format in my inputfile,you see a whitespace,it is just effect here ,in my compute,it is rather in order .i can enclose it if you could accept it . 2007/11/27, Tone Kokalj : > > > On Tue, 2007-11-27 at 21:19 +0800, xu yuehua wrote: > > ok .i will put my ***.rx.in here :i put it into XCRYSDEN.BUT > > THE ERROR SAYS::**,rx.in is not pwscf input file ! > > On my computer xcrysden says: > > invalid number: incomprehensible list input > apparent state: internal I/O > lately reading sequential formatted internal IO > > Luckily xcrysden is picky enough to complain against your input file, > because there is typo therein. In particular: > > > ATOMIC_POSITIONS { crystal } > > H 0.1102739014 -0.0774290299 0.6115517241 > > H 0.1389927901 -0.0435013455 1.0285862069 > > H -0.1054664601 0.0215451112 0.6018620690 > > H -0.1078330245 -0.0413062600 0.2584827586 > > H -0.1324189999 0.1241055395 1.0127586207 > > H - 0.0719001157 0.1095391668 0.5982068966 > > Notice the whitespace after minus sign for last x-coordinate: > - 0.0719001157 > > Regards, Tone > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/c39ad7d7/attachment.htm From raamesh123 at rediffmail.com Tue Nov 27 16:30:01 2007 From: raamesh123 at rediffmail.com (ramesh kumar) Date: 27 Nov 2007 15:30:01 -0000 Subject: [Pw_forum] Error msg while configure Message-ID: <20071127153001.14694.qmail@f5mail-237-203.rediffmail.com> Dear users i am new to pwscf when i try to configure it says success and it recognized all the lib files and compliers. but when i give the command make all it says erry_phase.f90:(.text+0x33): undefined reference to `ln_alloc_' berry_phase.f90:(.text+0x8f): undefined reference to `ln_set_' berry_phase.f90:(.text+0xa2): undefined reference to `ln_activate_' ../Modules/berry_phase.o: In function `__berry_phase__indi_of_ig': berry_phase.f90:(.text+0x220): undefined reference to `ln_ind_' berry_phase.f90:(.text+0x256): undefined reference to `ln_ind_' berry_phase.f90:(.text+0x28c): undefined reference to `ln_ind_' berry_phase.f90:(.text+0x340): undefined reference to `ln_ind_' berry_phase.f90:(.text+0x36d): undefined reference to `ln_ind_' ../Modules/berry_phase.o:berry_phase.f90:(.text+0x3b1): more undefined references to `ln_ind_' follow ../Modules/berry_phase.o: In function `__berry_phase__ln_closeup': berry_phase.f90:(.text+0xe2): undefined reference to `ln_dealloc_' ../Modules/fft_scalar.o: In function `__fft_scalar__cft_1z': fft_scalar.f90:(.text+0xf2): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0x118): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0x1fb): undefined reference to `fft_z_stick_' fft_scalar.f90:(.text+0x258): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x26c): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x2a0): undefined reference to `fft_z_stick_' ../Modules/fft_scalar.o: In function `__fft_scalar__cft_2xy': fft_scalar.f90:(.text+0x49c): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0x4c2): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0x516): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0x53c): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0x64c): undefined reference to `fft_x_stick_' fft_scalar.f90:(.text+0x6e1): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x6f5): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x709): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x71d): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x7b6): undefined reference to `fft_x_stick_' fft_scalar.f90:(.text+0x89d): undefined reference to `fft_y_stick_' fft_scalar.f90:(.text+0x903): undefined reference to `fft_y_stick_' ../Modules/fft_scalar.o: In function `__fft_scalar__cfft3d': fft_scalar.f90:(.text+0xa0e): undefined reference to `create_plan_3d_' fft_scalar.f90:(.text+0xa3f): undefined reference to `create_plan_3d_' fft_scalar.f90:(.text+0xae3): undefined reference to `fftw_inplace_drv_3d_' fft_scalar.f90:(.text+0xb1d): undefined reference to `destroy_plan_3d_' fft_scalar.f90:(.text+0xb31): undefined reference to `destroy_plan_3d_' fft_scalar.f90:(.text+0xbf4): undefined reference to `fftw_inplace_drv_3d_' ../Modules/fft_scalar.o: In function `__fft_scalar__cfft3ds': fft_scalar.f90:(.text+0xd8c): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0xdb5): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0xdde): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0xe07): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0xe30): undefined reference to `create_plan_1d_' ../Modules/fft_scalar.o:fft_scalar.f90:(.text+0xe59): more undefined references to `create_plan_1d_' follow ../Modules/fft_scalar.o: In function `__fft_scalar__cfft3ds': fft_scalar.f90:(.text+0xf51): undefined reference to `fftw_inplace_drv_1d_' fft_scalar.f90:(.text+0x1018): undefined reference to `fftw_inplace_drv_1d_' fft_scalar.f90:(.text+0x105c): undefined reference to `fftw_inplace_drv_1d_' fft_scalar.f90:(.text+0x1095): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x10bc): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x10e3): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x110a): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x1131): undefined reference to `destroy_plan_1d_' ../Modules/fft_scalar.o:fft_scalar.f90:(.text+0x1158): more undefined references to `destroy_plan_1d_' follow ../Modules/fft_scalar.o: In function `__fft_scalar__cfft3ds': fft_scalar.f90:(.text+0x1243): undefined reference to `fftw_inplace_drv_1d_' fft_scalar.f90:(.text+0x12d6): undefined reference to `fftw_inplace_drv_1d_' fft_scalar.f90:(.text+0x137b): undefined reference to `fftw_inplace_drv_1d_' ../Modules/fft_scalar.o: In function `__fft_scalar__cft_b': fft_scalar.f90:(.text+0x14eb): undefined reference to `create_plan_1d_' fft_scalar.f90:(.text+0x152c): undefined reference to `create_plan_2d_' fft_scalar.f90:(.text+0x15da): undefined reference to `fftw_inplace_drv_1d_' fft_scalar.f90:(.text+0x161c): undefined reference to `fftw_inplace_drv_2d_' fft_scalar.f90:(.text+0x163f): undefined reference to `destroy_plan_1d_' fft_scalar.f90:(.text+0x1653): undefined reference to `destroy_plan_2d_' ../Modules/xml_io_base.o: In function `__xml_io_base__create_directory': xml_io_base.f90:(.text+0x20a): undefined reference to `c_mkdir_' libpw.a(input.o): In function `verify_tmpdir_': input.f90:(.text+0x756): undefined reference to `c_mkdir_' libpw.a(bp_c_phase.o): In function `c_phase_': bp_c_phase.f90:(.text+0x107a): undefined reference to `ylm_q_' libpw.a(c_phase_field.o): In function `c_phase_field_': c_phase_field.f90:(.text+0x20cf): undefined reference to `ylm_q_' libpw.a(h_epsi_her.o): In function `h_epsi_her_': h_epsi_her.f90:(.text+0xdf5): undefined reference to `ylm_q_' h_epsi_her.f90:(.text+0x1341): undefined reference to `ylm_q_' collect2: ld returned 1 exit status make[1]: *** [pw.x] Error 1 make[1]: Leaving directory `/home/ramesh/espresso-3.2.3/PW' make: *** [pw] Error 2 can anybody help in that thanks in advance with rgds K.Ramesh Kumar Research scholar IIT Madras Chennai-36 India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/d14875a4/attachment-0001.htm From degironc at sissa.it Tue Nov 27 16:41:08 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Tue, 27 Nov 2007 16:41:08 +0100 Subject: [Pw_forum] Error msg while configure In-Reply-To: <20071127153001.14694.qmail@f5mail-237-203.rediffmail.com> References: <20071127153001.14694.qmail@f5mail-237-203.rediffmail.com> Message-ID: <474C3A94.8050701@sissa.it> ramesh kumar wrote: > > > > Dear users > > i am new to pwscf when i try to configure it says success and it > recognized all the lib files and compliers. > > but when i give the command make all > > it says > > erry_phase.f90:(.text+0x33): undefined reference to `ln_alloc_' > berry_phase.f90:(.text+0x8f): undefined refer > have you tried make clean; ./configure ; make all ? stefano From riccardo at fcs.it Tue Nov 27 16:40:54 2007 From: riccardo at fcs.it (Riccardo Sabatini) Date: Tue, 27 Nov 2007 16:40:54 +0100 Subject: [Pw_forum] Still problems on my intel mac .... again HELP !!!! Message-ID: <474C3A86.9050909@fcs.it> Hi everyone, first of all many many thanks for the help you gave me, but still something id wrong :((( I tried all the suggestions you gave me. I cleaned with make clean, i even deleted the directory and started from scratch several times but nothing different, still the problem at the library. I used the make.sys-mac-intel suggestions, i mixed them and my configuration (i don't have intel mkl). Still everything works great but when is the time of compiling pw.x ---------- ifort -o pw.x \ pwscf.o ../Modules/at{***CUTTED***}ml_io_base.o libpw.a ../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a -llapack -latlas ipo: warning #11023: file format not recognized for ../flib/flib.a ipo: warning #11023: file format not recognized for ../iotk/src/libiotk.a ld: warning empty table of contents: ../flib/flib.a (can't load from it) ld: warning empty table of contents: ../iotk/src/libiotk.a (can't load from it) ld: Undefined symbols: _MAIN__ make[1]: *** [pw.x] Error 1 make: *** [pw] Error 2 ---------- What i understood id that there is a problem either with ranlib or ar. I tried different solutions and configurations in make.sys. I tried AR = ar ARFLAGS = ruv RANLIB = ranlib No luck, same error. Then i tried (as in the make.sys-mac-intel) AR = ar ARFLAGS = ruv RANLIB = echo With this i recive a strange error, something about "ar error, cannot use the argument k". I found that that configuration doesn't work because no arguments are passed to ar, expect if i add ARFLAGS_DYNAMIC= ruv In this case the ar works fine but then again the same error !!!!!!!!!! Sorry but i'm getting crazy...........any help is very appreciated ! By the way, thanks again for the help, i rellay appreciate. Riccardo P.S.: I give you here somethig that can help. The error comes out right after ar creates the libraries at the first time. When ar runs both in iotk, libpw or any other one i have the same strange message. I write down the case for iotk (the main behaviour happens everytime ar runs) --------------- ar ruv libpw.a a2fmo***CUT***wsweight.o ar: creating archive libpw.a ***CUT*** a - a2fmod.o a - wsweight.o ranlib: file: libpw.a(a2fmod.o) has no symbols ranlib: file: libpw.a(add_bfield.o) has no symbols ***CUT*** ranlib: file: libpw.a(write_ns.o) has no symbols ranlib: file: libpw.a(wsweight.o) has no symbols echo libpw.a libpw.a --------------- in this case i used RANLIB = echo but seems ranlib runs anyway.....i'm starting to think something spooky is going on in my mac !!!!!!!! From paulatto at sissa.it Tue Nov 27 16:48:40 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Tue, 27 Nov 2007 16:48:40 +0100 (CET) Subject: [Pw_forum] Still problems on my intel mac .... again HELP !!!! In-Reply-To: <474C3A86.9050909@fcs.it> References: <474C3A86.9050909@fcs.it> Message-ID: <11510.147.122.5.182.1196178520.squirrel@webmail.sissa.it> On Tue, November 27, 2007 16:40, Riccardo Sabatini wrote: > ipo: warning #11023: file format not recognized for ../flib/flib.a > ipo: warning #11023: file format not recognized for ../iotk/src/libiotk.a In my experience ipo (Interprocedural Optimizations) is an endless source of problems, especially for ifort 8.0-8.1. You may try to disable it by adding the flag -no-ipo to the F90FLAGS in you make.sys file. -- Lorenzo Paulatto +39 040 3787 511 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From giannozz at nest.sns.it Tue Nov 27 16:56:33 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 Nov 2007 16:56:33 +0100 Subject: [Pw_forum] Still problems on my intel mac .... again HELP !!!! In-Reply-To: <474C3A86.9050909@fcs.it> References: <474C3A86.9050909@fcs.it> Message-ID: <4A9251DC-7B79-4780-B137-972FBA8D83E4@nest.sns.it> On Nov 27, 2007, at 16:40 , Riccardo Sabatini wrote: > ranlib: file: libpw.a(a2fmod.o) has no symbols ls -l a2fmod.o nm a2fmod.o > i'm starting to think something spooky is going on in my mac !!!!!!!! not impossible. I have a Mac and Q-E compiles perfectly on it with g95, gfortran, ifort, but I have to say that the reliability of its software is waaaay below my expectations. P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From katalin.gaal-nagy at physik.uni-regensburg.de Tue Nov 27 19:00:28 2007 From: katalin.gaal-nagy at physik.uni-regensburg.de (Katalin Gaal-Nagy) Date: Tue, 27 Nov 2007 19:00:28 +0100 (CET) Subject: [Pw_forum] Units of conv_thr and tr2_ph In-Reply-To: <4A9251DC-7B79-4780-B137-972FBA8D83E4@nest.sns.it> References: <474C3A86.9050909@fcs.it> <4A9251DC-7B79-4780-B137-972FBA8D83E4@nest.sns.it> Message-ID: Dear all, I have just a little question about the units of the parameters for the scf cycles: Are the units of conv_thr and tr2_ph Ry, meaning that the self consistency is checked agains the energy of the system or are they different? All the best, Katalin ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, , >>>>>>> Katalin Gaal-Nagy <<<<<<< , , Phys.Dept./INFM , , University of Milan , , Via Celoria 16 , , +39 02 50317377 , ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, , >>>>>>> katalin.gaal-nagy at physik.uni-regensburg.de <<<<<<< , ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, , >>>>>>> Don't dream your life, live your dream <<<<<<< , ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, From raamesh123 at rediffmail.com Tue Nov 27 18:56:25 2007 From: raamesh123 at rediffmail.com (ramesh kumar) Date: 27 Nov 2007 17:56:25 -0000 Subject: [Pw_forum] Using make clean Message-ID: <20071127175625.18506.qmail@f5mail-237-201.rediffmail.com> Dear sir, i have tried with make clean but i am getting the same error again. when i configure it says checking build system type... i686-pc-linux-gnu checking architecture... ia32 checking for ifort... no checking for ifc... no checking for pgf90... no checking for pathf95... no checking for g95... no checking for gfortran... gfortran checking for Fortran 77 compiler default output file name... a.out checking whether the Fortran 77 compiler works... yes checking whether we are cross compiling... yes checking for suffix of executables... checking for suffix of object files... o checking whether we are using the GNU Fortran 77 compiler... yes checking whether gfortran accepts -g... yes checking for mpif90... no checking for gfortran... gfortran checking whether we are using the GNU Fortran 77 compiler... yes checking whether gfortran accepts -g... yes checking version of gfortran... unknown setting F90... gfortran setting MPIF90... gfortran checking for cc... cc checking whether we are using the GNU C compiler... yes checking whether cc accepts -g... yes checking for cc option to accept ANSI C... none needed setting CC... cc checking how to run the C preprocessor... cc -E checking for egrep... grep -E checking for ANSI C header files... yes checking for sys/types.h... yes checking for sys/stat.h... yes checking for stdlib.h... yes checking for string.h... yes checking for memory.h... yes checking for strings.h... yes checking for inttypes.h... yes checking for stdint.h... yes checking for unistd.h... yes checking for int *... yes checking size of int *... 4 checking malloc.h usability... yes checking malloc.h presence... yes checking for malloc.h... yes checking for struct mallinfo.arena... yes checking for f77... f77 checking whether we are using the GNU Fortran 77 compiler... yes checking whether f77 accepts -g... yes setting F77... f77 setting FFLAGS... -O3 setting F90FLAGS... $(FFLAGS) -x f95-cpp-input setting FFLAGS_NOOPT... -O0 setting CFLAGS... -O3 setting CPP... cpp setting CPPFLAGS... -P -traditional setting LD... gfortran setting LDFLAGS... setting AR... ar setting ARFLAGS... ruv setting ARFLAGS_DYNAMIC...ruv checking whether make sets $(MAKE)... yes checking whether Fortran files must be preprocessed... no checking how to get verbose linking output from f77... -v checking for Fortran libraries of f77... -L/usr/lib/gcc/i386-redhat-linux/3.4.6 -L/usr/lib/gcc/i386-redhat-linux/3.4.6/../../.. -lfrtbegin -lg2c -lm -lgcc_s checking for dummy main to link with Fortran libraries... none checking for Fortran name-mangling scheme... lower case, underscore, extra underscore checking for library containing dgemm... no in /opt/intel/mkl/*/lib/32: checking for library containing dgemm... no in /opt/intel/mkl*/lib/32: checking for library containing dgemm... no in /cineca/prod/intel/lib: checking for library containing dgemm... no checking for library containing dgemm... no in /usr/local/lib: checking for library containing dgemm... no checking for library containing dgemm... -lblas checking for library containing zggev... -llapack setting BLAS_LIBS... -lblas setting LAPACK_LIBS... -llapack checking for library containing fftwnd... -lfftw setting FFT_LIBS... -lfftw checking for library containing fftwnd... (cached) -lfftw checking fftw.h usability... yes checking fftw.h presence... yes checking for fftw.h... yes setting MASS_LIBS... checking for library containing mpi_init... no setting MPI_LIBS... checking for library containing mpi_init... (cached) no setting DFLAGS... -D__FFTW setting IFLAGS... -I../include setting FDFLAGS... $(DFLAGS) checking for ranlib... ranlib setting RANLIB... ranlib configure: creating ./config.status config.status: creating include/fft_defs.h config.status: creating make.sys config.status: creating configure.msg config.status: creating include/c_defs.h config.status: include/c_defs.h is unchanged -------------------------------------------------------------------- ESPRESSO can take advantage of several optimized numerical libraries (essl, fftw, mkl...). This configure script attempts to find them, but may fail if they have been installed in non-standard locations. If a required library is not found, the local copy will be compiled. The following libraries have been found: BLAS_LIBS= -lblas LAPACK_LIBS= -llapack FFT_LIBS= -lfftw Please check if this is what you expect. If any libraries are missing, you may specify a list of directories to search and retry, as follows: ./configure LIBDIRS="list of directories, separated by spaces" Parallel environment not detected (is this a parallel machine?). Configured for compilation of serial executables. For more info, read the ESPRESSO User's Guide (Doc/users-guide.tex). -------------------------------------------------------------------- configure: success but make all giving the same error my system is having ia32, gfortran, fortran77 support, lapack, blas and fftw waiting for your reply sir with regards K. Ramesh Kumar On Tue, 27 Nov 2007 Stefano de Gironcoli wrote : >ramesh kumar wrote: >> >> >>Dear users >> >>i am new to pwscf when i try to configure it says success and it recognized all the lib files and compliers. >> >>but when i give the command make all >> >>it says >> >> erry_phase.f90:(.text+0x33): undefined reference to `ln_alloc_' >>berry_phase.f90:(.text+0x8f): undefined refer >> >have you tried make clean; ./configure ; make all ? >stefano > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/dbe93b55/attachment.htm From zalahmed at hotmail.com Tue Nov 27 19:05:44 2007 From: zalahmed at hotmail.com (ZEYAD ALAHMED) Date: Tue, 27 Nov 2007 18:05:44 +0000 Subject: [Pw_forum] Installation problem on Intel i32 ubuntu In-Reply-To: <7343.213.176.122.17.1196174294.squirrel@mail.iasbs.ac.ir> References: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> <0E0675AB-B151-4D71-ADDD-1B3F4752283F@nest.sns.it> <44F87DF5-10E8-4AEA-9EFC-9768473CAB47@nest.sns.it> <7343.213.176.122.17.1196174294.squirrel@mail.iasbs.ac.ir> Message-ID: Dear all, Both methods suggested below did not help. The problem has not been solved yet. Any idea! Zeyad Alahmed, PhD ---------------------------------------- > Date: Tue, 27 Nov 2007 14:38:14 +0000 > From: nazari at iasbs.ac.ir > To: pw_forum at pwscf.org > Subject: Re: [Pw_forum] Installation problem on Intel i32 ubuntu > >> On Nov 26, 2007, at 21:47 , ZEYAD ALAHMED wrote: >> >>> [...] -lmkl_lapack -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 - >>> lguide -lpthread -lfftw >>> /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined >>> reference to `mkl_serv_load_fun' >>> /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined >>> reference to `mkl_serv_load_dll' >> >> like states in the Lanczos chain, each new version of Intel software >> is orthogonal >> to all preceding versions. You might try Axel's advice for newer MKL: >> >> LDFLAGS=-i-static -openmp >> BLAS_LIBS= -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 >> (without -lguide -lpthread) >> >> If this doesn't work, please read the instructions on how to use your >> specific >> version of MKL (they should be included in the MKL library) and let >> us know >> what must be done so that we may try to update "configure" >> >> Paolo >> --- >> Paolo Giannozzi, Dept of Physics, University of Udine >> via delle Scienze 208, 33100 Udine, Italy >> Phone +39-0432-558216, fax +39-0432-558222 >> >> I just add the follwing lines in .bashrc and the problem lik this > > is solved. > source /opt/intel/cc/10.0.026/bin/iccvars.sh > source /opt/intel/fc/10.0.026/bin/ifortvars.sh > source /opt/intel/idb/10.0.026/bin/idbvars.sh > LD_LIBRARY_PATH=/opt/intel/mkl/9.1.023/lib/32 > > > Regards > Fariba Nazari > >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum _________________________________________________________________ Connect and share in new ways with Windows Live. http://www.windowslive.com/connect.html?ocid=TXT_TAGLM_Wave2_newways_112007 From giannozz at nest.sns.it Tue Nov 27 20:29:38 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 Nov 2007 20:29:38 +0100 Subject: [Pw_forum] Installation problem on Intel i32 ubuntu In-Reply-To: References: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> <0E0675AB-B151-4D71-ADDD-1B3F4752283F@nest.sns.it> <44F87DF5-10E8-4AEA-9EFC-9768473CAB47@nest.sns.it> <7343.213.176.122.17.1196174294.squirrel@mail.iasbs.ac.ir> Message-ID: <2199F5D5-FAC4-450B-A00B-B0C773303E3F@nest.sns.it> On Nov 27, 2007, at 19:05 , ZEYAD ALAHMED wrote: > Both methods suggested below did not help. The problem has not been > solved yet. >>> You might try Axel's advice for newer MKL: ...this was the first suggestion... >>> If this doesn't work, please read the instructions on how to use >>> your >>> specific version of MKL (they should be included in the MKL library) >>> and let us know what must be done so that we may try to update >>> "configure" ...and this was the second. Apparently your software installation has some weirdness that prevents linking of MKL libraries, using the syntax that is known to work in many other cases. If you do not find out how to link YOUR specific version of MKL on YOUR machine, it is very unlikely that anybody else can. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From jnkim at ncsa.uiuc.edu Tue Nov 27 20:36:58 2007 From: jnkim at ncsa.uiuc.edu (Jeongnim Kim) Date: Tue, 27 Nov 2007 13:36:58 -0600 Subject: [Pw_forum] Installation problem on Intel i32 ubuntu In-Reply-To: <7343.213.176.122.17.1196174294.squirrel@mail.iasbs.ac.ir> References: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> <0E0675AB-B151-4D71-ADDD-1B3F4752283F@nest.sns.it> <44F87DF5-10E8-4AEA-9EFC-9768473CAB47@nest.sns.it> <7343.213.176.122.17.1196174294.squirrel@mail.iasbs.ac.ir> Message-ID: <8e1b43770711271136uaae81cbi77424ddc71f9a28a@mail.gmail.com> Adding "-lmkl_sequential" may work. libmkl_sequential.so:0000000000145340 T mkl_serv_load_fun On Nov 27, 2007 8:38 AM, wrote: > > On Nov 26, 2007, at 21:47 , ZEYAD ALAHMED wrote: > > > >> [...] -lmkl_lapack -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 - > >> lguide -lpthread -lfftw > >> /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined > >> reference to `mkl_serv_load_fun' > >> /opt/intel/mkl/10.0.011/lib/32/libmkl_lapack.so: undefined > >> reference to `mkl_serv_load_dll' > > > > like states in the Lanczos chain, each new version of Intel software > > is orthogonal > > to all preceding versions. You might try Axel's advice for newer MKL: > > > > LDFLAGS=-i-static -openmp > > BLAS_LIBS= -L/opt/intel/mkl/10.0.011/lib/32 -lmkl_ia32 > > (without -lguide -lpthread) > > > > If this doesn't work, please read the instructions on how to use your > > specific > > version of MKL (they should be included in the MKL library) and let > > us know > > what must be done so that we may try to update "configure" > > > > Paolo > > --- > > Paolo Giannozzi, Dept of Physics, University of Udine > > via delle Scienze 208, 33100 Udine, Italy > > Phone +39-0432-558216, fax +39-0432-558222 > > > > I just add the follwing lines in .bashrc and the problem lik this > > is solved. > source /opt/intel/cc/10.0.026/bin/iccvars.sh > source /opt/intel/fc/10.0.026/bin/ifortvars.sh > source /opt/intel/idb/10.0.026/bin/idbvars.sh > LD_LIBRARY_PATH=/opt/intel/mkl/9.1.023/lib/32 > > > Regards > Fariba Nazari > > > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Dr. Jeongnim Kim NCSA and MCC, University of Illinois at Urbana-Champaign (Tel) 217-244-4971 From giannozz at nest.sns.it Tue Nov 27 20:39:49 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 Nov 2007 20:39:49 +0100 Subject: [Pw_forum] Using make clean In-Reply-To: <20071127175625.18506.qmail@f5mail-237-201.rediffmail.com> References: <20071127175625.18506.qmail@f5mail-237-201.rediffmail.com> Message-ID: On Nov 27, 2007, at 18:56 , ramesh kumar wrote: > checking for Fortran name-mangling scheme... lower case, > underscore, extra underscore go to include/c_defs.h . There should be two lines: #define F77_FUNC(name,NAME) name ## _ #define F77_FUNC_(name,NAME) name ## __ remove the second underscore. Not sure why, it looks like configure doesn't get the correct number of underscores to add (note that all missing symbols have one underscore in the name; some genius has decided that they deserve one more at the end, go figure why) --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Tue Nov 27 20:42:00 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 Nov 2007 20:42:00 +0100 Subject: [Pw_forum] Installation problem on Intel i32 ubuntu In-Reply-To: <8e1b43770711271136uaae81cbi77424ddc71f9a28a@mail.gmail.com> References: <59456.192.168.2.101.1196085686.squirrel@mail.fcslab.com> <0E0675AB-B151-4D71-ADDD-1B3F4752283F@nest.sns.it> <44F87DF5-10E8-4AEA-9EFC-9768473CAB47@nest.sns.it> <7343.213.176.122.17.1196174294.squirrel@mail.iasbs.ac.ir> <8e1b43770711271136uaae81cbi77424ddc71f9a28a@mail.gmail.com> Message-ID: On Nov 27, 2007, at 20:36 , Jeongnim Kim wrote: > Adding "-lmkl_sequential" may work. > > libmkl_sequential.so:0000000000145340 T mkl_serv_load_fun have fun loading mkl :-( --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at nest.sns.it Tue Nov 27 20:44:39 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 Nov 2007 20:44:39 +0100 Subject: [Pw_forum] Units of conv_thr and tr2_ph In-Reply-To: References: <474C3A86.9050909@fcs.it> <4A9251DC-7B79-4780-B137-972FBA8D83E4@nest.sns.it> Message-ID: <2C448878-5158-4F06-B50F-C3398813DA42@nest.sns.it> On Nov 27, 2007, at 19:00 , Katalin Gaal-Nagy wrote: > Are the units of conv_thr and tr2_ph Ry yes > meaning that the self consistency is checked agains the energy > of the system or are they different? this is a more complicated issue. conv_thr is an estimate of the error on the energy due to self-consistency. tr2_ph is an estimate of...I don't remember exactly what, but it is NOT (unfortunately) related to an error on any energy, but to the difference of input and output potentials Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From pqshen at gmail.com Wed Nov 28 00:10:54 2007 From: pqshen at gmail.com (PeiQing Sheng) Date: Tue, 27 Nov 2007 23:10:54 +0000 Subject: [Pw_forum] Compiling PWscf on AMD64 using PGI compilors Message-ID: <7218ea700711271510uc622149r1e85e87dcbd4e6a@mail.gmail.com> Hi, I tried to compile PWscf on a AMD64 cluster with PGI 6.2 and MPICH compilers installed. I used the following configurations: ./configure --enable-parallel ARCH=amd64 MPIF90=mpif90 F90=pgf90 MPIF77=mpif77 F77=pgf77 MPICC=mpicc CC=cc The configuration is successful. The libraries for BLAS and LAPACK are detected as acml on the amd64 Linux cluster, the detected information as following: arch = x86_64-unknown-linux-gnu BLAS_LIBS = -lamcl LAPACK_LIBS = -lacml To install the package, I add the following contents to the make.sys file" IFLAGS = -I../include -I/bigdisk/staff/pshen/FFTW2/include FFT_LIBS = -L/bigdisk/staff/pshen/FFTW2/lib -lfftw_mpi -lfftw MPI_LIBS = -L/opt/mpich.pgi64/lib -lmpich -lmpichf90 After running 'make all', the package was installed without error message. But it looks like I could only run pw.x on one processor, and it will give error message when running any other executables when I tried to run all those examples. I guess if PGI compilers and MPICH compilers give me trouble. Would you please let me know how to fix these problems? (see atached file: make.sysand example02_np1.log. Note: np1 means using 'mpirun -np 1' as prefix) Thanks a lot. Peiqing Shen -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/81372a39/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: example02_np1.log Type: text/x-log Size: 2950 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20071127/81372a39/attachment.bin -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: make.sys.text Url: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/81372a39/attachment.text From rcjhawk at gmail.com Wed Nov 28 02:12:52 2007 From: rcjhawk at gmail.com (Michael Mehl) Date: Tue, 27 Nov 2007 20:12:52 -0500 Subject: [Pw_forum] honeycomb lattices and shift in gamma point? In-Reply-To: <5bdab34e0711261904h2b0b3016p122b3c946452176e@mail.gmail.com> References: <5bdab34e0711261904h2b0b3016p122b3c946452176e@mail.gmail.com> Message-ID: <474CC094.3020607@gmail.com> Dev, I believe that if you look closely at the published band structures of graphene you'll find that many graphs are centered on the K point. This doesn't mean that the K point is now the origin, it just means that the K point is the center of the published graph. That's because K is the interesting part of the band structure. A quick Google (TM) of the web reveals more about graphene than I really wanted to know tonight. Here's a useful tutorial: http://www.ph.utexas.edu/~macdgrp/meetings/hongki-graphene(partial).pdf Note particularly the band structure on page 8, and compare it to the band structure on page 19. harman dev wrote: > Dear all, > > This is short extension with reference to my previous message. I > was wondering if this shift in the origin of the Brillouin Zone for > graphene is associated with the honeycomb symmetry. As mentioned, a > hcp cell such as that of Mg has the gamma point as the origin of its > Brillouin Zone whilst the point K seems to be the origin of the > Brillouin Zone for honeycomb lattices such as graphene and boron > nitride. Any comments on this would be great appreciated. Thank you. > > Cheers, > Dev > -- Mike Mehl Naval Research Laboratory Washington DC USA From cornycoww at gmail.com Wed Nov 28 03:48:29 2007 From: cornycoww at gmail.com (harman dev) Date: Wed, 28 Nov 2007 10:48:29 +0800 Subject: [Pw_forum] Pw_forum Digest, Vol 5, Issue 53 In-Reply-To: References: Message-ID: <5bdab34e0711271848sc8aa6a4wac6fd01dd94af923@mail.gmail.com> Thanks for your reply Mike. Yes I do realize that there are many instances where the reference is taken with respect to the high-symmetry point K. In those cases, the x-axis is usually taken as (k-K), where small case k is some k vector taken with respect to the k-vector that is associated with the point K found at the corner of the hexagonal Brillouin zone. However what I do not get is why in the pw.x program, the point K is not something like (1/2,1/sqrt(3))*2pi/a, in Cartesian coordinates. There are of course other possible combinations, depending on whether you have defined your real space lattice vectors using a gamma angle of 120 or 60 degrees. But that should not negate the fact that the true gamma point should have a k-vector of (0,0). where one would typically observe a large gap between the pi and pi* bands. Instead, and as mentioned earlier, the pwscf program (pw.x) calculates the band eigenvalues at each k-point and as it turns out, it is at and about the k-point (0,0) that the Dirac cone results. This is what I find strange and puzzling and I'm wondering if the program makes this artificial "shift" for computational efficiency purposes or for some other reason that has eluded me. Thanks to all for your time. Cheers, Dev > Dev, > > I believe that if you look closely at the published band structures of > graphene you'll find that many graphs are centered on the K > point. This doesn't mean that the K point is now the origin, it just > means that the K point is the center of the published graph. That's > because K is the interesting part of the band structure. > > A quick Google (TM) of the web reveals more about graphene than I really > wanted to know tonight. Here's a useful tutorial: > > http://www.ph.utexas.edu/~macdgrp/meetings/hongki-graphene(partial).pdf > > Note particularly the band structure on page 8, and compare it to the > band structure on page 19. > > > harman dev wrote: > > Dear all, > > > > This is short extension with reference to my previous message. I > > was wondering if this shift in the origin of the Brillouin Zone for > > graphene is associated with the honeycomb symmetry. As mentioned, a > > hcp cell such as that of Mg has the gamma point as the origin of its > > Brillouin Zone whilst the point K seems to be the origin of the > > Brillouin Zone for honeycomb lattices such as graphene and boron > > nitride. Any comments on this would be great appreciated. Thank you. > > > > Cheers, > > Dev > > > > > -- > Mike Mehl > Naval Research Laboratory > Washington DC USA > > > ------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > End of Pw_forum Digest, Vol 5, Issue 53 > *************************************** > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071128/74acf99e/attachment.htm From hushujun at mail.sdu.edu.cn Wed Nov 28 04:57:17 2007 From: hushujun at mail.sdu.edu.cn (=?gb2312?B?uvrK9778IFNodS1qdW4gSHU=?