From giannozz at nest.sns.it Mon Apr 2 11:14:39 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 2 Apr 2007 11:14:39 +0200 Subject: [Pw_forum] intel mac - compilation with ifort/icc and not gcc In-Reply-To: References: Message-ID: On Mar 31, 2007, at 22:04 , Gilles Hug wrote: > Does espresso search gcc in some other directory? no, but it is likely that since "configure" did not recognize your configuration, it assumed that you need explicit preprocessing of fortran files. Please try the following modified version of the "configure" script: http://www.fisica.uniud.it/~giannozz/public/configure It should recognize the intel compiler for macintosh as well. In order to recognize mkl the exact way to load mkl for macintos is needed Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From lukethulin at netscape.net Mon Apr 2 15:24:02 2007 From: lukethulin at netscape.net (lukethulin at netscape.net) Date: Mon, 2 Apr 2007 08:24:02 -0500 Subject: [Pw_forum] Electron affinity Message-ID: <001801c7752a$2e4357b0$6601a8c0@LukePC> Does anybody know how to choose a reference energy to track the change in electron affinity of a strained crystal? Thank you, Luke Thulin Nanoptek, Corp. 8 Clock Tower Place, Suite 120 Maynard, MA 01754 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070402/6ce8e6c4/attachment.htm From akohlmey at cmm.chem.upenn.edu Mon Apr 2 15:25:04 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 2 Apr 2007 09:25:04 -0400 (EDT) Subject: [Pw_forum] intel mac - compilation with ifort/icc and not gcc In-Reply-To: Message-ID: On Mon, 2 Apr 2007, Paolo Giannozzi wrote: PG> PG> On Mar 31, 2007, at 22:04 , Gilles Hug wrote: PG> PG> > Does espresso search gcc in some other directory? PG> PG> no, but it is likely that since "configure" did not recognize PG> your configuration, it assumed that you need explicit PG> preprocessing of fortran files. Please try the following one more remark. if that has happened once, you'll have to unpack the source archive again and take it from there. thanks to the (only?) braindead feature of the macos of having a casepreserving, yet case insensitive file system, one attempt of this kind of preprocessing can ruin some of the sources. axel. PG> modified version of the "configure" script: PG> http://www.fisica.uniud.it/~giannozz/public/configure PG> It should recognize the intel compiler for macintosh as PG> well. In order to recognize mkl the exact way to load PG> mkl for macintos is needed PG> PG> Paolo PG> --- PG> Paolo Giannozzi, Democritos and University of Udine, Italy PG> 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 baroni at sissa.it Mon Apr 2 19:07:01 2007 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 2 Apr 2007 19:07:01 +0200 Subject: [Pw_forum] Electron affinity In-Reply-To: <001801c7752a$2e4357b0$6601a8c0@LukePC> References: <001801c7752a$2e4357b0$6601a8c0@LukePC> Message-ID: <16C2E3D7-0855-4323-A925-35D284BA277D@sissa.it> I think it depends on how the electron affinity is supposed to be measured/used. The electron affinity (i.e. the absolute position of the bottom of the conduction band) is an ill defined concept in an infinite system. Its actual value depends on surface/interface effects, whose modeling may depends on the details of the experimental setup. Stefano On Apr 2, 2007, at 3:24 PM, lukethulin at netscape.net wrote: > Does anybody know how to choose a reference energy to track the > change in electron affinity of a strained crystal? > > Thank you, > > Luke Thulin > Nanoptek, Corp. > 8 Clock Tower Place, Suite 120 > Maynard, MA 01754 > --- 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: /pipermail/attachments/20070402/28ee4e52/attachment.htm From proffess at yandex.ru Mon Apr 2 21:17:35 2007 From: proffess at yandex.ru (Sergey Lisenkov) Date: Mon, 2 Apr 2007 23:17:35 +0400 (MSD) Subject: [Pw_forum] using *_extrapolation in pw In-Reply-To: References: Message-ID: <461156CF.000001.26516@webmail14.yandex.ru> Dear all, I noticed that when I use wfc(pot)_extrapolation='second_order' (according to tutorial, it should speed up calculations) in pwscf, job (in parallel) hangs somewhere in scf calculations, if I disable it (corresponds to 'none'), there are no such problems. Has anybody encountered such behavior? I am using pgi-7.0.2 on cray xd1, acml, -D__XD1 flag. Sergey From yaoxinxin at mail.sdu.edu.cn Tue Apr 3 12:08:42 2007 From: yaoxinxin at mail.sdu.edu.cn (Xinxin Yao) Date: Tue, 03 Apr 2007 18:08:42 +0800 Subject: [Pw_forum] error in band structure calculation Message-ID: <375594922.07374@mail.sdu.edu.cn> Hi users, I meet some problems when calculate the band structures. At first, I did an vc-relax calculation. After that, when I did the band structure calculation, I got an error as follows: nbndx = 640 nbnd = 160 natomwfc = 274 npwx = 20423 nelec = 278.00 nkb = 222 ngl = 1519 Check: negative/imaginary core charge= -0.000025 0.000000 The potential is recalculated from file : ge-mn.save/charge-density.xml %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from potinit : error # 1 starting and expected charges differ %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... Could someone suggest me something, please? Regards, Xinxin Yao My input for scf calculation is the following: &CONTROL title = ge-mn , calculation = 'vc-relax' , restart_mode = 'from_scratch' , wf_collect = .true. , outdir = './' , pseudo_dir = '../' , prefix = ge-mn , disk_io = 'minimal' , verbosity = 'high' , nstep = 1000 , tstress = .true. , tprnfor = .true. , dt = 100 , / &SYSTEM ibrav = 1, celldm(1) = 21.1940, nat = 64, ntyp = 3, ecutwfc = 40 , nbnd = 160, occupations = 'tetrahedra' , degauss = 0.005 , smearing = 'methfessel-paxton' , nspin = 2 , starting_magnetization(2) = 3, starting_magnetization(3) = -3, / &ELECTRONS electron_maxstep = 100, conv_thr = 1E-5 , startingpot = 'atomic' , startingwfc = 'atomic' , mixing_mode = 'plain' , mixing_beta = 0.2 , diagonalization = 'david' , / &IONS ion_dynamics = 'damp' , / &CELL cell_dynamics = 'damp-pr' , / ATOMIC_SPECIES Ge 72.64100 ge.ncpp Mn1 54.93800 Mn.pbe-sp-van.UPF Mn2 54.93800 Mn.pbe-sp-van.UPF ATOMIC_POSITIONS alat ...... K_POINTS automatic 4 4 4 1 1 1 The input for band structure calculation is the following: &CONTROL calculation = 'nscf' , restart_mode = 'from_scratch' , outdir = '/home/disk2/work/7.05/temp/' , pseudo_dir = '/home/neshque/work/PSEUDOPOTENTIAL/' , prefix = ge-mn , wf_collect = .true. , disk_io = 'minimal' , verbosity = 'high' , / &SYSTEM ibrav = 1, celldm(1) = 21.1940, nat = 64, ntyp = 3, ecutwfc = 40 , nbnd = 160, degauss = 0.005 , smearing = 'methfessel-paxton' , nspin = 2 , starting_magnetization(2) = 3 , starting_magnetization(3) = -3 , / &ELECTRONS conv_thr = 1E-5 , / ATOMIC_SPECIES Ge 72.64100 ge.ncpp Mn1 54.93800 Mn.pbe-sp-van.UPF Mn2 54.93800 Mn.pbe-sp-van.UPF ATOMIC_POSITIONS alat ...... K_POINTS tpiba 11 give manually From iori.federico at unimore.it Tue Apr 3 12:30:38 2007 From: iori.federico at unimore.it (Federico Iori) Date: Tue, 03 Apr 2007 12:30:38 +0200 Subject: [Pw_forum] error in band structure calculation In-Reply-To: <375594922.07374@mail.sdu.edu.cn> References: <375594922.07374@mail.sdu.edu.cn> Message-ID: <46122CCE.3080308@unimore.it> HI! I had the same problems.....and with the help of Paolo I finally resolved! 1) which version do you use? for PW > 3.0 maybe is better to use "bands" instead of "nscf" for a bands structure calculation. 2) Xheck the position also when doing the "bands" calculation: and assure that in the second run you are using exactly the relax position found in the relax calculation. Hope to be usefull this. Ciao! ff. Xinxin Yao ha scritto: > Hi users, > I meet some problems when calculate the band structures. > At first, I did an vc-relax calculation. After that, when I did the band > structure calculation, I got an error as follows: > > nbndx = 640 nbnd = 160 natomwfc = 274 npwx = 20423 > nelec = 278.00 nkb = 222 ngl = 1519 > > Check: negative/imaginary core charge= -0.000025 0.000000 > > The potential is recalculated from file : > ge-mn.save/charge-density.xml > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from potinit : error # 1 > starting and expected charges differ > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > > Could someone suggest me something, please? > > Regards, > Xinxin Yao > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Federico Iori - PhD Student - INFM-S3 and Dipartimento di Fisica Universit? degli Studi di Modena e Reggio Emilia via Campi, 213/A - 41100 Modena - Italia www.s3.infm.it www.fisica.unimo.it phone: +39 59 2055283 #1 fax n?: +39 059 37 47 94 #2 fax n?: +39 059 205 5651 email: iori.federico at unimore.it From giannozz at nest.sns.it Tue Apr 3 13:12:17 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 3 Apr 2007 13:12:17 +0200 Subject: [Pw_forum] error in band structure calculation In-Reply-To: <375594922.07374@mail.sdu.edu.cn> References: <375594922.07374@mail.sdu.edu.cn> Message-ID: <9AE6173E-32E8-4310-90B4-1F6501E83236@nest.sns.it> On Apr 3, 2007, at 12:08 , Xinxin Yao wrote: > prefix = ge-mn , prefix = 'something', not prefix = something Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Tue Apr 3 13:13:35 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 3 Apr 2007 13:13:35 +0200 Subject: [Pw_forum] using *_extrapolation in pw In-Reply-To: <461156CF.000001.26516@webmail14.yandex.ru> References: <461156CF.000001.26516@webmail14.yandex.ru> Message-ID: <8B71E08B-E3F4-42CD-B21F-950AE73A6209@nest.sns.it> On Apr 2, 2007, at 21:17 , Sergey Lisenkov wrote: > I noticed that when I use wfc(pot)_extrapolation='second_order' it should, but it seldom does P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From yaoxinxin at mail.sdu.edu.cn Tue Apr 3 16:31:01 2007 From: yaoxinxin at mail.sdu.edu.cn (Xinxin Yao) Date: Tue, 03 Apr 2007 22:31:01 +0800 Subject: [Pw_forum] Re:error in band structure calculation Message-ID: <375610661.31304@mail.sdu.edu.cn> Thank you for your help! I use version 3.2 . When calculation="nscf", I use the initial atom position of vc-relax, then I meet the error mentioned above, and the file "data-file.xml" in the output folder is missing. I change the atom position to the relax position, and restart. A new error appears as follows: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from setup : error # 1 problem reading ef from file /home/disk2/work/7.05/temp/ge-mn.save %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... So I copy the backup "data-file.xml" to the output folder. The second error is resolved, but the first one remains. When calculation="bands", I get the same results. My question is: 1)Why does the program delete the file "data-file.xml" when doing band structure calculation? I have to copy the backup file into the output folder after each calculation. 2)Whether I use "bands" or "nscf", and whether I use the initial atom position or the relax one when doing band calculation, I always meet the error "starting and expected charges differ". I delete the file "charge-density.xml" manually, and get message as follows: Check: negative/imaginary core charge= -0.000025 0.000000 Cannot read rho : file not found Initial potential from superposition of free atoms Check: negative starting charge=(component1): -12.900563 Check: negative starting charge=(component2): -12.927055 negative rho (up, down): 0.129E+02 0.129E+02 Starting wfc are atomic Then the calculation can finish. This is the only way I could use to finish the band calculation now. Is that right, or could you give me any other helpful suggestions? 3)During my use, I find the prefix = 'something' and prefix = something are equal. I think we don't have to distinguish them exactly. Thank you again! Xinxin Yao From giannozz at nest.sns.it Tue Apr 3 17:35:50 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 3 Apr 2007 17:35:50 +0200 Subject: [Pw_forum] Re:error in band structure calculation In-Reply-To: <375610661.31304@mail.sdu.edu.cn> References: <375610661.31304@mail.sdu.edu.cn> Message-ID: <7B75DFE5-580A-4C66-A331-D043080040F4@nest.sns.it> On Apr 3, 2007, at 16:31 , Xinxin Yao wrote: > When calculation="nscf", I use the initial atom position of vc-relax "nscf" (or "bands") means "perform a non-selfconsistent calculation, starting from a previously calculated charge density, with the atomic positions specified in input". It is your responsibility to use atomic positions that are consisten with the charge density. This is not the reason for the error but it is important to get a correct band structure. > 1)Why does the program delete the file "data-file.xml" when doing band > structure calculation? to ensure consistency of the output data. I don't remember the exact reason but there is one. > 2)Whether I use "bands" or "nscf", and whether I use the initial atom > position or the relax one when doing band calculation, I always meet > the error "starting and expected charges differ". also for the examples? in any event, install the 3.2.1 patch: it corrects a missing initialization that could lead to serious trouble in some cases > I delete the file "charge-density.xml" manually, and get > message as follows: > Cannot read rho : file not found > Initial potential from superposition of free atoms notice the messages > Check: negative starting charge=(component1): -12.900563 > Check: negative starting charge=(component2): -12.927055 -26 electrons? there is something wrong in the input data (or in the pseudopotential files). One thing that doesn't seem good is the value of "starting_magnetization" (+3 and -3) : this should be between +1 and -1 (it is not written in the documentation but it should be). Maybe this is the reason for so much negative charge. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From yaoxinxin at mail.sdu.edu.cn Wed Apr 4 09:26:24 2007 From: yaoxinxin at mail.sdu.edu.cn (Xinxin Yao) Date: Wed, 04 Apr 2007 15:26:24 +0800 Subject: [Pw_forum] Re:error in band structure calculation Message-ID: <375671584.17892@mail.sdu.edu.cn> Hi, Paolo I finally find my fault. I didn't use the relaxed celldm to calculate band structure. When I correct it, the error disappears. Thank you! Xinxin Yao From zhx1984 at gmail.com Wed Apr 4 09:45:20 2007 From: zhx1984 at gmail.com (Xi Zhu) Date: Wed, 4 Apr 2007 15:45:20 +0800 Subject: [Pw_forum] the magnetizan calculation Message-ID: <3036f19f0704040045n1bf80d83v8260370fa8a82b9d@mail.gmail.com> Dear pwscf users; I calculate one of metal-organic moleculer use vasp and pwscf, with the same gaussianoccupancy, but the pwscf can't give the net magmom,while vasp can. It seems the PWSCF is weak to deal with the metal-organic system.does anyone agree? thanks From baroni at sissa.it Wed Apr 4 10:43:45 2007 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 4 Apr 2007 10:43:45 +0200 Subject: [Pw_forum] the magnetizan calculation In-Reply-To: <3036f19f0704040045n1bf80d83v8260370fa8a82b9d@mail.gmail.com> References: <3036f19f0704040045n1bf80d83v8260370fa8a82b9d@mail.gmail.com> Message-ID: I don't - SB On Apr 4, 2007, at 9:45 AM, Xi Zhu wrote: > Dear pwscf users; > I calculate one of metal-organic moleculer use vasp and pwscf, > with the same gaussianoccupancy, but the pwscf can't give the net > magmom,while vasp can. > It seems the PWSCF is weak to deal with the metal-organic system.does > anyone agree? thanks > _______________________________________________ > 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: /pipermail/attachments/20070404/b7e79fbc/attachment.htm From eyvaz_isaev at yahoo.com Wed Apr 4 11:02:54 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 4 Apr 2007 02:02:54 -0700 (PDT) Subject: [Pw_forum] the magnetizan calculation Message-ID: <854120.48417.qm@web60323.mail.yahoo.com> Hi, I calculate one of metal-organic moleculer use vasp and pwscf, with the same gaussianoccupancy, Actually I did not understand what is "gaussianoccupancy". I can not remember such a kind keyword neither in VASP, nor in PWSCF. If you mean gaussian broadening, that is not correct, SIGMA in VASP is in eV, but degauss in PWSCF is in Ry. Restart your calculations using the tetrahedra method for both VASP and PWSCF. but the pwscf can't give the net magmom,while vasp can. It seems the PWSCF is weak to deal with the metal-organic system.does anyone agree? A while ago I had another case: there was a net magnetism using PWSCF, but not using VASP. It seems, finally they gave similar results. So, I think your fail is actually connected to details of your calculations. Because of that VASP and PWSCF are based on the same theoretical background (DFT, pseudopotentials) I do not agree that VASP can deal with metallo-organic systems, and PWSCF can not. Of course, there are specific properties, as example NMR, you can not calculate using ultrasoft pseudopotentials implemented in current Quantum Espresso (QE), but things are in progress and PAW potentials eventually will be implemented in QE, too. By the way, it was discussed before in the forum, that a user has to specify his/her name, at least, nickname. Unfortunately, this is not the case and I (presumably other participants of the forum) noticed there was lots of mails without a name/nickname. Bests, Eyvaz. ____________________________________________________________________________________ Looking for earth-friendly autos? Browse Top Cars by "Green Rating" at Yahoo! Autos' Green Center. http://autos.yahoo.com/green_center/ From wlyim at puccini.che.pitt.edu Wed Apr 4 10:19:22 2007 From: wlyim at puccini.che.pitt.edu (wlyim at puccini.che.pitt.edu) Date: Wed, 4 Apr 2007 04:19:22 -0400 (EDT) Subject: [Pw_forum] the magnetizan calculation In-Reply-To: <3036f19f0704040045n1bf80d83v8260370fa8a82b9d@mail.gmail.com> Message-ID: Can you provide your system setting for PWscf? There is not enough information to make your statement. In my experience, the net magmom from the two programs agree very well. On Wed, 4 Apr 2007, Xi Zhu wrote: > Dear pwscf users; > I calculate one of metal-organic moleculer use vasp and pwscf, > with the same gaussianoccupancy, but the pwscf can't give the net > magmom,while vasp can. > It seems the PWSCF is weak to deal with the metal-organic system.does > anyone agree? thanks > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Dr. Wai-Leung Yim Institut fuer Reine und Angewandte Chemie, Theoretische Chemie, Carl von Ossiezky Universtaet Oldenburg, 26129 Oldenburg, Germany Email: wlyim at puccini.che.pitt.edu Phone: +49-441-798-3950 (office) +49-441-798-5102 (home) Fax: +49-441-798-3964 From vshastra at gmail.com Wed Apr 4 12:51:35 2007 From: vshastra at gmail.com (Vastu Shastra) Date: Wed, 4 Apr 2007 05:51:35 -0500 Subject: [Pw_forum] compilation error Message-ID: Hello, I get the following error while compiling QE 3.2 on my machine. I am using ubuntu linux on PC. This error occurs while using 'make all' command. Any help is appreciated. . . . libpw.a(bp_radin.o): In function `radlg_':bp_radin.f:(.text+0x170): undefined reference to `s_wsl e' :bp_radin.f:(.text+0x194): undefined reference to `do_lio' :bp_radin.f:(.text+0x199): undefined reference to `e_wsle' :bp_radin.f:(.text+0x1a5): undefined reference to `s_wsle' :bp_radin.f:(.text+0x1c9): undefined reference to `do_lio' :bp_radin.f:(.text+0x1e9): undefined reference to `do_lio' :bp_radin.f:(.text+0x1ee): undefined reference to `e_wsle' :bp_radin.f:(.text+0x202): undefined reference to `s_stop' libpw.a(bp_radin.o): In function `radlg1_':bp_radin.f:(.text+0x2e4): undefined reference to `s_ws le' . . . . . :lapack_all.f:(.text+0x5d17): undefined reference to `do_fio' :lapack_all.f:(.text+0x5d2f): undefined reference to `do_fio' :lapack_all.f:(.text+0x5d34): undefined reference to `e_wsfe' :lapack_all.f:(.text+0x5d48): undefined reference to `s_stop' ../flib/lapack.a(dlamch.o): In function `dlamc2_':dlamch.f:(.text+0xde9): undefined reference to `s_wsfe' :dlamch.f:(.text+0xe05): undefined reference to `do_fio' :dlamch.f:(.text+0xe0a): undefined reference to `e_wsfe' make[1]: *** [pw.x] Error 1 make[1]: Leaving directory `/home/anshu/Desktop/espresso-3.2/PW' make: *** [pw] Error 2 Thanks, Anshu -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070404/a6055a4b/attachment.htm From ceresoli at sissa.it Wed Apr 4 12:59:53 2007 From: ceresoli at sissa.it (Davide Ceresoli) Date: Wed, 04 Apr 2007 12:59:53 +0200 Subject: [Pw_forum] compilation error In-Reply-To: References: Message-ID: <46138529.9060109@sissa.it> Vastu Shastra wrote: > Hello, > I get the following error while compiling QE 3.2 on my machine. I am > using ubuntu linux on PC. This error occurs while using 'make all' > command. Any help is appreciated. http://www.democritos.it/pipermail/pw_forum/2007-February/005863.html From eyvaz_isaev at yahoo.com Wed Apr 4 13:40:33 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 4 Apr 2007 04:40:33 -0700 (PDT) Subject: [Pw_forum] compilation error Message-ID: <524833.87449.qm@web60315.mail.yahoo.com> Hi Anshu, Unfortunately you did not tell us anything about a compiler you used. If you use g95 compiler follow David's advice. If not, add -lg2c or -lf2c flags to linker or to LIBS in make.sys. If you do not have f2c (g2c) library installed (check using ls /usr/lib/*g2c*) on your computer download it from www.netlib.org and compile it for yourself. Bests. Eyvaz. ----- Original Message ---- From: Vastu Shastra To: pw_forum at pwscf.org Sent: Wednesday, April 4, 2007 2:51:35 PM Subject: [Pw_forum] compilation error Hello, I get the following error while compiling QE 3.2 on my machine. I am using ubuntu linux on PC. This error occurs while using 'make all' command. Any help is appreciated. . . . libpw.a(bp_radin.o): In function `radlg_':bp_radin.f:(.text+0x170): undefined reference to `s_wsl e' :bp_radin.f:(.text+0x194): undefined reference to `do_lio' :bp_radin.f:(.text+0x199): undefined reference to `e_wsle' :bp_radin.f:(.text+0x1a5): undefined reference to `s_wsle' :bp_radin.f:(.text+0x1c9): undefined reference to `do_lio' :bp_radin.f:(.text+0x1e9): undefined reference to `do_lio' :bp_radin.f:(.text+0x1ee): undefined reference to `e_wsle' :bp_radin.f:(.text+0x202): undefined reference to `s_stop' libpw.a(bp_radin.o): In function `radlg1_':bp_radin.f:(.text+0x2e4): undefined reference to `s_ws le' . . . . . :lapack_all.f:(.text+0x5d17): undefined reference to `do_fio' :lapack_all.f:(.text+0x5d2f): undefined reference to `do_fio' :lapack_all.f:(.text+0x5d34): undefined reference to `e_wsfe' :lapack_all.f:(.text+0x5d48): undefined reference to `s_stop' ../flib/lapack.a(dlamch.o): In function `dlamc2_':dlamch.f:(.text+0xde9): undefined reference to `s_wsfe' :dlamch.f:(.text+0xe05): undefined reference to `do_fio' :dlamch.f:(.text+0xe0a): undefined reference to `e_wsfe' make[1]: *** [pw.x] Error 1 make[1]: Leaving directory `/home/anshu/Desktop/espresso-3.2/PW' make: *** [pw] Error 2 Thanks, Anshu ____________________________________________________________________________________ Be a PS3 game guru. Get your game face on with the latest PS3 news and previews at Yahoo! Games. http://videogames.yahoo.com/platform?platform=120121 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070404/c6fa8e9c/attachment.htm From akohlmey at cmm.chem.upenn.edu Wed Apr 4 13:10:07 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 4 Apr 2007 07:10:07 -0400 (EDT) Subject: [Pw_forum] the magnetizan calculation In-Reply-To: <3036f19f0704040045n1bf80d83v8260370fa8a82b9d@mail.gmail.com> Message-ID: On Wed, 4 Apr 2007, Xi Zhu wrote: XZ> Dear pwscf users; XZ> I calculate one of metal-organic moleculer use vasp and pwscf, XZ> with the same gaussianoccupancy, but the pwscf can't give the net XZ> magmom,while vasp can. XZ> It seems the PWSCF is weak to deal with the metal-organic system.does XZ> anyone agree? thanks in my experience with plane wave DFT codes so far, the vast majority of claims that a specific software package does not work well, could be traced down to incorrect input or the person not understanding the method at all. so please provide _proof_ for your claim. thanks, axel. XZ> _______________________________________________ XZ> Pw_forum mailing list XZ> Pw_forum at pwscf.org XZ> http://www.democritos.it/mailman/listinfo/pw_forum XZ> -- ======================================================================= 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 niuli1978 at yahoo.com.cn Wed Apr 4 15:48:43 2007 From: niuli1978 at yahoo.com.cn (li niu) Date: Wed, 4 Apr 2007 21:48:43 +0800 (CST) Subject: [Pw_forum] Does anyone have C.pz-mt.UPF ? Message-ID: <529777.3957.qm@web15007.mail.cnb.yahoo.com> Dear all, Does anyone have C.pz-mt.UPF ? Would you please send it to my mailbox: niuli1978 at yahoo.com.cn ? Thanks in advance! Best! Niu Li --------------------------------- ??????-3.5G???20M?? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070404/37ac19d6/attachment.htm From ss626 at cornell.edu Wed Apr 4 15:52:10 2007 From: ss626 at cornell.edu (Shriram Shivaraman) Date: Wed, 04 Apr 2007 09:52:10 -0400 Subject: [Pw_forum] fqha.x Message-ID: <4613AD8A.4090205@cornell.edu> Hi, I'm facing some trouble running this program and hope someone can help me out. I'm using the dos file generated in example 06 to calculate the free energy. But, when I give the input to fqha.x as alas.phdos, for any temperature, the free energy turns out to be zero. I'm not sure what is going wrong here. Can someone please help? Thanks, Shriram From marzari at MIT.EDU Wed Apr 4 15:59:39 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 04 Apr 2007 09:59:39 -0400 Subject: [Pw_forum] Does anyone have C.pz-mt.UPF ? In-Reply-To: <529777.3957.qm@web15007.mail.cnb.yahoo.com> References: <529777.3957.qm@web15007.mail.cnb.yahoo.com> Message-ID: <4613AF4B.8010007@mit.edu> Dear Niu Li, we have generated a LDA Troullier-Martins pseudopotential for C, using the Firtz-Haber code and its standard input for C, and then the fhi to upf converter. It's actually fairly useful for third-order calculations (e.g. electron-phonon), since for C LDA and GGA are of comparable accuracy, and LDA is more well behaved with respect to convergence of higher-order derivatives. In addition, the TM recipe makes it smoother than the currently available von-Barth-Car, on the pwscf database. We'll send you a copy, and send one to the database, to be posted. nicola li niu wrote: > Dear all, > > Does anyone have C.pz-mt.UPF ? > Would you please send it to my mailbox: niuli1978 at yahoo.com.cn > ? > Thanks in advance! > > Best! > > Niu Li > > ------------------------------------------------------------------------ > ??????-3.5G???20M?? -- --------------------------------------------------------------------- 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 eyvaz_isaev at yahoo.com Wed Apr 4 16:27:22 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 4 Apr 2007 07:27:22 -0700 (PDT) Subject: [Pw_forum] fqha.x Message-ID: <858371.21181.qm@web60311.mail.yahoo.com> Hi Shrihram, I'm facing some trouble running this program and hope someone can help me out. I'm using the dos file generated in example 06 to calculate the free energy. But, when I give the input to fqha.x as alas.phdos, for any temperature, the free energy turns out to be zero. I'm not sure what is going wrong here. Can someone please help? The answer is quite simple: as usual \omega(1) =0, and dos(1)=0, so, you can not calculate \ln(2sinh(\hbar\omega/2kT)). The starting point has to be 2. But I think if you find QHA code sent to the forum you will gain much more. Bests, Eyvaz. Thanks, Shriram _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum ____________________________________________________________________________________ Don't pick lemons. See all the new 2007 cars at Yahoo! Autos. http://autos.yahoo.com/new_cars.html From vshastra at gmail.com Wed Apr 4 14:44:02 2007 From: vshastra at gmail.com (Vastu Shastra) Date: Wed, 4 Apr 2007 07:44:02 -0500 Subject: [Pw_forum] compilation error In-Reply-To: <524833.87449.qm@web60315.mail.yahoo.com> References: <524833.87449.qm@web60315.mail.yahoo.com> Message-ID: Thanks Davide and Eyvaz. I am using g95 compiler and Davide's advice solved my problem. Anshu On 4/4/07, Eyvaz Isaev wrote: > > Hi Anshu, > > Unfortunately you did not tell us anything about a compiler you used. > If you use g95 compiler follow David's advice. If not, add -lg2c or > -lf2c flags to linker or to LIBS in make.sys. > If you do not have f2c (g2c) library installed (check using ls > /usr/lib/*g2c*) on your computer download it from www.netlib.org and > compile it for yourself. > > Bests. > Eyvaz. > > ----- Original Message ---- > From: Vastu Shastra > To: pw_forum at pwscf.org > Sent: Wednesday, April 4, 2007 2:51:35 PM > Subject: [Pw_forum] compilation error > > Hello, > I get the following error while compiling QE 3.2 on my machine. I am > using ubuntu linux on PC. This error occurs while using 'make all' command. > Any help is appreciated. > > . > . > . > libpw.a(bp_radin.o): In function `radlg_':bp_radin.f:(.text+0x170): > undefined reference to `s_wsl e' > :bp_radin.f:(.text+0x194): undefined reference to `do_lio' > :bp_radin.f:(.text+0x199): undefined reference to `e_wsle' > :bp_radin.f:(.text+0x1a5): undefined reference to `s_wsle' > :bp_radin.f:(.text+0x1c9): undefined reference to `do_lio' > :bp_radin.f:(.text+0x1e9): undefined reference to `do_lio' > :bp_radin.f:(.text+0x1ee): undefined reference to `e_wsle' > :bp_radin.f:(.text+0x202): undefined reference to `s_stop' > libpw.a(bp_radin.o): In function `radlg1_':bp_radin.f:(.text+0x2e4): > undefined reference to `s_ws le' > . > . > . > . > . > :lapack_all.f:(.text+0x5d17): undefined reference to `do_fio' > :lapack_all.f:(.text+0x5d2f): undefined reference to `do_fio' > :lapack_all.f:(.text+0x5d34): undefined reference to `e_wsfe' > :lapack_all.f:(.text+0x5d48): undefined reference to `s_stop' > ../flib/lapack.a(dlamch.o): In function `dlamc2_':dlamch.f:(.text+0xde9): > undefined reference to `s_wsfe' > :dlamch.f:(.text+0xe05): undefined reference to `do_fio' > :dlamch.f:(.text+0xe0a): undefined reference to `e_wsfe' > make[1]: *** [pw.x] Error 1 > make[1]: Leaving directory `/home/anshu/Desktop/espresso-3.2/PW' > make: *** [pw] Error 2 > > > Thanks, > Anshu > > > ------------------------------ > Food fight?Enjoy some healthy debate > in the Yahoo! Answers Food & Drink Q&A. > -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070404/8d8f9752/attachment.htm From amit76.india at gmail.com Wed Apr 4 16:48:50 2007 From: amit76.india at gmail.com (Amit Kumar) Date: Wed, 4 Apr 2007 20:18:50 +0530 Subject: [Pw_forum] larger k mesh Message-ID: Respected Users, This suggestion is to get convreged lambda for example 7. *>>> Repeat the calculation with a larger k mesh, for example, 32x32x32.* My confusion : Where I have to increase K-points ---- in al.scf.in or al.nscf2.in or in both input files?????? For new version, in al.scf.fit.in or al.scf.in or in both input files?????? Looking forward to your kind help. With best regards, Amit -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070404/77bf4ceb/attachment.htm From giannozz at nest.sns.it Wed Apr 4 17:10:20 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 4 Apr 2007 17:10:20 +0200 Subject: [Pw_forum] Does anyone have C.pz-mt.UPF ? In-Reply-To: <4613AF4B.8010007@mit.edu> References: <529777.3957.qm@web15007.mail.cnb.yahoo.com> <4613AF4B.8010007@mit.edu> Message-ID: <73674957-DA76-4AC9-817D-8FE6F23374F5@nest.sns.it> On Apr 4, 2007, at 15:59 , Nicola Marzari wrote: > we have generated a LDA Troullier-Martins pseudopotential for C, > using the Firtz-Haber code and its standard input for C bravi, bravi, si usa il codice dello Scheffler... :-) P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Wed Apr 4 17:15:13 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 4 Apr 2007 17:15:13 +0200 Subject: [Pw_forum] Does anyone have C.pz-mt.UPF ? In-Reply-To: <73674957-DA76-4AC9-817D-8FE6F23374F5@nest.sns.it> References: <529777.3957.qm@web15007.mail.cnb.yahoo.com> <4613AF4B.8010007@mit.edu> <73674957-DA76-4AC9-817D-8FE6F23374F5@nest.sns.it> Message-ID: <49E7D608-1814-4243-89E2-AF61A31239EB@nest.sns.it> Apologies for the previous message that was not intended to be sent to the mailing list ... --- Paolo Giannozzi, Democritos and University of Udine, Italy From j.antonio.montoya at gmail.com Wed Apr 4 17:31:23 2007 From: j.antonio.montoya at gmail.com (Javier Antonio Montoya) Date: Wed, 4 Apr 2007 17:31:23 +0200 Subject: [Pw_forum] vc-relax bug? Message-ID: <993c5eba0704040831x5e66866do15616ddd1002eee1@mail.gmail.com> Hello dear all, In recent days I have been trying to get a triclinic cell to have a diagonal stress by using vc-relax. The pw.x program reads data in the correct way, calculates the right starting volume for the cell given as input and also seems to do right the first electronic minimization but, as soon as the Wentzcovitch Damped Cell-Dynamics Minimization starts, it gives cell parameters and ion positions that are not consistent with the energies and stress that it prints (i.e. if you take those cells and ions to an independent scf run the final energy and stress differ from the ones in the vc-output) the problem is that in the end, when I get a nice diagonal stress after 60 or more dynamic steps, the difference between the reported energy in the vc-output and the energy of the system that you get with the cell and ion parameters shown is of almost 1 Ry and the pressures differ by more than 10GPa. Any idea? by the way the ph.x complain and doesn't get anything from such restart files. The output files are huge for sending by e-mail because my system is 72 atoms big, but I can provide them. cheers, JAvier. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070404/cbda2666/attachment.htm From ss626 at cornell.edu Wed Apr 4 17:46:59 2007 From: ss626 at cornell.edu (Shriram Shivaraman) Date: Wed, 04 Apr 2007 11:46:59 -0400 Subject: [Pw_forum] fqha.x In-Reply-To: <858371.21181.qm@web60311.mail.yahoo.com> References: <858371.21181.qm@web60311.mail.yahoo.com> Message-ID: <4613C873.8000607@cornell.edu> Thanks a lot, Eyvaz. I found the fqha.f90 code on the forum and compiled it with mpif77 to get the executable. I am not a physics major, so I don't understand the code completely. So, I have a question - please excuse me if it is grossly incorrect. I expect the free energy F to equal E at zero temperature. But, the F I obtain from the code is positive at T=0 and becomes negative at larger T. Could you please explain this? Thanks, Shriram Eyvaz Isaev wrote: > Hi Shrihram, > > I'm facing some trouble running this program and hope someone can help > me out. I'm using the dos file generated in example 06 to calculate the > free energy. But, when I give the input to fqha.x as alas.phdos, for any > temperature, the free energy turns out to be zero. I'm not sure what is > going wrong here. Can someone please help? > > The answer is quite simple: as usual \omega(1) =0, and dos(1)=0, so, you can not > calculate \ln(2sinh(\hbar\omega/2kT)). The starting point has to be 2. > > But I think if you find QHA code sent to the forum you will gain much more. > > Bests, > Eyvaz. > > > Thanks, > Shriram > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > ____________________________________________________________________________________ > Don't pick lemons. > See all the new 2007 cars at Yahoo! Autos. > http://autos.yahoo.com/new_cars.html > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From katalin.gaal-nagy at physik.uni-regensburg.de Wed Apr 4 18:25:33 2007 From: katalin.gaal-nagy at physik.uni-regensburg.de (Katalin Gaal-Nagy) Date: Wed, 4 Apr 2007 18:25:33 +0200 (CEST) Subject: [Pw_forum] fqha.x In-Reply-To: <4613C873.8000607@cornell.edu> References: <858371.21181.qm@web60311.mail.yahoo.com> <4613C873.8000607@cornell.edu> Message-ID: Dear Shriram, the difference at T=0 between E_tot and F is due to the inclusion of the zero-point motion. I just studied bulk silicon some years ago and I found the same behavior for higher temperature (See Comp. Mat. Sci 22, 49 (2001)). So it is not completely unexpected. But maybe you get some further hints here in the forum. Best wishes, Katalin On Wed, 4 Apr 2007, Shriram Shivaraman wrote: > Thanks a lot, Eyvaz. I found the fqha.f90 code on the forum and compiled it > with mpif77 to get the executable. I am not a physics major, so I don't > understand the code completely. So, I have a question - please excuse me if > it is grossly incorrect. I expect the free energy F to equal E at zero > temperature. But, the F I obtain from the code is positive at T=0 and becomes > negative at larger T. Could you please explain this? > > Thanks, > Shriram > > Eyvaz Isaev wrote: >> Hi Shrihram, >> I'm facing some trouble running this program and hope someone can help >> me out. I'm using the dos file generated in example 06 to calculate the >> free energy. But, when I give the input to fqha.x as alas.phdos, for any >> temperature, the free energy turns out to be zero. I'm not sure what is >> going wrong here. Can someone please help? >> >> The answer is quite simple: as usual \omega(1) =0, and dos(1)=0, so, you >> can not calculate \ln(2sinh(\hbar\omega/2kT)). The starting point has to >> be 2. >> >> But I think if you find QHA code sent to the forum you will gain much more. >> >> Bests, >> Eyvaz. >> >> >> Thanks, >> Shriram >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> >> >> >> ____________________________________________________________________________________ >> Don't pick lemons. >> See all the new 2007 cars at Yahoo! Autos. >> http://autos.yahoo.com/new_cars.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 > From nguyenhalvt at yahoo.com Wed Apr 4 18:47:56 2007 From: nguyenhalvt at yahoo.com (Nguyen Ngoc Ha) Date: Wed, 4 Apr 2007 09:47:56 -0700 (PDT) Subject: [Pw_forum] Help me about NEB Message-ID: <170612.53496.qm@web31012.mail.mud.yahoo.com> Dear PWscf Users, I used NEB to find TS of C4H10 adsorption on V2O5 slab. I used 2 input file with only different setup following 1. the first input ... &IONS num_of_images = 5 , first_last_opt = .true. , opt_scheme = 'quick-min' , use_freezing = .true. , ... The result here activation energy (->) = 0.000000 eV activation energy (<-) = 0.372113 eV image energy (eV) error (eV/A) frozen 1 -12971.5366188 0.024244 T 2 -12971.6145868 0.048834 T 3 -12971.7144516 0.043135 T 4 -12971.8455553 0.045283 T 5 -12971.9087318 0.020210 T neb: convergence achieved in 39 iterations 2. The second input ... &IONS num_of_images = 7 , first_last_opt = .true. , opt_scheme = 'quick-min' , use_freezing = .true. , CI_scheme = 'highest-TS' , ... The result here activation energy (->) = 0.118180 eV activation energy (<-) = 0.490293 eV image energy (eV) error (eV/A) frozen 1 -12971.5366188 0.024244 T 2 -12971.4184384 0.045218 T 3 -12971.6123457 0.048216 T 4 -12971.6908767 0.048515 T 5 -12971.7773555 0.034338 T 6 -12971.8715469 0.042795 T 7 -12971.9087318 0.020210 T climbing image = 2 neb: convergence achieved in 65 iterations My question: which input is correct? Ea = ? Thank you for your help Best Regards! Nguyen, Ngoc Ha Hanoi University of Education Faculty of Chemistry Department of Physical Chemistry Tel: Office: 04/8330842 Home: 04/7681083 Mobile: 0912129517 ____________________________________________________________________________________ Need Mail bonding? Go to the Yahoo! Mail Q&A for great tips from Yahoo! Answers users. http://answers.yahoo.com/dir/?link=list&sid=396546091 From eyvaz_isaev at yahoo.com Wed Apr 4 19:38:59 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 4 Apr 2007 10:38:59 -0700 (PDT) Subject: [Pw_forum] fqha.x Message-ID: <540973.14100.qm@web60318.mail.yahoo.com> Hi again, I found the fqha.f90 code on the forum and compiled it with mpif77 to get the executable. It is not the same I have been talking about, but it does not matter in this particular case. Besides, there is no need to compile it using mpif77 or mpif90. Use just a serial version. I am not a physics major, so I don't understand the code completely. In this case you will meet lots of troubles. I think it is worth reading some general purpose textbook on solid state physics. As example, textbook by Kittel, or Aschroft and Mermin, or Harrison, or Martin, or Marder etc. So, I have a question - please excuse me if it is grossly incorrect. I expect the free energy F to equal E at zero temperature. But, the F I obtain from the code is positive at T=0 and becomes negative at larger T. Could you please explain this? According to quantum-mechanical principles atoms in solid states are not frozen at T=0K, but vibrate due to the uncertinity principle of the quantum mechanics, so there is so called zero vibration energy which is positive by its definition (for a stable structure all frequences are positive). At higher T the phonon contribution to the free energy becomes negative. Bests, Eyvaz. Thanks, Shriram Eyvaz Isaev wrote: > Hi Shrihram, > > I'm facing some trouble running this program and hope someone can help > me out. I'm using the dos file generated in example 06 to calculate the > free energy. But, when I give the input to fqha.x as alas.phdos, for any > temperature, the free energy turns out to be zero. I'm not sure what is > going wrong here. Can someone please help? > > The answer is quite simple: as usual \omega(1) =0, and dos(1)=0, so, you can not > calculate \ln(2sinh(\hbar\omega/2kT)). The starting point has to be 2. > > But I think if you find QHA code sent to the forum you will gain much more. > > Bests, > Eyvaz. > > > Thanks, > Shriram > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > ____________________________________________________________________________________ > Don't pick lemons. > See all the new 2007 cars at Yahoo! Autos. > http://autos.yahoo.com/new_cars.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 ____________________________________________________________________________________ The fish are biting. Get more visitors on your site using Yahoo! Search Marketing. http://searchmarketing.yahoo.com/arp/sponsoredsearch_v2.php From lanhaiping at gmail.com Wed Apr 4 20:37:59 2007 From: lanhaiping at gmail.com (lan haiping) Date: Thu, 5 Apr 2007 02:37:59 +0800 Subject: [Pw_forum] A CRASH problem for ld1.x code Message-ID: Hi. I am trying to generate pseudopotential with ld1.x code . There is a crash problem occured when i modifid a inputfile provided by QE package, The CRASH information is below: " %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% task # 136949292 from ld1_readin : error # 1 US requires at least two energies per channel %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% " I really donot understand the physics revealed by this information, would you please tell me more ? the modified input file is following: &input title='Pt', zed=78.0, rel=2, iswitch=3, rlderiv=2.50, eminld=-4.0, emaxld=4.0, deld=0.02, nld=5, config='[Xe] 4f14. 5d8 6s2 6p0', dft='LDA' / &inputp pseudotype=3, lloc=0, file_pseudopw='Ptrel.RRKJ3', nlcc=.true., rcore=1.2, rcloc=2.5 / 7 5D 3 2 4.00 0.00 2.20 2.40 1.50 5D 3 2 0.00 -0.20 2.20 2.40 1.50 5D 3 2 4.00 0.00 2.20 2.40 2.50 5D 3 2 0.00 -0.20 2.20 2.40 2.50 6P 2 1 0.00 -0.00 3.30 3.60 0.50 6P 2 1 0.00 -0.00 3.40 3.60 1.50 6S 1 0 2.00 0.00 2.80 2.80 0.50 Regards, Hai Ping -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070405/61c41c6e/attachment.htm From konstantin_kudin at yahoo.com Wed Apr 4 22:40:09 2007 From: konstantin_kudin at yahoo.com (Konstantin Kudin) Date: Wed, 4 Apr 2007 13:40:09 -0700 (PDT) Subject: [Pw_forum] vc-relax bug? In-Reply-To: <993c5eba0704040831x5e66866do15616ddd1002eee1@mail.gmail.com> Message-ID: <614002.79921.qm@web57706.mail.re3.yahoo.com> Javier, First of all, I recommend NOT suggesting that there is a "bug" in the code right away when something does not seem to be clear to you personally. And even after one's expertise is vast, a significant percentage of seemingly weird related to the code things still turn out to be not actual bugs, but *intended* (though still weird!) features. With this important philosophical issue out of the way :-) , it looks like what you get in your calculations is an *intended* feature. With vc-relax the basis set is changing during the simulation, so then if the number of plane waves is kept fixed the cutoff effectively "floats" (that is what you seem to have been doing). Then, when you restart your calculation freshly from the input you get a *different* basis set. Hence the difference in energies, forces, and stresses. The basis set issue is an important and well-known one, and there is a mechanism to compensate for it. If for fixed cell calculations you use, for example: ecutwfc =80.0 then for vc-relax you need to use more parameters such as: ecutwfc = 95.0 ecfixed = 80.0 qcutz = 80.0 q2sigma = 9.0 [to developers: these values need better defaults than those in the manual: ecfixed REAL ( default = 0.0 ) qcutz REAL ( default = 0.0 ) q2sigma REAL ( default = 0.1 ) like ecutwfc, qcutz and q2sigma could all be certain percentages of ecfixed, triggered by "vc-relax".] These parameters add a buffer zone with planewaves, such that during a vc-relax run planewaves will be smoothly added or removed as needed, although still not giving the exact match with the fixed cutoff runs. I hope this helps. Kostya --- Javier Antonio Montoya wrote: > Hello dear all, > > In recent days I have been trying to get a triclinic cell to have a > diagonal > stress by using vc-relax. The pw.x program reads data in the correct > way, > calculates the right starting volume for the cell given as input and > also > seems to do right the first electronic minimization but, as soon as > the > Wentzcovitch Damped Cell-Dynamics Minimization starts, it gives cell > parameters and ion positions that are not consistent with the > energies and > stress that it prints (i.e. if you take those cells and ions to an > independent scf run the final energy and stress differ from the ones > in the > vc-output) the problem is that in the end, when I get a nice diagonal > stress > after 60 or more dynamic steps, the difference between the reported > energy > in the vc-output and the energy of the system that you get with the > cell and > ion parameters shown is of almost 1 Ry and the pressures differ by > more than > 10GPa. Any idea? by the way the ph.x complain and doesn't get > anything from > such restart files. > The output files are huge for sending by e-mail because my system is > 72 > atoms big, but I can provide them. > > cheers, > > JAvier. > ____________________________________________________________________________________ Don't pick lemons. See all the new 2007 cars at Yahoo! Autos. http://autos.yahoo.com/new_cars.html From adeagbo at thp.Uni-Duisburg.de Wed Apr 4 22:15:37 2007 From: adeagbo at thp.Uni-Duisburg.de (Adeagbo Waheed Adeniyi) Date: Wed, 4 Apr 2007 22:15:37 +0200 (CEST) Subject: [Pw_forum] fixed spin moment calculations In-Reply-To: <20070401055538.15219.57068.Mailman@democritos.sissa.it> References: <20070401055538.15219.57068.Mailman@democritos.sissa.it> Message-ID: Dear PWSCF users, I will like to do fixed spin moment calculations, .i.e. I want the magnetization (such as 5.0 uB) to be kept fixed in my calculations. Can anyone kindly suggest me the parameter(s) or how to specify the options in the manual? Regards Waheed From eyvaz_isaev at yahoo.com Wed Apr 4 23:14:39 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 4 Apr 2007 14:14:39 -0700 (PDT) Subject: [Pw_forum] fixed spin moment calculations Message-ID: <302080.64180.qm@web60320.mail.yahoo.com> Hi Waheed, I will like to do fixed spin moment calculations, .i.e. I want the magnetization (such as 5.0 uB) to be kept fixed in my calculations. Can anyone kindly suggest me the parameter(s) or how to specify the options in the manual? From cesards at msi.umn.edu Thu Apr 5 00:38:13 2007 From: cesards at msi.umn.edu (cesards at msi.umn.edu) Date: Wed, 4 Apr 2007 17:38:13 -0500 (CDT) Subject: [Pw_forum] vc-relax bug? In-Reply-To: <20070404154801.7356.35619.Mailman@democritos.sissa.it> References: <20070404154801.7356.35619.Mailman@democritos.sissa.it> Message-ID: <62082.128.101.191.226.1175726293.squirrel@www.msi.umn.edu> Hi Javier, After your first cell optimization, try starting again from scratch using the final configuration of the previous calculation. If the problem persists, increase the cutoff energies. Good luck. Cesar R.S. da Silva > > Message: 10 > Date: Wed, 4 Apr 2007 17:31:23 +0200 > From: "Javier Antonio Montoya" > To: pw_forum at pwscf.org > Subject: [Pw_forum] vc-relax bug? > Reply-To: pw_forum at pwscf.org > > ------=_Part_26133_24163750.1175700683893 > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > Content-Transfer-Encoding: 7bit > Content-Disposition: inline > > Hello dear all, > > In recent days I have been trying to get a triclinic cell to have a > diagonal > stress by using vc-relax. The pw.x program reads data in the correct way, > calculates the right starting volume for the cell given as input and also > seems to do right the first electronic minimization but, as soon as the > Wentzcovitch Damped Cell-Dynamics Minimization starts, it gives cell > parameters and ion positions that are not consistent with the energies and > stress that it prints (i.e. if you take those cells and ions to an > independent scf run the final energy and stress differ from the ones in > the > vc-output) the problem is that in the end, when I get a nice diagonal > stress > after 60 or more dynamic steps, the difference between the reported energy > in the vc-output and the energy of the system that you get with the cell > and > ion parameters shown is of almost 1 Ry and the pressures differ by more > than > 10GPa. Any idea? by the way the ph.x complain and doesn't get anything > from > such restart files. > The output files are huge for sending by e-mail because my system is 72 > atoms big, but I can provide them. > > cheers, > > JAvier. > From ss626 at cornell.edu Thu Apr 5 03:25:56 2007 From: ss626 at cornell.edu (Shriram Shivaraman) Date: Wed, 04 Apr 2007 21:25:56 -0400 Subject: [Pw_forum] fqha.x In-Reply-To: <540973.14100.qm@web60318.mail.yahoo.com> References: <540973.14100.qm@web60318.mail.yahoo.com> Message-ID: <46145024.9060706@cornell.edu> Thanks, Katalin and Eyvaz. Your answers have been helpful. Shriram Eyvaz Isaev wrote: > Hi again, > > I found the fqha.f90 code on the forum and compiled it with mpif77 to get the executable. > > It is not the same I have been talking about, but it does not matter in this particular case. > Besides, there is no need to compile it using mpif77 or mpif90. Use just a serial version. > > I am not a physics major, so I don't understand the code completely. > > In this case you will meet lots of troubles. I think it is worth reading some general purpose textbook on solid state physics. As example, textbook by Kittel, or Aschroft and Mermin, or Harrison, or Martin, or Marder etc. > > So, I have a question - please excuse me if it is grossly incorrect. I expect the free energy F to > equal E at zero temperature. But, the F I obtain from the code is positive at T=0 and becomes negative at larger T. Could you please explain this? > > According to quantum-mechanical principles atoms in solid states are not frozen at T=0K, but vibrate due to the uncertinity principle of the quantum mechanics, so there is so called zero vibration energy which is positive by its definition (for a stable structure all frequences are positive). At higher T the phonon contribution to the free energy becomes negative. > > Bests, > Eyvaz. > > > Thanks, > Shriram > > Eyvaz Isaev wrote: > >> Hi Shrihram, >> >> I'm facing some trouble running this program and hope someone can help >> me out. I'm using the dos file generated in example 06 to calculate the >> free energy. But, when I give the input to fqha.x as alas.phdos, for any >> temperature, the free energy turns out to be zero. I'm not sure what is >> going wrong here. Can someone please help? >> >> The answer is quite simple: as usual \omega(1) =0, and dos(1)=0, so, you can not >> calculate \ln(2sinh(\hbar\omega/2kT)). The starting point has to be 2. >> >> But I think if you find QHA code sent to the forum you will gain much more. >> >> Bests, >> Eyvaz. >> >> >> Thanks, >> Shriram >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> >> >> >> >> ____________________________________________________________________________________ >> Don't pick lemons. >> See all the new 2007 cars at Yahoo! Autos. >> http://autos.yahoo.com/new_cars.