=) Date: Wed, 28 Nov 2007 11:57:17 +0800 Subject: [Pw_forum] Strange integral magnetic moment Message-ID: <396222237.13630@mail.sdu.edu.cn> Dear all, When i calculate the doped magnetic semiconductor system Fe_2_In_30_O_47, it was found that the output total magnetic moment is integral, 8.00 Bohr, like: ****************************** ! total energy = -1743.92699154 Ry Harris-Foulkes estimate = -1743.92699155 Ry estimated scf accuracy < 0.00000542 Ry total magnetization = 8.00 Bohr mag/cell absolute magnetization = 9.21 Bohr mag/cell ****************************** By performing the consequent projwfc.x calculation, we found that the magnetic moment of Fe ions is not integral, like: ****************************** Atom # 1: total charge = 6.6980, s, p, d, f = 0.4324 0.0000 6.2657 spin up = 5.1402, s, p, d, f = 0.2145 0.0000 4.9258 spin down = 1.5578, s, p, d, f = 0.2179 0.0000 1.3399 polarization = 3.5825, s, p, d, f = -0.0034 0.0000 3.5859 Atom # 27: total charge = 6.7373, s, p, d, f = 0.4393 0.0000 6.2980 spin up = 5.1453, s, p, d, f = 0.2158 0.0000 4.9295 spin down = 1.5920, s, p, d, f = 0.2236 0.0000 1.3685 polarization = 3.5533, s, p, d, f = -0.0078 0.0000 3.5611 ****************************** Therefore it can be interpreted that the sum of the moment of all ions is integral. However, when we changed the doping site, ex-c functional (PZ to PBE) or chose the other doping element, the output magnetic is still integral (i.e. total magnetization = 4.00 Bohr for Co dopants). It's surprising! I don't think it is a coincidence. It seems that there would be some restricts to force the moment to be integral. Does anyone know why? Any reply is appreciated. Shujun Following is the input file: **********************input file******************************* &CONTROL title = ino , calculation = 'relax' , restart_mode = 'from_scratch' , wf_collect = .true. , outdir = '/home/work/7.05/temp/' , pseudo_dir = '/home/work/PSEUDOPOTENTIAL/' , prefix = ino , nstep = 1000 , dt = 10 , / &SYSTEM ibrav = 1, celldm(1) = 18.35, nat = 79, ntyp = 5, ecutwfc = 30 , nbnd = 205, occupations = 'tetrahedra' , degauss = 0.005 , smearing = 'methfessel-paxton' , nspin = 2 , starting_magnetization(4) = 0.3, starting_magnetization(5) = 0.3, / &ELECTRONS electron_maxstep = 500, conv_thr = 1E-5 , startingpot = 'atomic' , startingwfc = 'atomic' , mixing_mode = 'plain' , mixing_beta = 0.3 , diagonalization = 'david' , diago_thr_init = 1E-3 , / &IONS ion_dynamics = 'damp' , / &CELL cell_dynamics = 'damp-pr' , / ATOMIC_SPECIES O 15.99940 O.pz-rrkjus.UPF In1 114.8000 In.pz-bhs.UPF In2 114.8000 In.pz-bhs.UPF Fe1 55.8470 Fe.pz-nd-rrkjus.UPF Fe2 55.8470 Fe.pz-nd-rrkjus.UPF ATOMIC_POSITIONS alat Fe1 0.250000000 0.250000000 0.250000000 In1 0.250000000 0.750000000 0.750000000 In1 0.750000000 0.750000000 0.250000000 In1 0.750000000 0.250000000 0.750000000 In1 0.750000000 0.750000000 0.750000000 In1 0.750000000 0.250000000 0.250000000 In1 0.250000000 0.250000000 0.750000000 In1 0.250000000 0.750000000 0.250000000 In2 0.962352792 0.000000000 0.250000000 In2 0.537647208 0.000000000 0.750000000 In2 0.250000000 0.962352792 0.000000000 In2 0.750000000 0.537647208 0.000000000 In2 0.000000000 0.250000000 0.962352792 In2 0.000000000 0.750000000 0.537647208 In2 0.037647208 0.000000000 0.750000000 In2 0.462352792 0.000000000 0.250000000 In2 0.750000000 0.037647208 0.000000000 In2 0.250000000 0.462352792 0.000000000 In2 0.000000000 0.750000000 0.037647208 In2 0.000000000 0.250000000 0.462352792 In2 0.462352792 0.500000000 0.750000000 In2 0.037647208 0.500000000 0.250000000 In2 0.750000000 0.462352792 0.500000000 In2 0.250000000 0.037647208 0.500000000 In2 0.500000000 0.750000000 0.462352792 In2 0.500000000 0.250000000 0.037647208 Fe2 0.537647208 0.500000000 0.250000000 In2 0.962352792 0.500000000 0.750000000 In2 0.250000000 0.537647208 0.500000000 In2 0.750000000 0.962352792 0.500000000 In2 0.500000000 0.250000000 0.537647208 In2 0.500000000 0.750000000 0.962352792 O 0.390681098 0.155477591 0.382212636 O 0.109318902 0.844522409 0.882212636 O 0.609318902 0.655477591 0.117787364 O 0.890681098 0.344522409 0.617787364 O 0.882212636 0.109318902 0.844522409 O 0.117787364 0.609318902 0.655477591 O 0.617787364 0.890681098 0.344522409 O 0.155477591 0.382212636 0.390681098 O 0.844522409 0.882212636 0.109318902 O 0.655477591 0.117787364 0.609318902 O 0.344522409 0.617787364 0.890681098 O 0.609318902 0.844522409 0.617787364 O 0.890681098 0.155477591 0.117787364 O 0.390681098 0.344522409 0.882212636 O 0.109318902 0.655477591 0.382212636 O 0.617787364 0.609318902 0.844522409 O 0.117787364 0.890681098 0.155477591 O 0.882212636 0.390681098 0.344522409 O 0.382212636 0.109318902 0.655477591 O 0.844522409 0.617787364 0.609318902 O 0.155477591 0.117787364 0.890681098 O 0.344522409 0.882212636 0.390681098 O 0.655477591 0.382212636 0.109318902 O 0.890681098 0.655477591 0.882212636 O 0.609318902 0.344522409 0.382212636 O 0.109318902 0.155477591 0.617787364 O 0.390681098 0.844522409 0.117787364 O 0.882212636 0.890681098 0.655477591 O 0.382212636 0.609318902 0.344522409 O 0.617787364 0.109318902 0.155477591 O 0.117787364 0.390681098 0.844522409 O 0.655477591 0.882212636 0.890681098 O 0.344522409 0.382212636 0.609318902 O 0.155477591 0.617787364 0.109318902 O 0.844522409 0.117787364 0.390681098 O 0.109318902 0.344522409 0.117787364 O 0.390681098 0.655477591 0.617787364 O 0.890681098 0.844522409 0.382212636 O 0.609318902 0.155477591 0.882212636 O 0.117787364 0.109318902 0.344522409 O 0.617787364 0.390681098 0.655477591 O 0.382212636 0.890681098 0.844522409 O 0.882212636 0.609318902 0.155477591 O 0.344522409 0.117787364 0.109318902 O 0.655477591 0.617787364 0.390681098 O 0.844522409 0.382212636 0.890681098 O 0.155477591 0.882212636 0.609318902 K_POINTS automatic 2 2 2 1 1 1 ------------------------------ Shujun Hu e-mail: hushujun at 163.com Shandong university Phone: +86/0531-88375097 Jinan, Shandong Province, China, 250100 From meisam_a84 at yahoo.com Wed Nov 28 06:52:19 2007 From: meisam_a84 at yahoo.com (meisam aghtar) Date: Tue, 27 Nov 2007 21:52:19 -0800 (PST) Subject: [Pw_forum] an error during running pw input of zigzag nanotube Message-ID: <442016.55653.qm@web30606.mail.mud.yahoo.com> Hi first thanks hande Hande Ustunel for your response second I have a problem during running the pw input for a (5,0) nanotube. &CONTROL calculation = 'scf' , restart_mode = 'restart' , outdir = '/home/meisam/projhe/5-0/apoint/scratch/' , pseudo_dir = '/home/meisam/espresso-3.1.1/pseudo/' , prefix = 'cnt-scf' , / &SYSTEM ibrav = 4, celldm(1) = 20, celldm(3) = 0.282460, nat = 20, ntyp = 1, ecutwfc = 30 , / &ELECTRONS / ATOMIC_SPECIES C 12.00000 C.pz-vbc.UPF ATOMIC_POSITIONS angstrom C 1.957217600 0.000000000 0.000000000 C 1.583422300 1.150423600 0.710000000 C 0.604813500 1.861424500 0.000000000 C -0.604813500 1.861424500 0.710000000 C -1.583422300 1.150423600 0.000000000 C -1.957217600 0.000000000 0.710000000 C -1.583422300 -1.150423600 0.000000000 C -0.604813500 -1.861424500 0.710000000 C 0.604813500 -1.861424500 0.000000000 C 1.583422300 -1.150423600 0.710000000 C 1.583422300 1.150423600 2.130000000 C 0.604813500 1.861424500 2.840000000 C -0.604813500 1.861424500 2.130000000 C -1.583422300 1.150423600 2.840000000 C -1.957217600 0.000000000 2.130000000 C -1.583422300 -1.150423600 2.840000000 C -0.604813500 -1.861424500 2.130000000 C 0.604813500 -1.861424500 2.840000000 C 1.583422300 -1.150423600 2.130000000 C 1.957217600 0.000000000 2.840000000 K_POINTS automatic 1 1 14 1 1 1 error : Self-consistent Calculation iteration # 1 ecut= 30.00 ryd beta=0.70 Davidson diagonalization with overlap IOS = 36 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from davcio : error # 10 i/o error in davcio %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ____________________________________________________________________________________ Be a better pen pal. Text or chat with friends inside Yahoo! Mail. See how. http://overview.mail.yahoo.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071127/18b5e0c8/attachment.htm From hande at newton.physics.metu.edu.tr Wed Nov 28 07:45:53 2007 From: hande at newton.physics.metu.edu.tr (Hande Ustunel) Date: Wed, 28 Nov 2007 08:45:53 +0200 (EET) Subject: [Pw_forum] an error during running pw input of zigzag nanotube In-Reply-To: <442016.55653.qm@web30606.mail.mud.yahoo.com> Message-ID: Hi, As suggested in another post in reply to your previous question, the z coordinates of your atoms are incorrect. Your system is not really periodic and for this reason atoms coincide when the code imposes PBC. It's best to always visualize your input before starting the run, which you can do very conveniently by using xcrysden. Do xcrysden --pwi and then by typing shift-n bring up a dialog that allos you to adjust the number of units drawn in each direction. You'll see (and xcrysden will complain) that your atoms are on top of each other. Best of luck, Hande On Tue, 27 Nov 2007, meisam aghtar wrote: > Hi > first > thanks hande Hande Ustunel for your response > second > I have a problem during running the pw input for a (5,0) nanotube. > > > &CONTROL > calculation = 'scf' , > restart_mode = 'restart' , > outdir = '/home/meisam/projhe/5-0/apoint/scratch/' , > pseudo_dir = '/home/meisam/espresso-3.1.1/pseudo/' , > prefix = 'cnt-scf' , > / > &SYSTEM > ibrav = 4, > celldm(1) = 20, > celldm(3) = 0.282460, > nat = 20, > ntyp = 1, > ecutwfc = 30 , > / > &ELECTRONS > > > / > ATOMIC_SPECIES > C 12.00000 C.pz-vbc.UPF > ATOMIC_POSITIONS angstrom > C 1.957217600 0.000000000 0.000000000 > C 1.583422300 1.150423600 0.710000000 > C 0.604813500 1.861424500 0.000000000 > C -0.604813500 1.861424500 0.710000000 > C -1.583422300 1.150423600 0.000000000 > C -1.957217600 0.000000000 0.710000000 > C -1.583422300 -1.150423600 0.000000000 > C -0.604813500 -1.861424500 0.710000000 > C 0.604813500 -1.861424500 0.000000000 > C 1.583422300 -1.150423600 0.710000000 > C 1.583422300 1.150423600 2.130000000 > C 0.604813500 1.861424500 2.840000000 > C -0.604813500 1.861424500 2.130000000 > C -1.583422300 1.150423600 2.840000000 > C -1.957217600 0.000000000 2.130000000 > C -1.583422300 -1.150423600 2.840000000 > C -0.604813500 -1.861424500 2.130000000 > C 0.604813500 -1.861424500 2.840000000 > C 1.583422300 -1.150423600 2.130000000 > C 1.957217600 0.000000000 2.840000000 > K_POINTS automatic > 1 1 14 1 1 1 > error : > Self-consistent Calculation > iteration # 1 > ecut= 30.00 ryd beta=0.70 > > Davidson diagonalization with overlap > IOS = 36 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from davcio : error # 10 > i/o error in davcio > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > > ____________________________________________________________________________________ > Be a better pen pal. > Text or chat with friends inside Yahoo! Mail. See how. http://overview.mail.yahoo.com/ -- Hande Ustunel Department of Physics Office 439 Middle East Technical University Ankara 06531, Turkey Tel : +90 312 210 3264 http://www.physics.metu.edu.tr/~hande From katalin.gaal-nagy at physik.uni-regensburg.de Wed Nov 28 09:30:48 2007 From: katalin.gaal-nagy at physik.uni-regensburg.de (Katalin Gaal-Nagy) Date: Wed, 28 Nov 2007 09:30:48 +0100 (CET) Subject: [Pw_forum] Units of conv_thr and tr2_ph In-Reply-To: <2C448878-5158-4F06-B50F-C3398813DA42@nest.sns.it> References: <474C3A86.9050909@fcs.it> <4A9251DC-7B79-4780-B137-972FBA8D83E4@nest.sns.it> <2C448878-5158-4F06-B50F-C3398813DA42@nest.sns.it> Message-ID: Dear Paolo, thanks for your answer! All the best, Katalin On Tue, 27 Nov 2007, Paolo Giannozzi wrote: > > On Nov 27, 2007, at 19:00 , Katalin Gaal-Nagy wrote: > >> Are the units of conv_thr and tr2_ph Ry > > yes > >> meaning that the self consistency is checked agains the energy >> of the system or are they different? > > this is a more complicated issue. conv_thr is an estimate of > the error on the energy due to self-consistency. tr2_ph is > an estimate of...I don't remember exactly what, but it is > NOT (unfortunately) related to an error on any energy, but > to the difference of input and output potentials > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From giannozz at nest.sns.it Wed Nov 28 09:33:17 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 28 Nov 2007 09:33:17 +0100 Subject: [Pw_forum] Compiling PWscf on AMD64 using PGI compilors In-Reply-To: <7218ea700711271510uc622149r1e85e87dcbd4e6a@mail.gmail.com> References: <7218ea700711271510uc622149r1e85e87dcbd4e6a@mail.gmail.com> Message-ID: <9225FE7C-46A1-4B5C-B251-75A82CC181B3@nest.sns.it> On Nov 28, 2007, at 24:10 , PeiQing Sheng wrote: > ./configure --enable-parallel ARCH=amd64 MPIF90=mpif90 F90=pgf90 > MPIF77=mpif77 F77=pgf77 MPICC=mpicc CC=cc don't give arguments unless you need to > FFT_LIBS = -L/bigdisk/staff/pshen/FFTW2/lib -lfftw_mpi -lfftw do not use parallel FFT from FFTW: FFT are parallelized inside the code > But it looks like I could only run pw.x on one processor see here: http://www.quantum-espresso.org/wiki/index.php/ Frequently_Asked_Questions#Why_doesn. 27t_configure_recognize_that_I_have_a_parallel_machine.3F Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From zhufengcc at gmail.com Wed Nov 28 10:30:07 2007 From: zhufengcc at gmail.com (feng zhu) Date: Wed, 28 Nov 2007 17:30:07 +0800 Subject: [Pw_forum] Hi, I would like to know how to use the restart_mode='restart'? In-Reply-To: <20071123110014.nxt2xmw9gk40ksc0@wmail1.cc.ntu.edu.tw> References: <20071123110014.nxt2xmw9gk40ksc0@wmail1.cc.ntu.edu.tw> Message-ID: <5b273e100711280130u4e303aeg3f63dddaeae6dc15@mail.gmail.com> "restart" is to do calculation using the things already done. So it will read the geometry, charge, wfc, ... from the previous calculation. If there is no wfc file and you set "restart", you must set also "startingwfc='atomic' " . I think it is ok to "restart" even if not scf, but maybe this is not very safe. 2007/11/23, r95222066 at ntu.edu.tw : > > Hello, > I have read the user guide and the INPUT_PW doc file.I wolud like > to ask how to use > the restart_mode='restart'. Do I need to save the wfc file? what is > the process if I have > not scf yet,and I want to scf again by used the restart_mode="restart"? > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- ------------------ Zhu feng Department of Physics, Anhui University, Hefei, China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071128/ed24bf11/attachment.htm From meisam_a84 at yahoo.com Wed Nov 28 14:48:41 2007 From: meisam_a84 at yahoo.com (meisam aghtar) Date: Wed, 28 Nov 2007 05:48:41 -0800 (PST) Subject: [Pw_forum] a problem with pw input of (5,0) nanotube Message-ID: <362380.5338.qm@web30610.mail.mud.yahoo.com> Dear Hande you were write with previous input the atoms overlap each other , I correct this problem and it worked for a pw input with celldm(1)=19 during finding a point. but for the other celldms it made an error. another point is that I considered the c parameter equal 2.13 A but the atoms overlaped again but by cosidering the c as 2.84 , the problem was solved &CONTROL calculation = 'scf' , restart_mode = 'restart' , outdir = '/home/meisam/projhe/5-0/apoint/scratch/' , pseudo_dir = '/home/meisam/espresso-3.1.1/pseudo/' , prefix = 'cnt-scf' , / &SYSTEM ibrav = 4, celldm(1) = 21, celldm(3) = 0.255561615, nat = 15, ntyp = 1, ecutwfc = 30 , / &ELECTRONS / ATOMIC_SPECIES C 12.00000 C.pz-vbc.UPF ATOMIC_POSITIONS angstrom C 1.957217600 0.000000000 0.000000000 C 1.583422300 1.150423600 0.710000000 C 0.604813500 1.861424500 0.000000000 C -0.604813500 1.861424500 0.710000000 C -1.583422300 1.150423600 0.000000000 C -1.957217600 0.000000000 0.710000000 C -1.583422300 -1.150423600 0.000000000 C -0.604813500 -1.861424500 0.710000000 C 0.604813500 -1.861424500 0.000000000 C 1.583422300 -1.150423600 0.710000000 C 1.583422300 1.150423600 2.130000000 C -0.604813500 1.861424500 2.130000000 C -1.957217600 0.000000000 2.130000000 C -0.604813500 -1.861424500 2.130000000 C 1.583422300 -1.150423600 2.130000000 K_POINTS automatic 1 1 14 1 1 1 .............................................................. error: iteration # 1 ecut= 30.00 ryd beta=0.70 Davidson diagonalization with overlap %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from cdiaghg : error # 93 info =/= 0 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071128/b656faa7/attachment.htm From ezadshojaee at hotmail.com Wed Nov 28 14:14:25 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Wed, 28 Nov 2007 13:14:25 +0000 Subject: [Pw_forum] zone center phonons of Si lattice Message-ID: Hi It seems that if we look to Si lattice as an SC with 8 atoms per cell, the symmetries connected to the fractional translations will not be accepted, so if we use this scheme to obtain the zone-center phonons of Si, we will get frequencies rather different from the main one(FCC scheme), with different symmetries and degenerecies. How is it possible to obtain the right ones (zone-center phonons) of Si, with the right symmetries and degenerecies, by the SC scheme? Thanx _________________________________________________________________ Connect to the next generation of MSN Messenger? http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&source=wlmailtagline -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071128/3ea445dc/attachment.htm From giannozz at nest.sns.it Wed Nov 28 15:39:53 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 28 Nov 2007 15:39:53 +0100 Subject: [Pw_forum] zone center phonons of Si lattice In-Reply-To: References: Message-ID: <28A83832-A108-44BA-B03E-0CE56BCE8DE5@nest.sns.it> On Nov 28, 2007, at 14:14 , Ezad Shojaee wrote: > It seems that if we look to Si lattice as an SC with 8 atoms per cell, > the symmetries connected to the fractional translations will not be > accepted, so if we use this scheme to obtain the zone-center > phonons of Si, we will get frequencies rather different from the > main one(FCC scheme), with different symmetries and degenerecies. not at all. If you do things properly, you will get the modes you should get, with or without symmetry imposed, apart from minor losses of symmetry due to incomplete convergence Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From akohlmey at cmm.chem.upenn.edu Wed Nov 28 15:41:31 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 28 Nov 2007 09:41:31 -0500 Subject: [Pw_forum] a problem with pw input of (5,0) nanotube In-Reply-To: <362380.5338.qm@web30610.mail.mud.yahoo.com> References: <362380.5338.qm@web30610.mail.mud.yahoo.com> Message-ID: <7b6913e90711280641t2447921ci673685343d21ec4d@mail.gmail.com> meisam, please don't use restart_mode='restart' unless you are restarting from a _correct_ but interrupted calculation of the same input. in other words, use restart_mode='from_scratch' unless you are certain your input is correct. cheers, axel. On Nov 28, 2007 8:48 AM, meisam aghtar wrote: > > Dear Hande > > you were write with previous input the atoms overlap each other , I correct > this problem and it worked for a pw input with celldm(1)=19 during finding a > point. but for the other celldms it made an error. another point is that I > considered the c parameter equal 2.13 A but the atoms overlaped again but by > cosidering the c as 2.84 , the problem was solved > > > &CONTROL > calculation = 'scf' , > restart_mode = 'restart' , > outdir = '/home/meisam/projhe/5-0/apoint/scratch/' , > pseudo_dir = '/home/meisam/espresso-3.1.1/pseudo/' , > prefix = 'cnt-scf' , > / > &SYSTEM > ibrav = 4, > celldm(1) = 21, > celldm(3) = 0.255561615, > nat = 15, > ntyp = 1, > ecutwfc = 30 , > / > > &ELECTRONS > > / > > ATOMIC_SPECIES > C 12.00000 C.pz-vbc.UPF > ATOMIC_POSITIONS angstrom > C 1.957217600 0.000000000 0.000000000 > C 1.583422300 1.150423600 0.710000000 > C 0.604813500 1.861424500 0.000000000 > C -0.604813500 1.861424500 0.710000000 > C -1.583422300 1.150423600 0.000000000 > C -1.957217600 0.000000000 0.710000000 > C -1.583422300 -1.150423600 0.000000000 > C -0.604813500 -1.861424500 0.710000000 > C 0.604813500 -1.861424500 0.000000000 > C 1.583422300 -1.150423600 0.710000000 > C 1.583422300 1.150423600 2.130000000 > C -0.604813500 1.861424500 2.130000000 > C -1.957217600 0.000000000 2.130000000 > C -0.604813500 -1.861424500 2.130000000 > C 1.583422300 -1.150423600 2.130000000 > K_POINTS automatic > 1 1 14 1 1 1 > .............................................................. > error: > iteration # 1 ecut= 30.00 ryd beta=0.70 > Davidson diagonalization with overlap > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from cdiaghg : error # 93 > info =/= 0 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > > > > ________________________________ > Get easy, one-click access to your favorites. Make Yahoo! your homepage. > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From giannozz at nest.sns.it Wed Nov 28 15:42:28 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 28 Nov 2007 15:42:28 +0100 Subject: [Pw_forum] a problem with pw input of (5,0) nanotube In-Reply-To: <362380.5338.qm@web30610.mail.mud.yahoo.com> References: <362380.5338.qm@web30610.mail.mud.yahoo.com> Message-ID: <8715C689-5D17-49A1-9F35-A78702B54F36@nest.sns.it> On Nov 28, 2007, at 14:48 , meisam aghtar wrote: > from cdiaghg : error # 93 > info =/= 0 http://www.quantum-espresso.org/wiki/index.php/Troubleshooting_% 28PWscf%29#pw.x_stops_with_error_in_cdiaghg_or_rdiaghg --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From wmbmacam at lg.ehu.es Wed Nov 28 15:40:19 2007 From: wmbmacam at lg.ehu.es (=?UTF-8?B?TWlndWVsIE1hcnTDrW5leiBDYW5hbGVz?