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 > > > > > > > ____________________________________________________________________________________ > The fish are biting. > Get more visitors on your site using Yahoo! Search Marketing. > http://searchmarketing.yahoo.com/arp/sponsoredsearch_v2.php > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From giannozz at nest.sns.it Thu Apr 5 09:52:20 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 5 Apr 2007 09:52:20 +0200 Subject: [Pw_forum] fixed spin moment calculations In-Reply-To: <302080.64180.qm@web60320.mail.yahoo.com> References: <302080.64180.qm@web60320.mail.yahoo.com> Message-ID: <603E4596-AEBB-4862-894D-AB50674E6881@nest.sns.it> On Apr 4, 2007, at 23:14 , Eyvaz Isaev wrote: >> I will like to do fixed spin moment calculations, .i.e. I want the >> magnetization (such as 5.0 uB) to be kept fixed in my calculations. >> Can anyone kindly suggest me the parameter(s) or how to specify the >> options in the manual? >> From INPUT_PW > > fixed_magnetization(3) REAL (default 0.d0) > value of the total magnetization to be maintained > fixed when > constrained_magnetization='total' there are actually (2N+1) different ways to fix the magnetization. The one mentioned above is good for calculations with noncollinear magnetism. For LSDA calculations, the relevant variable is "total_magnetization", or alternatively "multiplicity" (also "nelup/ neldw", not recommanded). Documentation (sort of) in INPUT_PW Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Thu Apr 5 10:07:24 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 5 Apr 2007 10:07:24 +0200 Subject: [Pw_forum] A CRASH problem for ld1.x code In-Reply-To: References: Message-ID: On Apr 4, 2007, at 20:37 , lan haiping wrote: > US requires at least two energies per channel > I really do not understand the physics revealed by this information, > would you please tell me more ? the modified input file is > following: > pseudotype=3, ultrasoft > lloc=0, this is the local channel > 7 seven atomic waves are pseudized... > 5D 3 2 4.00 0.00 2.20 2.40 1.50 > 5D 3 2 0.00 -0.20 2.20 2.40 1.50 > 5D 3 2 4.00 0.00 2.20 2.40 2.50 > 5D 3 2 0.00 -0.20 2.20 2.40 2.50 4 d waves, 2 for j=l+s, 2 for j=l-s, with ultrasoft pseudization > 6P 2 1 0.00 -0.00 3.30 3.60 0.50 > 6P 2 1 0.00 -0.00 3.40 3.60 1.50 2 p waves, 1 for j=l+s, 1 for j=l-s, with ultrasoft pseudization. It is a good policy to supply at least two waves per l channel for ultrasoft pseudization, so the code complains. You can locate the check in the code and disable it if you want: the pseudization will go on. I don't see any point in using ultrasoft pseudization for p states in Pt, though. The example was using norm-conserving pseudization for the p states. > 6S 1 0 2.00 0.00 2.80 2.80 0.50 the s channel is the local channel (must be the last in the list) P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Thu Apr 5 10:14:13 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 5 Apr 2007 10:14:13 +0200 Subject: [Pw_forum] larger k mesh In-Reply-To: References: Message-ID: On Apr 4, 2007, at 16:48 , Amit Kumar wrote: > Where I have to increase K-points ---- in > > al.scf.in or al.nscf2.in or in both input files?????? typically, in the latter, since the calculation of the double-delta integral exhibits the slowest convergence, and this depends upon the k-point grid of the latter calculation > For new version, in > > al.scf.fit.in or al.scf.in or in both input files?????? in this case, in the former Please do not post html-formatted mail: they look like crap. P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Thu Apr 5 10:18:02 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 5 Apr 2007 10:18:02 +0200 Subject: [Pw_forum] EPC of Aluminium at X In-Reply-To: References: Message-ID: On Mar 30, 2007, at 21:15 , Amit Kumar wrote: > When individual matrix elements are not well converged (Please see > my last mail) then how come lambda value is so good even for small > broadening and small nq value (4 4 4)???? I don't think it is surprising that an integrated quantity converges better than the individual matrix elements P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Thu Apr 5 10:27:55 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 5 Apr 2007 10:27:55 +0200 Subject: [Pw_forum] vc-relax bug? In-Reply-To: <993c5eba0704040831x5e66866do15616ddd1002eee1@mail.gmail.com> References: <993c5eba0704040831x5e66866do15616ddd1002eee1@mail.gmail.com> Message-ID: On Apr 4, 2007, at 17:31 , Javier Antonio Montoya wrote: > by the way the ph.x complain and doesn't get anything from > such restart files. what is the complaint? P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From pnspeaks at sify.com Thu Apr 5 11:38:55 2007 From: pnspeaks at sify.com (Prithwish Nandi) Date: Thu, 05 Apr 2007 14:38:55 +0500 (IST) Subject: [Pw_forum] How to calculate the relaxation calculation by setting one of the atomic coordinate fixed??? Message-ID: <1175764135.4614bca74935e@mail.sify.com> Hi, I want to do a relaxation calculation of a 32 atom cell by keeping one atom at a fixed position and other 31 atoms to be relaxed.Is it possible to do the problem by setting X-iforce,Y-iforce and Z-iforce equals to 1 for 31 atoms and by not setting for the atom I want to be fixed ? Or, there is some other procedures for doing the job? Please help. Prithwish Nandi Research Scholar IGCAR, Kalpakkam, India -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070405/7f970484/attachment.htm From pnspeaks at sify.com Thu Apr 5 11:43:14 2007 From: pnspeaks at sify.com (pnspeaks at sify.com) Date: Thu, 05 Apr 2007 14:43:14 +0500 (IST) Subject: [Pw_forum] How to do the relaxation calculation by keeping one atom at a fixed coordinate??? Message-ID: <1175764394.4614bdaa9ef2e@mail.sify.com> Hi, I want to do a relaxation calculation of a 32 atom cell by keeping one atom at a fixed position and other 31 atoms to be relaxed.Is it possible to do the problem by setting X-iforce,Y-iforce and Z-iforce equals to 1 for 31 atoms and by not setting for the atom I want to be fixed ? Or, there is some other procedures for doing the job? Please help. Prithwish Nandi Research Scholar IGCAR, Kalpakkam, India -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070405/13cc08bb/attachment.htm From eyvaz_isaev at yahoo.com Thu Apr 5 11:34:18 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 5 Apr 2007 02:34:18 -0700 (PDT) Subject: [Pw_forum] How to calculate the relaxation calculation by setting one of the atomic coordinate fixed??? Message-ID: <450263.98359.qm@web60311.mail.yahoo.com> Hi Prithwish, You can do it easily. From INPUT_PW: X x y z {if_pos(1) if_pos(2) if_pos(3)} where : X Character: label of the atom as specified in ATOMIC_SPECIES x, y, z Real: atomic positions if_pos: Integer, optional ( default = 1 ): component i of the force for this atom is multiplied by if_pos(i), which must be either 0 or 1. Used to keep selected atoms and/or selected components fixed in meta-dynamics, neb, smd, MD dynamics or structural optimization run. I.e. like this where Al atom is supposed to be relaxed, but As atom not. Al 0.5 0.5 0.5 1 1 1 ........................... As 0 0 .5 0 0 0 Bests, Eyvaz. P.S. By the way there is no need to duplicate your message. ----- Original Message ---- From: Prithwish Nandi To: pw_forum at pwscf.org Sent: Thursday, April 5, 2007 1:38:55 PM Subject: [Pw_forum] How to calculate the relaxation calculation by setting one of the atomic coordinate fixed??? Hi, I want to do a relaxation calculation of a 32 atom cell by keeping one atom at a fixed position and other 31 atoms to be relaxed.Is it possible to do the problem by setting X-iforce,Y-iforce and Z-iforce equals to 1 for 31 atoms and by not setting for the atom I want to be fixed ? Or, there is some other procedures for doing the job? Please help. Prithwish Nandi Research Scholar IGCAR, Kalpakkam, India Now book your Railway Tickets by cash at Sify Iway. For more details contact our Customer Care Complete Coverage of the ICC World Cup '07! Log on to www.sify.com/khel for latest updates, expert columns, schedule, desktop scorecard, photo galleries and more! For the Expert view of the ICC World Cup log on to www.sify.com/khel. Read exclusive interviews with Sachin, Ganguly, Yuvraj, Sreesanth, Expert Columns by Gavaskar, Web chat with Dhoni and more! ____________________________________________________________________________________ Don't get soaked. Take a quick peek at the forecast with the Yahoo! Search weather shortcut. http://tools.search.yahoo.com/shortcuts/#loc_weather -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070405/ffdbda99/attachment.htm From lanhaiping at gmail.com Thu Apr 5 11:43:28 2007 From: lanhaiping at gmail.com (lan haiping) Date: Thu, 5 Apr 2007 17:43:28 +0800 Subject: [Pw_forum] A CRASH problem for ld1.x code In-Reply-To: References: Message-ID: Dear Paolo, Thank you ! I just modified input supplied with QE, since i am a newbie to pseudopotential construction.. Does it mean that norm-conserving pseudolization of p states causes this crash problem ? Actually , there are 2 waves per channel in the input settings, which is not consistent with crash information ... Regards, Hai-Ping On 4/5/07, Paolo Giannozzi wrote: > > > On Apr 4, 2007, at 20:37 , lan haiping wrote: > > > US requires at least two energies per channel > > > I really do not understand the physics revealed by this information, > > would you please tell me more ? the modified input file is > > following: > > > pseudotype=3, > > ultrasoft > > > lloc=0, > > this is the local channel > > > 7 > > seven atomic waves are pseudized... > > > 5D 3 2 4.00 0.00 2.20 2.40 1.50 > > 5D 3 2 0.00 -0.20 2.20 2.40 1.50 > > 5D 3 2 4.00 0.00 2.20 2.40 2.50 > > 5D 3 2 0.00 -0.20 2.20 2.40 2.50 > > 4 d waves, 2 for j=l+s, 2 for j=l-s, with ultrasoft pseudization > > > 6P 2 1 0.00 -0.00 3.30 3.60 0.50 > > 6P 2 1 0.00 -0.00 3.40 3.60 1.50 > > 2 p waves, 1 for j=l+s, 1 for j=l-s, with ultrasoft pseudization. > > It is a good policy to supply at least two waves per l channel > for ultrasoft pseudization, so the code complains. You can > locate the check in the code and disable it if you want: the > pseudization will go on. I don't see any point in using ultrasoft > pseudization for p states in Pt, though. The example was using > norm-conserving pseudization for the p states. > > > 6S 1 0 2.00 0.00 2.80 2.80 0.50 > > the s channel is the local channel (must be the last in the list) > > P. > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070405/636edd8e/attachment.htm From baroni at sissa.it Thu Apr 5 11:53:01 2007 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 5 Apr 2007 11:53:01 +0200 Subject: [Pw_forum] How to do the relaxation calculation by keeping one atom at a fixed coordinate??? In-Reply-To: <1175764394.4614bdaa9ef2e@mail.sify.com> References: <1175764394.4614bdaa9ef2e@mail.sify.com> Message-ID: You have to do nothing. Just relax the structure, and translate the relaxed structure so as the new position of the first atom coincides with the old one. Because of translational invariance, the position of any individual atom is arbitrary. SB On Apr 5, 2007, at 11:43 AM, pnspeaks at sify.com wrote: > Hi, > > I want to do a relaxation calculation of a 32 > > atom > cell by keeping one atom at a fixed position and other 31 atoms to > be > relaxed.Is it possible to do the problem by setting > X-iforce,Y-iforce and > Z-iforce equals to 1 for 31 atoms and by not > setting for the atom I want to be > fixed ? Or, there is some other > procedures for doing the job? Please help. > > > > > > Prithwish > Nandi > > Research Scholar > > IGCAR, Kalpakkam, India > > > > > > Now book your Railway Tickets by cash at Sify Iway. For more > details contact our Customer Care > > Complete Coverage of the ICC World Cup '07! Log on to www.sify.com/ > khel for latest updates, expert columns, schedule, desktop > scorecard, photo galleries and more! > > For the Expert view of the ICC World Cup log on to www.sify.com/ > khel. Read exclusive interviews with Sachin, Ganguly, Yuvraj, > Sreesanth, Expert Columns by Gavaskar, Web chat with Dhoni and more! --- 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: /pipermail/attachments/20070405/0e69e370/attachment.htm From pnspeaks at sify.com Thu Apr 5 11:25:34 2007 From: pnspeaks at sify.com (Prithwish Nandi) Date: Thu, 05 Apr 2007 15:25:34 +0600 (IST) Subject: [Pw_forum] Does any one have Gd and Ce Pseudopotentials??? Message-ID: <1175766934.4614c796e3e5b@mail.sify.com> Hi, Does anyone has the pseudo potentials for Gd and Ce ? With regards. Prithwish Nandi Research Scholar IGCAR, Kalpakkam, India -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070405/a0a4cf24/attachment.htm From giannozz at nest.sns.it Thu Apr 5 11:59:06 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 5 Apr 2007 11:59:06 +0200 Subject: [Pw_forum] A CRASH problem for ld1.x code In-Reply-To: References: Message-ID: <8BF43430-5B5E-4D13-82ED-C3635C48F205@nest.sns.it> On Apr 5, 2007, at 11:43 , lan haiping wrote: > Does it mean that norm-conserving pseudolization of p states > causes this crash problem ? no, it is the ULTRASOFT pseudization that you required. With norm-conserving pseudization it works. > Actually , there are 2 waves per channel in the input settings, > which is not consistent with crash information ... it is: the two waves are for different values of J=L+S P --- Paolo Giannozzi, Democritos and University of Udine, Italy From lanhaiping at gmail.com Thu Apr 5 14:42:19 2007 From: lanhaiping at gmail.com (lan haiping) Date: Thu, 5 Apr 2007 20:42:19 +0800 Subject: [Pw_forum] A CRASH problem for ld1.x code In-Reply-To: <8BF43430-5B5E-4D13-82ED-C3635C48F205@nest.sns.it> References: <8BF43430-5B5E-4D13-82ED-C3635C48F205@nest.sns.it> Message-ID: Thank you, Paolo.. There are some strange things for me to understand. I can generate Pt pseudopotential successfully by modifying p's states rcut to the condition below. *6P 2 1 0.00 -0.00 3.30 3.30 0.50 6P 2 1 0.00 -0.00 3.40 3.40 1.50* Why does code take such situation as 2 waves per channel ? Would you please explain more about this point ? By the way , i just found that Pt pseudotential generated by fully relativistic dirac equation at PWSCF pp's library has p state for pseudolizing, Regards Hai-Ping On 4/5/07, Paolo Giannozzi wrote: > > > On Apr 5, 2007, at 11:43 , lan haiping wrote: > > > Does it mean that norm-conserving pseudolization of p states > > causes this crash problem ? > > no, it is the ULTRASOFT pseudization that you required. > With norm-conserving pseudization it works. > > > Actually , there are 2 waves per channel in the input settings, > > which is not consistent with crash information ... > > it is: the two waves are for different values of J=L+S > > P > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070405/1970d7fb/attachment.htm From gilles.hug at onera.fr Fri Apr 6 14:45:24 2007 From: gilles.hug at onera.fr (Gilles Hug) Date: Fri, 6 Apr 2007 14:45:24 +0200 Subject: [Pw_forum] intel mac - compilation with ifort/icc and not gcc In-Reply-To: <22336FF2-ACB7-4463-8B85-04F66D2D8178@fisica.uniud.it> References: <22336FF2-ACB7-4463-8B85-04F66D2D8178@fisica.uniud.it> Message-ID: Le 4 avr. 07 ? 13:13, Paolo Giannozzi a ?crit : > Hi > >> Please try the following modified version of the "configure" script: >> http://www.fisica.uniud.it/~giannozz/public/configure >> It should recognize the intel compiler for macintosh as well. > > does it work? Yes, it works. At least his lanches the compilation. Unfortunately, I stamble now on another error. > matdyn.f90(1113) : (col. 6) remark: LOOP WAS VECTORIZED. > matdyn.f90(1143) : (col. 6) remark: LOOP WAS VECTORIZED. > ifort -o matdyn.x \ > matdyn.o rigid.o ../Modules/basic_algebra_routines.o ../ > Modules/cell_base.o ../Modules/constants.o ../Modules/ > control_flags.o ../Modules/io_global.o ../Modules/io_files.o ../ > Modules/ions_base.o ../Modules/fft_types.o ../Modules/ > fft_scalar.o ../Modules/kind.o ../Modules/mp.o ../Modules/ > mp_global.o ../Modules/parallel_include.o ../Modules/ > parameters.o ../Modules/parser.o ../Modules/path_formats.o ../ > Modules/printout_base.o ../Modules/random_numbers.o ../Modules/ > recvec.o ../Modules/splinelib.o ../Modules/timestep.o ../Modules/ > uspp.o ../PW/libpw.a ../flib/ptools.a ../flib/flib.a ../clib/ > clib.a ../iotk/src/libiotk.a -llapack -latlas > ld: warning multiple definitions of symbol _pow > /opt/intel/fc/9.1.029/lib/libimf.a(pow_stub.o) definition of _pow > in section (__TEXT,__text) > /usr/lib/gcc/i686-apple-darwin8/4.0.1/../../..//libdl.dylib > (xmm_power.o) definition of _pow > ld: Undefined symbols: > _noncollin_module_mp_noncolin_ > make[1]: *** [matdyn.x] Error 1 > make: *** [tools] Error 2 This seems to concern pwtool/matdyn but >grep noncolin pwtools/*.f90 gives no answer So I am stoped again! Thanks for your help Giles From fabris at democritos.it Fri Apr 6 16:08:27 2007 From: fabris at democritos.it (stefano fabris) Date: Fri, 6 Apr 2007 16:08:27 +0200 Subject: [Pw_forum] Does any one have Gd and Ce Pseudopotentials??? In-Reply-To: <1175766934.4614c796e3e5b@mail.sify.com> References: <1175766934.4614c796e3e5b@mail.sify.com> Message-ID: <896E301F-EDE7-4E70-B3D4-6FEB706D757B@democritos.it> Dear Prithwish, during next week, I will add to the pseudopotential library (http:// www.pwscf.org/pseudo.htm) the Ce pseudopotential we used to study cerium-oxide. As you know, in general, compounds based on rare earths may display quite intriguing properties and our pseudo has not been tested for systems other than cerias. You can refer to me if you have specific questions, I will be happy to contribute as far as I can. Yours sincerely, Stefano On 5 Apr , 2007, at 11:25 , Prithwish Nandi wrote: > > Hi, > Does anyone has the pseudo potentials for Gd and > Ce > ? > > With regards. > > Prithwish Nandi > Research Scholar > > IGCAR, Kalpakkam, India > > From akohlmey at cmm.chem.upenn.edu Fri Apr 6 15:57:07 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 6 Apr 2007 09:57:07 -0400 (EDT) Subject: [Pw_forum] intel mac - compilation with ifort/icc and not gcc In-Reply-To: Message-ID: On Fri, 6 Apr 2007, Gilles Hug wrote: GH> GH> Le 4 avr. 07 ? 13:13, Paolo Giannozzi a ?crit : GH> GH> > Hi GH> > GH> >> Please try the following modified version of the "configure" script: GH> >> http://www.fisica.uniud.it/~giannozz/public/configure GH> >> It should recognize the intel compiler for macintosh as well. GH> > GH> > does it work? GH> GH> Yes, it works. GH> At least his lanches the compilation. GH> Unfortunately, I stamble now on another error. GH> [...] GH> > clib.a ../iotk/src/libiotk.a -llapack -latlas GH> > ld: warning multiple definitions of symbol _pow GH> > /opt/intel/fc/9.1.029/lib/libimf.a(pow_stub.o) definition of _pow GH> > in section (__TEXT,__text) GH> > /usr/lib/gcc/i686-apple-darwin8/4.0.1/../../..//libdl.dylib GH> > (xmm_power.o) definition of _pow GH> > ld: Undefined symbols: GH> > _noncollin_module_mp_noncolin_ GH> > make[1]: *** [matdyn.x] Error 1 GH> > make: *** [tools] Error 2 GH> GH> This seems to concern pwtool/matdyn but GH> >grep noncolin pwtools/*.f90 GH> gives no answer GH> So I am stoped again! please try adding: -Wl,--allow-multiple-definition to LDFLAGS in make.sys and make again. salut, axel. GH> GH> Thanks for your help GH> Giles GH> GH> _______________________________________________ GH> Pw_forum mailing list GH> Pw_forum at pwscf.org GH> http://www.democritos.it/mailman/listinfo/pw_forum GH> -- ======================================================================= 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 pnspeaks at sify.com Sat Apr 7 13:20:31 2007 From: pnspeaks at sify.com (Prithwish Nandi) Date: Sat, 07 Apr 2007 17:20:31 +0600 (IST) Subject: [Pw_forum] Does any one have Gd and Ce Pseudopotentials??? Message-ID: <1175946631.461785875ae7c@mail.sify.com> Hi Stefano, Its really nice to hear from you. I am also interested in doped cerium-oxide and I want to do some spin polarized calculations.You have not specified whether your PP is LDA/GGA. Anyway, hope the PP will serve my purpose very well. Looking forward. Thanks a lot. With regards. Prithwish Nandi Research Scholar. IGCAR, Kalpakkam, India Quoting fabris at democritos.it: > Dear Prithwish, during next week, I will add to the pseudopotential library (http:// www.pwscf.org/pseudo.htm) the Ce pseudopotential we used to study cerium-oxide. As you know, in general, compounds based on rare earths may display quite intriguing properties and our pseudo has not been tested for systems other than cerias. You can refer to me if you have specific questions, I will be happy to contribute as far as I can. Yours sincerely, Stefano On 5 Apr , 2007, at 11:25 , Prithwish Nandi wrote: > > Hi, > Does anyone has the pseudo potentials for Gd and > Ce > ? > > With regards. > > Prithwish Nandi > Research Scholar > > IGCAR, Kalpakkam, India > > _______________________________________________ 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: /pipermail/attachments/20070407/e74692ac/attachment.htm From ezadshojaee at hotmail.com Sat Apr 7 15:33:10 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Sat, 07 Apr 2007 13:33:10 +0000 Subject: [Pw_forum] choose a grid for ph.x Message-ID: hi i want to know that how should i choose a grid ( for phonon calculation ) to have my system's symmetries? ( not to have problem with the q2r.x) thanx _________________________________________________________________ Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/ From nguyenhalvt at yahoo.com Sun Apr 8 03:15:59 2007 From: nguyenhalvt at yahoo.com (Nguyen Ngoc Ha) Date: Sat, 7 Apr 2007 18:15:59 -0700 (PDT) Subject: [Pw_forum] Please help me about NEB Message-ID: <322110.61418.qm@web31011.mail.mud.yahoo.com> Dear PWscf Users, I used NEB to find TS of C4H10 adsorption on V2O5 slab. I used 2 input file with only different setup following 1. the first input ... &IONS num_of_images = 5 , first_last_opt = .true. , opt_scheme = 'quick-min' , use_freezing = .true. , ... The result here activation energy (->) = 0.000000 eV activation energy (<-) = 0.372113 eV image energy (eV) error (eV/A) frozen 1 -12971.5366188 0.024244 T 2 -12971.6145868 0.048834 T 3 -12971.7144516 0.043135 T 4 -12971.8455553 0.045283 T 5 -12971.9087318 0.020210 T neb: convergence achieved in 39 iterations 2. The second input ... &IONS num_of_images = 7 , first_last_opt = .true. , opt_scheme = 'quick-min' , use_freezing = .true. , CI_scheme = 'highest-TS' , ... The result here activation energy (->) = 0.118180 eV activation energy (<-) = 0.490293 eV image energy (eV) error (eV/A) frozen 1 -12971.5366188 0.024244 T 2 -12971.4184384 0.045218 T 3 -12971.6123457 0.048216 T 4 -12971.6908767 0.048515 T 5 -12971.7773555 0.034338 T 6 -12971.8715469 0.042795 T 7 -12971.9087318 0.020210 T climbing image = 2 neb: convergence achieved in 65 iterations My question: which input is correct? Ea = ? Thank you for your help Best Regards! Nguyen, Ngoc Ha Hanoi University of Education Faculty of Chemistry Department of Physical Chemistry Tel: Office: 04/8330842 Home: 04/7681083 Mobile: 0912129517 ____________________________________________________________________________________ Now that's room service! Choose from over 150,000 hotels in 45,000 destinations on Yahoo! Travel to find your fit. http://farechase.yahoo.com/promo-generic-14795097 From w2agz at pacbell.net Mon Apr 9 08:06:24 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Sun, 8 Apr 2007 23:06:24 -0700 Subject: [Pw_forum] CPU Timing Message-ID: <000001c77a6d$34c35930$0301a8c0@w2agz> To All (especially Paolo): In February, Paolo sent me a "correction" to the standard distro of espresso-3.2 for cputime, listed below. I've now copied the new source of cputime.c to /clib in my espresso directory. I assume, of course, I have to re-"make," but do I have to re-"configure" as well? Needless to say, I'm kind of nervous about all this. /* Copyright (C) 2002-2006 Quantum-Espresso group This file is distributed under the terms of the GNU General Public License. See the file `License' in the root directory of the present distribution, or http://www.gnu.org/copyleft/gpl.txt . */ #include #include #include #include "c_defs.h" double F77_FUNC(cclock,CCLOCK)() /* Return the second elapsed since Epoch (00:00:00 UTC, January 1, 1970) */ { struct timeval tmp; double sec; gettimeofday( &tmp, (struct timezone *)0 ); sec = tmp.tv_sec + ((double)tmp.tv_usec)/1000000.0; return sec; } double F77_FUNC(scnds,SCNDS) ( ) /* Return the cpu time associated to the current process */ { static struct rusage T; getrusage(RUSAGE_SELF, &T); return ((double)T.ru_utime.tv_sec + ((double)T.ru_utime.tv_usec)/1000000.0); } 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: /pipermail/attachments/20070408/140552d6/attachment.htm From eyvaz_isaev at yahoo.com Mon Apr 9 11:44:41 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 9 Apr 2007 02:44:41 -0700 (PDT) Subject: [Pw_forum] CPU Timing Message-ID: <830901.94482.qm@web60321.mail.yahoo.com> Hi Paul, I do not think you need "re-configure", just do "re-make". Because of that you replaced cpitime.c you should recompile the code. So, relax. Bests, Eyvaz. ----- Original Message ---- From: Paul M. Grant To: pw_forum at pwscf.org Sent: Monday, April 9, 2007 10:06:24 AM Subject: [Pw_forum] CPU Timing To All (especially Paolo): In February, Paolo sent me a ?correction? to the standard distro of espresso-3.2 for cputime, listed below. I?ve now copied the new source of cputime.c to /clib in my espresso directory. I assume, of course, I have to re-?make,? but do I have to re-?configure? as well? Needless to say, I?m kind of nervous about all this. /* Copyright (C) 2002-2006 Quantum-Espresso group This file is distributed under the terms of the GNU General Public License. See the file `License' in the root directory of the present distribution, or http://www.gnu.org/copyleft/gpl.txt . */ #include #include #include #include "c_defs.h" double F77_FUNC(cclock,CCLOCK)() /* Return the second elapsed since Epoch (00:00:00 UTC, January 1, 1970) */ { struct timeval tmp; double sec; gettimeofday( &tmp, (struct timezone *)0 ); sec = tmp.tv_sec + ((double)tmp.tv_usec)/1000000.0; return sec; } double F77_FUNC(scnds,SCNDS) ( ) /* Return the cpu time associated to the current process */ { static struct rusage T; getrusage(RUSAGE_SELF, &T); return ((double)T.ru_utime.tv_sec + ((double)T.ru_utime.tv_usec)/1000000.0); } 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 ____________________________________________________________________________________ TV dinner still cooling? Check out "Tonight's Picks" on Yahoo! TV. http://tv.yahoo.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070409/e25cad83/attachment.htm From ezadshojaee at hotmail.com Mon Apr 9 13:56:33 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Mon, 09 Apr 2007 11:56:33 +0000 Subject: [Pw_forum] phonon DOS Message-ID: hi i want to know that how should i calculate phonon DOS in a special track in reciprocal lattice? (not by a grid) as the way we do for DOS when we define k_points ? thanx _________________________________________________________________ Don't just search. Find. Check out the new MSN Search! http://search.msn.click-url.com/go/onm00200636ave/direct/01/ From haskell.vivian.mcphail at gmail.com Mon Apr 9 15:49:47 2007 From: haskell.vivian.mcphail at gmail.com (Alexander McPhail) Date: Tue, 10 Apr 2007 01:49:47 +1200 Subject: [Pw_forum] Linux ia32 required additional flags and emits warnings Message-ID: <7b03b3c60704090649wc997ee7jaaa658aed3491c67@mail.gmail.com> Hi, When trying to compile the complete espresso distribution with patch level 3.2.1 I found that I had to include a couple of extra flags in the configure command IFLAGS= -I../include : so that in the compilation of iotk_error.f90 include files were found -I/usr/include/mpi : otherwise these headers were not found, even though libmpi was found and configured Also, during compilation there were many warnings like: In file cegterg.f90:349 Warning (155): Inconsistent types (COMPLEX(8)/REAL(8)) in actual argument lists at (1) and (2) Are these a problem? Thanks, Vivian -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070410/79619f95/attachment.htm From lawrence_lee_lee at yahoo.com.hk Mon Apr 9 16:12:58 2007 From: lawrence_lee_lee at yahoo.com.hk (Lawrence Lee) Date: Mon, 9 Apr 2007 22:12:58 +0800 (CST) Subject: [Pw_forum] Mg.pz-n-vbc.UPF O.pz-mt.UPF give negative pressure in MgO md calculation Message-ID: <20070409141258.70628.qmail@web54501.mail.re2.yahoo.com> Dear all, The pressure calculated by using the pseudo-potential (Mg.pz-n-vbc.UPF) and (O.pz-mt.UPF) is negative! These two pp were downloaded from www.pwscf.org. I've tried exactly the same configuration but using (Mg.pw91-np-van.UPF) and (O.pw91-van_ak.UPF) (also downloaded from www.pwscf.org), the results are fine. Here, I attach my simple input file. Please help see whether I am making mistakes or the pp have some problem. Thanks in advance for any comments. Lawrence &control calculation='md', restart_mode='from_scratch', nstep=3000, dt=20, tstress=.TRUE. tprnfor=.TRUE. pseudo_dir='~/espresso-3.1.1/pseudo', outdir='./tmp/', / &system ibrav=1, celldm(1)=7.80, nat=8 , ntyp=2, ecutwfc=30, ecutrho=240, nosym=.true. / &electrons conv_thr=1.d-8, mixing_mode='plain', mixing_beta=0.7d0, diagonalization='david', / &ions ion_dynamics='verlet', pot_extrapolation='second-order', wfc_extrapolation='second-order', ion_temperature='rescaling', tempw=300, delta_t=1, nraise=1 / ATOMIC_SPECIES Mg 24.305 Mg.pz-n-vbc.UPF O 16.000 O.pz-mt.UPF ATOMIC_POSITIONS {alat} Mg 0.00 0.00 0.00 Mg 0.50 0.50 0.00 Mg 0.50 0.00 0.50 Mg 0.00 0.50 0.50 O 0.50 0.00 0.00 O 1.00 0.50 0.00 O 1.00 0.00 0.50 O 0.50 0.50 0.50 K_POINTS {automatic} 2 2 2 0 0 0 _______________________________________ YM - ???? ??????????????????????????????????????????? http://messenger.yahoo.com.hk -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070409/af139350/attachment.htm From haskell.vivian.mcphail at gmail.com Mon Apr 9 16:27:11 2007 From: haskell.vivian.mcphail at gmail.com (Alexander McPhail) Date: Tue, 10 Apr 2007 02:27:11 +1200 Subject: [Pw_forum] Linux ia32 build error - MAIL_ not defined Message-ID: <7b03b3c60704090727r4d589700r2ae17680086f80d8@mail.gmail.com> When compliling with Debian g95, get the following error: make[1]: Entering directory `/usr/local/src/espresso-3.2/pwtools' test -n "" && ( cd .. ; make -w || exit 1) || : g95 -o dist.x \ dist.o ../Modules/basic_algebra_routines.o ../Modules/cell_base.o ../Modules/constants.o ../Modules/control_flags.o ../Modules/io_global.o ../Modules/io_files.o ../Modules/ions_base.o ../Modules/fft_types.o ../Modules/fft_scalar.o ../Modules/kind.o ../Modules/mp.o ../Modules/mp_global.o ../Modules/parallel_include.o ../Modules/parameters.o ../Modules/parser.o ../Modules/path_formats.o ../Modules/printout_base.o ../Modules/random_numbers.o ../Modules/recvec.o ../Modules/splinelib.o ../Modules/timestep.o ../Modules/uspp.o ../PW/libpw.a ../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a -llapack -lf77blas -latlas -lfftw -lmpi /usr/lib/g95/bin/../lib/gcc-lib/i686-pc-linux-gnu/4.0.3//libf95.a(main.o): In function `main': (.text+0x29): undefined reference to `MAIN_' make[1]: *** [dist.x] Error 1 make[1]: Leaving directory `/usr/local/src/espresso-3.2/pwtools' make: *** [tools] Error 2 Is this a problem with "dist.x" or with my fortran95 compiler? Thanks, Vivian -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070410/fd42aa8d/attachment.htm From marzari at MIT.EDU Mon Apr 9 16:27:26 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Mon, 09 Apr 2007 10:27:26 -0400 Subject: [Pw_forum] Mg.pz-n-vbc.UPF O.pz-mt.UPF give negative pressure in MgO md calculation In-Reply-To: <20070409141258.70628.qmail@web54501.mail.re2.yahoo.com> References: <20070409141258.70628.qmail@web54501.mail.re2.yahoo.com> Message-ID: <461A4D4E.6080302@mit.edu> Dear Lawrence, 1) please use your affiliation/address in these requests, if you'd like people like me to answer them (you do not have to, but also I do not have to answer) 2) the pressure will be zero at the equilibrium lattice parameter - so please tell us which are the equilibrium lattice parameters of your LDA and PW91 calculations, and how they differ from what is published in the literature. 3) calculate the eq. lattice parameters in these two ways: 3a) minimize the energy as a function of lattice parameter 3b) find the volume at which the trace of the stress tensor is zero 3a) and 3b) should give the same results, provided you use a large enough cutoff, but 3a will converge faster with respect to the cutoff 4) do not post until you have the results for 3a and 3b. Thanks, nicola Lawrence Lee wrote: > Dear all, > > The pressure calculated by using the pseudo-potential (Mg.pz-n-vbc.UPF) > and (O.pz-mt.UPF) is negative! These two pp were downloaded from > www.pwscf.org. --------------------------------------------------------------------- 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 lawrence_lee_lee at yahoo.com.hk Mon Apr 9 16:34:56 2007 From: lawrence_lee_lee at yahoo.com.hk (Lawrence Lee) Date: Mon, 9 Apr 2007 22:34:56 +0800 (CST) Subject: [Pw_forum] =?big5?q?=A6^=C2=D0=A1G=20Re:=20[Pw=5Fforum]=20How=20to=20calculate=20the?= =?big5?q?=20relaxation=20calculation=20by=20setting=20one=20of=20the=20atom?= =?big5?q?ic=20coordinate=20fixed=3F=3F=3F?= In-Reply-To: <450263.98359.qm@web60311.mail.yahoo.com> Message-ID: <587573.93243.qm@web54513.mail.yahoo.com> Hi Eyvaz., I've tried this method (doing relaxation while holding some atoms fixed) long ago when I started learning using PWSCF. But the problem was that the number of iterations needed in this case is much much more than if I set all atoms free!! I have this problem in my mind since then. Do any one know why? Thank you! Lawrence Eyvaz Isaev ?? Hi Prithwish, You can do it easily. From INPUT_PW: X x y z {if_pos(1) if_pos(2) if_pos(3)} where : X Character: label of the atom as specified in ATOMIC_SPECIES x, y, z Real: atomic positions if_pos: Integer, optional ( default = 1 ): component i of the force for this atom is multiplied by if_pos(i), which must be either 0 or 1. Used to keep selected atoms and/or selected components fixed in meta-dynamics, neb, smd, MD dynamics or structural optimization run. I.e. like this where Al atom is supposed to be relaxed, but As atom not. Al 0.5 0.5 0.5 1 1 1 ........................... As 0 0 .5 0 0 0 Bests, Eyvaz. P.S. By the way there is no need to duplicate your message. ----- Original Message ---- From: Prithwish Nandi To: pw_forum at pwscf.org Sent: Thursday, April 5, 2007 1:38:55 PM Subject: [Pw_forum] How to calculate the relaxation calculation by setting one of the atomic coordinate fixed??? Hi, I want to do a relaxation calculation of a 32 atom cell by keeping one atom at a fixed position and other 31 atoms to be relaxed.Is it possible to do the problem by setting X-iforce,Y-iforce and Z-iforce equals to 1 for 31 atoms and by not setting for the atom I want to be fixed ? Or, there is some other procedures for doing the job? Please help. Prithwish Nandi Research Scholar IGCAR, Kalpakkam, India Now book your Railway Tickets by cash at Sify Iway. For more details contact our Customer Care Complete Coverage of the ICC World Cup '07! Log on to www.sify.com/khel for latest updates, expert columns, schedule, desktop scorecard, photo galleries and more! For the Expert view of the ICC World Cup log on to www.sify.com/khel. Read exclusive interviews with Sachin, Ganguly, Yuvraj, Sreesanth, Expert Columns by Gavaskar, Web chat with Dhoni and more! --------------------------------- Sucker-punch spam with award-winning protection. Try the free Yahoo! Mail Beta. _______________________________________ YM - ???? ??????????????????????????????????????????? http://messenger.yahoo.com.hk -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070409/8289948a/attachment.htm From lawrence_lee_lee at yahoo.com.hk Mon Apr 9 17:36:56 2007 From: lawrence_lee_lee at yahoo.com.hk (Lawrence Lee) Date: Mon, 9 Apr 2007 23:36:56 +0800 (CST) Subject: [Pw_forum] =?big5?q?=A6^=C2=D0=A1G=20Re:=20[Pw=5Fforum]=20Mg.pz-n-vbc.UPF=20O.pz-mt.?= =?big5?q?UPF=20give=20negative=20pressure=20in=20MgO=20md=20calculation?= In-Reply-To: <461A4D4E.6080302@mit.edu> Message-ID: <20070409153656.10914.qmail@web54507.mail.re2.yahoo.com> dear Nicola, >>1) please use your affiliation/address in these requests, if you'd like people like me to answer them (you do not have to, but also I do not have to answer) Sorry! But do you mean those lines at the end of the mail, telling who I am and which university I am in? Or any thing to do with the setting of the sent mail? For the former case, I will add it from this time onwards. For the latter, I completely don't know what you refer to. Can you say clearly? I am an internet idiot! :-) >>2) the pressure will be zero at the equilibrium lattice parameter - so please tell us which are the equilibrium lattice parameters of your LDA and PW91 calculations, and how they differ from what is published in the literature. The literature value is somewhere around 7.9 (bohr radius). I cannot exactly recall it. I am doing my md run at 7.8 (very near equilibrium). I tried at least 3 other pp. One is that one I mention (the pw91 one). Another two are self-made. They give 18GPa, 12GPa, 0.7GPa, at the same 7.8 bohr radius of lattice parameter. But using (Mg.pz-n-vbc.UPF & O.pz-mt.UPF) give -160GPa !! A large negative value! >>3) Though I am still a freshman, I have been working on MgO for half a year. I have done those you have mentioned. 1)static lattice 0 K calculation to find eqm lattice parameter by minimization on both total energy and pressure. 2) testing on convergence of cutoff energy and k-points. The cutoff for Oxygen is a bit lower than acceptable, but it should be still fine for a rough trial run as that cutoff used in my other 3 calculations using different pp works well (upon fitting the results of different lattice parameter with Birch-Murnaghan 3rd order, agrees well with the literature value at least in the not too high pressure range). For safety, I am redoing the run using higher cutoff, because this pp for O is not ultrasoft, but the 3 used previouly by me are all ultrasoft. I know you may think that I am just a freshman posting questions without firstly do some detailed testing and thinking. I apologize for giving incomplete information on posting questions. Yet I have done many of those I shall check before asking here. I don't dare saying all of those, but at least those I can think of. Thank you very very much for giving me the advice :-). I will try to improve. Lawrence ---------------------------------- Lawrence Lee Shun Hang, The Chinese University of Hong Kong, Hong Kong _______________________________________ YM - ???? ??????????????????????????????????????????? http://messenger.yahoo.com.hk -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070409/09c8b6d1/attachment.htm From akohlmey at cmm.chem.upenn.edu Mon Apr 9 16:45:13 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 9 Apr 2007 10:45:13 -0400 (EDT) Subject: [Pw_forum] Linux ia32 required additional flags and emits warnings In-Reply-To: <7b03b3c60704090649wc997ee7jaaa658aed3491c67@mail.gmail.com> Message-ID: On Tue, 10 Apr 2007, Alexander McPhail wrote: dear vivian, (first off, please configure your mailer so that it shows the correct name. it is very confusing to start writing a reply to alexander and then see you're replying to vivian. thanks.). VM> Hi, VM> VM> When trying to compile the complete espresso distribution with patch level VM> 3.2.1 I found that I had to include a couple of extra flags in the configure VM> command VM> VM> IFLAGS= VM> -I../include : so that in the compilation of iotk_error.f90 include files VM> were found VM> -I/usr/include/mpi : otherwise these headers were not found, even though VM> libmpi was found and configured hmmm. that looks a lot like your MPI library is not correctly configured. the configure script has actually very little means to correctly detect a working fortran parallel programming environment (configure was mostly written to detect the (smallish) differences in unix installations). this is much more complicated on a linux machine because there are so many ways that linux machines can be set up or installed, not to mention the multitude of different fortran (and c) compilers. the problem with fortran is, that - in general - you cannot mix and match different compilers. so you have to make sure that your whole toolchain is compiled and set up consistently. configure only contains a number of 'educated guesses'. if you want to compile QE in parallel, you almost always have to adapt the resulting make.sys, in some cases even rewrite in large parts (see the files in the install directory that are templates for machines, where configure runs into big troubles or fails completely). VM> Also, during compilation there were many warnings like: VM> VM> In file cegterg.f90:349 VM> VM> Warning (155): Inconsistent types (COMPLEX(8)/REAL(8)) in actual argument VM> lists at (1) and (2) VM> VM> Are these a problem? g95 is a very picky compiler. it is quite common for fortran program to alias complex arrays to real arrays. cheers, axel. VM> VM> Thanks, VM> VM> Vivian VM> -- ======================================================================= 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 j.antonio.montoya at gmail.com Mon Apr 9 20:01:09 2007 From: j.antonio.montoya at gmail.com (Javier Antonio Montoya) Date: Mon, 9 Apr 2007 20:01:09 +0200 Subject: [Pw_forum] vc-relax bug? In-Reply-To: <62082.128.101.191.226.1175726293.squirrel@www.msi.umn.edu> References: <20070404154801.7356.35619.Mailman@democritos.sissa.it> <62082.128.101.191.226.1175726293.squirrel@www.msi.umn.edu> Message-ID: <993c5eba0704091101xf416897nda2e9fa98b5a7776@mail.gmail.com> Thanks a lot. On 4/5/07, cesards at msi.umn.edu wrote: > > Hi Javier, > > After your first cell optimization, try starting again from scratch > using the final configuration of the previous calculation. If the problem > persists, increase the cutoff energies. > > Good luck. > > Cesar R.S. da Silva > > > > > Message: 10 > > Date: Wed, 4 Apr 2007 17:31:23 +0200 > > From: "Javier Antonio Montoya" > > To: pw_forum at pwscf.org > > Subject: [Pw_forum] vc-relax bug? > > Reply-To: pw_forum at pwscf.org > > > > ------=_Part_26133_24163750.1175700683893 > > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > Content-Transfer-Encoding: 7bit > > Content-Disposition: inline > > > > Hello dear all, > > > > In recent days I have been trying to get a triclinic cell to have a > > diagonal > > stress by using vc-relax. The pw.x program reads data in the correct > way, > > calculates the right starting volume for the cell given as input and > also > > seems to do right the first electronic minimization but, as soon as the > > Wentzcovitch Damped Cell-Dynamics Minimization starts, it gives cell > > parameters and ion positions that are not consistent with the energies > and > > stress that it prints (i.e. if you take those cells and ions to an > > independent scf run the final energy and stress differ from the ones in > > the > > vc-output) the problem is that in the end, when I get a nice diagonal > > stress > > after 60 or more dynamic steps, the difference between the reported > energy > > in the vc-output and the energy of the system that you get with the cell > > and > > ion parameters shown is of almost 1 Ry and the pressures differ by more > > than > > 10GPa. Any idea? by the way the ph.x complain and doesn't get anything > > from > > such restart files. > > The output files are huge for sending by e-mail because my system is 72 > > atoms big, but I can provide them. > > > > cheers, > > > > JAvier. > > > > > _______________________________________________ > 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: /pipermail/attachments/20070409/89401c4b/attachment.htm From lawrence_lee_lee at yahoo.com.hk Tue Apr 10 08:35:04 2007 From: lawrence_lee_lee at yahoo.com.hk (Lawrence Lee) Date: Tue, 10 Apr 2007 14:35:04 +0800 (CST) Subject: [Pw_forum] Mg.pz-n-vbc.UPF O.pz-mt.UPF give negative pressure in MgO md calculation In-Reply-To: <20070409153656.10914.qmail@web54507.mail.re2.yahoo.com> Message-ID: <20070410063504.65870.qmail@web54515.mail.yahoo.com> Dear all, Concerning the matter that the two pp, Mg.pz-n-vbc.UPF and O.pz-mt.UPF give negative pressure in md calculation, but not my previous runs using other 3 pairs of pp, it was found to be because of this O.pz-mt.UPF is not an ultrasoft pp but the other pp I used previously are ulrasoft pp. When I use ecutwfc=30 and ecutrho=240 (Ryd), the ultrasoft pp gave reasonable results. But in this case the O pp is not ultrasoft, and ecutwfc=30 Ryd is not enough. Sorry for posting incorrect problem. Sorry. Lawrence ----------------------------------------------------- Lawrence Lee Shun Hang, The Chinese University of Hong Kong, Hong Kong --------------------------------- ?????????????????Yahoo! Mail ?????! -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070410/3d62a0ee/attachment.htm From baroni at sissa.it Tue Apr 10 10:05:54 2007 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 10 Apr 2007 10:05:54 +0200 Subject: [Pw_forum] Re: configure: error: linking to Fortran libraries from C fails In-Reply-To: <685054.13322.qm@web43141.mail.sp1.yahoo.com> References: <685054.13322.qm@web43141.mail.sp1.yahoo.com> Message-ID: <232DEB17-D9EF-4984-AADF-DAFFEF24E76F@sissa.it> Dear Reza: many thanks for your interest in the Quantum ESPRESSO software. Please report this kind of problems to the "pw_forum" mailing list that you can subscribe to from the pwscf.org home page. Best wishes Stefano Baroni\ On Apr 10, 2007, at 9:52 AM, jack lory wrote: > Dear Stefano Baroni > > I am reza shidpoor and nanotechnology course in tehran university. > (Iran) > > I have problem in installing espresso > > --------------------------------- > My System: > CPU intel core 2 dou > Main asus p 5 > > Linux fedora core 6 > ---------------------------------- > Error Massage > > "checking for ifort... ifort > checking whether we are using the GNU Fortran 77 compiler... no > checking whether ifort accepts -g... yes > setting F77... ifort > setting FFLAGS... -O2 -assume byterecl > setting F90FLAGS... $(FFLAGS) -nomodule > setting FFLAGS_NOOPT... -O0 -assume byterecl > setting CFLAGS... -O3 > setting CPP... cpp > setting CPPFLAGS... -P -traditional > setting LD... ifort > 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 ifort... -v > checking for Fortran libraries of ifort... -L/opt/intel/fc/9.0/lib > -L/usr/lib -lifport -lifcore -limf -lm -lirc -lirc_s > checking for dummy main to link with Fortran libraries... unknown > configure: error: linking to Fortran libraries from C fails > See `config.log' for more details." > -------------------------------------------------------- > > Regardly. > Reza.Shidpoor > > Don't get soaked. Take a quick peak at the forecast > with theYahoo! Search weather shortcut. --- 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: /pipermail/attachments/20070410/ce4c5e3b/attachment.htm From chen_shao_hua197 at yahoo.com.tw Tue Apr 10 20:18:30 2007 From: chen_shao_hua197 at yahoo.com.tw (=?big5?q?=B3=AF=20=A4=D6=B5=D8?=) Date: Wed, 11 Apr 2007 02:18:30 +0800 (CST) Subject: [Pw_forum] Does anyone know the Hubbard U parameter of InN ? Message-ID: <904014.10357.qm@web73005.mail.tp2.yahoo.com> Dear members, I am trying to calculate InN by LDA+U method(example25). Does anyone know the Hubbard U parameter of InN ? Thanks. max ?????????????????????????????????????????????????? http://tw.promo.yahoo.com/antispam/tools2.html From baroni at sissa.it Tue Apr 10 21:32:01 2007 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 10 Apr 2007 21:32:01 +0200 Subject: [Pw_forum] Does anyone know the Hubbard U parameter of InN ? In-Reply-To: <904014.10357.qm@web73005.mail.tp2.yahoo.com> References: <904014.10357.qm@web73005.mail.tp2.yahoo.com> Message-ID: For God's sake, why would you ever want to use LDA+U for InN? S. On Apr 10, 2007, at 8:18 PM, ? ?? wrote: > Dear members, > > I am trying to calculate InN by LDA+U > method(example25). > Does anyone know the Hubbard U parameter of InN ? > Thanks. > > max > > > > > ??????????????????????????????? > ??????????????????? > http://tw.promo.yahoo.com/antispam/tools2.html > _______________________________________________ > 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: /pipermail/attachments/20070410/56de7bde/attachment.htm From wierzbom at ts.infn.it Tue Apr 10 21:46:40 2007 From: wierzbom at ts.infn.it (Malgorzata Wierzbowska) Date: Tue, 10 Apr 2007 21:46:40 +0200 (CEST) Subject: [Pw_forum] Does anyone know the Hubbard U parameter of InN ? In-Reply-To: References: <904014.10357.qm@web73005.mail.tp2.yahoo.com> Message-ID: On Tue, 10 Apr 2007, Stefano Baroni wrote: > For God's sake, why would you ever want to use LDA+U for InN? > S. Maybe to move d-states of In (if PP contains them) into their experimental position below the top of valence band???? Gosia > > On Apr 10, 2007, at 8:18 PM, ?? ???? wrote: > >> Dear members, >> >> I am trying to calculate InN by LDA+U >> method(example25). >> Does anyone know the Hubbard U parameter of InN ? >> Thanks. >> >> max >> >> >> >> >> ?U???H?g?P?N???b?H???i?J?????X???e?A?N?z?w?q???U???H???H???????R???A?p???@???N?????i?O?U???H?????o?C >> http://tw.promo.yahoo.com/antispam/tools2.html >> _______________________________________________ >> 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 > > > > > -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From matteo at umn.edu Tue Apr 10 21:50:54 2007 From: matteo at umn.edu (Matteo Cococcioni) Date: Tue, 10 Apr 2007 14:50:54 -0500 Subject: [Pw_forum] Does anyone know the Hubbard U parameter of InN ? In-Reply-To: References: <904014.10357.qm@web73005.mail.tp2.yahoo.com> Message-ID: <461BEA9E.60602@umn.edu> Dear Max, yes, it's quite "unusual" to use LDA+U on atoms like In and N. Usually LDA+U is used with elements having open d or f shells. Anyway this is still possible. To do that you have to add some information in the code regarding the species you want to apply the LDA+U to. In particular you have to modify tabd.f90 and enter the number of electrons initially located on the "Hubbard" electronic states and set_hubbard_l.f90 to tell the code which l channel of the pseudopotential is going to be treated as Hubbard. For example Fe has 6 electrons on the d shell; so I will write that hubbard_l = 2 (d states) and initial occupation is 6. To compute U you can apply the procedure introduced in PRB 71 35105 (2005). Hope this helps. Matteo Stefano Baroni wrote: > For God's sake, why would you ever want to use LDA+U for InN? > S. > > On Apr 10, 2007, at 8:18 PM, ? ?? wrote: > >> Dear members, >> >> I am trying to calculate InN by LDA+U >> method(example25). >> Does anyone know the Hubbard U parameter of InN ? >> Thanks. >> >> max >> >> >> >> >> ??????????????????????????????? >> ??????????????????? >> http://tw.promo.yahoo.com/antispam/tools2.html >> _______________________________________________ >> 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 > > > From giannozz at nest.sns.it Tue Apr 10 22:05:32 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 10 Apr 2007 22:05:32 +0200 Subject: [Pw_forum] Linux ia32 required additional flags and emits warnings In-Reply-To: <7b03b3c60704090649wc997ee7jaaa658aed3491c67@mail.gmail.com> References: <7b03b3c60704090649wc997ee7jaaa658aed3491c67@mail.gmail.com> Message-ID: <6CC699C5-51D8-443E-8173-98C23B42AEEC@nest.sns.it> On Apr 9, 2007, at 15:49 , Alexander McPhail wrote: > When trying to compile the complete espresso distribution with > patch level 3.2.1 > I found that I had to include a couple of extra flags in the > configure command > > IFLAGS= > -I../include : so that in the compilation of iotk_error.f90 include > files were found It is there by default, unless something different for IFLAGS was specified to "configure" (which shouldn't be needed) Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Tue Apr 10 22:11:17 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 10 Apr 2007 22:11:17 +0200 Subject: [Pw_forum] Linux ia32 build error - MAIL_ not defined In-Reply-To: <7b03b3c60704090727r4d589700r2ae17680086f80d8@mail.gmail.com> References: <7b03b3c60704090727r4d589700r2ae17680086f80d8@mail.gmail.com> Message-ID: On Apr 9, 2007, at 16:27 , Alexander McPhail wrote: > (.text+0x29): undefined reference to `MAIN_' > [...] > Is this a problem with "dist.x" or with my fortran95 compiler? the latter. Try to change the program name from "dist" to something else (but it shouldn't be a reserved word) or to add "end program dist" at the end. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From chen_shao_hua197 at yahoo.com.tw Wed Apr 11 07:16:48 2007 From: chen_shao_hua197 at yahoo.com.tw (=?big5?q?=B3=AF=20=A4=D6=B5=D8?=) Date: Wed, 11 Apr 2007 13:16:48 +0800 (CST) Subject: [Pw_forum] Does anyone know the Hubbard U parameter of InN ? In-Reply-To: <461BEA9E.60602@umn.edu> Message-ID: <74264.83803.qm@web73008.mail.tp2.yahoo.com> Dear Matteo and ohters, Thanks for your suggestion. I had calculated the bandstructure of InN by LDA .But there is no band gap show up.Someone told me the LDA+U method could open the gap.That is why I ask this question. Best Regards max --- Matteo Cococcioni ?? > > Dear Max, > > yes, it's quite "unusual" to use LDA+U on atoms like > In and N. Usually > LDA+U is used with elements having open d or f > shells. > > Anyway this is still possible. To do that you have > to add some > information in the code regarding the species you > want to apply the > LDA+U to. > In particular you have to modify tabd.f90 and enter > the number of > electrons initially located on the "Hubbard" > electronic states and > set_hubbard_l.f90 to tell the code which l channel > of the > pseudopotential is going to be treated as Hubbard. > For example Fe has 6 > electrons on the d shell; so I will write that > hubbard_l = 2 (d states) > and initial occupation is 6. > > To compute U you can apply the procedure introduced > in PRB 71 35105 (2005). > > Hope this helps. > > Matteo > > > Stefano Baroni wrote: > > For God's sake, why would you ever want to use > LDA+U for InN? > > S. > > > > On Apr 10, 2007, at 8:18 PM, ? ?? wrote: > > > >> Dear members, > >> > >> I am trying to calculate InN by LDA+U > >> method(example25). > >> Does anyone know the Hubbard U parameter of InN ? > >> Thanks. > >> > >> max > >> > >> > >> > >> > >> > ??????????????????????????????? > > >> ??????????????????? > >> http://tw.promo.yahoo.com/antispam/tools2.html > >> _______________________________________________ > >> 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 > ?????????????????????????????Email???? ??????????????????????? http://tw.promo.yahoo.com/antispam/index.html From w2agz at pacbell.net Wed Apr 11 08:01:40 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Tue, 10 Apr 2007 23:01:40 -0700 Subject: [Pw_forum] CPU Timing Message-ID: <002c01c77bfe$e086aaa0$0301a8c0@w2agz> Good news. Paolo's fix to cptimer.c works. Grazi, Professore! 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: /pipermail/attachments/20070410/348d238c/attachment.htm From wierzbom at ts.infn.it Wed Apr 11 10:24:25 2007 From: wierzbom at ts.infn.it (Malgorzata Wierzbowska) Date: Wed, 11 Apr 2007 10:24:25 +0200 (CEST) Subject: "**Suspected SPAM**" Re: [Pw_forum] Does anyone know the Hubbard U parameter of InN ? In-Reply-To: <74264.83803.qm@web73008.mail.tp2.yahoo.com> References: <74264.83803.qm@web73008.mail.tp2.yahoo.com> Message-ID: On Wed, 11 Apr 2007, ?? ???? wrote: > Dear Matteo and ohters, > > Thanks for your suggestion. > > I had calculated the bandstructure of InN by LDA .But > there is no band gap show up.Someone told me the LDA+U > method could open the gap.That is why I ask this > question. > > Best Regards > max Yes someone (see the abstract below) said this. But I tried for ZnX (X=S,Se,Te) series - moving d-states of Zn into their experimental position (U was chosen by "numerical experiment") -- and I did not find much effect on the gap...... Maybe it depends also on implementation and the projectors for LDA+U. To open a gap you may try to correct pseudopotentials using "scissors" operator (see the second paper and citations of that paper). Gosia ---------------------------------------------------------- Title: Effects of cation d states on the structural and electronic properties of III-nitride and II-oxide wide-band-gap semiconductors Author(s): Janotti A, Segev D, Van de Walle CG Source: PHYSICAL REVIEW B 74 (4): Art. No. 045202 JUL 2006 Document Type: Article Language: English Cited References: 54 Times Cited: 1 Find Related Records Information Abstract: Using first-principles methods based on density functional theory within the local density approximation (LDA) we calculate the structural and electronic properties of wurtzite MgO, ZnO, and CdO, and discuss their similarities and dissimilarities with the corresponding Group-III nitrides AlN, GaN, and InN. We treat the semicore d states of Zn, Cd, Ga, and In explicitly as valence states in a pseudopotential approach, investigate the effects of including on-site Coulomb interaction for Zn, Cd, Ga, and In semicore d states within the LDA+U method, and propose a novel approach to calculate the parameter U. Our results show that the LDA+U approach systematically improves the LDA band gap by indirectly acting on both the valence-band maximum and conduction-band minimum. We also discuss the effects of the on-site Coulomb interaction on structural parameters and absolute deformation potentials of ZnO, CdO, GaN, and InN. ----------------------------------------------------------------------- Electronic structure of GaAs under strain PRB vol 30 p.5753 (1984) Results of self-consistent relativistic band calculations for GaAs under hydrostatic as well as uniaxial strain are presented. Deformation potentials related to the splitting of the valence-band edge (??15v) are calculated with and without inclusion of spin-orbit coupling. The trigonal-shear deformation potentials that agree with experiments correspond to an internal-strain parameter ??=0.6??0.1. The calculated values, 16-19 eV, of the optical deformation potential d0 are substantially smaller than the published experimental results (???41 eV). The E0 gap obtained in the local-density approximation is 0.25 eV. A method of correcting for this error and for calculating, self-consistently, the lowest s-like conduction band is described, and used to derive pressure dependences of the gaps and conduction-band masses. The parameters for this adjustment of the conduction band are determined for zero pressure, and can be kept pressure independent. We find (1/mc*)dmc*/dP=0.68??10-2 kbar-1. The pressure at which conduction-band inversion occurs is 30.5 kbar. The value calculated for shear deformation potential E2L is 19 eV for ??=0.6. The spin-orbit-induced splitting of the lowest conduction band for k??????[110] and the additional strain-induced splitting are calculated and related to experimental results for spin relaxation of photoexcited electrons. ------------------------------------------------------------------------- > --- Matteo Cococcioni ???G > >> >> Dear Max, >> >> yes, it's quite "unusual" to use LDA+U on atoms like >> In and N. Usually >> LDA+U is used with elements having open d or f >> shells. >> >> Anyway this is still possible. To do that you have >> to add some >> information in the code regarding the species you >> want to apply the >> LDA+U to. >> In particular you have to modify tabd.f90 and enter >> the number of >> electrons initially located on the "Hubbard" >> electronic states and >> set_hubbard_l.f90 to tell the code which l channel >> of the >> pseudopotential is going to be treated as Hubbard. >> For example Fe has 6 >> electrons on the d shell; so I will write that >> hubbard_l = 2 (d states) >> and initial occupation is 6. >> >> To compute U you can apply the procedure introduced >> in PRB 71 35105 (2005). >> >> Hope this helps. >> >> Matteo >> >> >> Stefano Baroni wrote: >>> For God's sake, why would you ever want to use >> LDA+U for InN? >>> S. >>> >>> On Apr 10, 2007, at 8:18 PM, ?? ???? wrote: >>> >>>> Dear members, >>>> >>>> I am trying to calculate InN by LDA+U >>>> method(example25). >>>> Does anyone know the Hubbard U parameter of InN ? >>>> Thanks. >>>> >>>> max >>>> >>>> >>>> >>>> >>>> >> > ?U???H?g?P?N???b?H???i?J?????X???e?A?N?z?w?q???U???H???H?????? >> >>>> ?R???A?p???@???N?????i?O?U???H?????o?C >>>> http://tw.promo.yahoo.com/antispam/tools2.html >>>> _______________________________________________ >>>> 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 >> > > > > ?O???J?u?????????v?????????L???|???_?i?H???????q?o?e???y??Email???F?z?C ?????p?????K?N?z???????b???M?K?X???S???o???H?C > http://tw.promo.yahoo.com/antispam/index.html > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From giannozz at nest.sns.it Wed Apr 11 10:37:26 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 11 Apr 2007 10:37:26 +0200 Subject: [Pw_forum] Does anyone know the Hubbard U parameter of InN ? In-Reply-To: References: <74264.83803.qm@web73008.mail.tp2.yahoo.com> Message-ID: On Apr 11, 2007, at 10:24 , Malgorzata Wierzbowska wrote: > On Wed, 11 Apr 2007, ?? ???? wrote: > >> I had calculated the bandstructure of InN by LDA .But >> there is no band gap show up.Someone told me the LDA+U >> method could open the gap.That is why I ask this >> question. > > Yes someone (see the abstract below) said this. also somebody else (not for InN, though): see http://www.scielo.br/pdf/bjp/v36n2a/a14v362a.pdf, but they use negative values of U for p states! Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From proffess at yandex.ru Wed Apr 11 18:01:23 2007 From: proffess at yandex.ru (Sergey Lisenkov) Date: Wed, 11 Apr 2007 20:01:23 +0400 (MSD) Subject: [Pw_forum] compiling error Message-ID: <461D0653.000007.31901@webmail9.yandex.ru> Dear developers, Pathscale compiler (3.0) version has some troubles to compile pw (cvs version). The message is following: ftn -march=opteron -O3 -ffast-math -funroll-loops -mcmodel=medium -ftpp --target=catamount -D__EKO -D__FFTW -D__FFTW_WITH_SIZE -D__MPI -D__PARA -D__QK_USER__ -I../include -I./ -I../Modules -I../iotk/src -I../PW -I../PH -I../CPV -c bfgs_module.f90 MODULE bfgs_module ^ pathf95-855 pathf90-3.0: ERROR BFGS_MODULE, File = bfgs_module.f90, Line = 9, Column = 8 The compiler has detected errors in module "BFGS_MODULE". No module information file will be created for this module. IF ( ( grad(:) .dot. step(:) ) > 0.D0 ) THEN ^ pathf95-318 pathf90-3.0: ERROR GDIIS_STEP, File = bfgs_module.f90, Line = 469, Column = 26 The operator "dot" has not been declared as a user defined operator. step(:) = - ( inv_hess(:,:) .times. grad(:) ) ^ pathf95-318 pathf90-3.0: ERROR GDIIS_STEP, File = bfgs_module.f90, Line = 474, Column = 42 The operator "times" has not been declared as a user defined operator. pathf95: PathScale(TM) Fortran Version 3.0 (f14) Wed Apr 11, 2007 10:59:48 pathf95: 1015 source lines pathf95: 3 Error(s), 0 Warning(s), 0 Other message(s), 0 ANSI(s) pathf95: "explain pathf95-message number" gives more information about each message make: *** [bfgs_module.o] Error 2 Is it a compiler bug? PGI (7.0) can compile this module. Thanks, Sergey From emenendez at macul.ciencias.uchile.cl Wed Apr 11 23:32:29 2007 From: emenendez at macul.ciencias.uchile.cl (Eduardo Ariel Menendez P) Date: Wed, 11 Apr 2007 17:32:29 -0400 (CLT) Subject: [Pw_forum] Xcrysden error Message-ID: Hi, I have the following error with xcrysden when trying to open any the pwscf output files, including the examples that come with xcryden. I found this error reported in he Forum by Diego Moreno some time ago, but I do not see the solution. The error hapens with verions 4.1 and 4.1.1. I am at ICTP/Trieste, where there are many users of the program, so please, help me. ERROR: Error reading PRIMVEC section ERROR: Error reading PRIMVEC section while executing "exec $system(BINDIR)/xsf2xsf $filedir $filedir.raw 3" ("eval" body line 1) invoked from within "eval {exec $system(BINDIR)/xsf2xsf $filedir $filedir.raw} $xcMisc(reduce_to)" (procedure "xsfOpen" line 26) invoked from within "xsfOpen $program_output .mesa" (procedure "openExtStruct" line 183) invoked from within "openExtStruct 3 crystal external [list sh $system(TOPDIR)/scripts/dummy.sh] pwo2xsf.xsf {PWSCF Output File} ANGS -preset pwOutputPreset -file $fil..." invoked from within ".menu.vmfile.menu.examples.pwscf invoke active" ("uplevel" body line 1) invoked from within "uplevel #0 [list $w invoke active]" (procedure "tk::MenuInvoke" line 47) invoked from within "tk::MenuInvoke .menu.vmfile.menu.examples.pwscf 1" (command bound to event) Thanks, Eduardo 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 http://fisica.ciencias.uchile.cl/~emenendez/ From w2agz at pacbell.net Thu Apr 12 07:06:57 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Wed, 11 Apr 2007 22:06:57 -0700 Subject: [Pw_forum] Lost My Subscription Message-ID: <002901c77cc0$658aaaa0$0301a8c0@w2agz> My subscription to pw_forum seems to have disappeared about five days ago. Help. 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: /pipermail/attachments/20070411/fe77f585/attachment.htm From paulatto at sissa.it Thu Apr 12 09:17:40 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Thu, 12 Apr 2007 09:17:40 +0200 (CEST) Subject: [Pw_forum] Xcrysden error In-Reply-To: References: Message-ID: <19890.82.54.232.225.1176362260.squirrel@webmail.sissa.it> On Wed, April 11, 2007 23:32, Eduardo Ariel Menendez P wrote: > "exec $system(BINDIR)/xsf2xsf $filedir $filedir.raw 3" I found the program xsf2xsf included in xcrysden to be the source of many problems: as far a s I remember it uses a very small buffer to read the input one line at a time; if one line is too long than it usually causes some nasty errors. You can try to run it by hand and see what happens, you will find it in the bin/ or script/ directory in XCD instalaltion. -- Lorenzo Paulatto +39 040 3787 312 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From tone.kokalj at ijs.si Thu Apr 12 11:08:06 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Thu, 12 Apr 2007 11:08:06 +0200 Subject: [Pw_forum] Xcrysden error In-Reply-To: <19890.82.54.232.225.1176362260.squirrel@webmail.sissa.it> References: <19890.82.54.232.225.1176362260.squirrel@webmail.sissa.it> Message-ID: <1176368886.2473.3.camel@localhost.localdomain> On Thu, 2007-04-12 at 09:17 +0200, Lorenzo Paulatto wrote: > I found the program xsf2xsf included in xcrysden to be the source of many > problems: as far a s I remember it uses a very small buffer to read the > input one line at a time; if one line is too long than it usually causes > some nasty errors. Good suggestion, thanks. This should be changed/improved. Regards, Tone PS: the xsf2xsf can be run manually as: $XCRYSDEN_TOPDIR/bin/xsf2xsf input.xsf output.xsf reduced_dim where reduced_dim = 3, 2, 1 or 0 (=crystal, slab, polymer, and molecule, respectively) From giannozz at nest.sns.it Thu Apr 12 10:48:43 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 12 Apr 2007 10:48:43 +0200 Subject: [Pw_forum] Lost My Subscription In-Reply-To: <002901c77cc0$658aaaa0$0301a8c0@w2agz> References: <002901c77cc0$658aaaa0$0301a8c0@w2agz> Message-ID: <4C77889E-99A3-47D9-ADCF-F45DCE0CBAF2@nest.sns.it> On Apr 12, 2007, at 7:06 , Paul M. Grant wrote: > My subscription to pw_forum seems to have disappeared about five > days ago. Help. if you can post, your subscription has not disappeared Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Thu Apr 12 11:30:38 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 12 Apr 2007 11:30:38 +0200 Subject: [Pw_forum] compiling error In-Reply-To: <461D0653.000007.31901@webmail9.yandex.ru> References: <461D0653.000007.31901@webmail9.yandex.ru> Message-ID: <6B1E4F99-B73B-47DD-AA88-C7A0BC9BE54D@nest.sns.it> On Apr 11, 2007, at 18:01 , Sergey Lisenkov wrote: > Pathscale compiler (3.0) version has some troubles to compile pw > (cvs version). where do you find all kind of strange machines and compilers??? > The operator "dot" has not been declared as a user defined operator. > [...] > Is it a compiler bug? very likely, since ".dot." is a user defined operator, defined in module "basic_algebra_routines" Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From lanhaiping at gmail.com Thu Apr 12 15:06:52 2007 From: lanhaiping at gmail.com (lan haiping) Date: Thu, 12 Apr 2007 21:06:52 +0800 Subject: [Pw_forum] About valence bands maximum Message-ID: Dear All, I want to determin the transition enery of a defect level. It is important to align the valence bands maximum of different systems under investigation. Many works has claimed valence bands maximum alignments were performed but no technique details are given. For my understanding, I thougt the valence bands maxium of charge defect systems could be determined by aligning its average potential to pure perfect system's potential, e.g \epsilon_{VBM}.(Defects,q) = \epsilon_{VBM}(0) + [ -]. But some person told me such expression has omitted something. So it really confused me much. As far as i know, the potential of system could be only determined to some constant . i thought the alignment of average potentials of two systems could take into account this arbitrariness.. I am not very certain about this VBM alignment right now Would you give me some comments and clarify my understanding ? I appreciate any suggestions and comments . Cheers , Hai-Ping -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070412/8258e440/attachment.htm From marzari at MIT.EDU Thu Apr 12 15:17:00 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 12 Apr 2007 09:17:00 -0400 Subject: [Pw_forum] Lost My Subscription In-Reply-To: <4C77889E-99A3-47D9-ADCF-F45DCE0CBAF2@nest.sns.it> References: <002901c77cc0$658aaaa0$0301a8c0@w2agz> <4C77889E-99A3-47D9-ADCF-F45DCE0CBAF2@nest.sns.it> Message-ID: <461E314C.2030005@mit.edu> Post, ergo sum ! nicola Paolo Giannozzi wrote: > > On Apr 12, 2007, at 7:06 , Paul M. Grant wrote: > >> My subscription to pw_forum seems to have disappeared about five days >> ago. Help. > > if you can post, your subscription has not disappeared > > Paolo > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- --------------------------------------------------------------------- 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 akohlmey at cmm.chem.upenn.edu Thu Apr 12 15:39:12 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 12 Apr 2007 09:39:12 -0400 (EDT) Subject: [Pw_forum] Lost My Subscription In-Reply-To: <002901c77cc0$658aaaa0$0301a8c0@w2agz> Message-ID: On Wed, 11 Apr 2007, Paul M. Grant wrote: PG> My subscription to pw_forum seems to have disappeared about five days ago. PG> Help. paul, you may want to check at the mailing list management webpage, whether mail delivery has been suspended. if your email address fails/bounces for a number of times, e-mail delivery usually gets disabled (you are still subsribed). i've had that happen to me a couple of times when we had repeated downtimes of our mailserver for more than 24 hours. cheers, axel. PG> PG> PG> PG> Paul M. Grant, PhD PG> PG> Principal, W2AGZ Technologies PG> PG> Visiting Scholar, Applied Physics, Stanford University PG> PG> EPRI Science Fellow (Retired) PG> PG> IBM Research Staff Member Emeritus PG> PG> w2agz at pacbell.net PG> PG> http://www.w2agz.com PG> PG> PG> PG> PG> PG> PG> PG> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From w2agz at pacbell.net Thu Apr 12 19:28:11 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Thu, 12 Apr 2007 10:28:11 -0700 Subject: [Pw_forum] Lost My Subscription Message-ID: <008401c77d27$f2b03820$0301a8c0@w2agz> Here's what happened. I have two levels of spam filters set, one on my e-mail server and the other on my e-mail client. About a week ago, the server "decided" to block everything from pw_forum. I suspect there's a "frequency of attack" algorithm they implement. I've now added a separate filter on the server to "clear" anything from pw_forum. We'll see. Thanks to all who responded.only Axel "got through" before I found this problem. -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 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070412/0638f1d5/attachment.htm From haskell.vivian.mcphail at gmail.com Fri Apr 13 06:53:53 2007 From: haskell.vivian.mcphail at gmail.com (Vivian McPhail) Date: Fri, 13 Apr 2007 16:53:53 +1200 Subject: [Pw_forum] Re: Linux ia32 build error - MAIL_ not defined (Paolo Giannozzi) Message-ID: <7b03b3c60704122153h166bd239mfe455478aefaf624@mail.gmail.com> > > From: Paolo Giannozzi > Subject: Re: [Pw_forum] Linux ia32 build error - MAIL_ not defined > Date: Tue, 10 Apr 2007 22:11:17 +0200 > To: pw_forum at pwscf.org > Reply-To: pw_forum at pwscf.org > > On Apr 9, 2007, at 16:27 , Alexander McPhail wrote: > > > (.text+0x29): undefined reference to `MAIN_' > > [...] > > Is this a problem with "dist.x" or with my fortran95 compiler? > > the latter. Try to change the program name from "dist" to > something else (but it shouldn't be a reserved word) or to > add "end program dist" at the end. > > Paolo > --- > Paolo Giannozzi, Democritos and University of Udine, Italy I tried (1) changing the program name (to foobar), (2) adding "program dist" after the last "end", and (3) adding "end program dist" after the last end. (1) and (2) result in the same error message as above. (3) produces: dist.f: In program `MAIN__': dist.f:336: end program dist 1 2 Program name at (2) superfluous, no PROGRAM statement specified at (1) make[1]: *** [dist.o] Error 1 make[1]: Leaving directory `/usr/local/src/espresso-3.2/pwtools' make: *** [tools] Error 2 Note that there are two underscores in MAIN__ of the error message. whereas in the other error message the linker was looking for a definition of MAIN_ with one underscore. Is there a compiler flag which governs the number of underscores? Thanks, Vivian -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070413/ba75e771/attachment.htm From ceresoli at sissa.it Fri Apr 13 09:41:19 2007 From: ceresoli at sissa.it (Davide Ceresoli) Date: Fri, 13 Apr 2007 09:41:19 +0200 Subject: [Pw_forum] Re: Linux ia32 build error - MAIL_ not defined (Paolo Giannozzi) In-Reply-To: <7b03b3c60704122153h166bd239mfe455478aefaf624@mail.gmail.com> References: <7b03b3c60704122153h166bd239mfe455478aefaf624@mail.gmail.com> Message-ID: <461F341F.7060603@sissa.it> Vivian McPhail wrote: > I tried (1) changing the program name (to foobar), (2) adding "program > dist" after the last "end", and (3) adding "end program dist" after the > last end. > > (1) and (2) result in the same error message as above. > (3) produces: > > dist.f: In program `MAIN__': > dist.f:336: > end program dist > 1 2 > Program name at (2) superfluous, no PROGRAM statement specified at (1) > make[1]: *** [dist.o] Error 1 > make[1]: Leaving directory `/usr/local/src/espresso-3.2/pwtools' > make: *** [tools] Error 2 > > > Note that there are two underscores in MAIN__ of the error message. > whereas in the other error message the linker was looking for a > definition of MAIN_ with one underscore. Is there a compiler flag which > governs the number of underscores? MAIN_ is the entry-point of the g95 runtime library, whereas MAIN__ is the entry-point of the g77 r.t. library, which are vastly incompatible. Try this: ./configure F77=g95 F90=g95 etc... see also: http://www.democritos.it/pipermail/pw_forum/2007-February/005859.html http://www.democritos.it/pipermail/pw_forum/2007-February/005863.html (why you didn't get also the same problems mentioned above?) From emenendez at macul.ciencias.uchile.cl Fri Apr 13 10:03:51 2007 From: emenendez at macul.ciencias.uchile.cl (Eduardo Ariel Menendez P) Date: Fri, 13 Apr 2007 04:03:51 -0400 (CLT) Subject: [Pw_forum] Re: Xcrysden error In-Reply-To: <20070413062615.19603.63458.Mailman@democritos.sissa.it> References: <20070413062615.19603.63458.Mailman@democritos.sissa.it> Message-ID: Hi, thanks for the replies. However, I want to visualize a .out file, not a .xsf. Please, note that the .inp files of PWSCF are visualized fine, the problem is with the .out files . May this be related to a particular configuration of PC-s at the ICTP? As far as I remember, I did not have this problem at home. Regards Eduardo > > On Wed, April 11, 2007 23:32, Eduardo Ariel Menendez P wrote: > > "exec $system(BINDIR)/xsf2xsf $filedir $filedir.raw 3" > > I found the program xsf2xsf included in xcrysden to be the source of many > problems: as far a s I remember it uses a very small buffer to read the > input one line at a time; if one line is too long than it usually causes > some nasty errors. > > You can try to run it by hand and see what happens, you will find it in > the bin/ or script/ directory in XCD instalaltion. > > -- > Lorenzo Paulatto > +39 040 3787 312 > http://people.sissa.it/~paulatto/ > > > On Thu, 2007-04-12 at 09:17 +0200, Lorenzo Paulatto wrote: > > I found the program xsf2xsf included in xcrysden to be the source of many > > problems: as far a s I remember it uses a very small buffer to read the > > input one line at a time; if one line is too long than it usually causes > > some nasty errors. > > Good suggestion, thanks. This should be changed/improved. > > Regards, Tone > > PS: the xsf2xsf can be run manually as: > > $XCRYSDEN_TOPDIR/bin/xsf2xsf input.xsf output.xsf reduced_dim > > where reduced_dim = 3, 2, 1 or 0 (=crystal, slab, polymer, and molecule, > respectively) > > > --__--__-- > From tone.kokalj at ijs.si Fri Apr 13 11:14:57 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Fri, 13 Apr 2007 11:14:57 +0200 Subject: [Pw_forum] Re: Xcrysden error In-Reply-To: References: <20070413062615.19603.63458.Mailman@democritos.sissa.it> Message-ID: <1176455697.2600.4.camel@localhost.localdomain> On Fri, 2007-04-13 at 04:03 -0400, Eduardo Ariel Menendez P wrote: > Hi, > thanks for the replies. However, I want to visualize a .out file, not a > .xsf. Any kind of file-format is converted to XSF file. The xsf2xsf is some internal xcrysden's filter. Because the any-to-xsf conversion is always there, the xsf2xsf filter is always used. You have the problem with xsf2xsf (don't know why), so please try to see manually what is going on, i.e.: pwo2xsf.sh -lc my_pwscf_output_file.out > file.xsf $XCRYSDEN_TOPDIR/bin/xsf2xsf file.xsf output.xsf 3 The first line converts pwscf output file to xsf file, and then you need to check what is going on with the xsf2xsf (2nd line). Regards, Tone From emenendez at macul.ciencias.uchile.cl Sat Apr 14 11:34:04 2007 From: emenendez at macul.ciencias.uchile.cl (Eduardo Ariel Menendez P) Date: Sat, 14 Apr 2007 05:34:04 -0400 (CLT) Subject: [Pw_forum] Re: Xcrysden error In-Reply-To: <20070414063652.31410.41668.Mailman@democritos.sissa.it> References: <20070414063652.31410.41668.Mailman@democritos.sissa.it> Message-ID: Hi Tone, I have just found that the computers at ICTP have defined the decimal point as ',' instead of '.'. This may affect numerical operations. In the case of the example CH3Rh111, I did pwo2xsf.sh -ic CH3Rh111.out > CH3Rh111.xsf Examining de file CH3Rh111.xsf I see this PRIMVEC 5,2917700000 0,0000000000 0,0000000000 0,0000000000 0,0000000000 0,0000000000 0,0000000000 0,0000000000 10,5835400000 They are numbers with decimal point ',' and the second lattice vector that was (-.5000, .8660,.0000) in the .out file, is converted into (0,0,0) I am afraid that I do not have control of the decimal point without root priviledges (correct me if I can change it), and that I would have to appeal to the UN Secratary to make the ICTP system manager to change this. Is there any solution at the level of my local copy of xcrysden? Thanks, Eduardo > > Any kind of file-format is converted to XSF file. The xsf2xsf is some > internal xcrysden's filter. Because the any-to-xsf conversion is always > there, the xsf2xsf filter is always used. You have the problem with > xsf2xsf (don't know why), so please try to see manually what is going > on, i.e.: > > pwo2xsf.sh -lc my_pwscf_output_file.out > file.xsf > $XCRYSDEN_TOPDIR/bin/xsf2xsf file.xsf output.xsf 3 > > The first line converts pwscf output file to xsf file, and then you need > to check what is going on with the xsf2xsf (2nd line). > > Regards, Tone > > > From lanhaiping at gmail.com Sat Apr 14 12:54:29 2007 From: lanhaiping at gmail.com (lan haiping) Date: Sat, 14 Apr 2007 18:54:29 +0800 Subject: [Pw_forum] About PH warning ! Message-ID: Dear All, I came to warnings below when performing IFC calculation with ph.x code : " Possibly too few bands at point 1 0.00000 0.00000 0.00000 Possibly too few bands at point 2 0.00000 0.00000 0.38151 Possibly too few bands at point 3 0.00000 0.00000 0.19076 Possibly too few bands at point 4 0.00000 0.00000 0.57227 Possibly too few bands at point 5 0.00000 0.00000 0.38151 Possibly too few bands at point 6 0.00000 0.00000 0.76302 Possibly too few bands at point 7 0.00000 0.00000 0.57227 Possibly too few bands at point 8 0.00000 0.00000 0.95378 Possibly too few bands at point 9 0.00000 0.00000 -0.76302 Possibly too few bands at point 10 0.00000 0.00000 -0.38151: " Would you please tell me how to suspress such warnings ? Regards, H.P -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070414/d016aef9/attachment.htm From eyvaz_isaev at yahoo.com Sat Apr 14 13:15:57 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sat, 14 Apr 2007 04:15:57 -0700 (PDT) Subject: [Pw_forum] About PH warning ! Message-ID: <509135.63872.qm@web60316.mail.yahoo.com> Hi , Just set up manually bands number "nbnd" in your scf-input file. The message means that nbnd used by default is not sufficient. If you use an old version set up nbnd in nscf-file. Bests, Eyvaz. ---------------------------------------------- Condensed Matter Theory Group, Uppsala University, Sweden Theoretical Physics Department, Moscow State Institute of Steel and Alloys, Russia ----- Original Message ---- From: lan haiping To: pw_forum Sent: Saturday, April 14, 2007 2:54:29 PM Subject: [Pw_forum] About PH warning ! Dear All, I came to warnings below when performing IFC calculation with ph.x code : " Possibly too few bands at point 1 0.00000 0.00000 0.00000 Possibly too few bands at point 2 0.00000 0.00000 0.38151 Possibly too few bands at point 3 0.00000 0.00000 0.19076 Possibly too few bands at point 4 0.00000 0.00000 0.57227 Possibly too few bands at point 5 0.00000 0.00000 0.38151 Possibly too few bands at point 6 0.00000 0.00000 0.76302 Possibly too few bands at point 7 0.00000 0.00000 0.57227 Possibly too few bands at point 8 0.00000 0.00000 0.95378 Possibly too few bands at point 9 0.00000 0.00000 -0.76302 Possibly too few bands at point 10 0.00000 0.00000 -0.38151: " Would you please tell me how to suspress such warnings ? Regards, H.P -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 __________________________________________________ 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: /pipermail/attachments/20070414/aaf1861b/attachment.htm From eyvaz_isaev at yahoo.com Sat Apr 14 13:27:32 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sat, 14 Apr 2007 04:27:32 -0700 (PDT) Subject: [Pw_forum] About PH warning ! Message-ID: <58959.66318.qm@web60319.mail.yahoo.com> Hi again, I mean increase the "nbnd" parameter. Have a look at scf.out file and find "nbnd=", and then increase it. Bests, Eyvaz. ----- Original Message ---- From: lan haiping To: pw_forum Sent: Saturday, April 14, 2007 2:54:29 PM Subject: [Pw_forum] About PH warning ! Dear All, I came to warnings below when performing IFC calculation with ph.x code : " Possibly too few bands at point 1 0.00000 0.00000 0.00000 Possibly too few bands at point 2 0.00000 0.00000 0.38151 Possibly too few bands at point 3 0.00000 0.00000 0.19076 Possibly too few bands at point 4 0.00000 0.00000 0.57227 Possibly too few bands at point 5 0.00000 0.00000 0.38151 Possibly too few bands at point 6 0.00000 0.00000 0.76302 Possibly too few bands at point 7 0.00000 0.00000 0.57227 Possibly too few bands at point 8 0.00000 0.00000 0.95378 Possibly too few bands at point 9 0.00000 0.00000 -0.76302 Possibly too few bands at point 10 0.00000 0.00000 -0.38151: " Would you please tell me how to suspress such warnings ? Regards, H.P -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 __________________________________________________ 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: /pipermail/attachments/20070414/d42926dd/attachment.htm From lanhaiping at gmail.com Sat Apr 14 13:51:17 2007 From: lanhaiping at gmail.com (lan haiping) Date: Sat, 14 Apr 2007 19:51:17 +0800 Subject: [Pw_forum] About PH warning ! In-Reply-To: <58959.66318.qm@web60319.mail.yahoo.com> References: <58959.66318.qm@web60319.mail.yahoo.com> Message-ID: Dear Eyvaz, Thank you ! My system for investigating is of 36 electrons , and my setting for nbnd is 24.. I thought this setting was enough for a metallic system .. Regards, H.P On 4/14/07, Eyvaz Isaev wrote: > > Hi again, > > I mean increase the "nbnd" parameter. Have a look at scf.out file and > find "nbnd=", and then increase it. > > Bests, > Eyvaz. > > ----- Original Message ---- > From: lan haiping > To: pw_forum > Sent: Saturday, April 14, 2007 2:54:29 PM > Subject: [Pw_forum] About PH warning ! > > Dear All, > > I came to warnings below when performing IFC calculation with ph.x code : > > " > > Possibly too few bands at point 1 0.00000 0.00000 0.00000 > > Possibly too few bands at point 2 0.00000 0.00000 0.38151 > > Possibly too few bands at point 3 0.00000 0.00000 0.19076 > > Possibly too few bands at point 4 0.00000 0.00000 0.57227 > > Possibly too few bands at point 5 0.00000 0.00000 0.38151 > > Possibly too few bands at point 6 0.00000 0.00000 0.76302 > > Possibly too few bands at point 7 0.00000 0.00000 0.57227 > > Possibly too few bands at point 8 0.00000 0.00000 0.95378 > > Possibly too few bands at point 9 0.00000 0.00000 -0.76302 > > Possibly too few bands at point 10 0.00000 0.00000 -0.38151: > " > Would you please tell me how to suspress such warnings ? > Regards, > > H.P > -- > Hai-Ping Lan > Department of Electronics , > Peking University , Bejing, 100871 > > > ------------------------------ > Ahhh...imagining that irresistible "new car" smell? > Check out new cars at Yahoo! Autos. > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070414/07f2ebc1/attachment.htm From akohlmey at cmm.chem.upenn.edu Sat Apr 14 16:08:31 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 14 Apr 2007 10:08:31 -0400 (EDT) Subject: [Pw_forum] Re: Xcrysden error In-Reply-To: Message-ID: On Sat, 14 Apr 2007, Eduardo Ariel Menendez P wrote: EAM> EAM> Hi Tone, EAM> EAM> I have just found that the computers at ICTP have defined the decimal EAM> point as ',' instead of '.'. This may affect numerical operations. aiiiih! another victim of NLS! why-oh-why do people insist on this nonsense?? anyway, enough ranting. i can only _strongly_ recommend to use either (depending on your shell) export LC_ALL=C or setenv LC_ALL C whenever you do scientific work. this is only a rather harmless example. not so long ago i came across an example, where a simulation program would give a completely wrong output depending on whether one would do the calculation in the native or the generic 'C' locale. tone, it might be a good idea to force LC_ALL=C inside of xcrysden before executing any ancillary scripts or executables (and the same would be true for PWGui). EAM> EAM> In the case of the example CH3Rh111, I did EAM> pwo2xsf.sh -ic CH3Rh111.out > CH3Rh111.xsf EAM> EAM> Examining de file CH3Rh111.xsf I see this EAM> PRIMVEC EAM> 5,2917700000 0,0000000000 0,0000000000 EAM> 0,0000000000 0,0000000000 0,0000000000 EAM> 0,0000000000 0,0000000000 10,5835400000 EAM> EAM> They are numbers with decimal point ',' and the second lattice vector that EAM> was (-.5000, .8660,.0000) in the .out file, is converted into (0,0,0) EAM> EAM> I am afraid that I do not have control of the decimal point without root EAM> priviledges (correct me if I can change it), and that I would have to you don't need root priviledges to change this. just set LC_ALL to the proper value (i.e. C in this case). it might be a good idea for the scientific (sic!) computer section of ICTP (cc'd) to think about this in a more general way. EAM> appeal to the UN Secratary to make the ICTP system manager to change this. EAM> Is there any solution at the level of my local copy of xcrysden? cheers, axel. EAM> EAM> Thanks, EAM> Eduardo EAM> EAM> > EAM> > Any kind of file-format is converted to XSF file. The xsf2xsf is some EAM> > internal xcrysden's filter. Because the any-to-xsf conversion is always EAM> > there, the xsf2xsf filter is always used. You have the problem with EAM> > xsf2xsf (don't know why), so please try to see manually what is going EAM> > on, i.e.: EAM> > EAM> > pwo2xsf.sh -lc my_pwscf_output_file.out > file.xsf EAM> > $XCRYSDEN_TOPDIR/bin/xsf2xsf file.xsf output.xsf 3 EAM> > EAM> > The first line converts pwscf output file to xsf file, and then you need EAM> > to check what is going on with the xsf2xsf (2nd line). EAM> > EAM> > Regards, Tone EAM> > EAM> > EAM> > EAM> _______________________________________________ EAM> Pw_forum mailing list EAM> Pw_forum at pwscf.org EAM> http://www.democritos.it/mailman/listinfo/pw_forum EAM> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From akohlmey at cmm.chem.upenn.edu Sat Apr 14 18:28:10 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 14 Apr 2007 12:28:10 -0400 Subject: [Pw_forum] Re: Xcrysden error In-Reply-To: <4620F99E.6090301@ictp.it> References: <4620F99E.6090301@ictp.it> Message-ID: <7b6913e90704140928v7257f443hf42f8fbf2060f3c@mail.gmail.com> FYI, LC_ALL sets all of them (i'm lazy). [akohlmey at vitriol]$ ssh ssh.ictp.it akohlmey at ssh.ictp.it's password: ssh-2:1> locale LANG=en_US LC_CTYPE=en_US LC_NUMERIC=it_IT LC_TIME=en_US LC_COLLATE=en_US LC_MONETARY=it_IT LC_MESSAGES=en_US LC_PAPER=it_IT LC_NAME=it_IT LC_ADDRESS=it_IT LC_TELEPHONE=it_IT LC_MEASUREMENT=it_IT LC_IDENTIFICATION=it_IT LC_ALL= ssh-2:2> env LC_ALL=C locale LANG=en_US LC_CTYPE="C" LC_NUMERIC="C" LC_TIME="C" LC_COLLATE="C" LC_MONETARY="C" LC_MESSAGES="C" LC_PAPER="C" LC_NAME="C" LC_ADDRESS="C" LC_TELEPHONE="C" LC_MEASUREMENT="C" LC_IDENTIFICATION="C" LC_ALL=C ssh-2:3> uname -a Linux ssh-2 2.4.22-41mdk #1 Wed Jan 19 23:11:32 MST 2005 i686 unknown unknown GNU/Linux cheers, axel. On 4/14/07, Onime Clement wrote: > > Sorry, > the right variable(s) to set instead of LC_ALL should be > LC_NUMERIC > LC_MEASUREMENT > LC_MONETARY > > Thanks > Clement > > Axel Kohlmeyer wrote: > > On Sat, 14 Apr 2007, Eduardo Ariel Menendez P wrote: > > > > EAM> > > EAM> Hi Tone, > > EAM> > > EAM> I have just found that the computers at ICTP have defined the decimal > > EAM> point as ',' instead of '.'