=) Date: Wed, 28 Nov 2007 15:40:19 +0100 Subject: [Pw_forum] zone center phonons of Si lattice In-Reply-To: References: Message-ID: <474D7DD3.4070704@lg.ehu.es> Hi Ezad, There's an example in Ashcroft's and Mermin's "Solid State Physics" (Ch. 22, p 433 in my copy) in which they take the parametres of a linear chain with a basis so that it's equivalent to solving the monoatomic linear chain. What you see in this case is similar to "folding" the phonon branch. Doing something similar for the bandstructure would also result in folding, since there is no bragg reflection to split the bands near the "zone boundary". I suppose one sees the same behaviour on "SC Si", but one should have to be very careful to select the same (or equivalent) q points (due to the different basis you're using). I hope this gives some hints. Regards, Miguel Ezad Shojaee escribi?: > Hi > It seems that if we look to Si lattice as an SC with 8 atoms per cell, > the symmetries connected to the fractional translations will not be > accepted, so if we use this scheme to obtain the zone-center phonons of > Si, we will get frequencies rather different from the main one(FCC > scheme), with different symmetries and degenerecies. > How is it possible to obtain the right ones (zone-center phonons) of Si, > with the right symmetries and degenerecies, by the SC scheme? > Thanx > > ------------------------------------------------------------------------ > Connect to the next generation of MSN Messenger Get it now! > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- ---------------------------------------- Miguel Mart?nez Canales Dto. F?sica de la Materia Condensada UPV/EHU Facultad de Ciencia y Tecnolog?a Apdo. 644 48080 Bilbao (Spain) Fax: +34 94 601 3500 Tlf: +34 94 601 5326 ---------------------------------------- "If you have an apple and I have an apple and we exchange these apples then you and I will still each have one apple. But if you have an idea and I have an idea and we exchange these ideas, then each of us will have two ideas." George Bernard Shaw From bnrj.rudra at yahoo.com Thu Nov 29 08:21:13 2007 From: bnrj.rudra at yahoo.com (Rudra Banerjee) Date: Thu, 29 Nov 2007 12:51:13 +0530 (IST) Subject: [Pw_forum] card K_POINTS{AUTOMATIC} ignored Message-ID: <485946.35643.qm@web94103.mail.in2.yahoo.com> dear Pw Users, i am do a "relax" with the given input: &control calculation='relax' pseudo_dir = '../../pseudo/', restart_mode = 'from_scratch' , outdir='.' title='Relax Water Hexamer Prism Structure' prefix='prism' tprnfor=.true., tstress=.true. nstep=100, / &system ibrav= 1, celldm(1) = 24., nat=18, ntyp= 2, ecutwfc = 60.0, ecutrho = 180.0 , nbnd= 52, xc_type = 'BLYP', / &electrons conv_thr = 1.d-3 / &ions ion_dynamics = 'bfgs' , upscale = 30.D0 , trust_radius_max = 5.0D-1 , trust_radius_min = 1.0D-10 , / ATOMIC_SPECIES O 16.0d0 O.BLYP.UPF H 1.0d0 H.fpmd.UPF ATOMIC_POSITIONS (bohr) O 0.275792E+01 -0.730311E+00 0.252973E+01 O 0.149999E+01 0.382366E+01 -0.402203E-01 O 0.305094E+01 -0.952937E+00 -0.267495E+01 O -0.345006E+01 0.235966E+01 -0.229934E-01 O -0.223541E+01 -0.243871E+01 -0.249855E+01 O -0.195463E+01 -0.219370E+01 0.272776E+01 H 0.953867E+00 -0.139641E+01 0.281176E+01 H 0.378173E+01 -0.130792E+01 0.394426E+01 H 0.212944E+01 0.264076E+01 0.126480E+01 H 0.216319E+01 0.305856E+01 -0.159340E+01 H 0.443203E+01 -0.181788E+01 -0.353014E+01 H 0.333225E+01 -0.120205E+01 -0.830040E+00 H -0.171332E+01 0.312962E+01 -0.587749E-01 H -0.463156E+01 0.376262E+01 -0.166814E+00 H -0.449834E+00 -0.210109E+01 -0.292679E+01 H -0.296092E+01 -0.736778E+00 -0..228072E+01 H -0.210758E+01 -0.276432E+01 0.926993E+00 H -0.286875E+01 -0.575556E+00 0.265126E+01 K_POINTS{automatic} 5 5 5 0 0 0 its giving a warnning: Warning: card K_POINTS{AUTOMATIC} ignored Warning: card 5 5 5 0 0 0 ignored WARNING: atomic wfc # 1 for atom type 2 has zero norm gamma-point specific algorithms are used can anybody pl. tell me whats wrong and why it is ignored? -- Rudra Have a Nice Time http://www.bose.res.in/~rudra Please, if possible, don't send me MS Word or PowerPoint attachments Why?See: http://www.gnu.org/philosophy/no-word-attachments.html Get the freedom to save as many mails as you wish. To know how, go to http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071129/3df1c050/attachment.htm From paulatto at sissa.it Thu Nov 29 11:01:10 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Thu, 29 Nov 2007 11:01:10 +0100 (CET) Subject: [Pw_forum] card K_POINTS{AUTOMATIC} ignored In-Reply-To: <485946.35643.qm@web94103.mail.in2.yahoo.com> References: <485946.35643.qm@web94103.mail.in2.yahoo.com> Message-ID: <29803.82.54.232.239.1196330470.squirrel@webmail.sissa.it> On Thu, November 29, 2007 08:21, Rudra Banerjee wrote: > K_POINTS{automatic} > ... > Warning: card K_POINTS{AUTOMATIC} ignored Dear Rundra, just leave a white space between KPOINTS and {automatic}. goodbye -- Lorenzo Paulatto +39 040 3787 511 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From ezadshojaee at hotmail.com Thu Nov 29 18:20:01 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Thu, 29 Nov 2007 17:20:01 +0000 Subject: [Pw_forum] about Si lattice Message-ID: Hi Thanx for your answer, besides i want to ask you one more thing! >> If you do things properly, you will get the modes youshould get, with or without symmetry imposed, apart fromminor losses of symmetry due to incomplete convergence what may cause this incomplete convergence(sorry if it's silly)? and about increasing the branches of phonon dispersion, using the SC scheme (like the linear chain described in Ashcrofht & Mermin), how many frequencies will you get of Si lattice at the experiment finally(the two schemes describe different # of modes at the Gamma point? thanx again _________________________________________________________________ Explore the seven wonders of the world http://search.msn.com/results.aspx?q=7+wonders+world&mkt=en-US&form=QBRE -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071129/371f2266/attachment.htm From glevita at units.it Thu Nov 29 19:01:12 2007 From: glevita at units.it (glevita at units.it) Date: Thu, 29 Nov 2007 19:01:12 +0100 Subject: [Pw_forum] Questions about plotting In-Reply-To: References: Message-ID: <20071129190112.oedjh66j2g8o48cw@webmail.units.it> Dear PW users, My name is Giacomo Levita, and I have a couple of questions, both quite naive. I tried to search in the Forum archive to get some hints, but I have to rely on your advice. I am trying to evaluate the LDOS (at different energies) of an isolated planar molecule (PVBA), for which I got a relaxed structure I am satisfied with. I am interested in calculating the LDOS only in certain points of the molecule. The first question is therefore: is it possible to have a numerical value of the LDOS around points I can choose in my supercell? I mean: is there any postprocessing tool in which by providing x,y,z,R and E, the LDOS within a sphere of radius R centred around a xyz point for the given energy E is obtained? Second: being unable to extract easily manageable numerical values out of the XSF file obtained for some |Psi^2|, I decided to create a series of one dimensional plots, related to vectors parallel to the molecular axis, out of which I'll extract some preliminary info. To test this tool, I tried to plot the charge density (as a test) on a certain direction. This was my input file: &inputpp prefix = 'PVBA' outdir = './' filplot = 'projlin' plot_num= 0 / &plot nfile = 1 filepp(1) = 'projlin' weight(1) = 1.0 iflag = 1 output_format = 3 fileout = 'projlin.ldos.xsf' e1(1)= 0.0, e1(2)= 0.0, e1(3)= 1.0, x0(1)= 0.0, x0(2)= 0.1, x0(3)= 0.0, nx= 100 / There must be something wrong with that, as the file 'projlin' is correctly generated, but the related XSF file is not. Actually, the job keeps running without any advance in writing the XSF file (it doesn't crash). These are the last lines of the output file (I submitted as an MPI parallel job on 64 processors): ... 61 5 292 64124 3 98 12354 24 1548 62 5 292 64124 3 98 12354 24 1548 63 5 292 64120 3 98 12354 24 1544 64 5 292 64116 3 98 12346 24 1544 0 360 186014103263 192 6205 789671 1555 98693 nbndx = 50 nbnd = 50 natomwfc = 79 npwx = 774 nelec = 84.00 nkb = 147 ngl = 8205 negative rho (up, down): 0.315E-01 0.000E+00 Calling punch_plot, plot_num = 0 Writing data to file projlin Reading header from file projlin Reading data from file projlin Writing data to be plotted to file projlin.ldos.xsf The file just ends like this, and nothing is written in the .XSF file. I am very suspicious about that line of the output starting with 0, at the end of the grid description. Needless to say, the same result is obtained when I change nx or e0 in my input file. Does anyone have any suggestion about this problem? I thank you very much for your help; sorry for this wordy message. Giacomo ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From chenhanghuipwscf at gmail.com Fri Nov 30 00:20:37 2007 From: chenhanghuipwscf at gmail.com (alan chen) Date: Thu, 29 Nov 2007 18:20:37 -0500 Subject: [Pw_forum] Warning in the 'nscf' calculation Message-ID: <22ae3ca40711291520rfd5cb65kc0e5b07a061fe056@mail.gmail.com> Dear PWSCF users, I have done a 'nscf' calculation to get the band structure. The number of bands is 280 and the number of irreducible k points is 10. The input file is attached below. In the output file I got some warning messages like: WARNING: 3 eigenvalues not converged WARNING: 3 eigenvalues not converged WARNING: 139 eigenvalues not converged WARNING: 2 eigenvalues not converged WARNING: 143 eigenvalues not converged WARNING: 3 eigenvalues not converged WARNING: 1 eigenvalues not converged WARNING: 1 eigenvalues not converged WARNING: 140 eigenvalues not converged WARNING: 13 eigenvalues not converged WARNING: 1 eigenvalues not converged The band structure is finally calculated at the end of output file but shall I believe in these E_{nk} values? How can I get rid of these warning messages? Thank you very much. The input file is as follows: &CONTROL calculation='nscf' wf_collect=.true. pseudo_dir = '../psp' outdir='.' wfcdir='/scratch' prefix='IF-relax' disk_io='low' verbosity='low' tprnfor = .true. tstress = .true. dt=80.D0 / &SYSTEM ibrav= 6 celldm(1) = 7.27 celldm(3) = 17.0 nat= 67 ntyp= 5 nbnd = 280 ecutwfc = 30.0 ecutrho = 180.0 occupations='smearing' smearing='gauss' degauss=0.03 / &ELECTRONS diagonalization='david' mixing_beta = 0.7D0 diago_david_ndim = 4 mixing_mode= 'local-TF' electron_maxstep = 300 conv_thr = 1.D-8 / ATOMIC_SPECIES La 138.91 057-La-ca-nsp-hanghui.uspp.format.UPF Al 26.98 013-Al-ca--hanghui.uspp.format.UPF Sr 87.62 038-Sr-ca-sp-vgrp.uspp.format.UPF Ti 47.90 022-Ti-ca-sp-vgrp.uspp.format.UPF O 16.00 008-O-ca--vgrp.uspp.format.UPF ATOMIC_POSITIONS alat Sr 0.000000000 0.000000000 0.0 0 0 0 O 0.500000000 0.500000000 0.0 0 0 0 Ti 0.500000000 0.500000000 0.5 0 0 0 O 0.500000000 0.000000000 0.5 0 0 0 O 0.000000000 0.500000000 0.5 0 0 0 Sr 0.000000000 0.000000000 1.0 0 0 0 O 0.500000000 0.500000000 1.0 0 0 0 Ti 0.500000000 0.500000000 1.5 0 0 0 O 0.500000000 0.000000000 1.5 0 0 0 O 0.000000000 0.500000000 1.5 0 0 0 Sr 0.000000000 0.000000000 2.004036406 1 1 1 O 0.500000000 0.500000000 2.007944928 1 1 1 Ti 0.500000000 0.500000000 2.508029381 1 1 1 O 0.500000000 0.000000000 2.513722783 1 1 1 O 0.000000000 0.500000000 2.513722783 1 1 1 Sr 0.000000000 0.000000000 3.011035804 1 1 1 O 0.500000000 0.500000000 3.017938555 1 1 1 Ti 0.500000000 0.500000000 3.515462896 1 1 1 O 0.500000000 0.000000000 3.523810273 1 1 1 O 0.000000000 0.500000000 3.523810273 1 1 1 ...... K_POINTS {automatic} 6 6 1 0 0 0 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071129/b6e05cde/attachment-0001.htm From somesh.kb at gmail.com Fri Nov 30 06:33:03 2007 From: somesh.kb at gmail.com (Somesh Kumar Bhattacharya) Date: Fri, 30 Nov 2007 11:03:03 +0530 Subject: [Pw_forum] Te pseudopotentials Message-ID: Dear Users, I wish to do some calculations on CdTe clusters using PWSCF. For this I need Vanderbilt USPP with PBE exchange correlation. For Cd it exist and I can get from the net. But for Te its not there. Can anyone please provide me with the required PP???? Regards, Somesh SOMESH KR. BHATTACHARYA Ph.no.- (O)+91- 020-25692678-426 JUNIOR RESEARCH FELLOW (M)+919921680580 DEPT. OF PHYSICS UNIV. OF PUNE GANESHKHIND PUNE-411007 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071130/4ea06810/attachment.htm From njuxuyuehua at gmail.com Fri Nov 30 08:10:44 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Fri, 30 Nov 2007 15:10:44 +0800 Subject: [Pw_forum] a simple question about the relax outfile Message-ID: hi all : i have a question about the out file of relax : for example :sometimes :the end of file is like: ......... convergence has been achieved Forces acting on atoms (Ry/au): atom 1 type 1 force = -0.00016739 0.00006788 0.00006825 atom 2 type 1 force = 0.00000361 0.00001673 -0.00004580 atom 3 type 1 force = 0.00007761 0.00005882 0.00004672 atom 4 type 1 force = 0.00003052 0.00004707 0.00002272 atom 5 type 1 force = 0.00001339 0.00015088 -0.00012165 atom 6 type 1 force = -0.00010455 -0.00002068 -0.00010771 atom 7 type 1 force = -0.00009487 -0.00001394 0.00006473 atom 8 type 1 force = -0.00000214 -0.00007630 0.00001052 atom 9 type 1 force = 0.00002701 0.00001038 0.00000479 atom 10 type 1 force = 0.00008230 0.00004823 0.00012935 atom 11 type 2 force = 0.00015243 -0.00016928 -0.00010298 atom 12 type 2 force = -0.00004574 -0.00002796 0.00000972 atom 13 type 2 force = 0.00003206 0.00001371 -0.00009648 atom 14 type 2 force = 0.00000462 -0.00000415 0.00003920 atom 15 type 2 force = -0.00000887 -0.00010141 0.00007861 Total force = 0.000506 Total SCF correction = 0.000024 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= - 0.60 -0.00001045 -0.00000009 0.00000000 -1.54 -0.01 0.00 -0.00000009 -0.00001053 -0.00000015 -0.01 -1.55 -0.02 0.00000000 -0.00000015 0.00000867 0.00 -0.02 1.27 The maximum number of steps has been reached. End of BFGS Geometry Optimization and sometimes the same part of the outfile: convergence has been achieved Forces acting on atoms (Ry/au): atom 1 type 1 force = 0.00000253 -0.00000639 -0.00002497 atom 2 type 1 force = -0.00000977 0.00007654 0.00003837 atom 3 type 1 force = 0.00001686 -0.00000393 -0.00003744 atom 4 type 1 force = -0.00002517 -0.00008976 -0.00005193 atom 5 type 1 force = 0.00002007 0.00000939 -0.00004732 atom 6 type 1 force = 0.00007598 -0.00004787 -0.00003895 atom 7 type 1 force = -0.00004187 -0.00005210 -0.00004987 atom 8 type 1 force = 0.00007160 0.00002922 -0.00003445 atom 9 type 1 force = -0.00002615 -0.00003529 0.00005795 atom 10 type 1 force = -0.00006129 0.00004927 0.00009998 atom 11 type 2 force = -0.00005381 0.00007155 0.00007934 atom 12 type 2 force = 0.00000162 0.00002190 -0.00004511 atom 13 type 2 force = 0.00000934 -0.00002334 -0.00004641 atom 14 type 2 force = 0.00003732 -0.00001936 0.00003124 atom 15 type 2 force = -0.00001727 0.00002015 0.00006958 Total force = 0.000311 Total SCF correction = 0.000021 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= - 1.13 -0.00001084 0.00000003 0.00000019 -1.59 0.01 0.03 0.00000003 -0.00001082 -0.00000037 0.01 -1.59 -0.05 0.00000019 -0.00000037 -0.00000140 0.03 -0.05 -0.21 bfgs converged in 27 scf cycles and 24 bfgs steps End of BFGS Geometry Optimization Final energy = -172.2464671441 Ry the red are different ,so i worried about the former is not convergence,because i set the step is 50. is that so ? i need your help . thank you -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071130/0756dda9/attachment.htm From zhufengcc at gmail.com Fri Nov 30 09:27:25 2007 From: zhufengcc at gmail.com (feng zhu) Date: Fri, 30 Nov 2007 16:27:25 +0800 Subject: [Pw_forum] a simple question about the relax outfile In-Reply-To: References: Message-ID: <5b273e100711300027v55ce6acfl32d9c6036ae5589a@mail.gmail.com> Yes. For the first case, you need to increase the nstep. 2007/11/30, xu yuehua : > > hi all : > i have a question about the out file of relax : > for example :sometimes :the end of file is like: > ......... > > convergence has been achieved > > Forces acting on atoms (Ry/au): > > atom 1 type 1 force = -0.00016739 0.00006788 0.00006825 > atom 2 type 1 force = 0.00000361 0.00001673 -0.00004580 > atom 3 type 1 force = 0.00007761 0.00005882 0.00004672 > atom 4 type 1 force = 0.00003052 0.00004707 0.00002272 > atom 5 type 1 force = 0.00001339 0.00015088 -0.00012165 > atom 6 type 1 force = -0.00010455 -0.00002068 - > 0.00010771 > atom 7 type 1 force = -0.00009487 -0.00001394 0.00006473 > atom 8 type 1 force = -0.00000214 -0.00007630 0.00001052 > atom 9 type 1 force = 0.00002701 0.00001038 0.00000479 > atom 10 type 1 force = 0.00008230 0.00004823 0.00012935 > atom 11 type 2 force = 0.00015243 -0.00016928 -0.00010298 > atom 12 type 2 force = -0.00004574 -0.00002796 > 0.00000972 > atom 13 type 2 force = 0.00003206 0.00001371 -0.00009648 > atom 14 type 2 force = 0.00000462 -0.00000415 0.00003920 > atom 15 type 2 force = - 0.00000887 -0.00010141 > 0.00007861 > > Total force = 0.000506 Total SCF correction = 0.000024 > > > entering subroutine stress ... > > total stress (Ry/bohr**3) (kbar) P= - > 0.60 > -0.00001045 -0.00000009 0.00000000 -1.54 -0.01 0.00 > -0.00000009 -0.00001053 -0.00000015 -0.01 - 1.55 -0.02 > 0.00000000 -0.00000015 0.00000867 0.00 -0.02 1.27 > > > The maximum number of steps has been reached. > > End of BFGS Geometry Optimization > > and sometimes the same part of the outfile: > > convergence has been achieved > > Forces acting on atoms (Ry/au): > > atom 1 type 1 force = 0.00000253 -0.00000639 -0.00002497 > atom 2 type 1 force = -0.00000977 0.00007654 0.00003837 > atom 3 type 1 force = 0.00001686 -0.00000393 - > 0.00003744 > atom 4 type 1 force = -0.00002517 -0.00008976 -0.00005193 > atom 5 type 1 force = 0.00002007 0.00000939 -0.00004732 > atom 6 type 1 force = 0.00007598 -0.00004787 -0.00003895 > atom 7 type 1 force = -0.00004187 -0.00005210 -0.00004987 > atom 8 type 1 force = 0.00007160 0.00002922 -0.00003445 > atom 9 type 1 force = - 0.00002615 -0.00003529 > 0.00005795 > atom 10 type 1 force = -0.00006129 0.00004927 0.00009998 > atom 11 type 2 force = -0.00005381 0.00007155 0.00007934 > atom 12 type 2 force = 0.00000162 0.00002190 -0.00004511 > atom 13 type 2 force = 0.00000934 -0.00002334 -0.00004641 > atom 14 type 2 force = 0.00003732 -0.00001936 0.00003124 > atom 15 type 2 force = - 0.00001727 0.00002015 > 0.00006958 > > Total force = 0.000311 Total SCF correction = 0.000021 > > > entering subroutine stress ... > > total stress (Ry/bohr**3) (kbar) P= - > 1.13 > -0.00001084 0.00000003 0.00000019 -1.59 0.01 0.03 > 0.00000003 -0.00001082 -0.00000037 0.01 -1.59 -0.05 > 0.00000019 -0.00000037 -0.00000140 0.03 -0.05 -0.21 > > > bfgs converged in 27 scf cycles and 24 bfgs steps > > End of BFGS Geometry Optimization > > Final energy = -172.2464671441 Ry > the red are different ,so i worried about the former is not > convergence,because i set the step is 50. > is that so ? > i need your help . > thank you > > -- > Xu Yuehua > physics Department of Nanjing university > China > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ------------------ Zhu feng Department of Physics, Anhui University, Hefei, China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071130/43596148/attachment.htm From tone.kokalj at ijs.si Fri Nov 30 12:05:14 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Fri, 30 Nov 2007 12:05:14 +0100 Subject: [Pw_forum] Questions about plotting In-Reply-To: <20071129190112.oedjh66j2g8o48cw@webmail.units.it> References: <20071129190112.oedjh66j2g8o48cw@webmail.units.it> Message-ID: <1196420714.2623.5.camel@localhost.localdomain> On Thu, 2007-11-29 at 19:01 +0100, glevita at units.it wrote: > Second: being unable to extract easily manageable numerical values out > of the XSF file obtained for some |Psi^2|, I decided to create a > series of one dimensional plots, related to vectors parallel to the > molecular axis, out of which I'll extract some preliminary info. To > test this tool, I tried to plot the charge density (as a test) on a > certain direction. This was my input file: > > &inputpp > prefix = 'PVBA' > outdir = './' > filplot = 'projlin' > plot_num= 0 > / > &plot > nfile = 1 > filepp(1) = 'projlin' > weight(1) = 1.0 > iflag = 1 > output_format = 3 > fileout = 'projlin.ldos.xsf' > e1(1)= 0.0, e1(2)= 0.0, e1(3)= 1.0, > x0(1)= 0.0, x0(2)= 0.1, x0(3)= 0.0, > nx= 100 > / > > There must be something wrong with that, as the file 'projlin' is > correctly generated, but the related XSF file is not. There is no 1D XSF file (it can be only 2D or 3D) ! As for the CVS version of espresso (that I am using), your input would result in kind of gnuplot data file, i.e., x & f(x) columns that is written to projlin.ldos.xsf. Plese change to, for example, output_format=0 for 1D plot (this will produce output suitable for gnuplot). Regards, Tone -- J. Stefan Institute, Ljubljana, Slovenia From giannozz at nest.sns.it Fri Nov 30 12:30:36 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 30 Nov 2007 12:30:36 +0100 Subject: [Pw_forum] Questions about