. This may affect numerical operations. > > > > aiiiih! > > > > another victim of NLS! > > why-oh-why do people insist on this nonsense?? > > > > anyway, enough ranting. i can only _strongly_ > > recommend to use either (depending on your shell) > > export LC_ALL=C > > or > > setenv LC_ALL C > > whenever you do scientific work. this is only > > a rather harmless example. not so long ago i > > came across an example, where a simulation program > > would give a completely wrong output depending on > > whether one would do the calculation in the native > > or the generic 'C' locale. > > > > tone, > > it might be a good idea to force LC_ALL=C inside of > > xcrysden before executing any ancillary scripts or > > executables (and the same would be true for PWGui). > > > > EAM> > > EAM> In the case of the example CH3Rh111, I did > > EAM> pwo2xsf.sh -ic CH3Rh111.out > CH3Rh111.xsf > > EAM> > > EAM> Examining de file CH3Rh111.xsf I see this > > EAM> PRIMVEC > > EAM> 5,2917700000 0,0000000000 0,0000000000 > > EAM> 0,0000000000 0,0000000000 0,0000000000 > > EAM> 0,0000000000 0,0000000000 10,5835400000 > > EAM> > > EAM> They are numbers with decimal point ',' and the second lattice vector that > > EAM> was (-.5000, .8660,.0000) in the .out file, is converted into (0,0,0) > > EAM> > > EAM> I am afraid that I do not have control of the decimal point without root > > EAM> priviledges (correct me if I can change it), and that I would have to > > > > you don't need root priviledges to change this. just set LC_ALL to the > > proper value (i.e. C in this case). it might be a good idea for the > > scientific (sic!) computer section of ICTP (cc'd) to think about this > > in a more general way. > > > > EAM> appeal to the UN Secratary to make the ICTP system manager to change this. > > EAM> Is there any solution at the level of my local copy of xcrysden? > > > > cheers, > > axel. > > > > > > EAM> > > EAM> Thanks, > > EAM> Eduardo > > EAM> > > EAM> > > > EAM> > Any kind of file-format is converted to XSF file. The xsf2xsf is some > > EAM> > internal xcrysden's filter. Because the any-to-xsf conversion is always > > EAM> > there, the xsf2xsf filter is always used. You have the problem with > > EAM> > xsf2xsf (don't know why), so please try to see manually what is going > > EAM> > on, i.e.: > > EAM> > > > EAM> > pwo2xsf.sh -lc my_pwscf_output_file.out > file.xsf > > EAM> > $XCRYSDEN_TOPDIR/bin/xsf2xsf file.xsf output.xsf 3 > > EAM> > > > EAM> > The first line converts pwscf output file to xsf file, and then you need > > EAM> > to check what is going on with the xsf2xsf (2nd line). > > EAM> > > > EAM> > Regards, Tone > > EAM> > > > EAM> > > > EAM> > > > EAM> _______________________________________________ > > EAM> Pw_forum mailing list > > EAM> Pw_forum at pwscf.org > > EAM> http://www.democritos.it/mailman/listinfo/pw_forum > > EAM> > > > > -- ======================================================================= 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 Sat Apr 14 18:02:15 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 14 Apr 2007 12:02:15 -0400 (EDT) Subject: [Pw_forum] Re: Xcrysden error In-Reply-To: <4620F725.9090505@ictp.it> Message-ID: On Sat, 14 Apr 2007, Onime Clement wrote: OC> Hmm, OC> OC> OC> > aiiiih! OC> > OC> > another victim of NLS! OC> > why-oh-why do people insist on this nonsense?? OC> > OC> The fundamental question is NLS or Unicode an issue for fortran or C OC> code's or some internal representation of numbers, I don't think so as OC> it is purely a visual problem? it _is_ an issue for scientific applications, since they write out and parse some (temporary) data in text format, usually to improve portability or to allow users to examine the results without the need for a specific viewer program. OC> Anyhow a bug report to the site of xcrysden is in order, I think the OC> developers could be encouraged to make a query to find out the right NLS OC> decimal separator or use NLS calls for handling numbers. this is plain undoable. most progam maintainers put in an insane effort to keep their packages runnable and up-to-date with the scientific problems. burdening them with making everything NLS clean on top of that would be insane. we're not talking about people that actually get paid to do this kind of job and even in that case, i'd rather pay for people to implement features that are useful instead of just following misunderstood userfriendlyness. i agree that for the 'normal user' having programs in their native language is helpful, but as soon as you start working with science and numbers, native language and locale support is more trouble for users than helping them and can even be a recipe for desaster (what if some mission critical software was developed and tested under one locale and is then installed and run on a machine with a different locale? there have been precedents already!). of course this is a generic, world wide problem and i can only urge everybody who is doing scientific work on a machine with a different than the C locale to always keep this problem in mind as a possible source of errors. cheers, axel. -- ======================================================================= 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 wlyim at puccini.che.pitt.edu Sat Apr 14 20:48:11 2007 From: wlyim at puccini.che.pitt.edu (wlyim at puccini.che.pitt.edu) Date: Sat, 14 Apr 2007 14:48:11 -0400 (EDT) Subject: [Pw_forum] ph.x v3.2 on NEC SX-8 In-Reply-To: <325F5EE3-9948-4475-AC50-6A4A48EF378B@nest.sns.it> Message-ID: Dear Paolo and the QE community, I downloaded and compiled the latest espresso-3.2.1. I used the Make.sxpara file, as provided in espresso-3.1. In NEC SX-8, the serial version works for your mentioned examples. I run example06 using parallel verion, but something strange happened. "mpirun -np 1 ./pw.x -in alas.scf.in" works. However, when using 2 cpus, the program crashed, which seemed to be related to the allocation problem in "atomic_rho.f90". ================================================================ . . nbndx = 16 nbnd = 4 natomwfc = 13 npwx = 157 nelec = 8.00 nkb = 8 ngl = 61 Initial potential from superposition of free atoms **** 90 Fatal exception PROG=atomic_rho ELN=116(40038f7f4) SIGSEGV: Segmentation violation Called from potinit ELN=202(4002b8b3c)f7f4) Called from init_h ELN=66(400254b90) SIGSEGV: Segmentation violation Called from init_run ELN=49(4002545f8) Called from potinit ELN=202(4002b8b3c) Called from pwscf ELN=94(400000f38) Called from init_h ELN=66(400254b90) Called from init_run ELN=49(4002545f8) Called from pwscf ELN=94(400000f38) ./pw.x(lang:f90): signal trap(SIGTERM: Software termination) v00: mpid: MPI process (universe = 0, rank = 1) terminated by exit(1) ========================================================================== I am not sure if this is a compiler problem, or compiler option problem, or something else. Any idea will be welcome. Thanks!! On Tue, 16 Jan 2007, Paolo Giannozzi wrote: > On Jan 4, 2007, at 14:00 , wlyim at puccini.che.pitt.edu wrote: > > > Unfortunately, ph.x still gave wrong results [...] Is there any > > benchmark test on NEC machine for the latest version of espresso? > > there are examples (examples 2, 5, 6 notably), that have been > reported to work with Nec sx-8. Do they work for you? > > P. > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Dr. Wai-Leung Yim Institut fuer Reine und Angewandte Chemie, Theoretische Chemie, Carl von Ossiezky Universtaet Oldenburg, 26129 Oldenburg, Germany Email: wlyim at puccini.che.pitt.edu Phone: +49-441-798-3950 (office) Fax: +49-441-798-3964 From emenendez at macul.ciencias.uchile.cl Sat Apr 14 21:01:14 2007 From: emenendez at macul.ciencias.uchile.cl (Eduardo Ariel Menendez P) Date: Sat, 14 Apr 2007 15:01:14 -0400 (CLT) Subject: [Pw_forum] Re: Xcrysden error In-Reply-To: References: Message-ID: Alex, The problem was solved with setenv LC_ALL C Thanks a lot Eduardo > On Sat, 14 Apr 2007, Eduardo Ariel Menendez P wrote: > > EAM> > EAM> Hi Tone, > EAM> > EAM> I have just found that the computers at ICTP have defined the decimal > EAM> point as ',' instead of '.'. This may affect numerical operations. > > aiiiih! > > another victim of NLS! > why-oh-why do people insist on this nonsense?? > > anyway, enough ranting. i can only _strongly_ > recommend to use either (depending on your shell) > export LC_ALL=C > or > setenv LC_ALL C > whenever you do scientific work. this is only > a rather harmless example. not so long ago i > came across an example, where a simulation program > would give a completely wrong output depending on > whether one would do the calculation in the native > or the generic 'C' locale. > > tone, > it might be a good idea to force LC_ALL=C inside of > xcrysden before executing any ancillary scripts or > executables (and the same would be true for PWGui). > > EAM> > EAM> In the case of the example CH3Rh111, I did > EAM> pwo2xsf.sh -ic CH3Rh111.out > CH3Rh111.xsf > EAM> > EAM> Examining de file CH3Rh111.xsf I see this > EAM> PRIMVEC > EAM> 5,2917700000 0,0000000000 0,0000000000 > EAM> 0,0000000000 0,0000000000 0,0000000000 > EAM> 0,0000000000 0,0000000000 10,5835400000 > EAM> > EAM> They are numbers with decimal point ',' and the second lattice vector that > EAM> was (-.5000, .8660,.0000) in the .out file, is converted into (0,0,0) > EAM> > EAM> I am afraid that I do not have control of the decimal point without root > EAM> priviledges (correct me if I can change it), and that I would have to > > you don't need root priviledges to change this. just set LC_ALL to the > proper value (i.e. C in this case). it might be a good idea for the > scientific (sic!) computer section of ICTP (cc'd) to think about this > in a more general way. > > EAM> appeal to the UN Secratary to make the ICTP system manager to change this. > EAM> Is there any solution at the level of my local copy of xcrysden? > > cheers, > axel. From babazade at gmail.com Sun Apr 15 12:24:22 2007 From: babazade at gmail.com (Amin Babazadeh) Date: Sun, 15 Apr 2007 13:54:22 +0330 Subject: [Pw_forum] about choosing density for solving scf Message-ID: Dear users For solving the self-consistent equation in Kohn-Sham method we need to select an ordinary (random) density to start the solving equation. i want to know in espresso how this density choose? sincerely yours -- AMIN -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070415/3d6d0fd1/attachment.htm From brad.malone at gmail.com Sun Apr 15 17:38:47 2007 From: brad.malone at gmail.com (Brad Malone) Date: Sun, 15 Apr 2007 08:38:47 -0700 Subject: [Pw_forum] Why does pw.x think that I have 76 k-points? Message-ID: Hi, Even though my input file has only 64 k-pts, something about the way I'm specifying the other information makes pw.x believe that there are 76 k-points. The only thing that I'm doing that's new compared to the last time I used pwSCF correctly is that I specified my lattice vectors by hand. Is part of my syntax regarding this incorrect? My input file (r8.scf.in) is below. Thanks in advance for any help you can give me. Best, Brad -------------------------------------------------------------------------------------------------------------- &control calculation='scf' title='r8' restart_mode='from_scratch', prefix='r8' pseudo_dir = '/usr/common/homes/b/bmalone/pwSCF/pseudo/', outdir='./' / &system ibrav= 0, celldm(1)=0, nat=8, ntyp= 1, nbnd=48, ecutwfc =40.0 / &electrons electron_maxstep = 100 conv_thr = 1.0d-8 mixing_beta = 0.7 / CELL_PARAMETERS 10.251078661674 -2.125522840388 -2.125522843283 -2.125522841113 10.251078660657 -2.125522845751 -2.125522845406 -2.125522847149 10.251078658233 ATOMIC_SPECIES Si 28.086 Si.vbc.UPF ATOMIC_POSITIONS crystal Si 0.464071693714 -0.032602013466 0.272678366246 Si -0.464071693714 0.032602013466 -0.272678366246 Si -0.032602013009 0.272678369157 0.464071693903 Si 0.032602013009 -0.272678369157 -0.464071693903 Si 0.272678366983 0.464071693535 -0.032602014399 Si -0.272678366983 -0.464071693535 0.032602014399 Si 0.282114632854 0.282114633625 0.282114633354 Si -0.282114632854 -0.282114633625 -0.282114633354 K_POINTS crystal 64 0.00000000 0.00000000 0.00000000 1.562500e-02 0.00000000 0.00000000 0.25000000 1.562500e-02 0.00000000 0.00000000 0.50000000 1.562500e-02 0.00000000 0.00000000 0.75000000 1.562500e-02 0.00000000 0.25000000 0.00000000 1.562500e-02 0.00000000 0.25000000 0.25000000 1.562500e-02 0.00000000 0.25000000 0.50000000 1.562500e-02 0.00000000 0.25000000 0.75000000 1.562500e-02 0.00000000 0.50000000 0.00000000 1.562500e-02 0.00000000 0.50000000 0.25000000 1.562500e-02 0.00000000 0.50000000 0.50000000 1.562500e-02 0.00000000 0.50000000 0.75000000 1.562500e-02 0.00000000 0.75000000 0.00000000 1.562500e-02 0.00000000 0.75000000 0.25000000 1.562500e-02 0.00000000 0.75000000 0.50000000 1.562500e-02 0.00000000 0.75000000 0.75000000 1.562500e-02 0.25000000 0.00000000 0.00000000 1.562500e-02 0.25000000 0.00000000 0.25000000 1.562500e-02 0.25000000 0.00000000 0.50000000 1.562500e-02 0.25000000 0.00000000 0.75000000 1.562500e-02 0.25000000 0.25000000 0.00000000 1.562500e-02 0.25000000 0.25000000 0.25000000 1.562500e-02 0.25000000 0.25000000 0.50000000 1.562500e-02 0.25000000 0.25000000 0.75000000 1.562500e-02 0.25000000 0.50000000 0.00000000 1.562500e-02 0.25000000 0.50000000 0.25000000 1.562500e-02 0.25000000 0.50000000 0.50000000 1.562500e-02 0.25000000 0.50000000 0.75000000 1.562500e-02 0.25000000 0.75000000 0.00000000 1.562500e-02 0.25000000 0.75000000 0.25000000 1.562500e-02 0.25000000 0.75000000 0.50000000 1.562500e-02 0.25000000 0.75000000 0.75000000 1.562500e-02 0.50000000 0.00000000 0.00000000 1.562500e-02 0.50000000 0.00000000 0.25000000 1.562500e-02 0.50000000 0.00000000 0.50000000 1.562500e-02 0.50000000 0.00000000 0.75000000 1.562500e-02 0.50000000 0.25000000 0.00000000 1.562500e-02 0.50000000 0.25000000 0.25000000 1.562500e-02 0.50000000 0.25000000 0.50000000 1.562500e-02 0.50000000 0.25000000 0.75000000 1.562500e-02 0.50000000 0.50000000 0.00000000 1.562500e-02 0.50000000 0.50000000 0.25000000 1.562500e-02 0.50000000 0.50000000 0.50000000 1.562500e-02 0.50000000 0.50000000 0.75000000 1.562500e-02 0.50000000 0.75000000 0.00000000 1.562500e-02 0.50000000 0.75000000 0.25000000 1.562500e-02 0.50000000 0.75000000 0.50000000 1.562500e-02 0.50000000 0.75000000 0.75000000 1.562500e-02 0.75000000 0.00000000 0.00000000 1.562500e-02 0.75000000 0.00000000 0.25000000 1.562500e-02 0.75000000 0.00000000 0.50000000 1.562500e-02 0.75000000 0.00000000 0.75000000 1.562500e-02 0.75000000 0.25000000 0.00000000 1.562500e-02 0.75000000 0.25000000 0.25000000 1.562500e-02 0.75000000 0.25000000 0.50000000 1.562500e-02 0.75000000 0.25000000 0.75000000 1.562500e-02 0.75000000 0.50000000 0.00000000 1.562500e-02 0.75000000 0.50000000 0.25000000 1.562500e-02 0.75000000 0.50000000 0.50000000 1.562500e-02 0.75000000 0.50000000 0.75000000 1.562500e-02 0.75000000 0.75000000 0.00000000 1.562500e-02 0.75000000 0.75000000 0.25000000 1.562500e-02 0.75000000 0.75000000 0.50000000 1.562500e-02 0.75000000 0.75000000 0.75000000 1.562500e-02 / -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070415/68e0e768/attachment.htm From wierzbom at ts.infn.it Sun Apr 15 17:49:08 2007 From: wierzbom at ts.infn.it (Malgorzata Wierzbowska) Date: Sun, 15 Apr 2007 17:49:08 +0200 (CEST) Subject: [Pw_forum] Why does pw.x think that I have 76 k-points? In-Reply-To: References: Message-ID: On Sun, 15 Apr 2007, Brad Malone wrote: Hi, look to example24: there is ibrav=0, celldm(1)=1., You have celldm(1)=0. Maybe this is the reason.... Gosia > Hi, > > Even though my input file has only 64 k-pts, something about the way I'm > specifying the other information makes pw.x believe that there are 76 > k-points. The only thing that I'm doing that's new compared to the last > time I used pwSCF correctly is that I specified my lattice vectors by hand. > Is part of my syntax regarding this incorrect? My input file (r8.scf.in) is > below. Thanks in advance for any help you can give me. > > Best, > Brad > > -------------------------------------------------------------------------------------------------------------- > &control > calculation='scf' > title='r8' > restart_mode='from_scratch', > prefix='r8' > pseudo_dir = '/usr/common/homes/b/bmalone/pwSCF/pseudo/', > outdir='./' > / > &system > ibrav= 0, celldm(1)=0, nat=8, ntyp= 1, nbnd=48, > ecutwfc =40.0 > > > / > &electrons > electron_maxstep = 100 > conv_thr = 1.0d-8 > mixing_beta = 0.7 > / > CELL_PARAMETERS > 10.251078661674 -2.125522840388 -2.125522843283 > -2.125522841113 10.251078660657 -2.125522845751 > -2.125522845406 -2.125522847149 10.251078658233 > ATOMIC_SPECIES > Si 28.086 Si.vbc.UPF > ATOMIC_POSITIONS crystal > Si 0.464071693714 -0.032602013466 0.272678366246 > Si -0.464071693714 0.032602013466 -0.272678366246 > Si -0.032602013009 0.272678369157 0.464071693903 > Si 0.032602013009 -0.272678369157 -0.464071693903 > Si 0.272678366983 0.464071693535 -0.032602014399 > Si -0.272678366983 -0.464071693535 0.032602014399 > Si 0.282114632854 0.282114633625 0.282114633354 > Si -0.282114632854 -0.282114633625 -0.282114633354 > > K_POINTS crystal > 64 > 0.00000000 0.00000000 0.00000000 1.562500e-02 > 0.00000000 0.00000000 0.25000000 1.562500e-02 > 0.00000000 0.00000000 0.50000000 1.562500e-02 > 0.00000000 0.00000000 0.75000000 1.562500e-02 > 0.00000000 0.25000000 0.00000000 1.562500e-02 > 0.00000000 0.25000000 0.25000000 1.562500e-02 > 0.00000000 0.25000000 0.50000000 1.562500e-02 > 0.00000000 0.25000000 0.75000000 1.562500e-02 > 0.00000000 0.50000000 0.00000000 1.562500e-02 > 0.00000000 0.50000000 0.25000000 1.562500e-02 > 0.00000000 0.50000000 0.50000000 1.562500e-02 > 0.00000000 0.50000000 0.75000000 1.562500e-02 > 0.00000000 0.75000000 0.00000000 1.562500e-02 > 0.00000000 0.75000000 0.25000000 1.562500e-02 > 0.00000000 0.75000000 0.50000000 1.562500e-02 > 0.00000000 0.75000000 0.75000000 1.562500e-02 > 0.25000000 0.00000000 0.00000000 1.562500e-02 > 0.25000000 0.00000000 0.25000000 1.562500e-02 > 0.25000000 0.00000000 0.50000000 1.562500e-02 > 0.25000000 0.00000000 0.75000000 1.562500e-02 > 0.25000000 0.25000000 0.00000000 1.562500e-02 > 0.25000000 0.25000000 0.25000000 1.562500e-02 > 0.25000000 0.25000000 0.50000000 1.562500e-02 > 0.25000000 0.25000000 0.75000000 1.562500e-02 > 0.25000000 0.50000000 0.00000000 1.562500e-02 > 0.25000000 0.50000000 0.25000000 1.562500e-02 > 0.25000000 0.50000000 0.50000000 1.562500e-02 > 0.25000000 0.50000000 0.75000000 1.562500e-02 > 0.25000000 0.75000000 0.00000000 1.562500e-02 > 0.25000000 0.75000000 0.25000000 1.562500e-02 > 0.25000000 0.75000000 0.50000000 1.562500e-02 > 0.25000000 0.75000000 0.75000000 1.562500e-02 > 0.50000000 0.00000000 0.00000000 1.562500e-02 > 0.50000000 0.00000000 0.25000000 1.562500e-02 > 0.50000000 0.00000000 0.50000000 1.562500e-02 > 0.50000000 0.00000000 0.75000000 1.562500e-02 > 0.50000000 0.25000000 0.00000000 1.562500e-02 > 0.50000000 0.25000000 0.25000000 1.562500e-02 > 0.50000000 0.25000000 0.50000000 1.562500e-02 > 0.50000000 0.25000000 0.75000000 1.562500e-02 > 0.50000000 0.50000000 0.00000000 1.562500e-02 > 0.50000000 0.50000000 0.25000000 1.562500e-02 > 0.50000000 0.50000000 0.50000000 1.562500e-02 > 0.50000000 0.50000000 0.75000000 1.562500e-02 > 0.50000000 0.75000000 0.00000000 1.562500e-02 > 0.50000000 0.75000000 0.25000000 1.562500e-02 > 0.50000000 0.75000000 0.50000000 1.562500e-02 > 0.50000000 0.75000000 0.75000000 1.562500e-02 > 0.75000000 0.00000000 0.00000000 1.562500e-02 > 0.75000000 0.00000000 0.25000000 1.562500e-02 > 0.75000000 0.00000000 0.50000000 1.562500e-02 > 0.75000000 0.00000000 0.75000000 1.562500e-02 > 0.75000000 0.25000000 0.00000000 1.562500e-02 > 0.75000000 0.25000000 0.25000000 1.562500e-02 > 0.75000000 0.25000000 0.50000000 1.562500e-02 > 0.75000000 0.25000000 0.75000000 1.562500e-02 > 0.75000000 0.50000000 0.00000000 1.562500e-02 > 0.75000000 0.50000000 0.25000000 1.562500e-02 > 0.75000000 0.50000000 0.50000000 1.562500e-02 > 0.75000000 0.50000000 0.75000000 1.562500e-02 > 0.75000000 0.75000000 0.00000000 1.562500e-02 > 0.75000000 0.75000000 0.25000000 1.562500e-02 > 0.75000000 0.75000000 0.50000000 1.562500e-02 > 0.75000000 0.75000000 0.75000000 1.562500e-02 > / > > -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From giannozz at nest.sns.it Sun Apr 15 19:52:28 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 15 Apr 2007 19:52:28 +0200 Subject: [Pw_forum] Why does pw.x think that I have 76 k-points? In-Reply-To: References: Message-ID: <357E6CE5-DBA1-44EC-B665-E5B13AF91C03@nest.sns.it> On Apr 15, 2007, at 17:38 , Brad Malone wrote > Even though my input file has only 64 k-pts, something about > the way I'm specifying the other information makes pw.x believe > that there are 76 k-points. The only thing that I'm doing that's new > compared to the last time I used pwSCF correctly is that I specified > my lattice vectors by hand. compare lattice vectors, atomic positions, symmetry operations in the 'new' case with the 'last time'. There will be some discrepancy. Notice that for hexagonal and trigonal lattices you should specify "hexagonal " after CELL_PARAMETERS to get the correct symmetries. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Sun Apr 15 19:56:58 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 15 Apr 2007 19:56:58 +0200 Subject: [Pw_forum] ph.x v3.2 on NEC SX-8 In-Reply-To: References: Message-ID: <9555BFF5-6BA3-4949-8B84-70FDE167203B@nest.sns.it> On Apr 14, 2007, at 20:48 , wlyim at puccini.che.pitt.edu wrote: > "mpirun -np 1 ./pw.x -in alas.scf.in" works. However, when using 2 > cpus, > the program crashed, which seemed to be related to the allocation > problem > in "atomic_rho.f90". > [...] > Initial potential from superposition of free atoms > **** 90 Fatal exception PROG=atomic_rho ELN=116(40038f7f4) > SIGSEGV: Segmentation violation > Called from potinit ELN=202(4002b8b3c)f7f4) > [...] I am not sure if this is a compiler problem, or compiler > option problem, > or something else. Any idea will be welcome. try to locate the exact point where the code crashes. Maybe it will give you some idea. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Sun Apr 15 20:01:14 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 15 Apr 2007 20:01:14 +0200 Subject: [Pw_forum] about choosing density for solving scf In-Reply-To: References: Message-ID: On Apr 15, 2007, at 12:24 , Amin Babazadeh wrote: > For solving the self-consistent equation in Kohn-Sham method > we need to select an ordinary (random) density to start the solving > equation. not necessarily: the default starting charge density is obtained from a sum of atomic (pseudo-)charges, stored in the pseudopotential files. A random starting density is, well, random, just normalized to yield the prescribed total charge. > i want to know in espresso how this density choose? "how is this density chosen" or " how to choose this density"? Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From wlyim at puccini.che.pitt.edu Sun Apr 15 20:38:29 2007 From: wlyim at puccini.che.pitt.edu (wlyim at puccini.che.pitt.edu) Date: Sun, 15 Apr 2007 14:38:29 -0400 (EDT) Subject: [Pw_forum] ph.x v3.2 on NEC SX-8 In-Reply-To: <9555BFF5-6BA3-4949-8B84-70FDE167203B@nest.sns.it> Message-ID: Dear Paolo, It crashed at the beginning of atom_rho subroutine. ===================================================== ! allocate work space (psic must already be allocated) ! allocate (rhocg( ngm, nspina)) <-- problematic line ndm = MAXVAL ( msh(1:ntyp) ) ===================================================== Best regards, William On Sun, 15 Apr 2007, Paolo Giannozzi wrote: > > On Apr 14, 2007, at 20:48 , wlyim at puccini.che.pitt.edu wrote: > > > "mpirun -np 1 ./pw.x -in alas.scf.in" works. However, when using 2 > > cpus, > > the program crashed, which seemed to be related to the allocation > > problem > > in "atomic_rho.f90". > > [...] > > Initial potential from superposition of free atoms > > **** 90 Fatal exception PROG=atomic_rho ELN=116(40038f7f4) > > SIGSEGV: Segmentation violation > > Called from potinit ELN=202(4002b8b3c)f7f4) > > > [...] I am not sure if this is a compiler problem, or compiler > > option problem, > > or something else. Any idea will be welcome. > > try to locate the exact point where the code crashes. Maybe it will give > you some idea. > > Paolo > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Dr. Wai-Leung Yim Institut fuer Reine und Angewandte Chemie, Theoretische Chemie, Carl von Ossiezky Universtaet Oldenburg, 26129 Oldenburg, Germany Email: wlyim at puccini.che.pitt.edu Phone: +49-441-798-3950 (office) Fax: +49-441-798-3964 From brad.malone at gmail.com Sun Apr 15 20:54:58 2007 From: brad.malone at gmail.com (Brad Malone) Date: Sun, 15 Apr 2007 11:54:58 -0700 Subject: [Pw_forum] Re: Why does pw.x think that I have 76 k-points? Message-ID: Gosia, I tried setting celldm(1)=1 and I still received 76 k-points, but thanks for the suggestion. Paolo, Sorry, I should have been more specific when I mentioned "last time". The last time was on a different crystal structure, so I obviously can't compare lattice vectors, atomic positions, and whatnot. I just mentioned it because I believe the other syntax is correct since it was left unchanged. I think my lattice vectors and atomic positions are correct because when I run pw2wannier.x everything checks except I get an error because in my wannier calculation I have 64 k-points but for the pwSCF run I have this strange thing with 76 k-points, so the pw2wannier calculation fails. But regardless, I still don't understand if in my input file I specify 64 k-points that my output file can have 76 k-points. I wouldn't think this would happen even if I had completely wrong lattice vectors, atomic positions, etc. Brad -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070415/3f6254c3/attachment.htm From brad.malone at gmail.com Sun Apr 15 21:15:06 2007 From: brad.malone at gmail.com (Brad Malone) Date: Sun, 15 Apr 2007 12:15:06 -0700 Subject: [Pw_forum] Re: Why does pw.x think that I have 76 k-points? In-Reply-To: References: Message-ID: Okay, I added the tag {hexagonal} and it worked. Thanks for the help. Best, Brad -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070415/ae9dd135/attachment.htm From tone.kokalj at ijs.si Mon Apr 16 11:12:24 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Mon, 16 Apr 2007 11:12:24 +0200 Subject: [Pw_forum] Re: Xcrysden error In-Reply-To: References: Message-ID: <1176714744.2470.2.camel@localhost.localdomain> On Sat, 2007-04-14 at 10:08 -0400, Axel Kohlmeyer wrote: > tone, > it might be a good idea to force LC_ALL=C inside of > xcrysden before executing any ancillary scripts or > executables (and the same would be true for PWGui). This is a very good suggestion, thanks! Tone From shaposh at isp.nsc.ru Tue Apr 17 13:43:49 2007 From: shaposh at isp.nsc.ru (Alexander Shaposhnikov) Date: Tue, 17 Apr 2007 18:43:49 +0700 Subject: [Pw_forum] Unit for "FREQUENCY" cart? Message-ID: <200704171843.49801.shaposh@isp.nsc.ru> Hi, i have a quick question: what is the unit for "FREQUENCY" cart for dynamical polarizability calculation in ph.x ? Best Regards, Alex From nhviet at sissa.it Tue Apr 17 14:28:16 2007 From: nhviet at sissa.it (Viet Huy Nguyen) Date: Tue, 17 Apr 2007 14:28:16 +0200 Subject: [Pw_forum] Unit for "FREQUENCY" cart? In-Reply-To: <200704171843.49801.shaposh@isp.nsc.ru> References: <200704171843.49801.shaposh@isp.nsc.ru> Message-ID: <1176812896.4624bd60c25f8@webmail.sissa.it> > i have a quick question: what is the unit for "FREQUENCY" > cart for dynamical polarizability calculation in ph.x ? > It is in Rydberg atomic unit. I would like to remind you that this stuff is still in experimental stage. Viet ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by Horde http://www.horde.org/ From nhviet at sissa.it Tue Apr 17 14:07:11 2007 From: nhviet at sissa.it (Huy-Viet Nguyen ) Date: Tue, 17 Apr 2007 14:07:11 +0200 Subject: [Pw_forum] Unit for "FREQUENCY" cart? In-Reply-To: <200704171843.49801.shaposh@isp.nsc.ru> References: <200704171843.49801.shaposh@isp.nsc.ru> Message-ID: > i have a quick question: what is the unit for "FREQUENCY" > cart for dynamical polarizability calculation in ph.x ? It is in Rydberg atomic unit. I would like to remind you that this stuff is still in experimental stage. Viet -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070417/f0872ec5/attachment.htm From degironc at sissa.it Tue Apr 17 16:13:34 2007 From: degironc at sissa.it (degironc) Date: Tue, 17 Apr 2007 16:13:34 +0200 Subject: [Pw_forum] Unit for "FREQUENCY" cart? In-Reply-To: <1176812896.4624bd60c25f8@webmail.sissa.it> References: <200704171843.49801.shaposh@isp.nsc.ru> <1176812896.4624bd60c25f8@webmail.sissa.it> Message-ID: <4624D60E.9020005@sissa.it> Notice also that the frequency is actually an "imaginary frequency". stefano Viet Huy Nguyen wrote: >>i have a quick question: what is the unit for "FREQUENCY" >>cart for dynamical polarizability calculation in ph.x ? >> >> >> > >It is in Rydberg atomic unit. >I would like to remind you that this stuff is still in experimental stage. > >Viet > > >---------------------------------------------------------------- > SISSA Webmail https://webmail.sissa.it/ > Powered by Horde http://www.horde.org/ > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From calderin at physics.queensu.ca Wed Apr 18 16:00:17 2007 From: calderin at physics.queensu.ca (Lazaro Calderin) Date: Wed, 18 Apr 2007 16:00:17 +0200 Subject: [Pw_forum] Relativistic GGA ? Message-ID: <1176904817.3081.41.camel@localhost.localdomain> Hi guys, Is it possible to do relativistic GGA calculations with pwscf? Thanks, Lazaro From dalcorso at sissa.it Wed Apr 18 16:10:05 2007 From: dalcorso at sissa.it (Andrea Dal Corso) Date: Wed, 18 Apr 2007 16:10:05 +0200 Subject: [Pw_forum] Relativistic GGA ? In-Reply-To: <1176904817.3081.41.camel@localhost.localdomain> References: <1176904817.3081.41.camel@localhost.localdomain> Message-ID: <1176905406.4168.36.camel@dhpc-5-05.sissa.it> It is possible but not well tested. It should work well for nonmagnetic systems, while it could have problems for magnetic systems. In any case if you find strange behavior please report it to the forum. Andrea On Wed, 2007-04-18 at 16:00 +0200, Lazaro Calderin wrote: > Hi guys, > > Is it possible to do relativistic GGA calculations with pwscf? > > Thanks, Lazaro > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- 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 From ghaderyan at gmail.com Wed Apr 18 23:15:51 2007 From: ghaderyan at gmail.com (saman ghaderyan) Date: Thu, 19 Apr 2007 00:45:51 +0330 Subject: [Pw_forum] pseudo potential Message-ID: hi i want to make pseudo potential ,how can i do it?(where can i found input data or example ) -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070419/455c58d7/attachment.htm From degironc at sissa.it Wed Apr 18 23:34:25 2007 From: degironc at sissa.it (degironc) Date: Wed, 18 Apr 2007 23:34:25 +0200 Subject: [Pw_forum] pseudo potential In-Reply-To: References: Message-ID: <46268EE1.9070305@sissa.it> saman ghaderyan wrote: > hi > i want to make pseudo potential ,how can i do it?(where can i found > input data or example ) have you considered the possibility to have a look in the atomic_doc subdirectory of the QE distribution ? stefano From peterwiney at gmail.com Wed Apr 18 23:28:38 2007 From: peterwiney at gmail.com (Peter Winey) Date: Wed, 18 Apr 2007 17:28:38 -0400 Subject: [Pw_forum] PWscf vs. CP Message-ID: Dear all, I was doing an electronic minimization using the CPV code cp.x. To make sure I did not make mistakes, I also did a similar calculation using the PWscf code pw.x. Well, when I compared the results, I found quite some difference, in particular, the forces acting on atoms are quite different while the total energy are close. My question is whether I should expect to get same results from the two codes? If so, there must be something wrong with the input files that I used (see below). Thanks in advance for any suggestions that you may have! -Peter --------------------------pw.x input file--------- C2H4 &control calculation='scf' restart_mode='from_scratch', wf_collect=.true., prefix='pwscf', pseudo_dir = '/espresso-3.0/pseudo', outdir='.', tprnfor=.true. / &system ibrav = 1 celldm(1) = 37.7945197733 nat=6, ntyp=2, ecutwfc = 30.0, occupations='fixed', / &electrons conv_thr = 1.0d-8 mixing_beta = 0.7 / ATOMIC_SPECIES C 12.0107 C.pbe-van_ak.UPF H 1.00794 H.pbe-van_ak.UPF K_POINTS {gamma} ATOMIC_POSITIONS {crystal} C 0.46664509 0.5 0.5 C 0.53335491 0.5 0.5 H 0.43793698 0.5 0.45351762 H 0.56206302 0.5 0.45351762 H 0.56206302 0.5 0.54648238 H 0.43793698 0.5 0.54648238 --------------------------cp.x input file--------- &control calculation='cp', restart_mode='from_scratch', ndr=52, ndw=53, nstep=400, iprint=10, isave=50, dt=5.0, etot_conv_thr = 1.d-9, ekin_conv_thr = 1.d-5, prefix = 'cp' pseudo_dir='/espresso-3.0/pseudo/', tprnfor = .TRUE. outdir='.', / &system ibrav = 1 celldm(1) = 37.7945197733 nat=6, ntyp=2, ecutwfc = 30.0, / &electrons electron_dynamics='damp' / &ions ion_dynamics='none', / ATOMIC_SPECIES C 12.0107 C.pbe-van_ak.UPF H 1.00794 H.pbe-van_ak.UPF ATOMIC_POSITIONS {crystal} C 0.46664509 0.5 0.5 C 0.53335491 0.5 0.5 H 0.43793698 0.5 0.45351762 H 0.56206302 0.5 0.45351762 H 0.56206302 0.5 0.54648238 H 0.43793698 0.5 0.54648238 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070418/31b047ba/attachment.htm From akohlmey at cmm.chem.upenn.edu Thu Apr 19 01:10:12 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 18 Apr 2007 19:10:12 -0400 (EDT) Subject: [Pw_forum] PWscf vs. CP In-Reply-To: Message-ID: On Wed, 18 Apr 2007, Peter Winey wrote: PW> Dear all, PW> PW> I was doing an electronic minimization using the CPV code cp.x. To make sure PW> I did PW> not make mistakes, I also did a similar calculation using the PWscf code PW> pw.x. PW> PW> Well, when I compared the results, I found quite some difference, PW> in particular, PW> the forces acting on atoms are quite different while the total energy are PW> close. there can be a number of reasons for this, but first please give us an example for what you consider a large difference in the forces. cheers, axel. PW> PW> My question is whether I should expect to get same results from the two PW> codes? PW> If so, there must be something wrong with the input files that I used (see PW> below). PW> PW> Thanks in advance for any suggestions that you may have! PW> PW> -Peter PW> PW> --------------------------pw.x input file--------- PW> C2H4 PW> &control PW> calculation='scf' PW> restart_mode='from_scratch', PW> wf_collect=.true., PW> prefix='pwscf', PW> pseudo_dir = '/espresso-3.0/pseudo', PW> outdir='.', PW> tprnfor=.true. PW> / PW> &system PW> ibrav = 1 PW> celldm(1) = 37.7945197733 PW> nat=6, ntyp=2, PW> ecutwfc = 30.0, PW> occupations='fixed', PW> / PW> &electrons PW> conv_thr = 1.0d-8 PW> mixing_beta = 0.7 PW> / PW> ATOMIC_SPECIES PW> C 12.0107 C.pbe-van_ak.UPF PW> H 1.00794 H.pbe-van_ak.UPF PW> K_POINTS {gamma} PW> ATOMIC_POSITIONS {crystal} PW> C 0.46664509 0.5 0.5 PW> C 0.53335491 0.5 0.5 PW> H 0.43793698 0.5 0.45351762 PW> H 0.56206302 0.5 0.45351762 PW> H 0.56206302 0.5 0.54648238 PW> H 0.43793698 0.5 0.54648238 PW> PW> --------------------------cp.x input file--------- PW> &control PW> calculation='cp', PW> restart_mode='from_scratch', PW> ndr=52, ndw=53, PW> nstep=400, iprint=10, isave=50, PW> dt=5.0, PW> etot_conv_thr = 1.d-9, PW> ekin_conv_thr = 1.d-5, PW> prefix = 'cp' PW> pseudo_dir='/espresso-3.0/pseudo/', PW> tprnfor = .TRUE. PW> outdir='.', PW> / PW> &system PW> ibrav = 1 PW> celldm(1) = 37.7945197733 PW> nat=6, ntyp=2, PW> ecutwfc = 30.0, PW> / PW> &electrons PW> electron_dynamics='damp' PW> / PW> &ions PW> ion_dynamics='none', PW> / PW> ATOMIC_SPECIES PW> C 12.0107 C.pbe-van_ak.UPF PW> H 1.00794 H.pbe-van_ak.UPF PW> ATOMIC_POSITIONS {crystal} PW> C 0.46664509 0.5 0.5 PW> C 0.53335491 0.5 0.5 PW> H 0.43793698 0.5 0.45351762 PW> H 0.56206302 0.5 0.45351762 PW> H 0.56206302 0.5 0.54648238 PW> H 0.43793698 0.5 0.54648238 PW> -- ======================================================================= 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 Apr 19 02:50:54 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 18 Apr 2007 20:50:54 -0400 Subject: [Pw_forum] PWscf vs. CP In-Reply-To: References: Message-ID: <4626BCEE.2030004@mit.edu> Dear Peter, results should be very close or almost identical. 1) make sure units are right - I'm pretty sure PWscf is Ry/bohr, and CP should be Ha/bohr (i.e. a factor of 2). 2) make sure you are very well converged. You could even try the conjugate gradient minimization on CP (but for this you might want to download the latest CVS version) 3) there is a bug nicola Peter Winey wrote: > Dear all, > > I was doing an electronic minimization using the CPV code cp.x. To make > sure I did > not make mistakes, I also did a similar calculation using the PWscf code > pw.x. > > Well, when I compared the results, I found quite some difference, > in particular, > the forces acting on atoms are quite different while the total energy > are close. > > My question is whether I should expect to get same results from the two > codes? > If so, there must be something wrong with the input files that I used > (see below). > > Thanks in advance for any suggestions that you may have! > > -Peter --------------------------------------------------------------------- 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 peterwiney at gmail.com Thu Apr 19 04:49:51 2007 From: peterwiney at gmail.com (Peter Winey) Date: Wed, 18 Apr 2007 22:49:51 -0400 Subject: [Pw_forum] PWscf vs. CP In-Reply-To: <4626BCEE.2030004@mit.edu> References: <4626BCEE.2030004@mit.edu> Message-ID: Dear Nicola and Axel, Thanks for the quick reply! I was aware of the unit difference, and both the CP and PWscf runs were well converged. The system is a simple C2H4 molecule in a supercell, so the calculations were very fast -- just took a couple of minutes on my PC cluster. My experience with CP is limited. Maybe there is something wrong with my CP input file (as shown in my initial message)? Thanks again! -Peter PS. 1). My pwscf run gave the following: ----------------------------------------------------------------------- ! total energy = -27.44577321 ryd estimated scf accuracy < 8.3E-09 ryd band energy sum = -10.24130443 ryd one-electron contribution = -93.97118129 ryd hartree contribution = 48.15240670 ryd xc contribution = -9.70116731 ryd ewald contribution = 28.07416868 ryd convergence has been achieved Forces acting on atoms (Ry/au): atom 1 type 1 force = 0.00001190 0.00000000 0.00000000 atom 2 type 1 force = -0.00001190 0.00000000 0.00000000 atom 3 type 2 force = -0.00009911 0.00000000 -0.00008380 atom 4 type 2 force = 0.00009911 0.00000000 -0.00008380 atom 5 type 2 force = 0.00009911 0.00000000 0.00008380 atom 6 type 2 force = -0.00009911 0.00000000 0.00008380 Total force = 0.000260 Total SCF correction = 0.000042 ----------------------------------------------------------------------- 2). My CP run gave quite difference results: ----------------------------------------------------------------------- Forces acting on atoms (au): C 0.644188E-01 0.411253E+00 0.458051E+00 C -0.644221E-01 0.411256E+00 0.458049E+00 H 0.640055E-02 0.147247E-01 0.834130E-02 H -0.640012E-02 0.147240E-01 0.834107E-02 H -0.702662E-02 0.146529E-01 -0.873173E-02 H 0.702701E-02 0.146514E-01 -0.873243E-02 Partial temperatures (for each ionic specie) Species Temp (K) Mean Square Displacement (a.u.) 1 0.00 0.0000 2 0.00 0.0000 nfi ekinc temph tempp etot enthal econs econt vnhh xnhh0 vnhp xnhp0 123 0.00000 0.0 0.0 -14.14165 -14.14165 -14.14165 -14.14165 0.0000 0.0000 0.0000 0.0000 MAIN: EKINC (thr) DETOT (thr) MAXFORCE (thr) MAIN: 0.135714D-08 0.1D-04 0.404064D-09 0.1D-08 0.000000D+00 0.1D+11 MAIN: convergence achieved for system relaxation ----------------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070418/c2a79832/attachment.htm From giannozz at nest.sns.it Thu Apr 19 09:09:32 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 19 Apr 2007 09:09:32 +0200 Subject: [Pw_forum] PWscf vs. CP In-Reply-To: References: <4626BCEE.2030004@mit.edu> Message-ID: <7874818F-A1F1-4765-A8CA-7B261459CB68@nest.sns.it> On Apr 19, 2007, at 4:49 , Peter Winey wrote: > My experience with CP is limited. Maybe there is something wrong > with my CP input file (as shown in my initial message)? or maybe something missing. With CP and ultrasoft PP you have to specify a "small box" (parameters nr1, nr2b, nr3b). I think there is a defaults but it is rather lousy. What the code uses is printed in the output (llokk for nr*b) Due to the different treatment of the augmentation charges, you will not get exactly the same numbers between PW and CP, but things are properly done, the results will be the same Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From roma at srmp.saclay.cea.fr Thu Apr 19 10:57:42 2007 From: roma at srmp.saclay.cea.fr (roma) Date: Thu, 19 Apr 2007 10:57:42 +0200 Subject: [Pw_forum] About valence bands maximum In-Reply-To: References: Message-ID: <1176973063.12937.44.camel@srmp19.saclay.cea.fr> Dear Ian Haiping, it is true that seldom papers about defects discuss this point, and I am probably guilty of this too. My understanding is that your expression for the valence band top of the defect is correct, but then it depends on how you use it. What has been omitted according to the person that told you this? The crucial quantity is [ -] (let us call it DeltaV(D,q) ). Long ago I did a small program to calculate the so called macroscopic average of the potential (Balderesci, Baroni,Resta, PRL v61, p734, 1988) in the supercell: it was ok for a defect in a simple metal (I could find the "bulklike" region) but for the system I was interested in, SiO2, the oscillations of the macroscopic average along the supercell made it useless. Supercell too small? Problems due to the symmetry of my supercell? I adopted a different strategy, I determined the energy shift that maximises the overlap (or minimises the square differences) of the density of states of the defected and perfect crystal. This shift is for me DeltaV(D,q). This, according to my experience, can be unambiguosly determined as long as the relaxations are good (and k-point sampling is sufficient). Then another question is to which extent the DeltaV term corrects for the image interaction of charged defects, on which I would appreciate also the feedback from others on the list. Best regards, Guido On thursday April 12 Ian Haiping wrote: > I want to determin the transition enery of a defect level. It is >important to align the valence bands maximum of different systems under >investigation. Many works has claimed valence bands maximum alignments >were performed but no technique details are given. >For my understanding, >I thougt the valence bands maxium of charge defect systems could be >determined by aligning its average potential >to pure perfect system's potential, e.g \epsilon_{VBM}.(Defects,q) = >\epsilon_{VBM}(0) + [ -]. But some person told me such >expression has omitted something. So it really confused me much. As >far as i know, the potential of system could be only determined to >some constant . i thought the alignment of average potentials of two >systems could take into account this arbitrariness.. I am not very >certain about this VBM alignment right now Would you give me some >comments and clarify my understanding ? > -- Guido Roma -- CEA-Saclay - DEN/DMN/SRMP Bat.520/13 Phone: [+33]-1-69081857 -- Fax: ...6867 -- Mobile: [+33]-6-20069085 From peterwiney at gmail.com Thu Apr 19 16:36:28 2007 From: peterwiney at gmail.com (Peter Winey) Date: Thu, 19 Apr 2007 10:36:28 -0400 Subject: [Pw_forum] PWscf vs. CP In-Reply-To: <7874818F-A1F1-4765-A8CA-7B261459CB68@nest.sns.it> References: <4626BCEE.2030004@mit.edu> <7874818F-A1F1-4765-A8CA-7B261459CB68@nest.sns.it> Message-ID: Thanks, Paolo! Indeed I forgot to specify nr1b, nr2b, nr3b. When I set them to 20, the CP result agrees reasonably well with the PW result. Now I just need to do more test and find the proper setting for "nr*b". Best, Peter > or maybe something missing. With CP and ultrasoft PP you have to > specify a "small box" (parameters nr1, nr2b, nr3b). I think there is > a defaults but it is rather lousy. What the code uses is printed in the > output (llokk for nr*b) > > Due to the different treatment of the augmentation charges, > you will not get exactly the same numbers between PW and CP, > but things are properly done, the results will be the same > > Paolo > -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070419/0c364a70/attachment.htm From th273 at cornell.edu Thu Apr 19 23:46:43 2007 From: th273 at cornell.edu (Tuo-Hung Hou) Date: Thu, 19 Apr 2007 17:46:43 -0400 Subject: [Pw_forum] Work function calculation of CNT Message-ID: <200704192145.l3JLjDXm005286@authusersmtp.mail.cornell.edu> Hi!, I am doing the work function calculation on carbon nanotubes. If I am not mistaken, the scf calculation should give me Fermi energy or midgap between HOMO and LUMO. Then, one can use pp.x plot_num 1 to generate the total potential profile. The work function is decided by the difference between the Fermi energy and the potential energy at the vaccum of the supercell. However, I am running into a problem that the Fermi energy from scf (1.65 ev) is higher than any potential inside the entire supercell (ranging from -10.4 to 0.4 eV). When I looked at both the charge and potential profiles, they seemed to suggest the Fermi energy should be somewhere around -4 eV, which is not even close to the value scf reported. I am very confused now. What is the Fermi energy I can use? Any thoughts on this? Thanks Tuo-Hung _______________________________________ Tuo-Hung Hou Cornell University School of Electrical and Computer Engineering 325 Phillips Hall Ithaca, NY 14853 phone: 607.255.4181 email : th273 at cornell.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070419/498be335/attachment.htm From marzari at MIT.EDU Fri Apr 20 00:54:20 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 19 Apr 2007 18:54:20 -0400 Subject: [Pw_forum] Work function calculation of CNT In-Reply-To: <200704192145.l3JLjDXm005286@authusersmtp.mail.cornell.edu> References: <200704192145.l3JLjDXm005286@authusersmtp.mail.cornell.edu> Message-ID: <4627F31C.2050808@mit.edu> Dear Tuo-Hung, most likely you are not looking at the electrostatic potential in the vacuum region, i.e. away from the nanotube axis. For a metallic nanotube, studied using pwscf, and treated as a metal (i.e. with smearing and fractional occupations), the code at self-consistency will give you the fermi energy - this is calculated so that the number of electrons between minus infinity and the fermi energy is exactly the number of electrons of your CNT, taking into account the fact that, at finite temperature/broadenings, the occupation function is not 1 below efermi and 0 above, but is smoothed out by the thermal distribution. Now that you have efermi, you need to know the vacuum level, i.e. the energy of a test charge very far away from the tube: the work function will be the difference between these two. PWSCF does not set automatically this level to zero, and it would actually be a fairly subtle exercise in electrostatics doing so. So, what you need to do is to plot the electrostatic potential in your unit cell, making sure that your unit cell is large enough so that your potential has become flat very far away from the tube. If your scf calculation is correct, the only thing that can go wrong is that you are not looking at the potential in the right places. nicola Tuo-Hung Hou wrote: > Hi!, > > I am doing the work function calculation on carbon nanotubes. If I am > not mistaken, the scf calculation should give me Fermi energy or midgap > between HOMO and LUMO. Then, one can use pp.x plot_num 1 to generate the > total potential profile. The work function is decided by the difference > between the Fermi energy and the potential energy at the vaccum of the > supercell. > > > > However, I am running into a problem that the Fermi energy from scf > (1.65 ev) is higher than any potential inside the entire supercell > (ranging from -10.4 to 0.4 eV). When I looked at both the charge and > potential profiles, they seemed to suggest the Fermi energy should be > somewhere around -4 eV, which is not even close to the value scf > reported. I am very confused now. What is the Fermi energy I can use? > Any thoughts on this? -- --------------------------------------------------------------------- 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 niuli1978 at yahoo.com.cn Fri Apr 20 01:10:43 2007 From: niuli1978 at yahoo.com.cn (li niu) Date: Fri, 20 Apr 2007 07:10:43 +0800 (CST) Subject: [Pw_forum] on Raman spectra of large disorder system Message-ID: <219420.61435.qm@web15014.mail.cnb.yahoo.com> Dear all PWscf users, I want to investigate Raman spectra in the case of large system. How to use the finite electric fields method by pwscf or the perturbative variational approach for calculation of dielectric tensors [in RAMAN SPECTRA OF DISORDERED OXIDES FROM FIRST PRINCIPLES by PAOLO UMARI]? I did not find those from the forum. Thank you in advance! Niu Li --------------------------------- ????????-3.5G???20M??? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070420/55f5c139/attachment.htm From hande at newton.physics.metu.edu.tr Fri Apr 20 08:59:37 2007 From: hande at newton.physics.metu.edu.tr (Hande Ustunel) Date: Fri, 20 Apr 2007 09:59:37 +0300 (EEST) Subject: [Pw_forum] Work function calculation of CNT In-Reply-To: <200704192145.l3JLjDXm005286@authusersmtp.mail.cornell.edu> Message-ID: Dear Tuo-Hung, > However, I am running into a problem that the Fermi energy from scf (1.65 > ev) is higher than any potential inside the entire supercell (ranging from > -10.4 to 0.4 eV). When I looked at both the charge and potential profiles, > they seemed to suggest the Fermi energy should be somewhere around -4 eV, > which is not even close to the value scf reported. I am very confused now. > What is the Fermi energy I can use? Any thoughts on this? You might also want to double-check the units. The Fermi energy is reported in the output in eV whereas the potential values are in Ryd unless I'm mistaken. Hande -- 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 lukethulin at netscape.net Fri Apr 20 19:58:05 2007 From: lukethulin at netscape.net (lukethulin at netscape.net) Date: Fri, 20 Apr 2007 12:58:05 -0500 Subject: [Pw_forum] Band structure paths Message-ID: <000e01c78375$725e8d80$6601a8c0@LukePC> Hello, Does anyone know of a good reference that lists conventional k-paths for full plotting band structures? I'm particularly interested in the bct lattice (c>a) and I have several papers that all seem to do it very differently. Thanks, Luke Thulin Nanoptek, Corp. 8 Clock Tower Place, Suite 120 Maynard, MA 01754 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070420/d48e5961/attachment.htm From th273 at cornell.edu Fri Apr 20 22:02:19 2007 From: th273 at cornell.edu (Tuo-Hung Hou) Date: Fri, 20 Apr 2007 16:02:19 -0400 Subject: [Pw_forum] Work function calculation of CNT In-Reply-To: Message-ID: <200704202000.l3KK0pNA008906@authusersmtp.mail.cornell.edu> Dear Hande, After some searching in the archives, I found the unit of the potential energy is eV not Ryd. Here is the link: http://www.democritos.it/pipermail/pw_forum/2006-August/004785.html Also based on the suggestion from Prof. Marzari, I found the calculation makes more sense with a larger unit cell. At least a 40 Bohr hexagonal unit cell may be required for a (8,0) tube. Thanks for all your inputs Tuo-Hung -----Original Message----- From: pw_forum-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] On Behalf Of Hande Ustunel Sent: Friday, April 20, 2007 3:00 AM To: pw_forum at pwscf.org Subject: Re: [Pw_forum] Work function calculation of CNT Dear Tuo-Hung, > However, I am running into a problem that the Fermi energy from scf (1.65 > ev) is higher than any potential inside the entire supercell (ranging from > -10.4 to 0.4 eV). When I looked at both the charge and potential profiles, > they seemed to suggest the Fermi energy should be somewhere around -4 eV, > which is not even close to the value scf reported. I am very confused now. > What is the Fermi energy I can use? Any thoughts on this? You might also want to double-check the units. The Fermi energy is reported in the output in eV whereas the potential values are in Ryd unless I'm mistaken. Hande -- 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 _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From degironc at sissa.it Fri Apr 20 22:11:19 2007 From: degironc at sissa.it (degironc) Date: Fri, 20 Apr 2007 22:11:19 +0200 Subject: [Pw_forum] Work function calculation of CNT In-Reply-To: <200704202000.l3KK0pNA008906@authusersmtp.mail.cornell.edu> References: <200704202000.l3KK0pNA008906@authusersmtp.mail.cornell.edu> Message-ID: <46291E67.4070407@sissa.it> dear Tuo-Hung Hou, >Dear Hande, > >After some searching in the archives, I found the unit of the potential >energy is eV not Ryd. Here is the link: > >http://www.democritos.it/pipermail/pw_forum/2006-August/004785.html > > if you READ the link you'll discover that the units for the "potential" (which is actually an energy since it is e*V) are Ry. stefano >Also based on the suggestion from Prof. Marzari, I found the calculation >makes more sense with a larger unit cell. At least a 40 Bohr hexagonal unit >cell may be required for a (8,0) tube. > >Thanks for all your inputs > >Tuo-Hung > >-----Original Message----- >From: pw_forum-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] On Behalf >Of Hande Ustunel >Sent: Friday, April 20, 2007 3:00 AM >To: pw_forum at pwscf.org >Subject: Re: [Pw_forum] Work function calculation of CNT > >Dear Tuo-Hung, > > > >>However, I am running into a problem that the Fermi energy from scf (1.65 >>ev) is higher than any potential inside the entire supercell (ranging from >>-10.4 to 0.4 eV). When I looked at both the charge and potential >> >> >profiles, > > >>they seemed to suggest the Fermi energy should be somewhere around -4 eV, >>which is not even close to the value scf reported. I am very confused now. >>What is the Fermi energy I can use? Any thoughts on this? >> >> > >You might also want to double-check the units. The Fermi energy is >reported in the output in eV whereas the potential values are in Ryd >unless I'm mistaken. > >Hande > > > From calderin at physics.queensu.ca Mon Apr 16 21:17:48 2007 From: calderin at physics.queensu.ca (Lazaro Calderin) Date: Mon, 16 Apr 2007 21:17:48 +0200 Subject: [Pw_forum] Relativistic GGA calculations Message-ID: <1176751069.3081.25.camel@localhost.localdomain> Hello every one, Could some one tell me if it is possible to perform full and/or scalar relativistic GGA calculations with PWscf? Thanks, Lazaro From rcjhawk at gmail.com Fri Apr 20 23:42:48 2007 From: rcjhawk at gmail.com (Michael Mehl) Date: Fri, 20 Apr 2007 17:42:48 -0400 Subject: [Pw_forum] Band structure paths In-Reply-To: <000e01c78375$725e8d80$6601a8c0@LukePC> References: <000e01c78375$725e8d80$6601a8c0@LukePC> Message-ID: <462933D8.90008@gmail.com> Not exactly what you are asking for, but the Bilbao Crystallographic Server lists the high symmetry K-points for each space group: http://www.cryst.ehu.es/cryst/get_kvec.html From there I usually just wing it :-) lukethulin at netscape.net wrote: > Hello, > > Does anyone know of a good reference that lists conventional k-paths for > full plotting band structures? I'm particularly interested in the bct > lattice (c>a) and I have several papers that all seem to do it very > differently. > > Thanks, > > Luke Thulin > Nanoptek, Corp. > 8 Clock Tower Place, Suite 120 > Maynard, MA 01754 -- mjm http://hawknotes.blogspot.com/ http://cst-www.nrl.navy.mil/lattice From Ihor.Kupchak at uoit.ca Sun Apr 22 00:21:47 2007 From: Ihor.Kupchak at uoit.ca (Ihor Kupchak) Date: Sat, 21 Apr 2007 18:21:47 -0400 Subject: [Pw_forum] Optimal number of CPUs Message-ID: Dear pwscf users! I run CP calculation with parallel version 3.2. Using 16 CPUs I got only 1.8 time/step improvement comparing with ?serial? mode and the same input. Is this improvement normal? Haw many CPUs I have to use to get an optimal result? In general, what is the principle haw to choose processors number? Thank you. Ihor Kupchak *********************** University of Ontario Institute of Technology 2000 Simcoe Street North, Oshawa, Ontario, Canada L1H 7K4 From akohlmey at cmm.chem.upenn.edu Sun Apr 22 00:21:17 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 21 Apr 2007 18:21:17 -0400 (EDT) Subject: [Pw_forum] Optimal number of CPUs In-Reply-To: Message-ID: On Sat, 21 Apr 2007, Ihor Kupchak wrote: IH> Dear pwscf users! dear ihor, IH> I run CP calculation with parallel version 3.2. Using 16 CPUs I got IH> only 1.8 time/step improvement comparing with ?serial? mode and the IH> same input. Is this improvement normal? Haw many CPUs I have to use IH> to get an optimal result? In general, what is the principle haw to IH> choose processors number? Thank you. good parallel performance with the cp.x code depends primarily on have a good low-latency network. if you have a gigabit ethernet, it is highly likely, that your scaling is limited. the best way to find the optimal number of processors, is to do a series of short jobs with increasing number of processors. there have been some discussions on benchmarks and performance on this list before, so you may want to search the mailing list archive. since the method implemented in cp.x and the parallelization strategy is very similar to the CPMD package, you may want to have a look at http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/cpmd-bench.html (it is quite outdated, but the essential conclusions are still valid, you just have different memory bandwidth properties on new cpus, particularly multi-core cpus, which are not very efficient with cp.x and similar codes). the optimal number will always depend on the combination of problem size, network, machine/memory speed, number of cpus/cores per node and urgency. cheers, axel. IH> IH> Ihor Kupchak IH> *********************** IH> University of Ontario Institute of Technology IH> 2000 Simcoe Street North, Oshawa, Ontario, Canada L1H 7K4 IH> _______________________________________________ IH> Pw_forum mailing list IH> Pw_forum at pwscf.org IH> http://www.democritos.it/mailman/listinfo/pw_forum IH> -- ======================================================================= 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 Sun Apr 22 04:46:35 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Sat, 21 Apr 2007 22:46:35 -0400 Subject: [Pw_forum] Optimal number of CPUs In-Reply-To: References: Message-ID: <462ACC8B.2020601@mit.edu> Ihor Kupchak wrote: > Dear pwscf users! > I run CP calculation with parallel version 3.2. Using 16 CPUs I got only 1.8 time/step > Ihor Kupchak Dear Ihor, have in particular a look at this, on scaling, latency, etc... http://www.democritos.it/pipermail/pw_forum/2006-December/005540.html 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 ongphuongvu at yahoo.com Sun Apr 22 17:01:39 2007 From: ongphuongvu at yahoo.com (vu ongphuong) Date: Sun, 22 Apr 2007 08:01:39 -0700 (PDT) Subject: [Pw_forum] Question on geometry relaxation In-Reply-To: <20070117063753.11518.57554.Mailman@democritos.sissa.it> Message-ID: <627132.28704.qm@web50604.mail.re2.yahoo.com> Dear users, I am trying to get the optimized geometry of monoclinic ZrO_{2} which have 4 parameters: a =/ b =/ c and angle \beta = (c,a) =/ 90. I don't know how to run a relaxation in which 4 parameters should be optimized at the same time. Is it right if I set the input file the same as the one for lattice constant calculation, but using calculation='md'? Please help me, Thank you very much, opvu, --------------------------------- Ahhh...imagining that irresistible "new car" smell? Check outnew cars at Yahoo! Autos. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070422/bf380022/attachment.htm From marcel at physik.tu-berlin.de Sun Apr 22 17:28:40 2007 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Sun, 22 Apr 2007 17:28:40 +0200 (CEST) Subject: [Pw_forum] Question on geometry relaxation In-Reply-To: <627132.28704.qm@web50604.mail.re2.yahoo.com> References: <627132.28704.qm@web50604.mail.re2.yahoo.com> Message-ID: <20070422172327.O89449@rosa.physik-pool.tu-berlin.de> Dear Opvu > I am trying to get the optimized geometry of monoclinic ZrO_{2} which have 4 parameters: > a =/ b =/ c and angle \beta = (c,a) =/ 90. > > I don't know how to run a relaxation in which 4 parameters should be optimized at the same time. Is it right if I set the input file the same as the one for lattice constant calculation, but using calculation='md'? I don't think that md is the right option for you, it is vc-relax. I think (but I am definitely not sure about it) the difference is that with vc-relax time is a bit more arbitray and motions are damped. I would repeat the relaxation from different starting geometries, to look, that the structure relaxes alway into the same geometry. Alternatevely, you could do a lot of scf-calculations, and slightly change parameters a,b,c, \beta. In this 4-parameter space you try to minimize the total energy by some formular. Cheers Marcel From ghaderyan at gmail.com Mon Apr 23 00:48:02 2007 From: ghaderyan at gmail.com (saman ghaderyan) Date: Mon, 23 Apr 2007 02:18:02 +0330 Subject: [Pw_forum] zns phonon freq Message-ID: hi when i calculation phonon freq for zns i look at q-vector (0 0 0) acoustic freq is in order 45 cm-1, i think it will be 0 cm-1? if that's right ,how can i changed it? but for another q-vectors almost are correct . thanks saman ghaderyan buali sina university -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070423/f107fcb5/attachment.htm From akohlmey at cmm.chem.upenn.edu Mon Apr 23 01:12:34 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sun, 22 Apr 2007 19:12:34 -0400 (EDT) Subject: [Pw_forum] zns phonon freq In-Reply-To: Message-ID: On Mon, 23 Apr 2007, saman ghaderyan wrote: SG> hi SG> when i calculation phonon freq for zns i look at q-vector (0 0 0) SG> acoustic freq is in order 45 cm-1, i think it will be 0 cm-1? SG> if that's right ,how can i changed it? please search the mailing list archives for similar discussions on this topics and make sure that your calculation is consistent with the recommendations given there. this kind of topic comes up quite oftern. regards, axel. SG> but for another q-vectors almost are correct . SG> thanks SG> saman ghaderyan SG> buali sina university SG> -- ======================================================================= 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 ongphuongvu at yahoo.com Mon Apr 23 03:24:23 2007 From: ongphuongvu at yahoo.com (vu ongphuong) Date: Sun, 22 Apr 2007 18:24:23 -0700 (PDT) Subject: [Pw_forum] Question on bond overlap population calculation In-Reply-To: <20070117063753.11518.57554.Mailman@democritos.sissa.it> Message-ID: <26175.9114.qm@web50602.mail.re2.yahoo.com> Dear users, Bond overlap population is defined as a measure of the shared electron density between 2 atoms. I find no PWSCF-examples calculating overlap population. My question is whether PWSCF can calculate this quantity. If yes, can someone teach me how to calculate the overlap population? Thank you very much, opvu --------------------------------- Ahhh...imagining that irresistible "new car" smell? Check outnew cars at Yahoo! Autos. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070422/71c58358/attachment.htm From yukihiro_okuno at fujifilm.co.jp Mon Apr 23 07:07:13 2007 From: yukihiro_okuno at fujifilm.co.jp (yukihiro_okuno at fujifilm.co.jp) Date: Mon, 23 Apr 2007 14:07:13 +0900 Subject: [Pw_forum] Not conversed force for relaxed calculation Message-ID: Dear PWSCF users. I'm working to calculate the monoclinic structure of PZT materials. In order to get relaxed atomic positions, I calculated the relax atomic structure but the Total force of system is not conversed and remains like, Total force = 0.006134 Total SCF correction = 0.000026 Total force = 0.006306 Total SCF correction = 0.000021 Total force = 0.006475 Total SCF correction = 0.000042 Total force = 0.006606 Total SCF correction = 0.000007 Total force = 0.006711 Total SCF correction = 0.000023 Total force = 0.006784 Total SCF correction = 0.000017 Total force = 0.006845 Total SCF correction = 0.000019 Total force = 0.006894 Total SCF correction = 0.000041 Total force = 0.006947 Total SCF correction = 0.000025 Total force = 0.006990 Total SCF correction = 0.000023 when such a not conversed total force has happened, how should we treat the problem ,and where we change the input parameters. Sincerely yours. My input file is as follow, &control calculation = 'relax', restart_mode = 'from_scratch', pseudo_dir = '../../pseudo/', outdir = './', prefix='PZT' forc_conv_thr = 1.0D-4 / &system ibrav=12 celldm(1)=7.57696068724 celldm(2)=2.09000418 celldm(3)=1.00 celldm(4)= 0.00199999600001 nat=10 ntyp=4 nbnd=44 ecutwfc=60.0 ecutrho = 600.0 occupations = 'fixed' degauss=0.00 / &electrons conv_thr = 1e-9, mixing_beta=0.3, / &IONS ion_dynamics='bfgs' bfgs_ndim=3 pot_extrapolation = "second_order" wfc_extrapolation = "second_order" trust_radius_max=5.0D-3 trust_radius_min=1.0D-4 / ATOMIC_SPECIES Pb 207.2 pb_ps.uspp.UPF Zr 91.224 zr_ps.uspp.UPF Ti 47.867 ti_ps.uspp.UPF O 15.9994 o_ps.uspp.UPF ATOMIC_POSITIONS {crystal} Pb 0.002397553 0.006008858 0.000000000 Pb 0.002030389 0.537841278 0.000000000 Zr 0.500010466 0.244178434 0.500000000 Ti 0.501896689 0.740659564 0.500000000 O 0.498716727 0.951077912 0.500000000 O 0.498709982 0.478318964 0.500000000 O 0.499093210 0.189008468 0.000000000 O -0.000233773 0.189010061 0.500000000 O 0.498627112 0.711878632 0.000000000 O -0.001248355 0.711877827 0.500000000 K_POINTS {automatic} 6 6 6 0 0 0 From w2agz at pacbell.net Mon Apr 23 08:08:46 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Sun, 22 Apr 2007 23:08:46 -0700 Subject: [Pw_forum] kpoint_grid & npk Message-ID: <003e01c7856d$db254210$0301a8c0@w2agz> To All.where is the default for npk set that is fed to kpoint_grid and what governs the choice? For a MP grid of 100^3 that's only a million! (just kidding ;).) Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford University EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070422/482dfd8e/attachment.htm From giannozz at nest.sns.it Mon Apr 23 09:29:29 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 23 Apr 2007 09:29:29 +0200 Subject: [Pw_forum] Question on bond overlap population calculation In-Reply-To: <26175.9114.qm@web50602.mail.re2.yahoo.com> References: <26175.9114.qm@web50602.mail.re2.yahoo.com> Message-ID: <94E05924-02A7-421B-ACE0-0F65D9F5049E@nest.sns.it> On Apr 23, 2007, at 3:24 , vu ongphuong wrote: > Bond overlap population is defined as a measure of the > shared electron density between 2 atoms. this is not a definition: this is what the bond overlap population is supposed to be. > I find no PWSCF-examples calculating overlap population. > My question is whether PWSCF can calculate this quantity it depends on the definition of "bond overlap population" Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Mon Apr 23 09:30:45 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 23 Apr 2007 09:30:45 +0200 Subject: [Pw_forum] kpoint_grid & npk In-Reply-To: <003e01c7856d$db254210$0301a8c0@w2agz> References: <003e01c7856d$db254210$0301a8c0@w2agz> Message-ID: On Apr 23, 2007, at 8:08 , Paul M. Grant wrote: > To All?where is the default for npk set that is fed to kpoint_grid > and what governs the choice? For a MP grid of 100^3 that?s only a > million! (just kidding ;)?) there is no "default" for the number of k-points; there is a maximum allowed number of k-points, defined in Modules/parameters.f90 Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From ongphuongvu at yahoo.com Mon Apr 23 10:31:17 2007 From: ongphuongvu at yahoo.com (vu ongphuong) Date: Mon, 23 Apr 2007 01:31:17 -0700 (PDT) Subject: [Pw_forum] Question on bond overlap population calculation Message-ID: <659664.49065.qm@web50601.mail.re2.yahoo.com> Dear Professor Paolo Giannozzi, Specifically, my system is SrTiO_{3}; It is well-known that there is hybridization between O-2p and Ti-3d. What I want to calculate is the number of shared electrons (or states) between Ti and Oxygen or any quantity that give information on how strong the bond is (that what I called, may be not exactly, "bond overlap population") Please tell me which kind of "bond overlap population" PWSCF can calculate, Thank you very much, Ong Phuong Vu ------------------------------------ On Apr 23, 2007, at 3:24 , vu ongphuong wrote: > Bond overlap population is defined as a measure of the > shared electron density between 2 atoms. this is not a definition: this is what the bond overlap population is supposed to be. > I find no PWSCF-examples calculating overlap population. > My question is whether PWSCF can calculate this quantity it depends on the definition of "bond overlap population" Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy --------------------------------- Ahhh...imagining that irresistible "new car" smell? Check outnew cars at Yahoo! Autos. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070423/d1086a81/attachment.htm From paulatto at sissa.it Mon Apr 23 12:53:47 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Mon, 23 Apr 2007 12:53:47 +0200 (CEST) Subject: [Pw_forum] Question on bond overlap population calculation In-Reply-To: <659664.49065.qm@web50601.mail.re2.yahoo.com> References: <659664.49065.qm@web50601.mail.re2.yahoo.com> Message-ID: <13008.82.54.224.73.1177325627.squirrel@webmail.sissa.it> On Mon, April 23, 2007 10:31, vu ongphuong wrote: > Dear Professor Paolo Giannozzi, > Dear Ong Phuong Vu, I'm not professor Giannozzi and my knowledge of PWSCF is only a small bit of Giannozzi's, still I hope I can give you a meaningful answer. > Specifically, my system is SrTiO_{3}; It is well-known that there is > hybridization between O-2p and Ti-3d. What I want to calculate is the > number of shared electrons (or states) between Ti and Oxygen or any > quantity that give information on how strong the bond is (that what I > called, may be not exactly, "bond overlap population") > I don't know if PWSCF can do what you wish: DFT electrons are not real electrons, but they only have the real charge density and total energy (+ some approximation). Furthermore electronic states are represented in on a plane wave basis in a periodic boundary condition formalism, so they are completely delocalized and it doesn't make much sense to ask to what atom do they belong. > Please tell me which kind of "bond overlap population" PWSCF can > calculate, On the other hand I think that Wannier functions formalism may help you: Wannier functions are actually localized and their position (and occupancy, for conductors) can give some qualitative informations on interatomic bonds. I think you should have a look at wannier90 or Wannier-transport and see if it fits your needs. > Thank you very much, > > Ong Phuong Vu Bye -- Lorenzo Paulatto +39 040 3787 312 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From asen at iitk.ac.in Mon Apr 23 15:03:36 2007 From: asen at iitk.ac.in (Sen) Date: Mon, 23 Apr 2007 18:33:36 +0530 (IST) Subject: [Pw_forum] geometry optimizaion-how to make it faster Message-ID: <50182.140.109.214.4.1177333416.squirrel@newwebmail.iitk.ac.in> Dear all, As I am trying to get the optimized geometry of an 192-atom system,my parallel-run gets slower as the total force reads the following: Total force = 194.872598 Total SCF correction = 0.008989 Total force = 186.362843 Total SCF correction = 0.007976 Total force = 184.690859 Total SCF correction = 0.004865 Total force = 179.229678 Total SCF correction = 0.005298 Total force = 173.830737 Total SCF correction = 0.005438 Though the run still continues, I fear, it would take months to reach something like "End of BFGS Geometry Optimization". I am using the following input (briefly given): &CONTROL calculation = "relax", restart_mode = 'from_scratch', prefix = "nanomol", pseudo_dir = "/work1/asen/pwscf/newwork1/pseudo", outdir = "/work1/asen/tmp", / &SYSTEM ibrav = 6, celldm(1) = 22.0, celldm(3) = 3.8, nat = 192, ntyp = 4, ecutwfc = 25.0, ecutrho = 200.0, occupations='smearing', smearing='Gaussian broadening', degauss=0.01, / &ELECTRONS diagonalization='cg', conv_thr = 1.0e-6, mixing_beta = 0.1, / &IONS pot_extrapolation = "second_order", wfc_extrapolation = "second_order", -------- Au 18.601360 16.000000 0.000000 Au 18.601360 16.000000 54.570496 --------------- -------------- Au 15.451947 13.457026 81.183342 ----------- Au 17.718653 14.047231 80.089668 ------------- -------------- S 15.998219 14.326591 31.279226 C 16.001303 16.565260 33.968327 H 17.675220 17.770269 33.820061 H 14.330640 17.774763 33.819920 --------------- The atomic positions are okay, however, the individual atom forcee are coming like this: atom 1 type 1 force = 8.78530224 1.13988509 -14.03979912 atom 2 type 1 force = 12.01091279 1.31202407 -13.48731135 atom 3 type 1 force = 9.69708845 11.26363626 -5.68058137 atom 4 type 1 force = 9.72514514 11.34982112 -4.79680588 atom 5 type 1 force = -2.54144525 8.89141511 -13.86762764 atom 6 type 1 force = -3.12157312 11.70097423 -12.31860197 atom 7 type 1 force = 10.21942251 9.52077429 -0.85911975 atom 8 type 1 force = 9.81543682 9.46159249 -0.73870247 atom 9 type 1 force = -1.32878929 12.88147452 2.62586847 atom 10 type 1 force = -1.32982873 14.40215166 1.69396997 atom 11 type 1 force = -13.51902374 6.88366314 -5.08150330 atom 12 type 1 force = -13.53147027 6.74037823 -5.25299321 ----and so on. I would greatly appreciate your valuable suggestions on what would be the most effecive way in handling such situations. Best regards, Sen From degironc at sissa.it Tue Apr 24 14:17:15 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Tue, 24 Apr 2007 14:17:15 +0200 (CEST) Subject: [Pw_forum] geometry optimizaion-how to make it faster In-Reply-To: <50182.140.109.214.4.1177333416.squirrel@newwebmail.iitk.ac.in> References: <50182.140.109.214.4.1177333416.squirrel@newwebmail.iitk.ac.in> Message-ID: On Mon, 23 Apr 2007, Sen wrote: ... > -------- > Au 18.601360 16.000000 0.000000 > Au 18.601360 16.000000 54.570496 > --------------- > -------------- > Au 15.451947 13.457026 81.183342 > ----------- > Au 17.718653 14.047231 80.089668 > ------------- > -------------- > S 15.998219 14.326591 31.279226 > C 16.001303 16.565260 33.968327 > H 17.675220 17.770269 33.820061 > H 14.330640 17.774763 33.819920 > --------------- > The atomic positions are okay, however, the individual atom forcee are > coming like this: > atom 1 type 1 force = 8.78530224 1.13988509 -14.03979912 > atom 2 type 1 force = 12.01091279 1.31202407 -13.48731135 > atom 3 type 1 force = 9.69708845 11.26363626 -5.68058137 > atom 4 type 1 force = 9.72514514 11.34982112 -4.79680588 > atom 5 type 1 force = -2.54144525 8.89141511 -13.86762764 > atom 6 type 1 force = -3.12157312 11.70097423 -12.31860197 > atom 7 type 1 force = 10.21942251 9.52077429 -0.85911975 > atom 8 type 1 force = 9.81543682 9.46159249 -0.73870247 > atom 9 type 1 force = -1.32878929 12.88147452 2.62586847 > atom 10 type 1 force = -1.32982873 14.40215166 1.69396997 > atom 11 type 1 force = -13.51902374 6.88366314 -5.08150330 > atom 12 type 1 force = -13.53147027 6.74037823 -5.25299321 > ----and so on. > I would greatly appreciate your valuable suggestions on what would be the > most effecive way in handling such situations. > Best regards, > Sen I would be very surprised if atomic positions that "are ok" produce such forces ... which tyme of ATOMIC_POSITIONS have you chosen (alat, crystal, bohr, angstrom) ? the default is alat If you have the xcrysden code installed you can visualize your structure by giving the command prompt> xcrysten --pwi your_input_file.in Hope this helps stefano > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From lanhaiping at gmail.com Tue Apr 24 14:53:39 2007 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 24 Apr 2007 20:53:39 +0800 Subject: [Pw_forum] About valence bands maximum In-Reply-To: <1176973063.12937.44.camel@srmp19.saclay.cea.fr> References: <1176973063.12937.44.camel@srmp19.saclay.cea.fr> Message-ID: Thank you, Guido. Regards, H.P On 4/19/07, roma wrote: > > Dear Ian Haiping, > > it is true that seldom papers about defects discuss this point, and I am > probably guilty of this too. My understanding is that your expression > for the valence band top of the defect is correct, but then it depends > on how you use it. What has been omitted according to the person that > told you this? > The crucial quantity is [ -] (let us call it > DeltaV(D,q) ). Long ago I did a small program to calculate the so called > macroscopic average of the potential (Balderesci, Baroni,Resta, PRL v61, > p734, 1988) in the supercell: it was ok for a defect in a simple metal > (I could find the "bulklike" region) but for the system I was interested > in, SiO2, the oscillations of the macroscopic average along the > supercell made it useless. Supercell too small? Problems due to the > symmetry of my supercell? I adopted a different strategy, I determined > the energy shift that maximises the overlap (or minimises the square > differences) of the density of states of the defected and perfect > crystal. This shift is for me DeltaV(D,q). > This, according to my experience, can be unambiguosly determined as long > as the relaxations are good (and k-point sampling is sufficient). > Then another question is to which extent the DeltaV term corrects for > the image interaction of charged defects, on which I would appreciate > also the feedback from others on the list. > Best regards, > > Guido > > > > > > On thursday April 12 Ian Haiping wrote: > > > I want to determin the transition enery of a defect level. It is > >important to align the valence bands maximum of different systems under > >investigation. Many works has claimed valence bands maximum alignments > >were performed but no technique details are given. > > > >For my understanding, > > > >I thougt the valence bands maxium of charge defect systems could be > >determined by aligning its average potential > > > >to pure perfect system's potential, e.g \epsilon_{VBM}.(Defects,q) = > >\epsilon_{VBM}(0) + [ -]. But some person told me such > >expression has omitted something. So it really confused me much. As > >far as i know, the potential of system could be only determined to > >some constant . i thought the alignment of average potentials of two > >systems could take into account this arbitrariness.. I am not very > >certain about this VBM alignment right now Would you give me some > >comments and clarify my understanding ? > > > > > -- > Guido Roma -- CEA-Saclay - DEN/DMN/SRMP Bat.520/13 > Phone: [+33]-1-69081857 -- Fax: ...6867 -- Mobile: [+33]-6-20069085 > > _______________________________________________ > 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070424/8a3f5975/attachment.htm From lanhaiping at gmail.com Tue Apr 24 15:13:19 2007 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 24 Apr 2007 21:13:19 +0800 Subject: [Pw_forum] Relativistic GGA calculations In-Reply-To: <1176751069.3081.25.camel@localhost.localdomain> References: <1176751069.3081.25.camel@localhost.localdomain> Message-ID: Dear Lazaro, You should do some search around the forum .... the answer to your question is "yes"... On 4/17/07, Lazaro Calderin wrote: > > Hello every one, > > Could some one tell me if it is possible to perform full and/or scalar > relativistic GGA calculations with PWscf? > > Thanks, Lazaro > > _______________________________________________ > 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070424/aa8af307/attachment.htm From hande at newton.physics.metu.edu.tr Tue Apr 24 15:34:58 2007 From: hande at newton.physics.metu.edu.tr (Hande Ustunel) Date: Tue, 24 Apr 2007 16:34:58 +0300 (EEST) Subject: [Pw_forum] magnetization and lowdin charges Message-ID: Dear pwscf users, I have a question concerning Lowdin analysis and the total magnetic moment. I read on the archives of the forum that one can obtain the polarization on each individiual atom by running projwfc.x and the reported polarization should be the magnetic moment on each atom. I was wondering how, in this case, the units work. Reading the forum I got the impression that polarization should directly give the magnetization (in Bohr magneton) without any unit conversion. ( Reading the code, it looks like the differences between up and down densities is multiplied by something like a volume (I think) to get the magnetization. ) If that is the case, then if I add up all the polarizations in the output of the Lowdin analysis, I should more or less get the "total magnetization" reported by the spin-polarized output of pw.x. Same argument should also apply to "absolute magnetization". It turns out for me that the sum does add up to the total and abs magnetizations in one of my calculations and it doesn't in two others. It's either the one that works that is just a fluke or there is a unit conversion issue that I couldn't see from the code or the archives. I hope I've made it clear enough. Thank you very much in advance for any help. Best regards, Hande -- 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 niuli1978 at yahoo.com.cn Tue Apr 24 16:08:34 2007 From: niuli1978 at yahoo.com.cn (li niu) Date: Tue, 24 Apr 2007 22:08:34 +0800 (CST) Subject: [Pw_forum] on the finite electric field method Message-ID: <74412.77822.qm@web15006.mail.cnb.yahoo.com> Dear all , I want to investigate Raman spectra in the case of large disordered system from first principles. I saw the exmple30 and 31, but they only gave the dielectric constant and Born effective charge by Moden polarizability theory. How can I obtain Raman coupling tensor by numerically calculating second derivatives of the atomic forces with respect to the electric field? [in RAMAN SPECTRA OF DISORDERED OXIDES FROM FIRST PRINCIPLES written by PAOLO UMARI]? I did not find those from the forum. Thank you in advance! Niu Li --------------------------------- ????????3.5G???20M??? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070424/debbda22/attachment.htm From akohlmey at cmm.chem.upenn.edu Tue Apr 24 17:06:07 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Tue, 24 Apr 2007 11:06:07 -0400 (EDT) Subject: [Pw_forum] magnetization and lowdin charges In-Reply-To: Message-ID: On Tue, 24 Apr 2007, Hande Ustunel wrote: HU> Dear pwscf users, dear hande, HU> I have a question concerning Lowdin analysis and the total magnetic HU> moment. I read on the archives of the forum that one can obtain the HU> polarization on each individiual atom by running projwfc.x and the HU> reported polarization should be the magnetic moment on each atom. I please note, that due the difference in nature of a plane wave basis set (which covers all space) and a local atomic basis set (you project on the pseudowavefunctions stored in the pseudopotentials), there will always be some part of the wavefunction unassigned, particularly if your wavefunction is very delocalized. HU> was wondering how, in this case, the units work. Reading the forum I HU> got the impression that polarization should directly give the HU> magnetization (in Bohr magneton) without any unit conversion. ( HU> Reading the code, it looks like the differences between up and down HU> densities is multiplied by something like a volume (I think) to get HU> the magnetization. ) If that is the case, then if I add up all the HU> polarizations in the output of the Lowdin analysis, I should more or HU> less get the "total magnetization" reported by the spin-polarized HU> output of pw.x. Same argument should also apply to "absolute HU> magnetization". It turns out for me that the sum does add up to the HU> total and abs magnetizations in one of my calculations and it doesn't HU> in two others. It's either the one that works that is just a fluke or HU> there is a unit conversion issue that I couldn't see from the code or no. the diference should be somewhat small, but is almost always noticable and it depends a lot on the type of system you are looking at. cheers, axel. HU> the archives. I hope I've made it clear enough. Thank you very much in HU> advance for any help. HU> HU> Best regards, HU> Hande HU> HU> -- ======================================================================= 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 calderin at physics.queensu.ca Tue Apr 24 20:34:43 2007 From: calderin at physics.queensu.ca (Lazaro Calderin) Date: Tue, 24 Apr 2007 20:34:43 +0200 Subject: [Pw_forum] Re: Relativistic GGA ? Message-ID: <1177439684.2623.110.camel@localhost.localdomain> Thanks Andrea. I have been trying ... but I haven't found a way to run relativistic GGA jobs with pwscf. Below are the input options and the corresponding errors that I get for some quick and dirty tests for fully relativistic GGA jobs. The following tests were done for Hg using TM and US fully relativistic GGA(PBE) pseudopotentials and PWSCF version 3.2. ------------------------------------------------------------------------ Using: GGA(PBE) fully relativistic Troullier-Martins pseudopotential for Hg 1) noncolin=.true., lspinorb=.true., starting_magnetization(1)=0.0, %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from stres_gradcorr : error # 1 noncollinear stress + GGA not implemented %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% NOTE: The same test (same input file) but using relativistic LDA psp runs without errors. 2) noncolin=.false., lspinorb=.true., starting_magnetization(1)=0.0, %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from setup : error # 1 wrong beta functions %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 3) noncolin=.false., lspinorb=.false., starting_magnetization(1)=0.0, %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from setup : error # 1 wrong beta functions %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 4) nspin=1, %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from setup : error # 1 wrong beta functions %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ----------------------------------------------------------------------- Using: GGA(PBE) fully relativistic ultra soft pseudopotential for Hg (the same happens with BLYP) 1) noncolin=.true., lspinorb=.true., starting_magnetization(1)=0.0, %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from stres_gradcorr : error # 1 noncollinear stress + GGA not implemented %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% NOTE: The same test (same input file) but using relativistic LDA psp runs without errors. 2) ncolin=.false., lspinorb=.true., starting_magnetization(1)=0.0, %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from setup : error # 1 US j-average not yet implemented %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 3) ncolin=.false., lspinorb=.false., starting_magnetization(1)=0.0, %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from setup : error # 1 US j-average not yet implemented %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% So far it seems that: I) Fully Relativistic GGA is not implemented yet for spin polarized systems. II) It is not implemented yet neither for non-spin polarized systems using US pseudopotentials. III) For Troullier-Martins kind of psps I get an error related to the KB projectors ( wrong beta functions ) for the non-spin polarized calculations. So it could be that it is possible to do the fully relativistic GGA in this case, but I must figure out first why the "wrong beta functions" error in this case. Is that the actual situation or I am missing something? I will do the same kind of tests for the scalar relativistic options soon. Thanks, L. >It is possible but not well tested. It should work well >for nonmagnetic systems, while it could have >problems for magnetic systems. In any case if you find >strange behavior please report it to the forum. > >Andrea > > >On Wed, 2007-04-18 at 16:00 +0200, Lazaro Calderin wrote: >> Hi guys, >> >> Is it possible to do relativistic GGA calculations with pwscf? >> >> Thanks, Lazaro From akohlmey at cmm.chem.upenn.edu Tue Apr 24 20:13:57 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Tue, 24 Apr 2007 14:13:57 -0400 (EDT) Subject: [Pw_forum] Re: Relativistic GGA ? In-Reply-To: <1177439684.2623.110.camel@localhost.localdomain> Message-ID: On Tue, 24 Apr 2007, Lazaro Calderin wrote: LC> Thanks Andrea. LC> LC> I have been trying ... but I haven't found a way to run relativistic GGA LC> jobs with pwscf. LC> LC> Below are the input options and the corresponding errors that I get for LC> some quick and dirty tests for fully relativistic GGA jobs. LC> LC> The following tests were done for Hg using TM and US fully relativistic LC> GGA(PBE) pseudopotentials and PWSCF version 3.2. LC> LC> ------------------------------------------------------------------------ LC> Using: GGA(PBE) fully relativistic Troullier-Martins pseudopotential for LC> Hg LC> LC> 1) LC> noncolin=.true., LC> lspinorb=.true., LC> starting_magnetization(1)=0.0, LC> LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> from stres_gradcorr : error # 1 LC> noncollinear stress + GGA not implemented LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% note, this message comes obviously from the subroutine calculating the stress tensor. so if you can go without, it might still work, cheers, axel. LC> LC> NOTE: The same test (same input file) but using relativistic LDA psp LC> runs without errors. LC> LC> 2) LC> noncolin=.false., LC> lspinorb=.true., LC> starting_magnetization(1)=0.0, LC> LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> from setup : error # 1 LC> wrong beta functions LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> LC> 3) LC> noncolin=.false., LC> lspinorb=.false., LC> starting_magnetization(1)=0.0, LC> LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> from setup : error # 1 LC> wrong beta functions LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> LC> 4) LC> nspin=1, LC> LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> from setup : error # 1 LC> wrong beta functions LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> LC> LC> ----------------------------------------------------------------------- LC> Using: GGA(PBE) fully relativistic ultra soft pseudopotential for Hg LC> (the same happens with BLYP) LC> LC> LC> 1) LC> noncolin=.true., LC> lspinorb=.true., LC> starting_magnetization(1)=0.0, LC> LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> from stres_gradcorr : error # 1 LC> noncollinear stress + GGA not implemented LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> LC> NOTE: The same test (same input file) but using relativistic LDA psp LC> runs without errors. LC> LC> 2) LC> ncolin=.false., LC> lspinorb=.true., LC> starting_magnetization(1)=0.0, LC> LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> from setup : error # 1 LC> US j-average not yet implemented LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> LC> LC> 3) LC> ncolin=.false., LC> lspinorb=.false., LC> starting_magnetization(1)=0.0, LC> LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> from setup : error # 1 LC> US j-average not yet implemented LC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LC> LC> LC> So far it seems that: LC> LC> I) Fully Relativistic GGA is not implemented yet for spin polarized LC> systems. LC> LC> II) It is not implemented yet neither for non-spin polarized systems LC> using US pseudopotentials. LC> LC> III) For Troullier-Martins kind of psps I get an error related to the KB LC> projectors ( wrong beta functions ) for the non-spin polarized LC> calculations. So it could be that it is possible to do the fully LC> relativistic GGA in this case, but I must figure out first why the LC> "wrong beta functions" error in this case. LC> LC> Is that the actual situation or I am missing something? LC> LC> I will do the same kind of tests for the scalar relativistic options LC> soon. LC> LC> Thanks, L. LC> LC> LC> >It is possible but not well tested. It should work well LC> >for nonmagnetic systems, while it could have LC> >problems for magnetic systems. In any case if you find LC> >strange behavior please report it to the forum. LC> > LC> >Andrea LC> > LC> > LC> >On Wed, 2007-04-18 at 16:00 +0200, Lazaro Calderin wrote: LC> >> Hi guys, LC> >> LC> >> Is it possible to do relativistic GGA calculations with pwscf? LC> >> LC> >> Thanks, Lazaro LC> LC> LC> _______________________________________________ LC> Pw_forum mailing list LC> Pw_forum at pwscf.org LC> http://www.democritos.it/mailman/listinfo/pw_forum LC> -- ======================================================================= 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 niuli1978 at yahoo.com.cn Wed Apr 25 00:24:11 2007 From: niuli1978 at yahoo.com.cn (li niu) Date: Wed, 25 Apr 2007 06:24:11 +0800 (CST) Subject: [Pw_forum] optimize the amorphour structure Message-ID: <252834.4929.qm@web15002.mail.cnb.yahoo.com> Dear all, I want to use a given amorphous network that generated by liquid quench method using Car-Parrinello molecular dynamics to investigate its vibrational properties.I know each calculation should be performed at the respective optimized lattice constant. Need I optimize this structure ? If the answer is yes, I use the relax, vc-relax or others? Thanks! Niu Liu --------------------------------- ????????3.5G???20M??? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070425/69484fcb/attachment.htm From ghaderyan at gmail.com Wed Apr 25 00:24:40 2007 From: ghaderyan at gmail.com (saman ghaderyan) Date: Wed, 25 Apr 2007 01:54:40 +0330 Subject: [Pw_forum] phonon freq Message-ID: thanks dear axel i searched and i found that i have to increase k-point,i will exam it, but why acoustic freq have effected ,(not optic freq at (0 0 0 ) thanks -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070425/fa0938e4/attachment.htm From hongyi.zhao at gmail.com Wed Apr 25 05:09:09 2007 From: hongyi.zhao at gmail.com (Hongyi Zhao) Date: Wed, 25 Apr 2007 11:09:09 +0800 Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. Message-ID: <20070425030909.GA14354@debian-sid-hy.net> Hi all, Now, I download the espresso-3.2.tar.gz and its patch. Tho I can successfully configure it, the make command will give me the following error messages: ------------- Fatal Error: File 'mpif.h' not found make[2]: *** [iotk_error.o] error 1 make[2]: Leaving directory `/home/hongyi/Desktop/espresso-3.2/iotk/src' make[1]: *** [libiotk.a] error 2 make[1]: Leaving directory `/home/hongyi/Desktop/espresso-3.2/iotk' make: *** [libiotk] error 2 ------------- Would anyone give some hints? P.S. My operating system is Debian Sid with gcc 4.1.2 20061115 and g95. Thanks in advance. --- Hongyi Zhao GnuPG DSA: 0xD108493B From cao at qtp.ufl.edu Wed Apr 25 06:03:33 2007 From: cao at qtp.ufl.edu (Chao Cao) Date: Wed, 25 Apr 2007 00:03:33 -0400 Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: <20070425030909.GA14354@debian-sid-hy.net> References: <20070425030909.GA14354@debian-sid-hy.net> Message-ID: <462ED315.1030809@qtp.ufl.edu> If you have mpif90, make sure in make.sys your FC is set to mpif90; otherwise issue "locate mpif.h" under console to find the path and then add the path to your FFLAGS like -I/PATH/TO/THE/FILE. In the later case, you might have bunch of error during linking, so I'll recommend you find mpif90 and mpicc. In case you need serial version, use ./configure --disable-parallel Good luck Chao Cao Hongyi Zhao wrote: > Hi all, > > Now, I download the espresso-3.2.tar.gz and its patch. Tho I can > successfully configure it, the make command will give me the following > error messages: > > ------------- > Fatal Error: File 'mpif.h' not found > make[2]: *** [iotk_error.o] error 1 > make[2]: Leaving directory `/home/hongyi/Desktop/espresso-3.2/iotk/src' > make[1]: *** [libiotk.a] error 2 > make[1]: Leaving directory `/home/hongyi/Desktop/espresso-3.2/iotk' > make: *** [libiotk] error 2 > ------------- > > Would anyone give some hints? > > P.S. My operating system is Debian Sid with gcc 4.1.2 20061115 and g95. > > Thanks in advance. > --- > Hongyi Zhao > GnuPG DSA: 0xD108493B > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From hongyi.zhao at gmail.com Wed Apr 25 06:38:37 2007 From: hongyi.zhao at gmail.com (Hongyi Zhao) Date: Wed, 25 Apr 2007 12:38:37 +0800 Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: <462ED315.1030809@qtp.ufl.edu> References: <20070425030909.GA14354@debian-sid-hy.net> <462ED315.1030809@qtp.ufl.edu> Message-ID: <20070425043837.GA26650@debian-sid-hy.net> On Wed, Apr 25, 2007 at 12:03:33AM -0400, Chao Cao wrote: > If you have mpif90, make sure in make.sys your FC is set to mpif90; I've the followings and so many others alike: -------- mpicc mpif77 mpif77.lam mpif77.mpich mpif77.mpich-mpd mpif77.mpich-shmem mpif77.openmpi mpif90.openmpi -------- According your advices, should I configure pwscf like this: ./configure FC=mpicc or ./configure FC=mpif90.openmpi In fact, in this way, I will meet the same error messages. > otherwise issue "locate mpif.h" under console to find the path and then > add the path to your FFLAGS like -I/PATH/TO/THE/FILE. I've mpif.h in many locations on my system, such as, /usr/include, /usr/lib/mpich/include/, but the *locate mpif.h* command give me nothing, why? P.S. I have poor knowledge about mpi. In my case, my machine is just a standalone computer, not a member of a cluster, should I use mpi or not? Thanks in advance. --- Hongyi Zhao GnuPG DSA: 0xD108493B From yukihiro_okuno at fujifilm.co.jp Wed Apr 25 07:32:38 2007 From: yukihiro_okuno at fujifilm.co.jp (yukihiro_okuno at fujifilm.co.jp) Date: Wed, 25 Apr 2007 14:32:38 +0900 Subject: [Pw_forum] what is w1 and w2 for bfgs relaxed calculation Message-ID: Dear PWscf users. I met not conserved force in relaxed calculation for slightly titled crystal structure for PZT in order to get e15 piezoelectric calculation. So I want to remedy this problem. In relaxed calculation in bfgs scheme, I set parametr w1 and w2 by default value, but what is this parameter means? and if this parameter is important for improve the convergence for force, how should i change this value when the force is not conserved. Sincerely, From cao at qtp.ufl.edu Wed Apr 25 07:49:45 2007 From: cao at qtp.ufl.edu (Chao Cao) Date: Wed, 25 Apr 2007 01:49:45 -0400 Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: <20070425043837.GA26650@debian-sid-hy.net> References: <20070425030909.GA14354@debian-sid-hy.net> <462ED315.1030809@qtp.ufl.edu> <20070425043837.GA26650@debian-sid-hy.net> Message-ID: <462EEBF9.9050105@qtp.ufl.edu> Hongyi Zhao wrote: > On Wed, Apr 25, 2007 at 12:03:33AM -0400, Chao Cao wrote: > >> If you have mpif90, make sure in make.sys your FC is set to mpif90; >> > > I've the followings and so many others alike: > > -------- > mpicc > mpif77 > mpif77.lam > mpif77.mpich > mpif77.mpich-mpd > mpif77.mpich-shmem > mpif77.openmpi > mpif90.openmpi > -------- > > According your advices, should I configure pwscf like this: > > ./configure FC=mpicc > > or > > ./configure FC=mpif90.openmpi > That depend on which MPI you are using. Find it out using "which mpirun" and you can see the PATH to the MPI version you are actually running, and your FC should be the one in that PATH. Also, from what you said, it seems to me that the only reasonable compiler is mpif90.openmpi since PWSCF is a fortran 90 code. Therefore I would recommend you use FC=mpif90.openmpi ./configure Or, you could simply modify the make.sys file so that FC is set to mpif90.openmpi in that file. > In fact, in this way, I will meet the same error messages. > If this does not solve your problem, it probably is because of your PATH/LD_LIBRARY_PATH environment variables. You may need to change them so that the default MPI environment is openmpi. However, it's still wierd. Put your result of "mpif90.openmpi -show" > >> otherwise issue "locate mpif.h" under console to find the path and then >> add the path to your FFLAGS like -I/PATH/TO/THE/FILE. >> > > I've mpif.h in many locations on my system, such as, /usr/include, > /usr/lib/mpich/include/, but the *locate mpif.h* command give me > nothing, why? > hmm, that's not your fault. I guess the system does not maintain a indexdb, or the admin forgot to put updatedb in crontab? > P.S. I have poor knowledge about mpi. In my case, my > machine is just a standalone computer, not a member of a cluster, should > I use mpi or not? > > That means do not use MPI. MPI is for cluster only. hehe, that solves everything. BTW, you should put this last sentence in front of everything...:p > Thanks in advance. > --- > Hongyi Zhao > GnuPG DSA: 0xD108493B > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > Best, Chao Cao From lanhaiping at gmail.com Wed Apr 25 08:14:35 2007 From: lanhaiping at gmail.com (lan haiping) Date: Wed, 25 Apr 2007 14:14:35 +0800 Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: <20070425030909.GA14354@debian-sid-hy.net> References: <20070425030909.GA14354@debian-sid-hy.net> Message-ID: Dear Hongyi, I thought your problem is related with FC setting.. For my experience, when intalling mpich package, we should set FC=ifort or FC=pgf90 depending on your compilor... Regards, H.P On 4/25/07, Hongyi Zhao wrote: > > Hi all, > > Now, I download the espresso-3.2.tar.gz and its patch. Tho I can > successfully configure it, the make command will give me the following > error messages: > > ------------- > Fatal Error: File 'mpif.h' not found > make[2]: *** [iotk_error.o] error 1 > make[2]: Leaving directory `/home/hongyi/Desktop/espresso-3.2/iotk/src' > make[1]: *** [libiotk.a] error 2 > make[1]: Leaving directory `/home/hongyi/Desktop/espresso-3.2/iotk' > make: *** [libiotk] error 2 > ------------- > > Would anyone give some hints? > > P.S. My operating system is Debian Sid with gcc 4.1.2 20061115 and g95. > > Thanks in advance. > --- > Hongyi Zhao > GnuPG DSA: 0xD108493B > _______________________________________________ > 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070425/a4c78cec/attachment.htm From hongyi.zhao at gmail.com Wed Apr 25 09:44:32 2007 From: hongyi.zhao at gmail.com (Hongyi Zhao) Date: Wed, 25 Apr 2007 15:44:32 +0800 Subject: [Pw_forum] Where can I get the latest cvs source of espresso and pwgui? Message-ID: <20070425074432.GA16931@debian-sid-hy.net> Hi all, I'm in a puzzle about the cvs source of the espresso and pwgui, I mean, where are the repositories? I just cann't find them. In addition, I use the espresso-3.2, but the pwgui only has the 3.1 version, are there some incompatibilities between them? Thanks in advance. --- Hongyi Zhao GnuPG DSA: 0xD108493B From marcel at physik.tu-berlin.de Wed Apr 25 10:30:39 2007 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Wed, 25 Apr 2007 10:30:39 +0200 (CEST) Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: <462EEBF9.9050105@qtp.ufl.edu> References: <20070425030909.GA14354@debian-sid-hy.net> <462ED315.1030809@qtp.ufl.edu> <20070425043837.GA26650@debian-sid-hy.net> <462EEBF9.9050105@qtp.ufl.edu> Message-ID: <20070425101932.Y79488@rosa.physik-pool.tu-berlin.de> >> > > That depend on which MPI you are using. Find it out using "which mpirun" and > you can see the PATH to the MPI version you are actually running, and your FC > should be the one in that PATH. > Also, from what you said, it seems to me that the only reasonable compiler is > mpif90.openmpi since PWSCF is a fortran 90 code. Therefore I would recommend > you use > FC=mpif90.openmpi ./configure > Or, you could simply modify the make.sys file so that FC is set to > mpif90.openmpi in that file. >> In fact, in this way, I will meet the same error messages. >> > If this does not solve your problem, it probably is because of your > PATH/LD_LIBRARY_PATH environment variables. You may need to change them so > that the default MPI environment is openmpi. However, it's still wierd. Put > your result of "mpif90.openmpi -show" >> >>> otherwise issue "locate mpif.h" under console to find the path and then >>> add the path to your FFLAGS like -I/PATH/TO/THE/FILE. >>> >> >> I've mpif.h in many locations on my system, such as, /usr/include, >> /usr/lib/mpich/include/, but the *locate mpif.h* command give me nothing, >> why? >> > hmm, that's not your fault. I guess the system does not maintain a indexdb, > or the admin forgot to put updatedb in crontab? >> P.S. I have poor knowledge about mpi. In my case, my machine is just a >> standalone computer, not a member of a cluster, should I use mpi or not? >> >> > That means do not use MPI. MPI is for cluster only. > hehe, that solves everything. > BTW, you should put this last sentence in front of everything...:p >> Thanks in advance. That statement is oversimplified ;-) nowadays there are multicore processors. We have here a dual processor dual core standalone computer, and parallizing performs quite well, at least better than over LAN. You can check if there are multiple by running one calculation intensive job, and then "top", if this job is only using 25 %, than there are 4 cores in your system. (I guess there are easier ways, e.g. "qhost" in SGE grid engine environments.) A small calculation intensive task for the linux calculator "bc" ----- $ bc define fac (x) { res = 1 while (x > 1) { res *=x; x -= 1; } return res; } fac (5000000) ------ This will keep the processor busy for a while :) Cheers Marcel >> --- >> Hongyi Zhao >> GnuPG DSA: 0xD108493B >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> > Best, > > > Chao Cao > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From jyraty at ulg.ac.be Wed Apr 25 10:25:33 2007 From: jyraty at ulg.ac.be (Jean-Yves Raty) Date: Wed, 25 Apr 2007 10:25:33 +0200 Subject: [Pw_forum] optimize the amorphour structure In-Reply-To: <252834.4929.qm@web15002.mail.cnb.yahoo.com> References: <252834.4929.qm@web15002.mail.cnb.yahoo.com> Message-ID: <002601c78713$4c27beb0$e4773c10$@ac.be> Hello, To my opinion, if you did not perform a constant-pressure quench of your liquid phase, it is necessary to relax the cell parameters of your ?frozen liquid? and the ionic positions. This is because usually the amorphous does not have the same density as the liquid. Now as you know, there is no guarantee the your frozen liquid will actually give you a reasonable model for the amorphous state, as the time scale for your quench is several orders of magnitude too short in comparison with a ?real quench?. Good luck JY ---------------------------------------------------------------------------- ------------------- Jean-Yves Raty FNRS - Universit? de Li?ge Tel +32 4 3663747 Condensed Matter Physics B5 Fax +32 4 3662990 4000 Sart-Tilman Belgium ---------------------------------------------------------------------------- ------------------- From: pw_forum-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] On Behalf Of li niu Sent: mercredi 25 avril 2007 0:24 To: pwscf Subject: [Pw_forum] optimize the amorphour structure Dear all, I want to use a given amorphous network that generated by liquid quench method using Car-Parrinello molecular dynamics to investigate its vibrational properties.I know each calculation should be performed at the respective optimized lattice constant. Need I optimize this structure ? If the answer is yes, I use the relax, vc-relax or others? Thanks! Niu Liu _____ ????????3.5G???20M??? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070425/f2d9964e/attachment.htm From marcel at physik.tu-berlin.de Wed Apr 25 10:33:04 2007 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Wed, 25 Apr 2007 10:33:04 +0200 (CEST) Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: References: <20070425030909.GA14354@debian-sid-hy.net> Message-ID: <20070425103133.S79488@rosa.physik-pool.tu-berlin.de> Dear Hongyi i forget one little thing in the top output, you have too look for this line Tasks: 69 total, 3 running, 66 sleeping, 0 stopped, 0 zombie Cpu(s): 49.0% us, 1.1% sy, 0.0% ni, 49.9% id, 0.0% wa, 0.0% hi, 0.0% si ^^^^^^^^^^^^^^^^^ Mem: 8311244k total, 1326936k used, 6984308k free, 8156k buffers Swap: 0k total, 0k used, 0k free, 1061072k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 8367 marcel 25 0 2612 1560 708 R 99.9 0.0 1:20.28 bc 3253 marcel 25 0 220m 209m 2968 R 99.9 2.6 3682:20 pw.x so here we have 4 cpus in one computer! Cheers Marcel On Wed, 25 Apr 2007, lan haiping wrote: > Dear Hongyi, > > I thought your problem is related with FC setting.. > For my experience, when intalling mpich package, we should set FC=ifort or > FC=pgf90 depending on your compilor... > > Regards, > H.P > > > On 4/25/07, Hongyi Zhao wrote: >> >> Hi all, >> >> Now, I download the espresso-3.2.tar.gz and its patch. Tho I can >> successfully configure it, the make command will give me the following >> error messages: >> >> ------------- >> Fatal Error: File 'mpif.h' not found >> make[2]: *** [iotk_error.o] error 1 >> make[2]: Leaving directory `/home/hongyi/Desktop/espresso-3.2/iotk/src' >> make[1]: *** [libiotk.a] error 2 >> make[1]: Leaving directory `/home/hongyi/Desktop/espresso-3.2/iotk' >> make: *** [libiotk] error 2 >> ------------- >> >> Would anyone give some hints? >> >> P.S. My operating system is Debian Sid with gcc 4.1.2 20061115 and g95. >> >> Thanks in advance. >> --- >> Hongyi Zhao >> GnuPG DSA: 0xD108493B >> _______________________________________________ >> 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 > From hongyi.zhao at gmail.com Wed Apr 25 11:27:57 2007 From: hongyi.zhao at gmail.com (Hongyi Zhao) Date: Wed, 25 Apr 2007 17:27:57 +0800 Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: <20070425101932.Y79488@rosa.physik-pool.tu-berlin.de> References: <20070425030909.GA14354@debian-sid-hy.net> <462ED315.1030809@qtp.ufl.edu> <20070425043837.GA26650@debian-sid-hy.net> <462EEBF9.9050105@qtp.ufl.edu> <20070425101932.Y79488@rosa.physik-pool.tu-berlin.de> Message-ID: <20070425092756.GA2248@debian-sid-hy.net> > A small calculation intensive task for the linux calculator "bc" > > ----- > $ bc > define fac (x) > { > res = 1 > while (x > 1) > { > res *=x; > x -= 1; > } > return res; > } > fac (5000000) > ------ Some hints on performing the above codes, will you? --- Hongyi Zhao GnuPG DSA: 0xD108493B From marcel at physik.tu-berlin.de Wed Apr 25 11:34:02 2007 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Wed, 25 Apr 2007 11:34:02 +0200 (CEST) Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: <20070425092756.GA2248@debian-sid-hy.net> References: <20070425030909.GA14354@debian-sid-hy.net> <462ED315.1030809@qtp.ufl.edu> <20070425043837.GA26650@debian-sid-hy.net> <462EEBF9.9050105@qtp.ufl.edu> <20070425101932.Y79488@rosa.physik-pool.tu-berlin.de> <20070425092756.GA2248@debian-sid-hy.net> Message-ID: <20070425113107.X80934@rosa.physik-pool.tu-berlin.de> It just calculates the faculty of 5000000 to keep the processor busy. So you can estimate the number of processor (see my mail before). Just type 'bc' in a shell and copy the lines below. >> A small calculation intensive task for the linux calculator "bc" >> >> ----- >> $ bc >> define fac (x) >> { >> res = 1 >> while (x > 1) >> { >> res *=x; >> x -= 1; >> } >> return res; >> } >> fac (5000000) >> ------ > > Some hints on performing the above codes, will you? > --- > Hongyi Zhao > GnuPG DSA: 0xD108493B > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From wlyim at puccini.che.pitt.edu Wed Apr 25 11:38:39 2007 From: wlyim at puccini.che.pitt.edu (wlyim at puccini.che.pitt.edu) Date: Wed, 25 Apr 2007 05:38:39 -0400 (EDT) Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: <20070425113107.X80934@rosa.physik-pool.tu-berlin.de> Message-ID: If you are running linux system, check "/proc/cpuinfo" William On Wed, 25 Apr 2007, Marcel Mohr wrote: > It just calculates the faculty of 5000000 to keep the processor busy. So > you can estimate the number of processor (see my mail before). > Just type 'bc' in a shell and copy the lines below. > > >> A small calculation intensive task for the linux calculator "bc" > >> > >> ----- > >> $ bc > >> define fac (x) > >> { > >> res = 1 > >> while (x > 1) > >> { > >> res *=x; > >> x -= 1; > >> } > >> return res; > >> } > >> fac (5000000) > >> ------ > > > > Some hints on performing the above codes, will you? > > --- > > Hongyi Zhao > > GnuPG DSA: 0xD108493B > > _______________________________________________ > > 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. Wai-Leung Yim Institut fuer Reine und Angewandte Chemie, Theoretische Chemie, Carl von Ossiezky Universtaet Oldenburg, 26129 Oldenburg, Germany Email: wlyim at puccini.che.pitt.edu Phone: +49-441-798-3950 (office) Fax: +49-441-798-3964 From hyello25 at gmail.com Wed Apr 25 11:31:56 2007 From: hyello25 at gmail.com (hyello25 Qgrui) Date: Wed, 25 Apr 2007 17:31:56 +0800 Subject: [Pw_forum] help: about the unit of force Message-ID: hi, what is the unit of "Total force"in the pwscf, and what does it mean? and does Ry/au equal ryd/bohr? for example: ! total energy = -15.19528681 ryd estimated scf accuracy < 0.00000010 ryd band energy sum = -3.00278508 ryd one-electron contribution = -1.90959454 ryd hartree contribution = 1.92327670 ryd xc contribution = -3.79218415 ryd ewald contribution = -11.41678483 ryd convergence has been achieved Forces acting on atoms (Ry/au): atom 1 type 1 force = -0.00024330 0.00000000 0.00000000 atom 2 type 1 force = 0.00024330 0.00000000 0.00000000 Total force = 0.000344 Total SCF correction = 0.000001 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070425/5c38e964/attachment.htm From jhashemifar at gmail.com Wed Apr 25 12:13:01 2007 From: jhashemifar at gmail.com (Javad hashemifar) Date: Wed, 25 Apr 2007 12:13:01 +0200 Subject: [Pw_forum] help: about the unit of force In-Reply-To: References: Message-ID: <2fb18d9f0704250313v1d405627j6464f857d67b87b7@mail.gmail.com> It seems that the Total force is the square root of the sum of the squares of all forces ! Total Force = sqrt ( sigme_i ( force_i ^ 2 ) ) Regards Javad Hashemifar On 4/25/07, hyello25 Qgrui wrote: > hi, > what is the unit of "Total force"in the pwscf, and what does it mean? > and does Ry/au equal ryd/bohr? > > for example: > > ! total energy = -15.19528681 ryd > estimated scf accuracy < 0.00000010 ryd > > band energy sum = -3.00278508 ryd > one-electron contribution = -1.90959454 ryd > hartree contribution = 1.92327670 ryd > xc contribution = -3.79218415 ryd > ewald contribution = -11.41678483 ryd > > convergence has been achieved > > Forces acting on atoms (Ry/au): > > atom 1 type 1 force = -0.00024330 0.00000000 0.00000000 > atom 2 type 1 force = 0.00024330 0.00000000 0.00000000 > > Total force = 0.000344 Total SCF correction = 0.000001 > > > -- ====================================== Seyed Javad Hashemifar, Ph.D. [current:] Tel:+49-203-3794743 Fax:+49-203-3794742 Fachbereich Physik (AG Kratzer), Universitat Duisburg-Essen 47048 Dusiburg, Germany [permanent:] Tel: +98-311-3912375 Fax: +98-311-3912376 Physics Department, Isfahan University of Technology 84156 Isfahan, Iran --------------------------------------------------------------------------- From marzari at MIT.EDU Wed Apr 25 14:42:37 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 25 Apr 2007 08:42:37 -0400 Subject: [Pw_forum] help: about the unit of force In-Reply-To: References: Message-ID: <462F4CBD.1010702@mit.edu> hyello25 Qgrui wrote: > and does Ry/au equal ryd/bohr? Yes ! 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 niuli1978 at yahoo.com.cn Wed Apr 25 15:11:19 2007 From: niuli1978 at yahoo.com.cn (li niu) Date: Wed, 25 Apr 2007 21:11:19 +0800 (CST) Subject: [Pw_forum] on optimize amorphous structure In-Reply-To: <20070425094703.5346.42199.Mailman@democritos.sissa.it> Message-ID: <329900.31097.qm@web15013.mail.cnb.yahoo.com> Dear Jean-Yves Raty, Thanks for your help! I don't know if I express myself clearly. I use a model structure, generated by CPMD. This model contains 64 atoms in a periodically repeated cubic cell, at the experimental density. The electric structure and optical properties have been previously calculated for the same model structure. I want to calculate vibrational properties for the same structure. In this case, need I optimize this structure ? If the answer is yes, do I use the relax, vc-relax or others? Thanks! Best! Niu Li --------------------------------- ????????3.5G???20M??? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070425/844bce5c/attachment.htm From clie at aphy.iphy.ac.cn Wed Apr 25 15:16:56 2007 From: clie at aphy.iphy.ac.cn (clie) Date: Wed, 25 Apr 2007 13:16:56 GMT Subject: [Pw_forum] problems about electron-phonon coupling In-Reply-To: <20070425094703.5346.42199.Mailman@democritos.sissa.it> References: <20070425094703.5346.42199.Mailman@democritos.sissa.it> Message-ID: <20070425131656.5485.eqmail@aphy.iphy.ac.cn> Dear professor, i have some difficulty about calculation electron-phonon couplin g constance. It seems that there are somethings wrong in k-point sample. The pro gram stoped and told "running the el-ph calculation....2". This is my input file. could you check it for me, is the k-point sample proper o r how to get proper k-point sample, thanks. cat > GeH4.scf.fit.in << EOF &control calculation='scf' restart_mode='from_scratch', prefix='GeH4', pseudo_dir = '$PSEUDO_DIR/', outdir='$TMP_DIR/' / &system ibrav= 8, celldm(1) =7.478,celldm(2)=0.76, celldm(3)=0.72, nat= 10, ntyp= 2, ecutwfc =60.0, occupations='smearing', smearing='methfessel-paxton', degauss=0.05, la2F = .true., / &electrons conv_thr = 1.0d-8 mixing_beta = 0.7 / ATOMIC_SPECIES Ge 72.61 Ge.pz-bhs.UPF H 1.00794 H.pz-van_ak.UPF ATOMIC_POSITIONS Ge 0.500000 0.000000 0.500000 Ge 0.000000 0.000000 0.000000 H 0.000000 0.388640 0.688119 H 0.500000 0.611360 0.188119 H 0.500000 0.388640 0.811881 H 0.000000 0.611360 0.311811 H 0.250000 0.357636 0.250000 H 0.250000 0.642364 0.750000 H 0.750000 0.642364 0.750000 H 0.750000 0.357636 0.250000 K_POINTS {automatic} 36 48 48 0 0 0 EOF $ECHO " running the scf calculation with dense k-point grid...\c" $PW_COMMAND < GeH4.scf.fit.in > GeH4.scf.fit.out $ECHO " done" # # SCF at k-mesh good enough for phonons # cat > GeH4.scf.in << EOF &control calculation='scf' restart_mode='from_scratch', prefix='GeH4', pseudo_dir = '$PSEUDO_DIR/', outdir='$TMP_DIR/' / &system ibrav= 8, celldm(1) =7.478,celldm(2)=0.76, celldm(3)=0.72, nat= 10, ntyp= 2, ecutwfc =60.0, occupations='smearing', smearing='methfessel-paxton', degauss=0.05 / &electrons conv_thr = 1.0d-8 mixing_beta = 0.7 / ATOMIC_SPECIES Ge 72.61 Ge.pz-bhs.UPF H 1.00794 H.pz-van_ak.UPF ATOMIC_POSITIONS Ge 0.500000 0.000000 0.500000 Ge 0.000000 0.000000 0.000000 H 0.000000 0.388640 0.688119 H 0.500000 0.611360 0.188119 H 0.500000 0.388640 0.811881 H 0.000000 0.611360 0.311811 H 0.250000 0.357636 0.250000 H 0.250000 0.642364 0.750000 H 0.750000 0.642364 0.750000 H 0.750000 0.357636 0.250000 K_POINTS {automatic} 12 16 16 0 0 0 EOF $ECHO " running the scf calculation...\c" $PW_COMMAND < GeH4.scf.in > GeH4.scf.out $ECHO " done" # cat > GeH4.elph.in << EOF Electron-phonon coefficients for GeH4 &inputph tr2_ph=1.0d-10, prefix='GeH4', fildvscf='GeH4dv', amass(1)=72.61, amass(2)=1.00794, outdir='$TMP_DIR/', fildyn='GeH4.dyn', elph=.true., trans=.true., ldisp=.true. nq1=3, nq2=4, nq3=4 / EOF $ECHO " running the el-ph calculation...\c" $PH_COMMAND < GeH4.elph.in > GeH4.elph.out $ECHO " done" # # q2r and matdyn # cat > q2r.in << EOF &input fildyn='GeH4.dyn', flfrc='GeH4444.fc', la2F=.true. / EOF $ECHO " running q2r...\c" $Q2R_COMMAND < q2r.in > q2r.out $ECHO " done" # # # cat > matdyn.in.freq << EOF &input amass(1)=72.61,amass(2)=1.00794, flfrc='GeH4444.fc', flfrq='GeH4444.freq', la2F=.true., dos=.false. / 8 0.00000 0.00000 0.00000 0.00000 0.00000 -0.69444 0.00000 -0.65789 0.00000 0.00000 -0.65789 -0.69444 -0.50000 0.00000 0.00000 -0.50000 0.00000 -0.69444 -0.50000 -0.65789 0.00000 -0.50000 -0.65789 -0.69444 EOF $ECHO " running matdyn for frequency calculation...\c" $MATDYN_COMMAND < matdyn.in.freq > matdyn.out.freq $ECHO " done" # # # cat > matdyn.in.dos << EOF &input amass(1)=72.61, amass(2)=1.00794, flfrc='GeH4444.fc', flfrq='GeH4444.freq', la2F=.true., dos=.true. fldos='phonon.dos', nk1=10, nk2=10, nk3=10, ndos=50 / EOF $ECHO " running matdyn for a2F(omega) calculation...\c" $MATDYN_COMMAND < matdyn.in.dos > matdyn.out.dos $ECHO " done" From matteo at umn.edu Wed Apr 25 16:42:43 2007 From: matteo at umn.edu (Matteo Cococcioni) Date: Wed, 25 Apr 2007 09:42:43 -0500 Subject: [Pw_forum] on optimize amorphous structure In-Reply-To: <329900.31097.qm@web15013.mail.cnb.yahoo.com> References: <329900.31097.qm@web15013.mail.cnb.yahoo.com> Message-ID: <462F68E3.4050006@umn.edu> Dear Li I totally agree with what Jean-Yves suggested. To realize how far your structure is from equilibrium a simple thing to do is looking at forces and stresses you can compute at the end of a scf run. Then, on the base of these results you will decide what to do. Matteo li niu wrote: > Dear Jean-Yves Raty, > > Thanks for your help! > I don't know if I express myself clearly. I use a model structure, > generated by CPMD. This model contains 64 atoms in a periodically > repeated cubic cell, at the experimental density. The electric > structure and optical properties have been previously calculated for > the same model structure. > I want to calculate vibrational properties for the same structure. In > this case, need I optimize this structure ? If the answer is yes, do I > use the relax, vc-relax or others? > Thanks! > > Best! > Niu Li > > ????????3.5G???20M??? From cao at qtp.ufl.edu Wed Apr 25 17:24:20 2007 From: cao at qtp.ufl.edu (Chao Cao) Date: Wed, 25 Apr 2007 11:24:20 -0400 Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: References: Message-ID: <462F72A4.1010004@qtp.ufl.edu> That's good point... Another question is, for multi-core system, shouldn't we use OpenMP over MPI, or we could easily by-pass MPI by enable thread in compilation? I didn't do test though... Chao Cao wlyim at puccini.che.pitt.edu wrote: > If you are running linux system, check "/proc/cpuinfo" > > William > > On Wed, 25 Apr 2007, Marcel Mohr wrote: > > >> It just calculates the faculty of 5000000 to keep the processor busy. So >> you can estimate the number of processor (see my mail before). >> Just type 'bc' in a shell and copy the lines below. >> >> >>>> A small calculation intensive task for the linux calculator "bc" >>>> >>>> ----- >>>> $ bc >>>> define fac (x) >>>> { >>>> res = 1 >>>> while (x > 1) >>>> { >>>> res *=x; >>>> x -= 1; >>>> } >>>> return res; >>>> } >>>> fac (5000000) >>>> ------ >>>> >>> Some hints on performing the above codes, will you? >>> --- >>> Hongyi Zhao >>> GnuPG DSA: 0xD108493B >>> _______________________________________________ >>> 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 akohlmey at cmm.chem.upenn.edu Wed Apr 25 17:13:03 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 25 Apr 2007 11:13:03 -0400 (EDT) Subject: [Pw_forum] Fatal Error: File 'mpif.h' not found. In-Reply-To: <462F72A4.1010004@qtp.ufl.edu> Message-ID: On Wed, 25 Apr 2007, Chao Cao wrote: CC> That's good point... CC> Another question is, for multi-core system, shouldn't we use OpenMP over CC> MPI, or we could easily by-pass MPI by enable thread in compilation? I CC> didn't do test though... quantum espresso has no direct openmp support. unless you have a _very_ much memory bandwidth impaired machine (e.g. a old pentium4 based dual core dual xeon processor) you will benefit _much_ more from MPI. i have not done any tests recently with quantum espresso, but with another plane-wave pseudopotential code on a dual-processor dual-core intel (woodcrest) machine, it turned out that using MPI with three (3!) mpi processes (i.e. let one processor core be available to the operating system) was the most efficient way to utilize the machine. the only way to enable openmp for QE is by using a threaded BLAS/LAPACK. i've done a few tests many years ago and it was successful on a cluster of dual athlon MP machines. for the rest, just running plain MPI was the best. this will change in the future when well be have quad/oct/???? core cpus. where making efficient use of the cache (through multiple threads doing overlapping access to the same block of physical memory will become an important optimization strategy). cheers, axel. CC> CC> Chao Cao CC> CC> wlyim at puccini.che.pitt.edu wrote: CC> > If you are running linux system, check "/proc/cpuinfo" CC> > CC> > William CC> > CC> > On Wed, 25 Apr 2007, Marcel Mohr wrote: CC> > CC> > CC> >> It just calculates the faculty of 5000000 to keep the processor busy. So CC> >> you can estimate the number of processor (see my mail before). CC> >> Just type 'bc' in a shell and copy the lines below. CC> >> CC> >> CC> >>>> A small calculation intensive task for the linux calculator "bc" CC> >>>> CC> >>>> ----- CC> >>>> $ bc CC> >>>> define fac (x) CC> >>>> { CC> >>>> res = 1 CC> >>>> while (x > 1) CC> >>>> { CC> >>>> res *=x; CC> >>>> x -= 1; CC> >>>> } CC> >>>> return res; CC> >>>> } CC> >>>> fac (5000000) CC> >>>> ------ CC> >>>> CC> >>> Some hints on performing the above codes, will you? CC> >>> --- CC> >>> Hongyi Zhao CC> >>> GnuPG DSA: 0xD108493B CC> >>> _______________________________________________ CC> >>> Pw_forum mailing list CC> >>> Pw_forum at pwscf.org CC> >>> http://www.democritos.it/mailman/listinfo/pw_forum CC> >>> CC> >>> CC> >> _______________________________________________ CC> >> Pw_forum mailing list CC> >> Pw_forum at pwscf.org CC> >> http://www.democritos.it/mailman/listinfo/pw_forum CC> >> CC> >> CC> > CC> > CC> CC> _______________________________________________ CC> Pw_forum mailing list CC> Pw_forum at pwscf.org CC> http://www.democritos.it/mailman/listinfo/pw_forum CC> -- ======================================================================= 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 Apr 25 18:56:51 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 25 Apr 2007 18:56:51 +0200 Subject: [Pw_forum] Where can I get the latest cvs source of espresso and pwgui? In-Reply-To: <20070425074432.GA16931@debian-sid-hy.net> References: <20070425074432.GA16931@debian-sid-hy.net> Message-ID: On Apr 25, 2007, at 9:44 , Hongyi Zhao wrote: > I'm in a puzzle about the cvs source of the espresso and pwgui, > I mean, where are the repositories? explained in the wiki documentation: http://www.quantum-espresso.org/wiki/index.php/ Developer_Manual#Using_CVS > In addition, I use the espresso-3.2, but the pwgui only has the 3.1 > version, are there some incompatibilities between them? there are a few, but unless you use exotic options, it shouldn't be a problem Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From hashem.yamani at gmail.com Wed Apr 25 20:10:01 2007 From: hashem.yamani at gmail.com (Hashem Al-Yamani) Date: Wed, 25 Apr 2007 16:10:01 -0200 Subject: [Pw_forum] iotk make error Message-ID: <82a593400704251110r7590dac4u882070ca8282b850@mail.gmail.com> Dear Sir; I've configured PWscf successfully ... but while doing the "make all" command I had the following error : Error: Dummy argument 'dat' at (1) cannot be INTENT(OUT) In file iotk_dat.spp:656 Main-server:/home/physlab/espresso-3.2 # make all test -d bin || mkdir bin if test -d iotk ; then \ ( cd iotk ; if test "make" = "" ; then make TLDEPS= libiotk.a ; \ else make TLDEPS= libiotk.a ; fi ) ; fi make[1]: Entering directory `/home/physlab/espresso-3.2/iotk' cd src ; make libiotk.a make[2]: Entering directory `/home/physlab/espresso-3.2/iotk/src' mpif90 -O3 -x f95-cpp-input -D__FFTW -D__USE_INTERNAL_FFTW -D__MPI -D__PARA -I../include -I./ -I../Modules -I../iotk/src -I../PW -I../PH -I../CPV -c iotk_dat+CHARACTER1_0.f90 In file iotk_dat.spp:656 CHARACTER (kind=iotk_defkind_character,len=len(dat)), allocatable :: tmpdat(:) 1 Error: Dummy argument 'dat' at (1) cannot be INTENT(OUT) In file iotk_dat.spp:656 CHARACTER (kind=iotk_defkind_character,len=len(dat)), allocatable :: tmpdat(:) 1 Error: Dummy argument 'dat' at (1) cannot be INTENT(OUT) In file iotk_dat.spp:656 CHARACTER (kind=iotk_defkind_character,len=len(dat)), allocatable :: tmpdat(:) 1 Error: Dummy argument 'dat' at (1) cannot be INTENT(OUT) make[2]: *** [iotk_dat+CHARACTER1_0.o] Error 1 make[2]: Leaving directory `/home/physlab/espresso-3.2/iotk/src' make[1]: *** [libiotk.a] Error 2 make[1]: Leaving directory `/home/physlab/espresso-3.2/iotk' make: *** [libiotk] Error 2 Main-server:/home/physlab/espresso-3.2 # What Shall I do ?? any help would be very appreciaple .................... thanks Regards; Hashem -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070425/a3da8467/attachment.htm From hashem.yamani at gmail.com Wed Apr 25 20:31:17 2007 From: hashem.yamani at gmail.com (Hashem Al-Yamani) Date: Wed, 25 Apr 2007 20:31:17 +0200 Subject: [Pw_forum] mpiexec parallel error Message-ID: <82a593400704251131k108885c7ia110b90a27cbd996@mail.gmail.com> Dear Sir ; What does it mean when I try to use PWSCF in parallel using mpiexec ... it gives the following massege : ----------------------------------------------------------------------------- *** Oops -- I cannot open the LAM help file. *** I tried looking for it in the following places: *** *** $HOME/lam-helpfile *** $HOME/lam-7.0.6-helpfile *** $HOME/etc/lam-helpfile *** $HOME/etc/lam-7.0.6-helpfile *** $LAMHELPDIR/lam-helpfile *** $LAMHELPDIR/lam-7.0.6-helpfile *** $LAMHOME/etc/lam-helpfile *** $LAMHOME/etc/lam-7.0.6-helpfile *** $SYSCONFDIR/lam-helpfile *** $SYSCONFDIR/lam-7.0.6-helpfile *** *** You were supposed to get help on the program "MPI" *** about the topic "no-lamd" *** *** Sorry! ----------------------------------------------------------------------------- ----------------------------------------------------------------------------- *** Oops -- I cannot open the LAM help file. *** I tried looking for it in the following places: *** *** $HOME/lam-helpfile *** $HOME/lam-7.0.6-helpfile *** $HOME/etc/lam-helpfile *** $HOME/etc/lam- 7.0.6-helpfile *** $LAMHELPDIR/lam-helpfile *** $LAMHELPDIR/lam-7.0.6-helpfile *** $LAMHOME/etc/lam-helpfile *** $LAMHOME/etc/lam-7.0.6-helpfile *** $SYSCONFDIR/lam-helpfile *** $SYSCONFDIR/lam- 7.0.6-helpfile *** *** You were supposed to get help on the program "MPI" *** about the topic "no-lamd" *** *** Sorry! ----------------------------------------------------------------------------- ----------------------------------------------------------------------------- It seems that there is no lamd running on the host Main-server. This indicates that the LAM/MPI runtime environment is not operating. The LAM/MPI runtime environment is necessary for MPI programs to run (the MPI program tired to invoke the "MPI_Init" function). Please run the "lamboot" command the start the LAM/MPI runtime environment. See the LAM/MPI documentation for how to invoke "lamboot" across multiple machines. ----------------------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070425/0dbb2cf1/attachment.htm From akohlmey at cmm.chem.upenn.edu Wed Apr 25 19:45:02 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 25 Apr 2007 13:45:02 -0400 (EDT) Subject: [Pw_forum] mpiexec parallel error In-Reply-To: <82a593400704251131k108885c7ia110b90a27cbd996@mail.gmail.com> Message-ID: On Wed, 25 Apr 2007, Hashem Al-Yamani wrote: HAY> Dear Sir ; HAY> HAY> What does it mean when I try to use PWSCF in parallel using mpiexec ... it HAY> gives the following massege : this has nothing to do with QE. check out the LAM/MPI documentation... before trying out to compile application programs with MPI, you should make sure that your MPI installation actually works. cheers, axel. HAY> HAY> HAY> HAY> ----------------------------------------------------------------------------- HAY> *** Oops -- I cannot open the LAM help file. HAY> *** I tried looking for it in the following places: HAY> *** HAY> *** $HOME/lam-helpfile HAY> *** $HOME/lam-7.0.6-helpfile HAY> *** $HOME/etc/lam-helpfile HAY> *** $HOME/etc/lam-7.0.6-helpfile HAY> *** $LAMHELPDIR/lam-helpfile HAY> *** $LAMHELPDIR/lam-7.0.6-helpfile HAY> *** $LAMHOME/etc/lam-helpfile HAY> *** $LAMHOME/etc/lam-7.0.6-helpfile HAY> *** $SYSCONFDIR/lam-helpfile HAY> *** $SYSCONFDIR/lam-7.0.6-helpfile HAY> *** HAY> *** You were supposed to get help on the program "MPI" HAY> *** about the topic "no-lamd" HAY> *** HAY> *** Sorry! HAY> ----------------------------------------------------------------------------- HAY> ----------------------------------------------------------------------------- HAY> HAY> *** Oops -- I cannot open the LAM help file. HAY> *** I tried looking for it in the following places: HAY> *** HAY> *** $HOME/lam-helpfile HAY> *** $HOME/lam-7.0.6-helpfile HAY> *** $HOME/etc/lam-helpfile HAY> *** $HOME/etc/lam- 7.0.6-helpfile HAY> *** $LAMHELPDIR/lam-helpfile HAY> *** $LAMHELPDIR/lam-7.0.6-helpfile HAY> *** $LAMHOME/etc/lam-helpfile HAY> *** $LAMHOME/etc/lam-7.0.6-helpfile HAY> *** $SYSCONFDIR/lam-helpfile HAY> *** $SYSCONFDIR/lam- 7.0.6-helpfile HAY> *** HAY> *** You were supposed to get help on the program "MPI" HAY> *** about the topic "no-lamd" HAY> *** HAY> *** Sorry! HAY> ----------------------------------------------------------------------------- HAY> HAY> ----------------------------------------------------------------------------- HAY> HAY> HAY> It seems that there is no lamd running on the host Main-server. HAY> HAY> This indicates that the LAM/MPI runtime environment is not operating. HAY> The LAM/MPI runtime environment is necessary for MPI programs to run HAY> (the MPI program tired to invoke the "MPI_Init" function). HAY> HAY> Please run the "lamboot" command the start the LAM/MPI runtime HAY> environment. See the LAM/MPI documentation for how to invoke HAY> "lamboot" across multiple machines. HAY> ----------------------------------------------------------------------------- HAY> -- ======================================================================= 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 ferretti.andrea at unimore.it Wed Apr 25 22:26:33 2007 From: ferretti.andrea at unimore.it (Andrea Ferretti) Date: Wed, 25 Apr 2007 22:26:33 +0200 (CEST) Subject: [Pw_forum] iotk make error In-Reply-To: <82a593400704251110r7590dac4u882070ca8282b850@mail.gmail.com> References: <82a593400704251110r7590dac4u882070ca8282b850@mail.gmail.com> Message-ID: Dear Hasehm, > mpif90 -O3 -x f95-cpp-input -D__FFTW -D__USE_INTERNAL_FFTW -D__MPI > -D__PARA -I../include? -I./? > -I../Modules? -I../iotk/src -I../PW? -I../PH? -I../CPV -c > iotk_dat+CHARACTER1_0.f90 > In file iotk_dat.spp:656 > > CHARACTER (kind=iotk_defkind_character,len=len(dat)), allocatable :: > tmpdat(:) > 1 > Error: Dummy argument 'dat' at (1) cannot be INTENT(OUT) > In file iotk_dat.spp:656 > looking in the file ~espresso/iotk/src/iotk_dat.spp it seems that this misbehaviour is already workarounded by a suitable precompiler macro (__IOTK_WORKAROUND6)... this macros are swtiched on in the file ~espresso/iotk/include/iotk_config.h which setis the workarounds for every platform... unfortunately, it seems that your machine was not recognized and no precompiler flag was set... this makes confusion also in iotk_config.h the problem in iotk can be solved editing ~/espresso/iotk/include/iotk_config.h (around line 66) substitute ... #elif defined(__SX6) # define __IOTK_WORKAROUND5 # define __IOTK_WORKAROUND7 #endif with #elif defined(__SX6) # define __IOTK_WORKAROUND5 # define __IOTK_WORKAROUND7 #else # define __IOTK_WORKAROUND1 # define __IOTK_WORKAROUND2 # define __IOTK_WORKAROUND3 # define __IOTK_WORKAROUND4 # define __IOTK_WORKAROUND5 # define __IOTK_WORKAROUND6 # define __IOTK_WORKAROUND7 #endif ============================ Anyway, the best way to fix all this would be to properly configure espresso in such a way that all the needed macros are automatically set. by the way, could you please provide the make.sys file generated by configure ?? Cheers Andrea From w2agz at pacbell.net Thu Apr 26 05:41:06 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Wed, 25 Apr 2007 20:41:06 -0700 Subject: [Pw_forum] Lattice Constant Optimization of Aluminum Message-ID: <008501c787b4$b9716d30$0301a8c0@w2agz> To All: (BTW: PG: My last note on the "default" value of npk should have been termed "default maximum." Sorry, but we Americans haven't spoken English since 1776. Anyway, thanks for pointing me to the right Fortran routine.) I'm trying to do a very simple homework problem with PWscf.the optimization of the lattice constant of aluminum. Even with a wildly wrong first estimate of celldm(1) of 7.2 (the experimental value at RT is 7.665 au), I get zero net individual atomic forces and thus no optimization. I've gotten "relax" in PWscf to work with a lot more complex polymer monoclinic unit cell containing eight atoms and two different elements. So, I must be doing something really dumb (stupid).maybe I should be using vc-relax instead? Anyway, here's the I/O (scusi e grazie): INPUT &control title = '$prefix_root$s$i.rx.in' calculation = 'relax' restart_mode = 'from_scratch', prefix = '$prefix_root$s$i', pseudo_dir = '/home/pmpgrant/espresso-3.2/pseudo/', outdir = '$home_root', verbosity = 'minimal', tprnfor = .false., tstress = .true., / &system ibrav = 2, celldm(1) = 7.2, nat = 1, ntyp = 1, ecutwfc = 15, occupations = 'smearing', smearing = 'methfessel-paxton', degauss = 0.05, la2F = .false., / &electrons diagonalization = 'david', conv_thr = 1.0d-12, mixing_beta = 0.7, / &ions ion_dynamics = 'bfgs', pot_extrapolation = 'second-order', wfc_extrapolation = 'second-order', / ATOMIC_SPECIES Al 26.98 Al.vbc.UPF ATOMIC_POSITIONS Al 0.00 0.00 0.00 K_POINTS {automatic} 32 32 32 0 0 0 OUTPUT The total energy is the sum of the following terms: one-electron contribution = 3.23304593 Ry hartree contribution = 0.01297127 Ry xc contribution = -1.69806267 Ry ewald contribution = -5.73107767 Ry smearing contrib. (-TS) = 0.00011533 Ry convergence has been achieved Forces acting on atoms (Ry/au): atom 1 type 1 force = 0.00000000 0.00000000 0.00000000 Total force = 0.000000 Total SCF correction = 0.000000 SCF correction compared to forces is too large, reduce conv_thr entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 110.51 0.00075124 0.00000000 0.00000000 110.51 0.00 0.00 0.00000000 0.00075124 0.00000000 0.00 110.51 0.00 0.00000000 0.00000000 0.00075124 0.00 0.00 110.51 BFGS Geometry Optimization bfgs converged in 1 scf cycles and 0 bfgs steps End of BFGS Geometry Optimization Final energy = -4.1830078084 Ry CELL_PARAMETERS (alat) -0.500000000 0.000000000 0.500000000 0.000000000 0.500000000 0.500000000 -0.500000000 0.500000000 0.000000000 ATOMIC_POSITIONS (alat) Al 0.000000000 0.000000000 0.000000000 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: /pipermail/attachments/20070425/915a3253/attachment.htm From prasenjit at jncasr.ac.in Thu Apr 26 06:21:26 2007 From: prasenjit at jncasr.ac.in (PRASENJIT GHOSH) Date: Thu, 26 Apr 2007 09:51:26 +0530 (IST) Subject: [Pw_forum] Lattice Constant Optimization of Aluminum In-Reply-To: <008501c787b4$b9716d30$0301a8c0@w2agz> References: <008501c787b4$b9716d30$0301a8c0@w2agz> Message-ID: <54318.172.16.1.1.1177561286.squirrel@172.16.1.1> Dear Paul, You are getting zero forces because u have set tprnfor= .false. As a result the code is not at all calculating forces on the atoms. For relaxation cal. u should have set it to .true. Moreover ur unit cell has only one atom. Hence I don't think there is any point in using the 'relax' option. Rather u can do a simple scf run and still get the forces. With 'relax' option u can only relax the coordinates of the atoms, not the unit cell. With regards, Prasenjit. On Thu, April 26, 2007 9:11, Paul M. Grant said: > To All: > > (BTW: PG: My last note on the "default" value of npk should have been > termed > "default maximum." Sorry, but we Americans haven't spoken English since > 1776. Anyway, thanks for pointing me to the right Fortran routine.) > > I'm trying to do a very simple homework problem with PWscf.the > optimization > of the lattice constant of aluminum. > > Even with a wildly wrong first estimate of celldm(1) of 7.2 (the > experimental value at RT is 7.665 au), I get zero net individual atomic > forces and thus no optimization. I've gotten "relax" in PWscf to work > with > a lot more complex polymer monoclinic unit cell containing eight atoms and > two different elements. So, I must be doing something really dumb > (stupid).maybe I should be using vc-relax instead? Anyway, here's the I/O > (scusi e grazie): > > INPUT > > &control > > title = '$prefix_root$s$i.rx.in' > > calculation = 'relax' > > restart_mode = 'from_scratch', > > prefix = '$prefix_root$s$i', > > pseudo_dir = '/home/pmpgrant/espresso-3.2/pseudo/', > > outdir = '$home_root', > > verbosity = 'minimal', > > tprnfor = .false., > > tstress = .true., > > / > > &system > > ibrav = 2, > > celldm(1) = 7.2, > > nat = 1, > > ntyp = 1, > > ecutwfc = 15, > > occupations = 'smearing', > > smearing = 'methfessel-paxton', > > degauss = 0.05, > > la2F = .false., > > / > > &electrons > > diagonalization = 'david', > > conv_thr = 1.0d-12, > > mixing_beta = 0.7, > > / > > &ions > > ion_dynamics = 'bfgs', > > pot_extrapolation = 'second-order', > > wfc_extrapolation = 'second-order', > > / > > ATOMIC_SPECIES > > Al 26.98 Al.vbc.UPF > > ATOMIC_POSITIONS > > Al 0.00 0.00 0.00 > > K_POINTS {automatic} > > 32 32 32 0 0 0 > > OUTPUT > > The total energy is the sum of the following terms: > > > > one-electron contribution = 3.23304593 Ry > > hartree contribution = 0.01297127 Ry > > xc contribution = -1.69806267 Ry > > ewald contribution = -5.73107767 Ry > > smearing contrib. (-TS) = 0.00011533 Ry > > > > convergence has been achieved > > > > Forces acting on atoms (Ry/au): > > > > atom 1 type 1 force = 0.00000000 0.00000000 0.00000000 > > > > Total force = 0.000000 Total SCF correction = 0.000000 > > SCF correction compared to forces is too large, reduce conv_thr > > > > > > entering subroutine stress ... > > > > total stress (Ry/bohr**3) (kbar) P= > 110.51 > > 0.00075124 0.00000000 0.00000000 110.51 0.00 0.00 > > 0.00000000 0.00075124 0.00000000 0.00 110.51 0.00 > > 0.00000000 0.00000000 0.00075124 0.00 0.00 110.51 > > > > > > BFGS Geometry Optimization > > > > bfgs converged in 1 scf cycles and 0 bfgs steps > > > > End of BFGS Geometry Optimization > > > > Final energy = -4.1830078084 Ry > > > > CELL_PARAMETERS (alat) > > -0.500000000 0.000000000 0.500000000 > > 0.000000000 0.500000000 0.500000000 > > -0.500000000 0.500000000 0.000000000 > > > > ATOMIC_POSITIONS (alat) > > Al 0.000000000 0.000000000 0.000000000 > > > > > > 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 > > > > > > > > PRASENJIT GHOSH, Ph. D STUDENT, THEORETICAL SCIENCES UNIT, JAWAHARLAL NEHRU CENTRE, JAKKUR P. O., BANGALORE - 560064, INDIA. PHONE:+91-80-22082835 (OFFICE) +91 9880519401 (MOBILE) From w2agz at pacbell.net Thu Apr 26 08:10:15 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Wed, 25 Apr 2007 23:10:15 -0700 Subject: [Pw_forum] Lattice Constant Optimization of Aluminum In-Reply-To: <54318.172.16.1.1.1177561286.squirrel@172.16.1.1> References: <008501c787b4$b9716d30$0301a8c0@w2agz> <54318.172.16.1.1.1177561286.squirrel@172.16.1.1> Message-ID: <009001c787c9$8f20b490$0301a8c0@w2agz> Prasenjit, thanks for getting back so quick. In the "stress" of battle, I had indeed reversed tstress and tprnfor, but tprnfor merely prints the forces and setting it to .true. made no difference. I suspect you may be correct in that the aluminum structure is too "simple" for "relax." I've been trying "vc-relax" but am still floundering around trying to get that to work as well. Incidentally, before trying to use the "relax" or "vc-relax" feature of PWscf, I did indeed "optimize by hand" searching for the minimal total energy by just varying celldm(1) and of course it works. I was testing the "automatic" route. -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-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] On Behalf Of PRASENJIT GHOSH Sent: Wednesday, April 25, 2007 9:21 PM To: pw_forum at pwscf.org Subject: Re: [Pw_forum] Lattice Constant Optimization of Aluminum Dear Paul, You are getting zero forces because u have set tprnfor= .false. As a result the code is not at all calculating forces on the atoms. For relaxation cal. u should have set it to .true. Moreover ur unit cell has only one atom. Hence I don't think there is any point in using the 'relax' option. Rather u can do a simple scf run and still get the forces. With 'relax' option u can only relax the coordinates of the atoms, not the unit cell. With regards, Prasenjit. On Thu, April 26, 2007 9:11, Paul M. Grant said: > To All: > > (BTW: PG: My last note on the "default" value of npk should have been > termed > "default maximum." Sorry, but we Americans haven't spoken English since > 1776. Anyway, thanks for pointing me to the right Fortran routine.) > > I'm trying to do a very simple homework problem with PWscf.the > optimization > of the lattice constant of aluminum. > > Even with a wildly wrong first estimate of celldm(1) of 7.2 (the > experimental value at RT is 7.665 au), I get zero net individual atomic > forces and thus no optimization. I've gotten "relax" in PWscf to work > with > a lot more complex polymer monoclinic unit cell containing eight atoms and > two different elements. So, I must be doing something really dumb > (stupid).maybe I should be using vc-relax instead? Anyway, here's the I/O > (scusi e grazie): > > INPUT > > &control > > title = '$prefix_root$s$i.rx.in' > > calculation = 'relax' > > restart_mode = 'from_scratch', > > prefix = '$prefix_root$s$i', > > pseudo_dir = '/home/pmpgrant/espresso-3.2/pseudo/', > > outdir = '$home_root', > > verbosity = 'minimal', > > tprnfor = .false., > > tstress = .true., > > / > > &system > > ibrav = 2, > > celldm(1) = 7.2, > > nat = 1, > > ntyp = 1, > > ecutwfc = 15, > > occupations = 'smearing', > > smearing = 'methfessel-paxton', > > degauss = 0.05, > > la2F = .false., > > / > > &electrons > > diagonalization = 'david', > > conv_thr = 1.0d-12, > > mixing_beta = 0.7, > > / > > &ions > > ion_dynamics = 'bfgs', > > pot_extrapolation = 'second-order', > > wfc_extrapolation = 'second-order', > > / > > ATOMIC_SPECIES > > Al 26.98 Al.vbc.UPF > > ATOMIC_POSITIONS > > Al 0.00 0.00 0.00 > > K_POINTS {automatic} > > 32 32 32 0 0 0 > > OUTPUT > > The total energy is the sum of the following terms: > > > > one-electron contribution = 3.23304593 Ry > > hartree contribution = 0.01297127 Ry > > xc contribution = -1.69806267 Ry > > ewald contribution = -5.73107767 Ry > > smearing contrib. (-TS) = 0.00011533 Ry > > > > convergence has been achieved > > > > Forces acting on atoms (Ry/au): > > > > atom 1 type 1 force = 0.00000000 0.00000000 0.00000000 > > > > Total force = 0.000000 Total SCF correction = 0.000000 > > SCF correction compared to forces is too large, reduce conv_thr > > > > > > entering subroutine stress ... > > > > total stress (Ry/bohr**3) (kbar) P= > 110.51 > > 0.00075124 0.00000000 0.00000000 110.51 0.00 0.00 > > 0.00000000 0.00075124 0.00000000 0.00 110.51 0.00 > > 0.00000000 0.00000000 0.00075124 0.00 0.00 110.51 > > > > > > BFGS Geometry Optimization > > > > bfgs converged in 1 scf cycles and 0 bfgs steps > > > > End of BFGS Geometry Optimization > > > > Final energy = -4.1830078084 Ry > > > > CELL_PARAMETERS (alat) > > -0.500000000 0.000000000 0.500000000 > > 0.000000000 0.500000000 0.500000000 > > -0.500000000 0.500000000 0.000000000 > > > > ATOMIC_POSITIONS (alat) > > Al 0.000000000 0.000000000 0.000000000 > > > > > > 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 > > > > > > > > PRASENJIT GHOSH, Ph. D STUDENT, THEORETICAL SCIENCES UNIT, JAWAHARLAL NEHRU CENTRE, JAKKUR P. O., BANGALORE - 560064, INDIA. PHONE:+91-80-22082835 (OFFICE) +91 9880519401 (MOBILE) _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From giannozz at nest.sns.it Thu Apr 26 09:05:15 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 26 Apr 2007 09:05:15 +0200 Subject: [Pw_forum] Lattice Constant Optimization of Aluminum In-Reply-To: <008501c787b4$b9716d30$0301a8c0@w2agz> References: <008501c787b4$b9716d30$0301a8c0@w2agz> Message-ID: <7C989E8A-1AD9-4A52-A4DF-FA58E46BFC38@nest.sns.it> On Apr 26, 2007, at 5:41 , Paul M. Grant wrote: > Even with a wildly wrong first estimate of celldm(1) of 7.2 (the > experimental value at RT is 7.665 au), I get zero net individual > atomic forces and thus no optimization. "optimization wrt internal positions" != "optimization of the lattice parameter(s)". Forces allow the former, stresses the latter. The simplest way to optimize the lattice parameter of a simple crystal is to fit a few E(V) points to an equation of state, using for instance "ev.f90" Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From clie at aphy.iphy.ac.cn Thu Apr 26 09:11:24 2007 From: clie at aphy.iphy.ac.cn (clie) Date: Thu, 26 Apr 2007 07:11:24 GMT Subject: [Pw_forum] Re: Pw_forum digest, Vol 1 #1444 - 2 msgs In-Reply-To: <20070426062738.22273.26737.Mailman@democritos.sissa.it> References: <20070426062738.22273.26737.Mailman@democritos.sissa.it> Message-ID: <20070426071124.16930.eqmail@aphy.iphy.ac.cn> Dear all, i come across problems when i calculate electron-phonon coupling of G eH4. It seems that there are somethings wrong in my k-points sample. The progra m stops and tells "running the el-ph calculation.....2". Could you check my in put file and give some suggestions. Thank you vevy much. # SCF at dense k-mesh, good enough for electronic DOS # cat > GeH4.scf.fit.in << EOF &control calculation='scf' restart_mode='from_scratch', prefix='GeH4', pseudo_dir = '$PSEUDO_DIR/', outdir='$TMP_DIR/' / &system ibrav= 8, celldm(1) =7.478,celldm(2)=0.76, celldm(3)=0.72, nat= 10, ntyp= 2, ecutwfc =60.0, occupations='smearing', smearing='methfessel-paxton', degauss=0.05, la2F = .true., / &electrons conv_thr = 1.0d-8 mixing_beta = 0.7 / ATOMIC_SPECIES Ge 72.61 Ge.pz-bhs.UPF H 1.00794 H.pz-van_ak.UPF ATOMIC_POSITIONS Ge 0.500000 0.000000 0.500000 Ge 0.000000 0.000000 0.000000 H 0.000000 0.388640 0.688119 H 0.500000 0.611360 0.188119 H 0.500000 0.388640 0.811881 H 0.000000 0.611360 0.311811 H 0.250000 0.357636 0.250000 H 0.250000 0.642364 0.750000 H 0.750000 0.642364 0.750000 H 0.750000 0.357636 0.250000 K_POINTS {automatic} 36 48 48 0 0 0 EOF $ECHO " running the scf calculation with dense k-point grid...