plotting In-Reply-To: <20071129190112.oedjh66j2g8o48cw@webmail.units.it> References: <20071129190112.oedjh66j2g8o48cw@webmail.units.it> Message-ID: On Nov 29, 2007, at 19:01 , glevita at units.it wrote: > is there any postprocessing tool in which by providing x,y,z,R and E, > the LDOS within a sphere of radius R centred around a xyz point for > the given energy E is obtained? no, presently there isn't Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From njuxuyuehua at gmail.com Fri Nov 30 12:36:00 2007 From: njuxuyuehua at gmail.com (xu yuehua) Date: Fri, 30 Nov 2007 19:36:00 +0800 Subject: [Pw_forum] how to continue my relax job Message-ID: hi everyone here : i have another question about how to continue my relax job.for i set the nstep is 50, too small ,and can not acheive convergence,and i want to continue by adding the nstep,how to do with it ? is it to add the restart option in input file? in general ,how many step i should to add? i wish your help thank you ! -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20071130/cef61e6a/attachment.htm From giannozz at nest.sns.it Fri Nov 30 13:21:05 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 30 Nov 2007 13:21:05 +0100 Subject: [Pw_forum] Warning in the 'nscf' calculation In-Reply-To: <22ae3ca40711291520rfd5cb65kc0e5b07a061fe056@mail.gmail.com> References: <22ae3ca40711291520rfd5cb65kc0e5b07a061fe056@mail.gmail.com> Message-ID: On Nov 30, 2007, at 24:20 , alan chen wrote: > WARNING: 3 eigenvalues not converged [...] > The band structure is finally calculated at the end of output file > but shall I believe in these E_{nk} values? the code counts unconverged eigenvalues and makes another tour of iterative diagonalization, until the desired number of eigenvalues (all of them for non-scf calculations) have converged: see function test_exit_cond. So yes, you should believe these values, or more exactly, you shouldn't be any more suspicious that for all other numbers. > How can I get rid of these warning messages? try the "see no evil" approach: comment the line yielding the warning message. The alternative is the "computer geek" approach: find a better algorithm. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From paulatto at sissa.it Fri Nov 30 14:09:38 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Fri, 30 Nov 2007 14:09:38 +0100 (CET) Subject: [Pw_forum] how to continue my relax job In-Reply-To: References: Message-ID: <4690.147.122.5.182.1196428178.squirrel@webmail.sissa.it> On Fri, November 30, 2007 12:36, xu yuehua wrote: > hi everyone here : > i have another question about how to continue my relax job.for i set the > nstep is 50, too small ,and can not acheive convergence,and i want to > continue > by adding the nstep,how to do with it ? > is it to add the restart option in input file? > in general ,how many step i should to add? > i wish your help > thank you ! dear Xu Yuehua, you should put restart_mode = 'restart' in the &control section of your input and rerun the calculation. I'm not sure how it works for relax, so, to be sure, I would replace the initial atomic positions with the lower energy ones I obtained during the incomplete relaxation (they are listed in the output). If for some reason restart doesn't work you can do it by hand, replacing atomic positions with the better ones in your input and starting from scratch. You will loose a bit of time during the first few iteration, but it always work. goodbye -- Lorenzo Paulatto +39 040 3787 511 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From gaoguoying at gmail.com Fri Nov 30 14:36:44 2007 From: gaoguoying at gmail.com (Guoying Gao) Date: Fri, 30 Nov 2007 21:36:44 +0800 Subject: [Pw_forum] how to calculate phonon spectra with q2r.x for ibrav=0 Message-ID: <6234c0c60711300536y204d2f30n6fa7ac334e0f5bb2@mail.gmail.com> Dear everyone, I have calculated the phonon frequencies of special q points through defining ibrav=0, however, when I interpolate the force constants for the phonon spectra with q2r.x, it outputs nothing in q2r.out file. I am using Quantum-espresso-2.1.5 version. This ciumstance only happens for ibrav=0, not for others. So does anyone kown the reason for the circumstance of ibrav=0, and how to deal with this case? Any suggestions will be appreciated. Thanks in advance. Gao From riccardo at fcs.it Fri Nov 30 14:35:02 2007 From: riccardo at fcs.it (Riccardo Sabatini) Date: Fri, 30 Nov 2007 14:35:02 +0100 Subject: [Pw_forum] Good news about intel mac installation Message-ID: <47501186.1040007@fcs.it> Hi everyone, i don't know if it's the case but i'm writing to inform about the final good compilation on my mac intel !!! In some previous mail i asked help (and i really had a gret one) from the list to install QE on my intel mac. There have been several problems, mainly about the linker and ranlib. Well, it seems the problems were generated from the presence of several incompatible versions of libraries/compilers/linkers. For other users of intel mac i would suggest some important notes: uninstall darwin ports, fink and developer tools. The presence of all of those at the same time generates many spooky events in the compilation procedure. I installed just the developer tools from apple, the intel fortran compiler and everything went on great. I then installed all the other things i removed before. I got to install everything and even XCrysDen, thanks to the patch found in the list. Now i'm all set and ready to start the hard part, learning QE !!!!!!!!!!!! Really really thanks to all of you, Riccardo From glevita at units.it Fri Nov 30 16:12:30 2007 From: glevita at units.it (glevita at units.it) Date: Fri, 30 Nov 2007 16:12:30 +0100 Subject: [Pw_forum] Questions about plotting In-Reply-To: <1196420714.2623.5.camel@localhost.localdomain> References: <20071129190112.oedjh66j2g8o48cw@webmail.units.it> <1196420714.2623.5.camel@localhost.localdomain> Message-ID: <20071130161230.40gbd9j093s4cg40@webmail.units.it> Quoting Tone Kokalj : > > On Thu, 2007-11-29 at 19:01 +0100, glevita at units.it wrote: > >> Second: being unable to extract easily manageable numerical values out >> of the XSF file obtained for some |Psi^2|, I decided to create a >> series of one dimensional plots, related to vectors parallel to the >> molecular axis, out of which I'll extract some preliminary info. To >> test this tool, I tried to plot the charge density (as a test) on a >> certain direction. This was my input file: >> >> &inputpp >> prefix = 'PVBA' >> outdir = './' >> filplot = 'projlin' >> plot_num= 0 >> / >> &plot >> nfile = 1 >> filepp(1) = 'projlin' >> weight(1) = 1.0 >> iflag = 1 >> output_format = 3 >> fileout = 'projlin.ldos.xsf' >> e1(1)= 0.0, e1(2)= 0.0, e1(3)= 1.0, >> x0(1)= 0.0, x0(2)= 0.1, x0(3)= 0.0, >> nx= 100 >> / >> >> There must be something wrong with that, as the file 'projlin' is >> correctly generated, but the related XSF file is not. > > There is no 1D XSF file (it can be only 2D or 3D) ! > > As for the CVS version of espresso (that I am using), your input would > result in > kind of gnuplot data file, i.e., x & f(x) columns that is written to > projlin.ldos.xsf. > > Plese change to, for example, output_format=0 for 1D plot (this will > produce output suitable for gnuplot). > > Regards, Tone > -- > J. Stefan Institute, Ljubljana, Slovenia > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > Dear Tone, I understand your point. Still, I cannot overcome the problem. In the INPUT_PP example it is reported that the output_format selection is ignored when 1D plots are chosen. (moreover, the output_format= 3 option is reported to be suited also to 1D Xcrysden). In fact, I failed both when output_format was 0 and 3, and when iflag was either 0 or 1. No plot files were generated, even if the first PP output file was OK. I used both a 3.2.2 version of Espresso and a 3.0 one. Any suggestion welcome! Giacomo ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From akohlmey at cmm.chem.upenn.edu Fri Nov 30 16:36:59 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 30 Nov 2007 10:36:59 -0500 (EST) Subject: [Pw_forum] how to calculate phonon spectra with q2r.x for ibrav=0 In-Reply-To: <6234c0c60711300536y204d2f30n6fa7ac334e0f5bb2@mail.gmail.com> References: <6234c0c60711300536y204d2f30n6fa7ac334e0f5bb2@mail.gmail.com> Message-ID: On Fri, 30 Nov 2007, Guoying Gao wrote: GG> Dear everyone, dear gao, GG> I have calculated the phonon frequencies of special q points GG> through defining ibrav=0, however, when I interpolate the force GG> constants for the phonon spectra with q2r.x, it outputs nothing in GG> q2r.out file. I am using Quantum-espresso-2.1.5 version. This without looking at your problem, please note that the current release of quantum espresso is version 3.2.3 and 2.1.5 is over two years old. given the pace of development in quantum espresso and the amounts of improvements that have been put into the code since, i would not expect anybody to be interested in fixing a problem with a now obsolete version of the code (if your problem has not already been fixed). GG> ciumstance only happens for ibrav=0, not for others. So does anyone GG> kown the reason for the circumstance of ibrav=0, and how to deal with GG> this case? Any suggestions will be appreciated. two suggestions: 1) try the current version of QE, 2) you could translate your system so that it matches ibrav=14 or any other appropriate bravais lattice. cheers, axel. GG> Thanks in advance. GG> GG> Gao GG> _______________________________________________ GG> Pw_forum mailing list GG> Pw_forum at pwscf.org GG> http://www.democritos.it/mailman/listinfo/pw_forum GG> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From giannozz at nest.sns.it Fri Nov 30 16:50:15 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 30 Nov 2007 16:50:15 +0100 Subject: [Pw_forum] Questions about plotting In-Reply-To: <20071130161230.40gbd9j093s4cg40@webmail.units.it> References: <20071129190112.oedjh66j2g8o48cw@webmail.units.it> <1196420714.2623.5.camel@localhost.localdomain> <20071130161230.40gbd9j093s4cg40@webmail.units.it> Message-ID: On Nov 30, 2007, at 16:12 , glevita at units.it wrote: > In the INPUT_PP example it is reported that the output_format > selection is ignored when 1D plots are chosen. (moreover, the > output_format= 3 option is reported to be suited also to 1D Xcrysden). > In fact, I failed both when output_format was 0 and 3, and when iflag > was either 0 or 1. No plot files were generated, even if the first PP > output file was OK. there are way too many options in the postprocessing codes to keep track of all of them. If it works in the serial case, it is for sure a problem of parallel execution. > I used both a 3.2.2 version of Espresso and a 3.0 one. it is not unlikely that your problem may have been solved meanwhile. See http://www.democritos.it:8888/O-sesame/chngview?cn=5321 and http://www.democritos.it:8888/O-sesame/chngview?cn=4894 Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222