\c" $PW_COMMAND < GeH4.scf.fit.in > GeH4.scf.fit.out $ECHO " done" # # SCF at k-mesh good enough for phonons # cat > GeH4.scf.in << EOF &control calculation='scf' restart_mode='from_scratch', prefix='GeH4', pseudo_dir = '$PSEUDO_DIR/', outdir='$TMP_DIR/' / &system ibrav= 8, celldm(1) =7.478,celldm(2)=0.76, celldm(3)=0.72, nat= 10, ntyp= 2, ecutwfc =60.0, occupations='smearing', smearing='methfessel-paxton', degauss=0.05 / &electrons conv_thr = 1.0d-8 mixing_beta = 0.7 / ATOMIC_SPECIES Ge 72.61 Ge.pz-bhs.UPF H 1.00794 H.pz-van_ak.UPF ATOMIC_POSITIONS Ge 0.500000 0.000000 0.500000 Ge 0.000000 0.000000 0.000000 H 0.000000 0.388640 0.688119 H 0.500000 0.611360 0.188119 H 0.500000 0.388640 0.811881 H 0.000000 0.611360 0.311811 H 0.250000 0.357636 0.250000 H 0.250000 0.642364 0.750000 H 0.750000 0.642364 0.750000 H 0.750000 0.357636 0.250000 K_POINTS {automatic} 12 16 16 0 0 0 EOF $ECHO " running the scf calculation...\c" $PW_COMMAND < GeH4.scf.in > GeH4.scf.out $ECHO " done" # cat > GeH4.elph.in << EOF Electron-phonon coefficients for GeH4 &inputph tr2_ph=1.0d-10, prefix='GeH4', fildvscf='GeH4dv', amass(1)=72.61, amass(2)=1.00794, outdir='$TMP_DIR/', fildyn='GeH4.dyn', elph=.true., trans=.true., ldisp=.true. nq1=3, nq2=4, nq3=4 / EOF $ECHO " running the el-ph calculation...\c" $PH_COMMAND < GeH4.elph.in > GeH4.elph.out $ECHO " done" # # q2r and matdyn # cat > q2r.in << EOF &input fildyn='GeH4.dyn', flfrc='GeH4444.fc', la2F=.true. / EOF $ECHO " running q2r...\c" $Q2R_COMMAND < q2r.in > q2r.out $ECHO " done" # # # cat > matdyn.in.freq << EOF &input amass(1)=72.61,amass(2)=1.00794, flfrc='GeH4444.fc', flfrq='GeH4444.freq', la2F=.true., dos=.false. / 8 0.00000 0.00000 0.00000 0.00000 0.00000 -0.69444 0.00000 -0.65789 0.00000 0.00000 -0.65789 -0.69444 -0.50000 0.00000 0.00000 -0.50000 0.00000 -0.69444 -0.50000 -0.65789 0.00000 -0.50000 -0.65789 -0.69444 EOF $ECHO " running matdyn for frequency calculation...\c" $MATDYN_COMMAND < matdyn.in.freq > matdyn.out.freq $ECHO " done" # # # cat > matdyn.in.dos << EOF &input amass(1)=72.61, amass(2)=1.00794, flfrc='GeH4444.fc', flfrq='GeH4444.freq', la2F=.true., dos=.true. fldos='phonon.dos', nk1=10, nk2=10, nk3=10, ndos=50 / EOF $ECHO " running matdyn for a2F(omega) calculation...\c" $MATDYN_COMMAND < matdyn.in.dos > matdyn.out.dos $ECHO " done" -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070426/d1f02551/attachment.htm From giannozz at nest.sns.it Thu Apr 26 10:05:51 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 26 Apr 2007 10:05:51 +0200 Subject: [Pw_forum] Re: Pw_forum digest, Vol 1 #1444 - 2 msgs In-Reply-To: <20070426071124.16930.eqmail@aphy.iphy.ac.cn> References: <20070426062738.22273.26737.Mailman@democritos.sissa.it> <20070426071124.16930.eqmail@aphy.iphy.ac.cn> Message-ID: <9E4F2B78-EAFC-407C-BEA0-6C35298D16E9@nest.sns.it> On Apr 26, 2007, at 7:11 , clie wrote: > Dear all, i come across problems when i calculate electron-phonon > coupling of GeH4. It seems that there are somethings wrong in my k- > points sample. The program stops and tells "running the el-ph > calculation.....2". How can you say that there is something wrong in your k-point sample on the basis of a simple "2" ? it tells more than this --- Paolo Giannozzi, Democritos and University of Udine, Italy From wmbmacam at lg.ehu.es Thu Apr 26 10:21:04 2007 From: wmbmacam at lg.ehu.es (=?ISO-8859-1?Q?Miguel_Mart=EDnez_Canales?=) Date: Thu, 26 Apr 2007 10:21:04 +0200 Subject: [Pw_forum] Re: Pw_forum digest, Vol 1 #1444 - 2 msgs In-Reply-To: <9E4F2B78-EAFC-407C-BEA0-6C35298D16E9@nest.sns.it> References: <20070426062738.22273.26737.Mailman@democritos.sissa.it> <20070426071124.16930.eqmail@aphy.iphy.ac.cn> <9E4F2B78-EAFC-407C-BEA0-6C35298D16E9@nest.sns.it> Message-ID: <463060F0.5000102@lg.ehu.es> Dear clie, You should look, at least, at the last lines of GeH4.elph.out. It should give a more detailed message than "2". Maybe some hints on starting density files or q-point choice. On Apr 26, 2007, at 7:11 , clie wrote: > > Dear all, i come across problems when i calculate electron-phonon > coupling of GeH4. It seems that there are somethings wrong in my > k-points sample. The program stops and tells "running the el-ph > calculation.....2". -- ---------------------------------------- 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 5437 ---------------------------------------- "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 giannozz at nest.sns.it Thu Apr 26 13:42:59 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 26 Apr 2007 13:42:59 +0200 Subject: [Pw_forum] what is w1 and w2 for bfgs relaxed calculation In-Reply-To: References: Message-ID: <35D43845-54C5-47D8-91EC-E7E28A494FFF@nest.sns.it> On Apr 25, 2007, at 7:32 , yukihiro_okuno at fujifilm.co.jp wrote: > I met not conserved force "not converged", I guess. > in relaxed calculation for slightly titled crystal structure for PZT Are you referring to the following: http://www.democritos.it/pipermail/pw_forum/2007-April/006311.html ? This behavour might simply mean that the structural optimization is not converged; or that the scf convergence towards the end of the structural optimization is not sufficient to get forces with high accuracy; finally, it might also signal some inconsistency between the calculated energy and the calculated forces. You may test for the latter by calculating forces as numerical derivatives of the energy with respect to atomic displacements (it is better to try this in simple systems). If you notice serious discrepancies, please report them to the list Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From umari at democritos.it Thu Apr 26 14:54:57 2007 From: umari at democritos.it (umari at democritos.it) Date: Thu, 26 Apr 2007 14:54:57 +0200 Subject: [Pw_forum] RE: on the finite electric field method Message-ID: <20070426145457.l58snv21g1wk4kc4@mail.democritos.it> Dear Liu, For calculating Raman spectra of large model structure I suggest the method described in : P. Umari, A. Pasquarello ?Infrared and Raman spectra of disordered materials from first principles?, DIAMOND AND RELATED MATERIALS 14, 1255 (2005). The idea is that the Raman tensor is calculated by taking finite differences of atomic forces in the presence of an electric field For the calculation of the whole tensor, two different electric fields are required. In the code cp.x (see the documentation file: INPUT_CP ) the possibility of up two different electric fields is implemented. Please consider that: 1)atomic forces and electric fields are implemented for the US pseudopotentials only in the cp code 2)for parallel calculations onlt the option gdir=3 is supported Both limitations will be overcome with a new version in about one-month time. Best wishes, PaoloU > Dear all , > > I want to investigate Raman spectra in the case of large > disordered system >from > first principles. I saw the exmple30 and 31, but they only gave the > dielectric > constant and Born effective charge by Moden polarizability theory. > How can I obtain Raman coupling tensor by numerically calculating second > derivatives of the atomic forces with respect to the electric field? > [in RAMAN SPECTRA OF DISORDERED OXIDES FROM FIRST PRINCIPLES > written by >PAOLO > UMARI]? > I did not find those from the forum. > Thank you in advance! > Niu Li ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From niuli1978 at yahoo.com.cn Thu Apr 26 15:23:39 2007 From: niuli1978 at yahoo.com.cn (li niu) Date: Thu, 26 Apr 2007 21:23:39 +0800 (CST) Subject: [Pw_forum] vc-relax In-Reply-To: <20070425181102.13627.1119.Mailman@democritos.sissa.it> Message-ID: <860693.20081.qm@web15014.mail.cnb.yahoo.com> Dear Matteo, Thanks for your help. I'm now running a vc-relax calculation and the input is follows. Is it right? Best! Niu Liu ----------------------------------- # self-consistent calculation cat > tac.vcrx.in << EOF &CONTROL calculation = "vc-relax", prefix = "ta-c", pseudo_dir = "$PSEUDO_DIR", outdir = "$TMP_DIR", tstress = .true. , tprnfor = .true. , nstep = 65 , etot_conv_thr = 1.0E-4 , forc_conv_thr = 1.0D-3 , iprint = 1 , max_seconds = 36000 , dt = 100 , / &SYSTEM ibrav = 1, celldm(1) = 13.9123, nat= 64, ntyp= 1, ecutwfc =55.0 / &ELECTRONS conv_thr = 1.D-9, mixing_beta = 0.7D0, / &IONS pot_extrapolation = "second_order", wfc_extrapolation = "second_order", / &CELL cell_dynamics = 'damp-w' , press = 0.0 , wmass = 0.010 , / ATOMIC_SPECIES C 12.0107 C.pz-mt.UPF ATOMIC_POSITIONS {angstrom} C 0.335930 -0.353760 -0.911160 C 1.410315 -0.756669 3.705328 C -0.462432 3.049465 -0.168292 C -0.022431 3.092983 3.870317 C 4.202728 0.738363 -0.193781 ... K_POINTS {Gamma} --------------------------------- ??????3.5G???20M??? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070426/ca2fbe88/attachment.htm From niuli1978 at yahoo.com.cn Thu Apr 26 15:34:17 2007 From: niuli1978 at yahoo.com.cn (li niu) Date: Thu, 26 Apr 2007 21:34:17 +0800 (CST) Subject: [Pw_forum] on finite electric field method Message-ID: <420054.33894.qm@web15002.mail.cnb.yahoo.com> Dear all , I want to investigate Raman spectra in the case of large disordered system from first principles. I saw the exmple30 and 31, but they only gave the dielectric constant and Born effective charge by Moden polarizability theory. Can the espresso-3.2 code be used to calculate Rman scattering by numerically calculating second derivatives of the atomic forces with respect to the electric field? Where can I obtain CPV code? I did not find those from the forum. Thank you in advance! Niu Li --------------------------------- ????????3.5G???20M??? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070426/aa4c670d/attachment.htm From mverissi at ictp.it Thu Apr 26 15:41:25 2007 From: mverissi at ictp.it (Marcos Verissimo Alves) Date: Thu, 26 Apr 2007 15:41:25 +0200 (CEST) Subject: [Pw_forum] Off-topic: memory usage issues In-Reply-To: <35D43845-54C5-47D8-91EC-E7E28A494FFF@nest.sns.it> References: <35D43845-54C5-47D8-91EC-E7E28A494FFF@nest.sns.it> Message-ID: <49540.10.50.40.120.1177594885.squirrel@webmail1.ictp.trieste.it> Hi all, This is not a question specifically aimed at espresso issues, but since we have quite a few knowledgeable people in the list, I'm posting it. I have compiled pw.x in a dual-core Turion with 2GB RAM under linux (OpenSuse 10.2), installed and compiled mpich 2-1.0.5 with ifort 9.1.041, and then compiled pw.x (from Espresso 3.0) with mpich, mkl 8.0 and fftw, and I have been running it successfully since then in parallel. However, looking at the memory usage using top, when running pw.x, I see that, although the memory from top is at most 250 MB per processor (from the part where the individual processes' cpu and memory usage is displayed), almost the whole of physical memory is taken (from the "Mem: total, used, free, buffers" line). Right after the booting of the machine, the total memory taken by the system processes is 230 MB (I'm running it without the X-windows system to make the most possible memory available to the calculations). When I run pw.x in the two processors, each process takes 298 MB of virtual memory, according to top. Thus, there should be only around 830 MB (596 + 230 MB) used, but about 1650 MB are taken. Running pw.x in serial mode (that is, the same compilation, but without invoking mpirun or mpiexec) has the same effect: the total virtual memory is 532 MB, but 367 MB. The same kind of thing happens when I run another DFT code (siesta), so I would guess this could be an MPICH-related issue or, worse, a kernel- or compiler-related issue. Does anyone know if there are any memory-usage issues for MPICH, for the newer versions of the linux kernel or for the ifort 9.1.041/mkl 8.0? Has anybody seen something similar when running pw.x on dual-processor machines? Thanks, Marcos -- Dr. Marcos Verissimo Alves Post-Doctoral Fellow Condensed Matter and Statistical Physics Sector International Centre for Theoretical Physics Trieste, Italy -------- I have become so addicted to vi that I try to exit OpenOffice by typing :wq! From nchamel at ulb.ac.be Thu Apr 26 15:23:51 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Thu, 26 Apr 2007 15:23:51 +0200 (CEST) Subject: [Pw_forum] looking for a routine to symmetrize matrix elements Message-ID: <20070426132351.EAC7840D@bonito.ulb.ac.be> Dear users, I'm not working on electronic structure calculations but on an analogous problem in nuclear physics where neutrons play the role of electrons and ions are replaced by nuclei. I would like to use some parts of the code for my own calculations but I'm a bit lost in the code. For instance I would like to use a routine for the symmetrization of matrix elements for a given crystal structure, let's say bcc. I would really appreciate if someone could guide me though the code. Nicolas Chamel Institut d'Astronomie et d'Astrophysique Universite Libre de Bruxelles CP 226 Boulevard du Triomphe B-1050 Brussels BELGIUM tel. +32 2 650 35 72 fax. +32 2 650 28 64 From marzari at MIT.EDU Thu Apr 26 15:58:00 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 26 Apr 2007 09:58:00 -0400 Subject: [Pw_forum] looking for a routine to symmetrize matrix elements In-Reply-To: <20070426132351.EAC7840D@bonito.ulb.ac.be> References: <20070426132351.EAC7840D@bonito.ulb.ac.be> Message-ID: <4630AFE8.1080703@mit.edu> Dear Nicolas, I think it would be far easier for you to write the symmetrization by hand, rather than reverse engineering any code. nicola Nicolas Chamel wrote: > Dear users, > > I'm not working on electronic structure calculations but on an analogous problem in nuclear physics where neutrons play the role of electrons and ions are replaced by nuclei. I would like to use some parts of the code for my own calculations but I'm a bit lost in the code. For instance I would like to use a routine for the symmetrization of matrix elements for a given crystal structure, let's say bcc. I would really appreciate if someone could guide me though the code. > > Nicolas Chamel > > Institut d'Astronomie et d'Astrophysique > Universite Libre de Bruxelles > CP 226 > Boulevard du Triomphe > B-1050 Brussels > BELGIUM > > tel. +32 2 650 35 72 > fax. +32 2 650 28 64 > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- --------------------------------------------------------------------- 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 giannozz at nest.sns.it Thu Apr 26 16:01:10 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 26 Apr 2007 16:01:10 +0200 Subject: [Pw_forum] Off-topic: memory usage issues In-Reply-To: <49540.10.50.40.120.1177594885.squirrel@webmail1.ictp.trieste.it> References: <35D43845-54C5-47D8-91EC-E7E28A494FFF@nest.sns.it> <49540.10.50.40.120.1177594885.squirrel@webmail1.ictp.trieste.it> Message-ID: <0EC1D86C-F04F-466A-A11A-6D3E25C5CBEB@nest.sns.it> On Apr 26, 2007, at 15:41 , Marcos Verissimo Alves wrote: > looking at the memory usage using top, when running pw.x, I see that, > although the memory from top is at most 250 MB per processor (from the > part where the individual processes' cpu and memory usage is > displayed), > almost the whole of physical memory is taken (from the "Mem: total, > used, > free, buffers" line). unix always uses the physical memory as much as possibile to store files, buffers, temporary data and whatnot. For instance: if enough RAM is available, "writing to file" will produce no physical writing (the data is stored into memory) unless the file is closed or the unit flushed or the memory reclaimed by some other process with higher priority Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From akohlmey at cmm.chem.upenn.edu Thu Apr 26 15:37:37 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 26 Apr 2007 09:37:37 -0400 (EDT) Subject: [Pw_forum] Off-topic: memory usage issues In-Reply-To: <49540.10.50.40.120.1177594885.squirrel@webmail1.ictp.trieste.it> Message-ID: On Thu, 26 Apr 2007, Marcos Verissimo Alves wrote: MVA> Hi all, hi marcos, MVA> This is not a question specifically aimed at espresso issues, but since we MVA> have quite a few knowledgeable people in the list, I'm posting it. actually, i think this is not so off-topic as you say, since this gives an opportunity to clear up some confusions. MVA> I have compiled pw.x in a dual-core Turion with 2GB RAM under linux MVA> (OpenSuse 10.2), installed and compiled mpich 2-1.0.5 with ifort 9.1.041, MVA> and then compiled pw.x (from Espresso 3.0) with mpich, mkl 8.0 and fftw, MVA> and I have been running it successfully since then in parallel. However, MVA> looking at the memory usage using top, when running pw.x, I see that, MVA> although the memory from top is at most 250 MB per processor (from the MVA> part where the individual processes' cpu and memory usage is displayed), first of all you have to realize that there are multiple entries for the memory usage with different meaning. none of them gives you the 'real' memory usage, since this is not a well defined entity. under linux you have multiple fields and it depends on the configuration of top what you can see. if you type 'f' you can add or remove fields and then save the config with 'W' to ~/.toprc. in particular there is: VIRT: the total address space (= the maximum amount of memory that is reserved for the process, even if it is not used or in swap) RES: the resident set size (= the amount of _physical_ memory used, the amount of actively used memory minus the part in swap). SHR: shared memory contribution (= how much sharing is going on, usually due to shared libraries) SWAP: the amount of memory in swap (= for technical reason this can also include 'mapped' memory, e.g. the memory for a video card, or network card) CODE: the part of the executable that is loaded into memory DATA: the amount of memory allocated in the 'data segment' so what to watch out for? for applications with dynamical memory management, the memory usage is somewhere between VIRT and RES (which are usually both displayed), VIRT will always be larger than the real amount of memory used, but RES will always stay within the limits of your machine, as it does not show the amount of swap space used by this process. only if %CPU is low and the process is mostly in state 'D' (for waiting on 'D'isk, as opposed to 'R'unning or 'S'leeping) than you are probably swapping a lot. however performance suffers significantly, if there is too much swap usage (which is usually not displayed). so RES is actually not a good measure. DATA is probably better, but it does not include CODE or SHR. and SHR is undefined since you cannot know how much each process contributes to the shared memory pool (well, technically you can, but that would slow down the kernel a lot). bottom line: don't get fooled by VIRT or RES, but if VIRT fits into your physical memory, you are on the safe side. MVA> almost the whole of physical memory is taken (from the "Mem: total, used, MVA> free, buffers" line). to quote linus torvalds: unused memory is wasted memory, so linux uses 'free' memory for file system caching and i/o buffering. if an application needs more memory, that is then claimed from the 'cached' and/or 'buffered' pool of memory which has a delicately tune strategy of ageing buffered pages to re-use those first that are used the least. if you are lazy (like me), you can get the 'corrected' numbers from 'free -t', e.g.: [akohlmey at vitriol ~]$ free -t total used free shared buffers cached Mem: 16399936 15969172 430764 0 358728 14729608 -/+ buffers/cache: 880836 15519100 Swap: 4096524 346008 3750516 Total: 20496460 16315180 4181280 [...] MVA> The same kind of thing happens when I run another DFT code (siesta), so I MVA> would guess this could be an MPICH-related issue or, worse, a kernel- or MVA> compiler-related issue. Does anyone know if there are any memory-usage MVA> issues for MPICH, for the newer versions of the linux kernel or for the no, but for a local (shared memory) communication, MPICH needs to allocate additional buffers and some of them are actually SYSV shared memory segments and should thus show up as used and shared. one thing to be noted here is, that when you run a machine for a very long time, you may actually run out of shared memory segments and lose memory due to crashed jobs that did not have a chance to clean up the shared memory segment first. with some smart, but not overly portable programming this can be avoided, but for a number of reasons this is not always done. (this also affects other programs that use shared memory like audio/video players). you can see the sysv ipc status with 'ipcs' and if you have too many shared memory segments with 'nattach' 0, then you should clean up. e.g. with: [akohlmey at vitriol ~]$ cat ~/bin/clearipc.sh #!/bin/sh for s in `ipcs -m | grep $USER | cut -d \ -f 2` do \ ipcrm -m $s done for s in `ipcs -s | grep $USER | cut -d \ -f 2` do \ ipcrm -s $s done for s in `ipcs -q | grep $USER | cut -d \ -f 2` do \ ipcrm -q $s done MVA> ifort 9.1.041/mkl 8.0? Has anybody seen something similar when running MVA> pw.x on dual-processor machines? hope that explains some of the oddities. cheers, axel. MVA> MVA> Thanks, MVA> MVA> Marcos MVA> MVA> -- ======================================================================= 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 yaoxinxin at mail.sdu.edu.cn Thu Apr 26 16:40:02 2007 From: yaoxinxin at mail.sdu.edu.cn (Xinxin Yao) Date: Thu, 26 Apr 2007 22:40:02 +0800 Subject: [Pw_forum] problem in band structure calculation Message-ID: <377598402.11884@mail.sdu.edu.cn> Dear all, I meet a problem when doing band structure calculation. The message is: nbndx = 140 nbnd = 140 natomwfc = 256 npwx = 40800 nelec = 256.00 nkb = 192 ngl = 1519 Check: negative/imaginary core charge= -0.000024 0.000000 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from punch_band : error # 11 increase maxdeg %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... Any advice will be appreciated! Xinxin Yao From Giovanni.Cantele at na.infn.it Thu Apr 26 16:49:48 2007 From: Giovanni.Cantele at na.infn.it (Giovanni Cantele) Date: Thu, 26 Apr 2007 16:49:48 +0200 Subject: [Pw_forum] problem in band structure calculation In-Reply-To: <377598402.11884@mail.sdu.edu.cn> References: <377598402.11884@mail.sdu.edu.cn> Message-ID: <4630BC0C.2040307@na.infn.it> Xinxin Yao wrote: > Dear all, > I meet a problem when doing band structure calculation. The message is: > > nbndx = 140 nbnd = 140 natomwfc = 256 npwx = 40800 > nelec = 256.00 nkb = 192 ngl = 1519 > > Check: negative/imaginary core charge= -0.000024 0.000000 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from punch_band : error # 11 > increase maxdeg > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > Any advice will be appreciated! > > Xinxin Yao > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > Hi, I think the problem has been addressed one or more times in the forum. See, for example http://www.democritos.it/pipermail/pw_forum/2006-July/004504.html 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 hyello25 at gmail.com Thu Apr 26 17:59:50 2007 From: hyello25 at gmail.com (hyello25 Qgrui) Date: Thu, 26 Apr 2007 23:59:50 +0800 Subject: [Pw_forum] about the ibrav=0 Message-ID: hi, * When I use ibrav=0 to calculate the force of Si, I encounter some problems. Are there some thing wrong in my codes?* &system ibrav = 0, nat = 2, ntyp = 1, ecutwfc = 50, A = 5.39780000 , B = 5.39780000 , C = 5.39780000 , cosAB = 0.500000 , cosAC = 0.500000 , cosBC = 0.500000 , / &electrons mixing_beta = 0.7 conv_thr = 1.0d-7 / ATOMIC_SPECIES Si 28.086 Si.vbc.UPF ATOMIC_POSITIONS crystal Si 0.00000 0.00000 0.00000 Si 0.25000 0.25000 0.25000 CELL_PARAMETERS cubic 0.288675135 -0.500000000 0.816496581 0.288675135 0.500000000 0.816496581 -0.577350269 0.000000000 0.816496581 K_POINTS automatic 4 4 4 1 1 1 * The result is below, why does force exit in this situation?* ! total energy = -15.46091639 ryd estimated scf accuracy < 0.00000003 ryd band energy sum = -2.48914635 ryd one-electron contribution = -1.72655466 ryd hartree contribution = 2.24142346 ryd xc contribution = -4.02628209 ryd ewald contribution = -11.94950310 ryd convergence has been achieved Forces acting on atoms (Ry/au): * ** atom 1 type 1 force = 0.00314138 0.00000000 - 0.00856639 atom 2 type 1 force = -0.00314138 0.00000000 0.00856639* Total force = 0.012904 Total SCF correction = 0.000110 entering subroutine stress ... total stress (ryd/bohr**3) (kbar) P= - 86.51 -0.00058668 0.00000000 0.00000976 -86.30 0.00 1.44 0.00000000 -0.00058838 0.00000000 0.00 -86.55 0.00 0.00000976 0.00000000 -0.00058914 1.44 0.00 -86.67 *Another question :whether we need not ot set cosAB, cosAC, cosBC when we use ibrav=0?* thanks. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070426/1e3dc9dd/attachment.htm From w2agz at pacbell.net Thu Apr 26 18:11:21 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Thu, 26 Apr 2007 09:11:21 -0700 Subject: [Pw_forum] Lattice Constant Optimization of Aluminum In-Reply-To: <7C989E8A-1AD9-4A52-A4DF-FA58E46BFC38@nest.sns.it> References: <008501c787b4$b9716d30$0301a8c0@w2agz> <7C989E8A-1AD9-4A52-A4DF-FA58E46BFC38@nest.sns.it> Message-ID: <00af01c7881d$88107ec0$0301a8c0@w2agz> Got it...thanks. 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-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] On Behalf Of Paolo Giannozzi Sent: Thursday, April 26, 2007 12:05 AM To: pw_forum at pwscf.org Subject: Re: [Pw_forum] Lattice Constant Optimization of Aluminum On Apr 26, 2007, at 5:41 , Paul M. Grant wrote: > Even with a wildly wrong first estimate of celldm(1) of 7.2 (the > experimental value at RT is 7.665 au), I get zero net individual > atomic forces and thus no optimization. "optimization wrt internal positions" != "optimization of the lattice parameter(s)". Forces allow the former, stresses the latter. The simplest way to optimize the lattice parameter of a simple crystal is to fit a few E(V) points to an equation of state, using for instance "ev.f90" Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From nchamel at ulb.ac.be Thu Apr 26 18:47:10 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Thu, 26 Apr 2007 18:47:10 +0200 (CEST) Subject: [Pw_forum] Re: looking for a routine to symmetrize matrix elements Message-ID: <20070426164710.5DF3E4EE@bonito.ulb.ac.be> Thank you for your answer. Since I'm not an expert in this field, could you provide me some pedagogical references? Nicolas. Do you know some re >From: Nicolas Chamel >To: pw_forum at pwscf.org >Date: Thu, 26 Apr 2007 15:23:51 +0200 (CEST) >Subject: [Pw_forum] looking for a routine to symmetrize matrix elements >Reply-To: pw_forum at pwscf.org > >Dear users, > >I'm not working on electronic structure calculations but on an analogous problem in nuclear physics where neutrons play the role of electrons and ions are replaced by nuclei. I would like to use some parts of the code for my own calculations but I'm a bit lost in the code. For instance I would like to use a routine for the symmetrization of matrix elements for a given crystal structure, let's say bcc. I would really appreciate if someone could guide me though the code. > >Nicolas Chamel > >Institut d'Astronomie et d'Astrophysique >Universite Libre de Bruxelles >CP 226 >Boulevard du Triomphe >B-1050 Brussels >BELGIUM > >tel. +32 2 650 35 72 >fax. +32 2 650 28 64 > >--__--__-- > >Message: 4 >Date: Thu, 26 Apr 2007 09:58:00 -0400 >From: Nicola Marzari >Organization: Massachusetts Institute of Technology >To: pw_forum at pwscf.org >Subject: Re: [Pw_forum] looking for a routine to symmetrize matrix elements >Reply-To: pw_forum at pwscf.org > > > >Dear Nicolas, > > >I think it would be far easier for you to write the >symmetrization by hand, rather than reverse engineering >any code. > > nicola From baroni at sissa.it Thu Apr 26 19:04:04 2007 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 26 Apr 2007 18:04:04 +0100 Subject: [Pw_forum] looking for a routine to symmetrize matrix elements In-Reply-To: <4630AFE8.1080703@mit.edu> References: <20070426132351.EAC7840D@bonito.ulb.ac.be> <4630AFE8.1080703@mit.edu> Message-ID: <2D867A08-0663-45B2-A535-5F742F585499@sissa.it> also, no matrix elements are actually calculated, let alone symmetrized. what is simmetrized is the electron charge density. have fun - Stefano On Apr 26, 2007, at 2:58 PM, Nicola Marzari wrote: > > > Dear Nicolas, > > > I think it would be far easier for you to write the > symmetrization by hand, rather than reverse engineering > any code. > > nicola > > > Nicolas Chamel wrote: >> Dear users, >> I'm not working on electronic structure calculations but on an >> analogous problem in nuclear physics where neutrons play the role >> of electrons and ions are replaced by nuclei. I would like to use >> some parts of the code for my own calculations but I'm a bit lost >> in the code. For instance I would like to use a routine for the >> symmetrization of matrix elements for a given crystal structure, >> let's say bcc. I would really appreciate if someone could guide me >> though the code. Nicolas Chamel >> Institut d'Astronomie et d'Astrophysique >> Universite Libre de Bruxelles >> CP 226 >> Boulevard du Triomphe >> B-1050 Brussels >> BELGIUM >> tel. +32 2 650 35 72 >> fax. +32 2 650 28 64 >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum > > > -- > --------------------------------------------------------------------- > 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 --- 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: /pipermail/attachments/20070426/26d55e27/attachment.htm From mverissi at ictp.it Thu Apr 26 19:42:07 2007 From: mverissi at ictp.it (Marcos Verissimo Alves) Date: Thu, 26 Apr 2007 19:42:07 +0200 (CEST) Subject: [Pw_forum] Off-topic: memory usage issues In-Reply-To: References: <49540.10.50.40.120.1177594885.squirrel@webmail1.ictp.trieste.it> Message-ID: <38061.10.50.40.120.1177609327.squirrel@webmail1.ictp.trieste.it> Dear Paolo and Axel, Thanks for your replies. Axel: this was a quite enlightening explanation; maybe this topic could be included in the next School on Computational Methods? Cheers, Marcos -- Dr. Marcos Verissimo Alves Post-Doctoral Fellow Condensed Matter and Statistical Physics Sector International Centre for Theoretical Physics Trieste, Italy -------- I have become so addicted to vi that I try to exit OpenOffice by typing :wq! From nchamel at ulb.ac.be Thu Apr 26 20:05:09 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Thu, 26 Apr 2007 20:05:09 +0200 (CEST) Subject: [Pw_forum] 2D special points Message-ID: <20070426180509.AC7AF445@bonito.ulb.ac.be> Dear all, I'm looking for a set of special points for Brillouin zone integrations in 2D lattices. Does someone has references? Thanks in advance, Nicolas Chamel. Institut d'Astronomie et d'Astrophysique Universite Libre de Bruxelles CP 226 Boulevard du Triomphe B-1050 Brussels BELGIUM tel. +32 2 650 35 72 fax. +32 2 650 28 64 From baroni at sissa.it Thu Apr 26 20:25:42 2007 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 26 Apr 2007 19:25:42 +0100 Subject: [Pw_forum] 2D special points In-Reply-To: <20070426180509.AC7AF445@bonito.ulb.ac.be> References: <20070426180509.AC7AF445@bonito.ulb.ac.be> Message-ID: take the xy components of the special points of any crystal structure which has ona lattice vector orthogonal to the other two. this situation is the rule, e.g., for supercell surface- or interface- supercell calculations. SB On Apr 26, 2007, at 7:05 PM, Nicolas Chamel wrote: > Dear all, > > I'm looking for a set of special points for Brillouin zone > integrations in 2D lattices. Does someone has references? > > Thanks in advance, > Nicolas Chamel. > > Institut d'Astronomie et d'Astrophysique > Universite Libre de Bruxelles > CP 226 > Boulevard du Triomphe > B-1050 Brussels > BELGIUM > > tel. +32 2 650 35 72 > fax. +32 2 650 28 64 > _______________________________________________ > 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: /pipermail/attachments/20070426/89929213/attachment.htm From naromero at gmail.com Thu Apr 26 20:26:43 2007 From: naromero at gmail.com (Nichols A. Romero) Date: Thu, 26 Apr 2007 14:26:43 -0400 Subject: [Pw_forum] 2D special points In-Reply-To: <20070426180509.AC7AF445@bonito.ulb.ac.be> References: <20070426180509.AC7AF445@bonito.ulb.ac.be> Message-ID: <6ac064b60704261126m4ee676e5v634fb914db95ea7e@mail.gmail.com> This web page and the references contained there in might be useful: http://cst-www.nrl.navy.mil/bind/kpts/ On 4/26/07, Nicolas Chamel wrote: > > Dear all, > > I'm looking for a set of special points for Brillouin zone integrations in > 2D lattices. Does someone has references? > > Thanks in advance, > Nicolas Chamel. > > Institut d'Astronomie et d'Astrophysique > Universite Libre de Bruxelles > CP 226 > Boulevard du Triomphe > B-1050 Brussels > BELGIUM > > tel. +32 2 650 35 72 > fax. +32 2 650 28 64 > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Nichols A. Romero, Ph.D. 1613 Denise Dr. Apt. D Forest Hill, MD 21050 443-567-8328 (C) 410-306-0709 (O) -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070426/5a28e1fe/attachment.htm From marzari at MIT.EDU Thu Apr 26 20:45:13 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 26 Apr 2007 14:45:13 -0400 Subject: [Pw_forum] Re: looking for a routine to symmetrize matrix elements In-Reply-To: <20070426164710.5DF3E4EE@bonito.ulb.ac.be> References: <20070426164710.5DF3E4EE@bonito.ulb.ac.be> Message-ID: <4630F339.50503@mit.edu> Nicolas Chamel wrote: > Thank you for your answer. Since I'm not an expert in this field, could you provide me some pedagogical references? > > Nicolas. > A few people here like a lot "Group theory and its applications in physics" by Teturo Inui 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 baroni at sissa.it Thu Apr 26 21:00:17 2007 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 26 Apr 2007 20:00:17 +0100 Subject: [Pw_forum] Lattice Constant Optimization of Aluminum In-Reply-To: <008501c787b4$b9716d30$0301a8c0@w2agz> References: <008501c787b4$b9716d30$0301a8c0@w2agz> Message-ID: Dear Paul: you are not doing anything wrong, but expecting something the code cannot do (not even if it was a Perfect Schr?dinger Solver). Let me make a simple example that may convince you of what is missing. Take a monoatomic, periodic, linear chain out of equilibrium: what would the force acting on each atom be? 0, of course, by symmetry! How can it be, for a non-equilibrium configuration? Forces are zero by symmetry because you are considering (the code is considering) an infinite system. If the system was finite instead, forces on atoms near the surfaces would be nonvanishing, and the relaxation would (slowly) propagate from near the surface to deep into the bulk. The point is, in an infinite system vanishing forces are a necessary condition for equilibrium, but not a sufficient one. Vanishing forces AND stresses are both necessary and sufficient conditions for equilibrium to occur. If forces are already zero (by symmetry in your Al case, because each atomic site has cubic symmetry: notice that they are zero to machine precision, not just small) then you have to search for a configuration of zero pressure (which is large for your non- equilibrium configuration). This is achieved by vc-relax, or by fitting the E vs. volume relation to some equation of state, as suggested by somebody else in this forum recently. The latter is actually the preferred method for very simple systems (i.e. for systems whith very few degrees of freedom). Hope I did not misunderstood your question and that this clarifies. Yours - Stefano On Apr 26, 2007, at 4:41 AM, Paul M. Grant wrote: > To All: > > (BTW: PG: My last note on the ?default? value of npk should have > been termed ?default maximum.? Sorry, but we Americans haven?t > spoken English since 1776. Anyway, thanks for pointing me to the > right Fortran routine.) > > I?m trying to do a very simple homework problem with PWscf?the > optimization of the lattice constant of aluminum. > > Even with a wildly wrong first estimate of celldm(1) of 7.2 (the > experimental value at RT is 7.665 au), I get zero net individual > atomic forces and thus no optimization. I?ve gotten ?relax? in > PWscf to work with a lot more complex polymer monoclinic unit cell > containing eight atoms and two different elements. So, I must be > doing something really dumb (stupid)?maybe I should be using vc- > relax instead? Anyway, here?s the I/O (scusi e grazie): --- 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: /pipermail/attachments/20070426/63a284d3/attachment.htm From eyvaz_isaev at yahoo.com Thu Apr 26 21:01:54 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 26 Apr 2007 12:01:54 -0700 (PDT) Subject: [Pw_forum] 2D special points In-Reply-To: Message-ID: <756372.75262.qm@web60320.mail.yahoo.com> Hi, In addition to Stefano's theoretical introduction please have a look at the next papers: 1. Special points in the two-dimensional Brillouin zone, S. L. Cunningham, Phys. Rev. B 10, 4988 (1974) 2. Brillouin-zone integration by Fourier quadrature: Special points for superlattice and supercell calculations, Sverre Froyen. Phys. Rev. B 39, 3168 (1989) Bests, Eyvaz. --- Stefano Baroni wrote: > take the xy components of the special points of any > crystal structure which has ona lattice vector > orthogonal to the other two. this > situation is the rule, e.g., for supercell surface- > or interface- > supercell calculations. SB > > On Apr 26, 2007, at 7:05 PM, Nicolas Chamel wrote: > > > Dear all, > > > > I'm looking for a set of special points for > Brillouin zone > > integrations in 2D lattices. Does someone has > references? > > > > Thanks in advance, > > Nicolas Chamel. > > > > Institut d'Astronomie et d'Astrophysique > > Universite Libre de Bruxelles > > CP 226 > > Boulevard du Triomphe > > B-1050 Brussels > > BELGIUM > > > > tel. +32 2 650 35 72 > > fax. +32 2 650 28 64 > > _______________________________________________ > > 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 > > > > __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From baroni at sissa.it Thu Apr 26 21:15:35 2007 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 26 Apr 2007 20:15:35 +0100 Subject: [Pw_forum] 2D special points In-Reply-To: <756372.75262.qm@web60320.mail.yahoo.com> References: <756372.75262.qm@web60320.mail.yahoo.com> Message-ID: <3535503D-407E-4C42-A959-013A853758A7@sissa.it> On Apr 26, 2007, at 8:01 PM, Eyvaz Isaev wrote: > Hi, > > In addition to Stefano's theoretical introduction actually, I did not mean to be any "theoratical", but rather praticat instead ;-) ... S. > please have a look at the next papers: > > 1. Special points in the two-dimensional Brillouin > zone, S. L. Cunningham, Phys. Rev. B 10, 4988 (1974) > > 2. Brillouin-zone integration by Fourier quadrature: > Special points for superlattice and supercell > calculations, Sverre Froyen. Phys. Rev. B 39, 3168 > (1989) > > Bests, > Eyvaz. > > > --- Stefano Baroni wrote: > >> take the xy components of the special points of any >> crystal structure which has ona lattice vector >> orthogonal to the other two. this >> situation is the rule, e.g., for supercell surface- >> or interface- >> supercell calculations. SB >> >> On Apr 26, 2007, at 7:05 PM, Nicolas Chamel wrote: >> >>> Dear all, >>> >>> I'm looking for a set of special points for >> Brillouin zone >>> integrations in 2D lattices. Does someone has >> references? >>> >>> Thanks in advance, >>> Nicolas Chamel. >>> >>> Institut d'Astronomie et d'Astrophysique >>> Universite Libre de Bruxelles >>> CP 226 >>> Boulevard du Triomphe >>> B-1050 Brussels >>> BELGIUM >>> >>> tel. +32 2 650 35 72 >>> fax. +32 2 650 28 64 >>> _______________________________________________ >>> 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 >> >> >> >> > > > __________________________________________________ > 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 --- 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: /pipermail/attachments/20070426/a6611bab/attachment.htm From nchamel at ulb.ac.be Thu Apr 26 22:22:53 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Thu, 26 Apr 2007 22:22:53 +0200 (CEST) Subject: [Pw_forum] 2D special points Message-ID: <20070426202253.078D941D@bonito.ulb.ac.be> >take the xy components of the special points of any crystal structure >which has ona lattice vector orthogonal to the other two. this >situation is the rule, e.g., for supercell surface- or interface- >supercell calculations. SB Sounds like a pretty smart idea! And the weights are still the same, right? Nicolas. From nchamel at ulb.ac.be Thu Apr 26 22:27:20 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Thu, 26 Apr 2007 22:27:20 +0200 (CEST) Subject: [Pw_forum] 2D special points Message-ID: <20070426202720.44590492@bonito.ulb.ac.be> >This web page and the references contained there in might be useful: >http://cst-www.nrl.navy.mil/bind/kpts/ Thank you very much. Actually for 3D calculations I'm currently using the formulae of reference J. Hama, M. Watanabe, General formulae for the special points and their weighting factors in k-space integration, J. Phys.: Condens. Matter 4 (1992) 4583?4594. For someone like me who doesn't know much how to determine the weights of the special points, these formulae are providential! Nicolas Chamel. Institut d'Astronomie et d'Astrophysique Universite Libre de Bruxelles CP 226 Boulevard du Triomphe B-1050 Brussels BELGIUM tel. +32 2 650 35 72 fax. +32 2 650 28 64 From nchamel at ulb.ac.be Thu Apr 26 22:28:26 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Thu, 26 Apr 2007 22:28:26 +0200 (CEST) Subject: [Pw_forum] Re: looking for a routine to symmetrize matrix elements Message-ID: <20070426202826.EA3D9451@bonito.ulb.ac.be> >A few people here like a lot >"Group theory and its applications in physics" by Teturo Inui Thank you very much! Nicolas Chamel. Institut d'Astronomie et d'Astrophysique Universite Libre de Bruxelles CP 226 Boulevard du Triomphe B-1050 Brussels BELGIUM tel. +32 2 650 35 72 fax. +32 2 650 28 64 From nchamel at ulb.ac.be Thu Apr 26 22:38:49 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Thu, 26 Apr 2007 22:38:49 +0200 (CEST) Subject: [Pw_forum] 2D special points Message-ID: <20070426203849.9047F497@bonito.ulb.ac.be> Thank you for your answer. Actually I know these references but to me it's a bit cumbersome to use in practice. I don't understand very well how to find the associated weigths in a systematic way with such recursive methods. Nicolas Chamel Institut d'Astronomie et d'Astrophysique Universite Libre de Bruxelles CP 226 Boulevard du Triomphe B-1050 Brussels BELGIUM tel. +32 2 650 35 72 fax. +32 2 650 28 64 >Hi, > >In addition to Stefano's theoretical introduction >please have a look at the next papers: > >1. Special points in the two-dimensional Brillouin >zone, S. L. Cunningham, Phys. Rev. B 10, 4988 (1974) > >2. Brillouin-zone integration by Fourier quadrature: >Special points for superlattice and supercell >calculations, Sverre Froyen. Phys. Rev. B 39, 3168 >(1989) > >Bests, >Eyvaz. From nchamel at ulb.ac.be Thu Apr 26 22:46:00 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Thu, 26 Apr 2007 22:46:00 +0200 (CEST) Subject: [Pw_forum] mean value point for arbitrary lattices Message-ID: <20070426204600.2FF5A4B1@bonito.ulb.ac.be> The mean value point method of Baldereschi for Brillouin zone integrations (A. Baldereschi, Mean value point in the Brillouin zone, Phys. Rev. B 7 (12) (1973) 5212?5215.) turns out to be surprisingly accurate for some quantities like the total energy density for instance. Does someone knows if this mean value point has been determined for 3D and 2D lattices apart from the points given by Baldereschi himself? Nicolas Chamel. Institut d'Astronomie et d'Astrophysique Universite Libre de Bruxelles CP 226 Boulevard du Triomphe B-1050 Brussels BELGIUM tel. +32 2 650 35 72 fax. +32 2 650 28 64 From w2agz at pacbell.net Fri Apr 27 03:41:40 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Thu, 26 Apr 2007 18:41:40 -0700 Subject: [Pw_forum] Lattice Constant Optimization of Aluminum In-Reply-To: References: <008501c787b4$b9716d30$0301a8c0@w2agz> Message-ID: <00d001c7886d$34965a30$0301a8c0@w2agz> Wow! Thanks for thinking so hard about my predicament! OK, let?s see. I follow your points about the code. As usual, my blunder was not looking at the source behind ?relax.? It would be great if there were a ?guide to the source? a la the way PW_INPUT serves for the namelists, that is, which ?*.f90? does what? Under the covers, I find the PWscf source well commented, though, when I can find it. Now, it?s well known that a linear, infinite, periodic metallic chain is unstable to dimerization (Peiels-Froehlich or static CDW instability) inasmuch as gapping the planar Fermi surface by commensurate nesting with respect to the 1D Brillouin zone boundary lowers the ground state energy of the system. That is, all (quasi)-1D metals sit on top of a metastable hill in the local configurational ground state space and will fall into an insulating state under an infinitesimal change in lattice constant, even if the new equilibrium is a quasiperiodic chain (e.g., ?Fibonacci sequence?), if indeed such a thing can exist. On the other hand, in 2D or 3D metals, we?re saved by the fact that most Fermi surfaces have a high overall degree of curvature with respect to their polygonal BZs, thus suppressing a tendency to nest (the A15 compounds are somewhat an exception that?s why they?re strong-coupled superconductors). At the end of the day, at T = 0, all Landau-Fermi metals are ?gapped? by some arbitrarily weak electron-electron interaction only superconductors, antiferromagnets, and possibly ferromagnets can exist (John Hubbard once told me he didn?t think itinerant ferromagnets were ?real metals.?). So, what you?re telling me is that pw.x with ?relax? will not find the Peierls state of a periodic chain of, say, sodium atoms, or polyacetylene with equal carbon-carbon bond lengths, but presumably ?vc-relax? will? (I tried playing with vc-relax last night on Al, but fell asleep on my keyboard before I could get it to work I?ll try again this evening). Naturally, I ?manually optimized? Al with trial ?scf? runs before all of this (I should have thought about using Paolo?s equation of state fit, but I?d didn?t know it existed within PCscf), and believed ?relax? sort of worked the same way. Up at Stanford I use the CASTEP geometry optimization package which on simple structures ?jiggles? all atoms in a symmetry preserving way (i.e., cubic cells stay cubic) in its search for local configurational total energy minima. I see now that PWscf separates this task into ?relax? and ?vc-relax.? By the way, PWscf is a MUCH better computational tool than CASTEP, moreover it?s free, but what?s most important, as any reader of pw_forum can see, encompasses a far more collegial user support community. Sorry for the above polemic, Stefano, but I spent a good deal of my life trying to turn quasi-1D metals into superconductors! 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 _____ From: pw_forum-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] On Behalf Of Stefano Baroni Sent: Thursday, April 26, 2007 12:00 PM To: pw_forum at pwscf.org Subject: Re: [Pw_forum] Lattice Constant Optimization of Aluminum Dear Paul: you are not doing anything wrong, but expecting something the code cannot do (not even if it was a Perfect Schr?dinger Solver). Let me make a simple example that may convince you of what is missing. Take a monoatomic, periodic, linear chain out of equilibrium: what would the force acting on each atom be? 0, of course, by symmetry! How can it be, for a non-equilibrium configuration? Forces are zero by symmetry because you are considering (the code is considering) an infinite system. If the system was finite instead, forces on atoms near the surfaces would be nonvanishing, and the relaxation would (slowly) propagate from near the surface to deep into the bulk. The point is, in an infinite system vanishing forces are a necessary condition for equilibrium, but not a sufficient one. Vanishing forces AND stresses are both necessary and sufficient conditions for equilibrium to occur. If forces are already zero (by symmetry in your Al case, because each atomic site has cubic symmetry: notice that they are zero to machine precision, not just small) then you have to search for a configuration of zero pressure (which is large for your non-equilibrium configuration). This is achieved by vc-relax, or by fitting the E vs. volume relation to some equation of state, as suggested by somebody else in this forum recently. The latter is actually the preferred method for very simple systems (i.e. for systems whith very few degrees of freedom). Hope I did not misunderstood your question and that this clarifies. Yours - Stefano On Apr 26, 2007, at 4:41 AM, Paul M. Grant wrote: To All: (BTW: PG: My last note on the ?default? value of npk should have been termed ?default maximum.? Sorry, but we Americans haven?t spoken English since 1776. Anyway, thanks for pointing me to the right Fortran routine.) I?m trying to do a very simple homework problem with PWscf the optimization of the lattice constant of aluminum. Even with a wildly wrong first estimate of celldm(1) of 7.2 (the experimental value at RT is 7.665 au), I get zero net individual atomic forces and thus no optimization. I?ve gotten ?relax? in PWscf to work with a lot more complex polymer monoclinic unit cell containing eight atoms and two different elements. So, I must be doing something really dumb (stupid) maybe I should be using vc-relax instead? Anyway, here?s the I/O (scusi e grazie): --- 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: /pipermail/attachments/20070426/37a3a340/attachment.htm From w2agz at pacbell.net Fri Apr 27 04:07:59 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Thu, 26 Apr 2007 19:07:59 -0700 Subject: [Pw_forum] iotk_scan_dat & dos.x Message-ID: <00d501c78870$e1535cc0$0301a8c0@w2agz> I'm having trouble doing something simple (again).the density of states of silicon. I keep getting during execution of dos.x the following error (the scf calculation is fine): # UNRECOVERABLE ERROR (ierr=1) # ERROR IN: iotk_scan_dat (iotk_dat.spp:688) # CVS Revision: 1.3 With input &inputpp outdir = '/home/pmpgrant/Research_Projects/DFT_Calculations/Silicon/2007_4/' prefix = 'Si_2007_4_K32_0' fildos = '/home/pmpgrant/Research_Projects/DFT_Calculations/Silicon/2007_4/Si_2007_4_ K32_0.dos' / After running the script: cat > Si_2007_4_K32_0.doscalc.in << EOF &inputpp outdir = '/home/pmpgrant/Research_Projects/DFT_Calculations/Silicon/2007_4/' prefix = 'Si_2007_4_K32_0' fildos = '/home/pmpgrant/Research_Projects/DFT_Calculations/Silicon/2007_4/Si_2007_4_ K32_0.dos' / EOF /home/pmpgrant/espresso-3.2/bin/dos.x < Si_2007_4_K32_0.doscalc.in > Si_2007_4_K32_0.doscalc.out Drilling and surfing around the forum archives, it seems this type of error may be associated with multiple MB processors (I have a dual Xeon board) or clusters. 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: /pipermail/attachments/20070426/869346ea/attachment.htm From rcjhawk at gmail.com Fri Apr 27 04:23:01 2007 From: rcjhawk at gmail.com (Michael Mehl) Date: Thu, 26 Apr 2007 22:23:01 -0400 Subject: [Pw_forum] 2D special points In-Reply-To: <6ac064b60704261126m4ee676e5v634fb914db95ea7e@mail.gmail.com> References: <20070426180509.AC7AF445@bonito.ulb.ac.be> <6ac064b60704261126m4ee676e5v634fb914db95ea7e@mail.gmail.com> Message-ID: <46315E85.2060609@gmail.com> Nichols A. Romero wrote: > This web page and the references contained there in might be useful: > http://cst-www.nrl.navy.mil/bind/kpts/ > Unfortunately, I didn't include any two dimensional cells on that page. Stefano outlined the correct procedure. -- mjm http://hawknotes.blogspot.com/ http://cst-www.nrl.navy.mil/lattice From hongyi.zhao at gmail.com Fri Apr 27 04:38:07 2007 From: hongyi.zhao at gmail.com (Hongyi Zhao) Date: Fri, 27 Apr 2007 10:38:07 +0800 Subject: [Pw_forum] What's meaning of the origin of the first BZ and which planes are corresponding to it in the real lattice space? Message-ID: <20070427023807.GA4012@debian-sid-hy.net> Hi all, I can understand the concept of the first Brillouin Zone, but I cann't figure out the meaning of the origin of the first BZ and which planes are corresponding to it in the real lattice space. I mean, the origin coordinate of the first Brillouin Zone is (0, 0, 0), so the corresponding plane(s) to it in the real lattice space should have the index (/infty /infty /infty ). But what does this plane index mean, I just cann't understand. Would anyone give me some hints on this? Thanks in advance. --- Hongyi Zhao GnuPG DSA: 0xD108493B From clie at aphy.iphy.ac.cn Fri Apr 27 05:41:21 2007 From: clie at aphy.iphy.ac.cn (clie) Date: Fri, 27 Apr 2007 03:41:21 GMT Subject: [Pw_forum] Re: Pw_forum digest, Vol 1 #1445 - 7 msgs In-Reply-To: <20070426132402.766.56502.Mailman@democritos.sissa.it> References: <20070426132402.766.56502.Mailman@democritos.sissa.it> Message-ID: <20070427034121.14392.eqmail@aphy.iphy.ac.cn> At the end of GeH4.elph.out file, it says: from lint: error# 261 can't remap grid on k-point list Dear clie, You should look, at least, at the last lines of GeH4.elph.out. It should give a more detailed message than "2". Maybe some hints on starting density files or q-point choice. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070427/ca783e00/attachment.htm From degironc at sissa.it Fri Apr 27 10:10:50 2007 From: degironc at sissa.it (degironc) Date: Fri, 27 Apr 2007 10:10:50 +0200 Subject: [Pw_forum] What's meaning of the origin of the first BZ and which planes are corresponding to it in the real lattice space? In-Reply-To: <20070427023807.GA4012@debian-sid-hy.net> References: <20070427023807.GA4012@debian-sid-hy.net> Message-ID: <4631B00A.5070500@sissa.it> BZ has to do with the symmetry properties of a lattice (described by a infinite discrete-translation group). k points in the BZ label possible periodicity of wfc, phonons, or other function defined for this lattice. They give you the complex phase factor involved in a discrete lattice translation. Gamma points means that all these phase faqctors are equal to 1... why do you want to see it as a plane in real space? Any how if you want to see it pictorially the question is: find the set of points in real space such that the accumulated phase shift is pi ! Start with a small k vector and approach gamma ... what happens to the plane? If I rememeber correctly this subject is covered in a basic geometry or linear algebra course as "projective geometry" ... at least it was so 25 yrs ago... hope this helps, stefano Hongyi Zhao wrote: >Hi all, > >I can understand the concept of the first Brillouin Zone, but I cann't >figure out the meaning of the origin of the first BZ and which planes >are corresponding to it in the real lattice space. I mean, the origin >coordinate of the first Brillouin Zone is (0, 0, 0), so the >corresponding plane(s) to it in the real lattice space should have the >index (/infty /infty /infty ). But what does this plane index mean, I >just cann't understand. Would anyone give me some hints on this? > >Thanks in advance. > >--- >Hongyi Zhao >GnuPG DSA: 0xD108493B > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From paulatto at sissa.it Fri Apr 27 11:13:19 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Fri, 27 Apr 2007 11:13:19 +0200 (CEST) Subject: [Pw_forum] mean value point for arbitrary lattices In-Reply-To: <20070426204600.2FF5A4B1@bonito.ulb.ac.be> References: <20070426204600.2FF5A4B1@bonito.ulb.ac.be> Message-ID: <10113.147.122.5.194.1177665199.squirrel@webmail.sissa.it> On Thu, April 26, 2007 22:46, Nicolas Chamel wrote: > Does someone knows > if this mean value point has been determined for 3D and 2D lattices apart > from the points given by Baldereschi himself? Dear Nicolas, I am not sure I understood you question correctly, I guess you are thinking of H.J. Mokhorst & J.D. Pack, Phys. Rev. B 13, 5188 (1976). -- Lorenzo Paulatto +39 040 3787 312 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From baroni at sissa.it Fri Apr 27 13:45:45 2007 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 27 Apr 2007 12:45:45 +0100 Subject: [Pw_forum] Lattice Constant Optimization of Aluminum In-Reply-To: <00d001c7886d$34965a30$0301a8c0@w2agz> References: <008501c787b4$b9716d30$0301a8c0@w2agz> <00d001c7886d$34965a30$0301a8c0@w2agz> Message-ID: On Apr 27, 2007, at 2:41 AM, Paul M. Grant wrote: > ... > Now, it?s well known that a linear, infinite, periodic metallic > chain is unstable to dimerization (Peiels-Froehlich or static CDW > instability) inasmuch as gapping the planar Fermi surface by > commensurate nesting with respect to the 1D Brillouin zone boundary > lowers the ground state energy of the system. That is, all > (quasi)-1D metals sit on top of a metastable hill in the local > configurational ground state space and will fall into an insulating > state under an infinitesimal change in lattice constant, even if > the new equilibrium is a quasiperiodic chain (e.g., ?Fibonacci > sequence?), if indeed such a thing can exist. On the other hand, > in 2D or 3D metals, we?re saved by the fact that most Fermi > surfaces have a high overall degree of curvature with respect to > their polygonal BZs, thus suppressing a tendency to nest (the A15 > compounds are somewhat an exception?that?s why they?re strong- > coupled superconductors). At the end of the day, at T = 0, all > Landau-Fermi metals are ?gapped? by some arbitrarily weak electron- > electron interaction?only superconductors, antiferromagnets, and > possibly ferromagnets can exist (John Hubbard once told me he > didn?t think itinerant ferromagnets were ?real metals.?). > > > > So, what you?re telling me is that pw.x with ?relax? will not find > the Peierls state of a periodic chain of, say, sodium atoms, or > polyacetylene with equal carbon-carbon bond lengths, but presumably > ?vc-relax? will? (I tried playing with vc-relax last night on Al, > but fell asleep on my keyboard before I could get it to work?I?ll > try again this evening). NO, I was not saying this. In order to find dimerization (to have the code find it) you have to double the unit cell and slightly displace the atoms from their mono-atomic chain equilibrium positions. In order to relax to the equilibrium lattice constant, you have to allow the latter to vary (which "cv-relax" does, while "relax" does not). If you do "relax" on a stable monoatomic chain off equilibrium or on a Peierls-unstable chain, atoms would not move from their positions, but for two different reasons (at least, different in my understanding). > Naturally, I ?manually optimized? Al with trial ?scf? runs before > all of this (I should have thought about using Paolo?s equation of > state fit, but I?d didn?t know it existed within PCscf), and > believed ?relax? sort of worked the same way. Up at Stanford I use > the CASTEP geometry optimization package which on simple structures > ?jiggles? all atoms in a symmetry preserving way (i.e., cubic cells > stay cubic) in its search for local configurational total energy > minima. I see now that PWscf separates this task into ?relax? and > ?vc-relax.? By the way, PWscf is a MUCH better computational tool > than CASTEP, moreover it?s free, but what?s most important, as any > reader of pw_forum can see, encompasses a far more collegial user > support community. We do appreciate your praise, thanks. > Sorry for the above polemic, Stefano, but I spent a good deal of > my life trying to turn quasi-1D metals into superconductors! I know, Paul. I was, how to say? a bit too pedagogic? too prolix? not for your sake, but the kids' sake. Hope I was not too wrong! Ciao - Stefano --- 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: /pipermail/attachments/20070427/93146524/attachment.htm From baroni at sissa.it Fri Apr 27 13:47:59 2007 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 27 Apr 2007 12:47:59 +0100 Subject: [Pw_forum] 2D special points In-Reply-To: <20070426202253.078D941D@bonito.ulb.ac.be> References: <20070426202253.078D941D@bonito.ulb.ac.be> Message-ID: <75F531D2-A32E-48AE-84F3-6201398072AC@sissa.it> yes. the RELATIVE weight of the 2D k points are the same as those of the 3D k point with a same value of the z component. S. On Apr 26, 2007, at 9:22 PM, Nicolas Chamel wrote: >> take the xy components of the special points of any crystal structure >> which has ona lattice vector orthogonal to the other two. this >> situation is the rule, e.g., for supercell surface- or interface- >> supercell calculations. SB > > Sounds like a pretty smart idea! And the weights are still the > same, right? > > Nicolas. > _______________________________________________ > 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: /pipermail/attachments/20070427/c4959d32/attachment.htm From nchamel at ulb.ac.be Fri Apr 27 14:45:44 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Fri, 27 Apr 2007 14:45:44 +0200 (CEST) Subject: [Pw_forum] mean value point for arbitrary lattices Message-ID: <20070427124544.39C7B672@bonito.ulb.ac.be> >Dear Nicolas, >I am not sure I understood you question correctly, I guess you are >thinking of H.J. Mokhorst & J.D. Pack, Phys. Rev. B 13, 5188 (1976). > >-- >Lorenzo Paulatto >+39 040 3787 312 >http://people.sissa.it/~paulatto/ Thanks for your reply. The scheme proposed by Monkhorst and Pack is different from that of Baldereschi (who was looking for just one point) and is not optimized to find only one good special point. In the method of Monkhorst and Pack, considering only one special point leads to k=0 for any lattice. I'm wondering if the mean value point of Baldereschi has been calculated for any kinds of lattices and tabulated somewhere as a function of the lattice constants. But probably it's faster to find it myself. Nicolas Chamel Institut d'Astronomie et d'Astrophysique Universite Libre de Bruxelles CP 226 Boulevard du Triomphe B-1050 Brussels BELGIUM tel. +32 2 650 35 72 fax. +32 2 650 28 64 From degironc at sissa.it Fri Apr 27 14:59:53 2007 From: degironc at sissa.it (degironc) Date: Fri, 27 Apr 2007 14:59:53 +0200 Subject: [Pw_forum] mean value point for arbitrary lattices In-Reply-To: <20070427124544.39C7B672@bonito.ulb.ac.be> References: <20070427124544.39C7B672@bonito.ulb.ac.be> Message-ID: <4631F3C9.5050506@sissa.it> The paper by Baldereschi on the mean value point [Phys. Rev. B 7, 5212 - 5215 (1973)] is the "father" of all schemes on BZ integration based on "mean value" or "special" points. A generalization of Baldereschi's idea to more than just one point is given by Chadi and Cohen, Phys. Rev. B 8, 5747 - 5753 (1973). Monkhorst and Pack give a simpler (in my opinion) and more flexible generation scheme. For cubic lattices the two schemes lead to the same set of points. Accurate integration with just a few points is only possible when the system is highly symmetric. stefano Nicolas Chamel wrote: >>Dear Nicolas, >>I am not sure I understood you question correctly, I guess you are >>thinking of H.J. Mokhorst & J.D. Pack, Phys. Rev. B 13, 5188 (1976). >> >>-- >>Lorenzo Paulatto >>+39 040 3787 312 >>http://people.sissa.it/~paulatto/ >> >> > >Thanks for your reply. The scheme proposed by Monkhorst and Pack is different from that of Baldereschi (who was looking for just one point) and is not optimized to find only one good special point. In the method of Monkhorst and Pack, considering only one special point leads to k=0 for any lattice. > >I'm wondering if the mean value point of Baldereschi has been calculated for any kinds of lattices and tabulated somewhere as a function of the lattice constants. But probably it's faster to find it myself. > >Nicolas Chamel > >Institut d'Astronomie et d'Astrophysique >Universite Libre de Bruxelles >CP 226 >Boulevard du Triomphe >B-1050 Brussels >BELGIUM > >tel. +32 2 650 35 72 >fax. +32 2 650 28 64 >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From paulatto at sissa.it Fri Apr 27 15:05:08 2007 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Fri, 27 Apr 2007 15:05:08 +0200 (CEST) Subject: [Pw_forum] mean value point for arbitrary lattices In-Reply-To: <20070427124544.39C7B672@bonito.ulb.ac.be> References: <20070427124544.39C7B672@bonito.ulb.ac.be> Message-ID: <17539.147.122.5.194.1177679108.squirrel@webmail.sissa.it> On Fri, April 27, 2007 14:45, Nicolas Chamel wrote: > Thanks for your reply. The scheme proposed by Monkhorst and Pack is > different from that of Baldereschi (who was looking for just one point) > and is not optimized to find only one good special point. In the method of > Monkhorst and Pack, considering only one special point leads to k=0 for > any lattice. I think that the work from Chadi and Cohen (PRB, 1974) may provide some hint for you; as far as I know there was no systematic classification of mean value points for "less common" lattices. Bye -- Lorenzo Paulatto +39 040 3787 312 http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From ferretti.andrea at unimore.it Fri Apr 27 17:35:39 2007 From: ferretti.andrea at unimore.it (Andrea Ferretti) Date: Fri, 27 Apr 2007 17:35:39 +0200 (CEST) Subject: [Pw_forum] iotk_scan_dat & dos.x In-Reply-To: <00d501c78870$e1535cc0$0301a8c0@w2agz> References: <00d501c78870$e1535cc0$0301a8c0@w2agz> Message-ID: Hi Paul, I tried locally on my machine with a silicon test case and for me it works... as you are saying, it seems therefore it is a machine dependent stuff... did you try also other postprocessing or nscf calculations ?? could you please also post the make.sys file generated during configuration ?? (and eventually the config.log file) andrea On Thu, 26 Apr 2007, Paul M. Grant wrote: > I'm having trouble doing something simple (again).the density of states of > silicon. I keep getting during execution of dos.x the following error (the > scf calculation is fine): > > > > # UNRECOVERABLE ERROR (ierr=1) > # ERROR IN: iotk_scan_dat (iotk_dat.spp:688) > # CVS Revision: 1.3 > > > Drilling and surfing around the forum archives, it seems this type of error > may be associated with multiple MB processors (I have a dual Xeon board) or > clusters. > From w2agz at pacbell.net Fri Apr 27 20:41:34 2007 From: w2agz at pacbell.net (Paul M. Grant) Date: Fri, 27 Apr 2007 11:41:34 -0700 Subject: [Pw_forum] iotk_scan_dat & dos.x In-Reply-To: References: <00d501c78870$e1535cc0$0301a8c0@w2agz> Message-ID: <002701c788fb$af613260$0301a8c0@w2agz> Hi Andrea, Yeah, I've done a fair amount of nscf post-processing with the PWscf package...DOS, band structures, Fermi surfaces, e-p coupling...without problems, but all on metals so far. This is my first "homework problem" on a semiconductor (outside the examples, which all ran fine...at least four or five months ago). Let me investigate what's going on with my silicon exercise a little more and I'll get back. 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-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] On Behalf Of Andrea Ferretti Sent: Friday, April 27, 2007 8:36 AM To: pw_forum at pwscf.org Subject: Re: [Pw_forum] iotk_scan_dat & dos.x Hi Paul, I tried locally on my machine with a silicon test case and for me it works... as you are saying, it seems therefore it is a machine dependent stuff... did you try also other postprocessing or nscf calculations ?? could you please also post the make.sys file generated during configuration ?? (and eventually the config.log file) andrea On Thu, 26 Apr 2007, Paul M. Grant wrote: > I'm having trouble doing something simple (again).the density of states of > silicon. I keep getting during execution of dos.x the following error (the > scf calculation is fine): > > > > # UNRECOVERABLE ERROR (ierr=1) > # ERROR IN: iotk_scan_dat (iotk_dat.spp:688) > # CVS Revision: 1.3 > > > Drilling and surfing around the forum archives, it seems this type of error > may be associated with multiple MB processors (I have a dual Xeon board) or > clusters. > _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From niuli1978 at yahoo.com.cn Sat Apr 28 01:22:44 2007 From: niuli1978 at yahoo.com.cn (li niu) Date: Sat, 28 Apr 2007 07:22:44 +0800 (CST) Subject: [Pw_forum] finite electric fields In-Reply-To: <20070426132402.766.56502.Mailman@democritos.sissa.it> Message-ID: <317951.14774.qm@web15003.mail.cnb.yahoo.com> Dear Paolo U, Thanks for your reply. I'm very interested for your finite electric field method. The dielectric polarizability tensor can be calculated by taking dinite differences of atomic forces. I have several questions to ask you. 1. We obtained the diagonal terms, by considering values for the electric field of -2h,-h, 0, h and 2h in the cp.x code [tefield=.true. tcg = .true., epol = 1 or 2 or 3 efield = -2h,-h, 0, h or 2h], and by using the five-point formular . Is it right? 2. For the calculation of the off-diagonal terms, two different electric fields are required. This need apply electric fields of intensity -2h/1.414,-h/1.414, 0, h/1.414 and 2h/1.414 along the three different couples of cartesian directions, and apply eauation. But how can I apply simultaneously finite fields along two different Cartesian directions in cp.x code?[ESSPRESSO-3.2] 3. In this way, we could obtain all the dielectric polarizability tensor. Then the Raman susceptibilities associated to the normal mode n can be obtianed according to the equation. How to implement it ? By a small code? Best! I'm counting on the new version. Niu Li --------------------------------- ????????3.5G???20M??? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070428/eeb0e762/attachment.htm From nchamel at ulb.ac.be Sat Apr 28 11:27:45 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Sat, 28 Apr 2007 11:27:45 +0200 (CEST) Subject: [Pw_forum] mean value point for arbitrary lattices Message-ID: <20070428092745.C7A06485@bonito.ulb.ac.be> >The paper by Baldereschi on the mean value point [Phys. Rev. B 7, 5212 >- 5215 (1973)] is the "father" of all schemes on BZ integration based >on "mean value" or "special" points. >A generalization of Baldereschi's idea to more than just one point is >given by Chadi and Cohen, Phys. Rev. B 8, 5747 - 5753 (1973). >Monkhorst and Pack give a simpler (in my opinion) and more flexible >generation scheme. For cubic lattices the two schemes lead to the same >set of points. >Accurate integration with just a few points is only possible when the >system is highly symmetric. Actually I'm using for my calculations the formulae of Hama et al (I gave the reference in an earlier message) which provide the coordinates and the weigths of an arbitrary number of special points. The code thus computes BZ integrals by increasing the number of points until a given precision is reached. This guarantees that all quantities are computed with the same precision. This is however computationnaly very expensive and in some cases (computation of Fermi energies for instance) rather unnecessary. This is the reason why I'm also interested in the method of Baldereschi. Nicolas Chamel. Institut d'Astronomie et d'Astrophysique Universite Libre de Bruxelles CP 226 Boulevard du Triomphe B-1050 Brussels BELGIUM tel. +32 2 650 35 72 fax. +32 2 650 28 64 From nchamel at ulb.ac.be Sat Apr 28 11:32:02 2007 From: nchamel at ulb.ac.be (Nicolas Chamel) Date: Sat, 28 Apr 2007 11:32:02 +0200 (CEST) Subject: [Pw_forum] (no subject) Message-ID: <20070428093202.4DA784AF@bonito.ulb.ac.be> >I think that the work from Chadi and Cohen (PRB, 1974) may provide some >hint for you; as far as I know there was no systematic classification of >mean value points for "less common" lattices. > >Bye Thanks for your reply. In fact I was looking for the coordinates of the mean value point in 2D lattices. By reading back the paper of Cunningham, Physical Review B, vol. 10, Issue 12, pp. 4988-4994 (1974), I found that he gave the coordinates of special points as well as the mean value point! :-) Nicolas Chamel. Institut d'Astronomie et d'Astrophysique Universite Libre de Bruxelles CP 226 Boulevard du Triomphe B-1050 Brussels BELGIUM tel. +32 2 650 35 72 fax. +32 2 650 28 64 From sagars_daya at yahoo.co.in Sun Apr 29 00:30:54 2007 From: sagars_daya at yahoo.co.in (Daya sagar) Date: Sun, 29 Apr 2007 04:00:54 +0530 (IST) Subject: [Pw_forum] Can anyone help me with bandstructure calculations of GaN Message-ID: <373763.52967.qm@web8327.mail.in.yahoo.com> Hi all, I am doing the band structure calculations of GaN nanowires. I am using VASP for the calculations. I have no idea of the KPOINTS in brillouin zone for calculating the band structures. Can anyone help me with the KPOINTS file for GaN nanowire. Thank in advance Regards, Dayasagar. Send a FREE SMS to your friend's mobile from Yahoo! Messenger. Get it now at http://in.messenger.yahoo.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070429/60091e47/attachment.htm From akohlmey at cmm.chem.upenn.edu Sun Apr 29 01:48:06 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 28 Apr 2007 19:48:06 -0400 Subject: [Pw_forum] Can anyone help me with bandstructure calculations of GaN In-Reply-To: <373763.52967.qm@web8327.mail.in.yahoo.com> References: <373763.52967.qm@web8327.mail.in.yahoo.com> Message-ID: <7b6913e90704281648x6f38466bka3425947849ae64e@mail.gmail.com> On 4/28/07, Daya sagar wrote: > > Hi all, > > I am doing the band structure calculations of GaN nanowires. I am using > VASP for the calculations. I have no idea of the KPOINTS in brillouin zone since you want to use VASP, how about asking in the corresponding VASP forum or switching to quantum espresso? cheers, axel. > for calculating the band structures. Can anyone help me with the KPOINTS > file for GaN nanowire. > > Thank in advance > > Regards, > Dayasagar. > ________________________________ > Check out what you're missing if you're not on Yahoo! Messenger -- ======================================================================= 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 pbsorokin at gmail.com Sun Apr 29 13:36:51 2007 From: pbsorokin at gmail.com (Pavel B. Sorokin) Date: Sun, 29 Apr 2007 15:36:51 +0400 Subject: [Pw_forum] Can anyone help me with bandstructure calculations of GaN In-Reply-To: <373763.52967.qm@web8327.mail.in.yahoo.com> References: <373763.52967.qm@web8327.mail.in.yahoo.com> Message-ID: Anyway, read carefully the url http://cms.mpi.univie.ac.at/vasp/vasp/node57.html and keep in mind that in your case only one direction is periodical thus only in this one you should input number of kpoints more than one 2007/4/29, Daya sagar : > > Hi all, > > I am doing the band structure calculations of GaN nanowires. I am using > VASP for the calculations. I have no idea of the KPOINTS in brillouin zone > for calculating the band structures. Can anyone help me with the KPOINTS > file for GaN nanowire. > > Thank in advance > > Regards, > Dayasagar. > ________________________________ > Check out what you're missing if you're not on Yahoo! Messenger -- Sincerely yours, Pavel B Sorokin From ihsanas at yahoo.com Sun Apr 29 13:42:03 2007 From: ihsanas at yahoo.com (احسان عريقات) Date: Sun, 29 Apr 2007 04:42:03 -0700 (PDT) Subject: [Pw_forum] transition state Message-ID: <130578.14058.qm@web52303.mail.re2.yahoo.com> Hi, transition state structures can be determined by searching for first order saddle points on the potential energy surface ,how we can do this using pwscf without using neb . thanks __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From baroni at sissa.it Sun Apr 29 13:49:30 2007 From: baroni at sissa.it (Stefano Baroni) Date: Sun, 29 Apr 2007 12:49:30 +0100 Subject: [Pw_forum] Can anyone help me with bandstructure calculations of GaN In-Reply-To: <7b6913e90704281648x6f38466bka3425947849ae64e@mail.gmail.com> References: <373763.52967.qm@web8327.mail.in.yahoo.com> <7b6913e90704281648x6f38466bka3425947849ae64e@mail.gmail.com> Message-ID: <71711E08-FA27-4171-805B-8D499E43220D@sissa.it> Also, I think that doing an electronic-structure calculation for a periodic system having (verbatim) "no idea of the KPOINTS in the Brillouin zone" is not a very good idea. SB On Apr 29, 2007, at 12:48 AM, Axel Kohlmeyer wrote: > On 4/28/07, Daya sagar wrote: >> >> Hi all, >> >> I am doing the band structure calculations of GaN nanowires. I >> am using >> VASP for the calculations. I have no idea of the KPOINTS in >> brillouin zone > > since you want to use VASP, how about asking in the corresponding > VASP forum or switching to quantum espresso? > > cheers, > axel. > >> for calculating the band structures. Can anyone help me with the >> KPOINTS >> file for GaN nanowire. >> >> Thank in advance >> >> Regards, >> Dayasagar. >> ________________________________ >> Check out what you're missing if you're not on Yahoo! Messenger > > > -- > ====================================================================== > = > 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. > _______________________________________________ > 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: /pipermail/attachments/20070429/aaae5ff7/attachment.htm From baroni at sissa.it Sun Apr 29 17:51:10 2007 From: baroni at sissa.it (Stefano Baroni) Date: Sun, 29 Apr 2007 16:51:10 +0100 Subject: [Pw_forum] transition state In-Reply-To: <130578.14058.qm@web52303.mail.re2.yahoo.com> References: <130578.14058.qm@web52303.mail.re2.yahoo.com> Message-ID: <9F3F75CE-8E95-4C01-A179-0FBC921E112A@sissa.it> Why don't you want to use NEB? Which other method would you like to use? SB (too much of a hassle signing your correspondence?) On Apr 29, 2007, at 12:42 PM, ????? ?????? wrote: > Hi, > transition state structures can be determined by > searching for first order saddle points on the > potential energy surface ,how we can do this using > pwscf without using neb . > thanks > > __________________________________________________ > 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 --- 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: /pipermail/attachments/20070429/e6f0f793/attachment.htm From merlin.meheut at impmc.jussieu.fr Sun Apr 29 19:59:59 2007 From: merlin.meheut at impmc.jussieu.fr (merlin meheut) Date: Sun, 29 Apr 2007 19:59:59 +0200 Subject: [Pw_forum] Compilation trouble with g95 on AMD64 Athlon 3500 Message-ID: <20070429195959.3pmkhvvxq884wc8c@www-ext.impmc.jussieu.fr> Dear all, I tried to compile espresso-3.2 on an AMD64 Athlon 3.5 GHz, using g95, and I have the following error running the example01: from create_directory : error #4294967295 unable to create directory ./silicon.save I have seen somewhere that it may be due to a problem with iotk: ( http://www.democritos.it/pipermail/pw_forum/2006-October/005151.html ) > > this is a kind inconsistency in call to intrinsic functions; > > such problems have been fixed in the cvs version of espresso > substituting lines like > > control = modulo(header,iotk_ncontrol+1) > > with > > control = modulo(header,int(iotk_ncontrol+1,kind=kind(header))) > > > you can try to repeat the trick for all the lines where the compiler > complains, otherwise try to you replace the espresso-3.1.1/iotk/ with > espresso_CVS/iotk, it may probably work without any further problem.. > > andrea > But then I make : bash at meheut:results$ grep "control = modulo" ../../../*/*/* ../../../iotk/src/iotk_scan.f90: control = modulo(header,int(iotk_ncontrol+1,kind=kind(header))) ../../../iotk/src/iotk_scan.spp: control = modulo(header,int(iotk_ncontrol+1,kind=kind(header))) I have apparently the good version . May you have any idea on how to fix this? ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From ferretti.andrea at unimore.it Sun Apr 29 20:46:04 2007 From: ferretti.andrea at unimore.it (Andrea Ferretti) Date: Sun, 29 Apr 2007 20:46:04 +0200 (CEST) Subject: [Pw_forum] Compilation trouble with g95 on AMD64 Athlon 3500 In-Reply-To: <20070429195959.3pmkhvvxq884wc8c@www-ext.impmc.jussieu.fr> References: <20070429195959.3pmkhvvxq884wc8c@www-ext.impmc.jussieu.fr> Message-ID: Dear Merlin, > > I tried to compile espresso-3.2 on an AMD64 Athlon 3.5 GHz, using g95, and I > have the following error running the example01: > > from create_directory : error #4294967295 > unable to create directory ./silicon.save the error you are experiencing is, I think, a different one with respect to that from the post below... the present one is a run-time error while the one posted was due to a complain of the compiler at compile-time... it seems this is not due to iotk, but rather to the usage of the mkdir c-function as you can see from the error number (which sounds as random), something nasty is occurring... basically I think the problem is the following: in the create_directory routine, the error code (which causes the crash if different from zero), is directly taken from the call IF ( ionode ) ierr = c_mkdir( TRIM( dirname ), LEN_TRIM( dirname ) ) that is, it is returned from a wrapper to a c-lib function (mkdir) since you are on a 64bit architecture, default integers maybe 4bytes (32bit) or 8bytes (64bit) long (only pointers are alsways 8bytes long! )... which one depends on the fortran compiler... I suspect you are using 8bytes default integers, while mkdir is returning a 4bytes integer... this may cause the almost random error number and the following crash of the code.... if this is the case, try to compile espresso with a flag forcing 32 bit integers (maybe something like -i4, to be set manually in FFLAGS, make.sys file) otherwise, download the g95 version with 32bit integers as default... hope it helps andrea > > I have seen somewhere that it may be due to a problem with iotk: > ( http://www.democritos.it/pipermail/pw_forum/2006-October/005151.html ) > > > > > this is a kind inconsistency in call to intrinsic functions; > > > > such problems have been fixed in the cvs version of espresso > > substituting lines like > > > > control = modulo(header,iotk_ncontrol+1) > > > > with > > > > control = modulo(header,int(iotk_ncontrol+1,kind=kind(header))) > > > > you can try to repeat the trick for all the lines where the compiler > > complains, otherwise try to you replace the espresso-3.1.1/iotk/ with > > espresso_CVS/iotk, it may probably work without any further problem.. > > > -- Andrea Ferretti National Research Center S3, CNR-INFM ( http://s3.infm.it ) Dipartimento di Fisica, Universita' di Modena e Reggio Emilia Via Campi 213/A I-41100 Modena, Italy Tel: +39 059 2055301 Fax: +39 059 374794 Skype: andrea_ferretti URL: http://www.nanoscience.unimo.it Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html From niuli1978 at yahoo.com.cn Mon Apr 30 00:31:48 2007 From: niuli1978 at yahoo.com.cn (li niu) Date: Mon, 30 Apr 2007 06:31:48 +0800 (CST) Subject: [Pw_forum] finite dielectric methods Message-ID: <689740.44162.qm@web15004.mail.cnb.yahoo.com> Dear Paolo U, Thanks for your reply.I'm very interested for your finite electric field method. The dielectric polarizability tensor can be calculated by taking dinite differences of atomic forces. I have several questions to ask you. 1. We obtained the diagonal terms, by considering values for the electric field of -2h,-h, 0, h and 2h in the cp.x code [like example 30 tefield=.true. tcg = .true., epol = 1 or 2 or 3 efield = -2h,-h, 0, h or 2h], and by using the five-point formular . Is it right? 2. For the calculation of the off-diagonal mixedterms, two different electric fields are required. This need apply electric fields of intensity - 2h/1.414 , -h/1.414, 0, h/1.414 and 2h/1.414 along the three different couples of cartesian directions, and apply equations. But how can I apply simultaneously finite fields along two different Cartesian directions in cp.x code?[ESSPRESSO-3.2] 3. In this way, we could obtain all the dielectric polarizability tensor. Then the Raman susceptibilities associated to the normal mode n can be obtianed according to the equations. How to implement it ? By a small code? Best! I'm counting on the new version. Niu Li --------------------------------- ????????3.5G???20M??? -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20070430/b750a6bf/attachment.htm From merlin.meheut at impmc.jussieu.fr Mon Apr 30 01:39:57 2007 From: merlin.meheut at impmc.jussieu.fr (merlin meheut) Date: Mon, 30 Apr 2007 01:39:57 +0200 Subject: [Pw_forum] Compilation trouble with g95 on AMD64 Athlon 3500 In-Reply-To: References: <20070429195959.3pmkhvvxq884wc8c@www-ext.impmc.jussieu.fr> Message-ID: <20070430013957.3nbnfv08gswscwkg@www-ext.impmc.jussieu.fr> Dear Andrea, downloading the g95 version with 32bit integers as default was the good solution. Thank you very much, Merlin Quoting Andrea Ferretti : > > Dear Merlin, > > >> >> I tried to compile espresso-3.2 on an AMD64 Athlon 3.5 GHz, using g95, and I >> have the following error running the example01: >> >> from create_directory : error #4294967295 >> unable to create directory ./silicon.save > > > the error you are experiencing is, I think, a different one with respect > to that from the post below... > > the present one is a run-time error while the one posted was due to a > complain of the compiler at compile-time... > > it seems this is not due to iotk, but rather to the usage of the mkdir > c-function > > as you can see from the error number (which sounds as random), something > nasty is occurring... > basically I think the problem is the following: > > in the create_directory routine, the error code (which causes the crash if > different from zero), is directly taken from the call > > IF ( ionode ) ierr = c_mkdir( TRIM( dirname ), LEN_TRIM( dirname ) ) > > that is, it is returned from a wrapper to a c-lib function (mkdir) > since you are on a 64bit architecture, default integers maybe 4bytes > (32bit) or > 8bytes (64bit) long (only pointers are alsways 8bytes long! )... > which one depends on the fortran compiler... > > I suspect you are using 8bytes default integers, while mkdir is returning > a 4bytes integer... this may cause the almost random error number and the > following crash of the code.... > > if this is the case, try to compile espresso with a flag forcing > 32 bit integers (maybe something like -i4, to be set manually in FFLAGS, > make.sys file) > > otherwise, download the g95 version with 32bit integers as default... > > hope it helps > > andrea > >> >> I have seen somewhere that it may be due to a problem with iotk: >> ( http://www.democritos.it/pipermail/pw_forum/2006-October/005151.html ) >> >> > >> > this is a kind inconsistency in call to intrinsic functions; >> > >> > such problems have been fixed in the cvs version of espresso >> > substituting lines like >> > >> > control = modulo(header,iotk_ncontrol+1) >> > >> > with >> > >> > control = modulo(header,int(iotk_ncontrol+1,kind=kind(header))) >> > >> > you can try to repeat the trick for all the lines where the compiler >> > complains, otherwise try to you replace the espresso-3.1.1/iotk/ with >> > espresso_CVS/iotk, it may probably work without any further problem.. >> > >> > > > -- > Andrea Ferretti > National Research Center S3, CNR-INFM ( http://s3.infm.it ) > Dipartimento di Fisica, Universita' di Modena e Reggio Emilia > Via Campi 213/A I-41100 Modena, Italy > Tel: +39 059 2055301 Fax: +39 059 374794 > Skype: andrea_ferretti > URL: http://www.nanoscience.unimo.it > > 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 > ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From baroni at sissa.it Mon Apr 30 11:16:39 2007 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 30 Apr 2007 10:16:39 +0100 Subject: [Pw_forum] finite dielectric methods In-Reply-To: <689740.44162.qm@web15004.mail.cnb.yahoo.com> References: <689740.44162.qm@web15004.mail.cnb.yahoo.com> Message-ID: On Apr 29, 2007, at 11:31 PM, li niu wrote: > Dear Paolo U, > > Thanks for your reply.I'm very interested for your finite electric > field method. > The dielectric polarizability tensor can be calculated by taking > dinite differences of atomic forces. NO: forces have nothing to do with the dielectric tensor. If you want to calculate the dielectric constant using finite electric fields (why? linear response is there exactly to avoid that), you will have to calculate finite differences of electric polarizations corresponding to different diractions of the applied field. > I have several questions to ask you. > ... > 2. For the calculation of the off-diagonal mixedterms, two > different electric fields are required. This need apply electric > fields of intensity - 2h/1.414 , > -h/1.414, 0, h/1.414 and 2h/1.414 along the three different couples > of cartesian directions, and apply equations. > But how can I apply simultaneously finite fields along two > different Cartesian directions in cp.x code?[ESSPRESSO-3.2] here, I can hardly follow you. Two finite (or infinitesimal, for what matters) fileds applied along different directions make ONE electrc field directed along the the sum of the two. Aren't electric fields VECTORS? SB --- 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: /pipermail/attachments/20070430/ae92a59c/attachment.htm From roccad at sissa.it Mon Apr 30 14:40:00 2007 From: roccad at sissa.it (Dario Rocca) Date: Mon, 30 Apr 2007 14:40:00 +0200 Subject: [Pw_forum] Titanium rrkjus pseudopotential In-Reply-To: References: <689740.44162.qm@web15004.mail.cnb.yahoo.com> Message-ID: <1177936800.4635e3a04453d@webmail.sissa.it> Does anyone have a Rabe-Rappe-Kaxiras-Joannopoulos ultrasoft for titanium? Thank you Dario ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by Horde http://www.horde.org/ From kazempoor2000 at yahoo.com Mon Apr 30 15:09:03 2007 From: kazempoor2000 at yahoo.com (ali kazempoor) Date: Mon, 30 Apr 2007 06:09:03 -0700 (PDT) Subject: [Pw_forum] nanowire Message-ID: <662776.31335.qm@web33112.mail.mud.yahoo.com> Dear all I am phd student in isfahan and subject of my thesis is about ab initio properties of Hexagonal freestanding MnAs nanowire on GaAs(001).before this time i worked on interface of this system, but i don't know about simulation of this nanowire in espresso .please comment me in making supercell and optimization of this structure. thanks __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From ihsanas at yahoo.com Mon Apr 30 19:16:54 2007 From: ihsanas at yahoo.com (احسان عريقات) Date: Mon, 30 Apr 2007 10:16:54 -0700 (PDT) Subject: [Pw_forum] transition state Message-ID: <81867.45049.qm@web52306.mail.re2.yahoo.com> Thank you for replaying ,NEB take much time I`m trying to built potential surface by taking many scf calculations for different position for reactant by changing the bond length ,angles and distance ,and then I`ll try to find saddle point .Is this way is correct ,how could I know that the atomic position are wrong . Thanks Ihsan __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From baroni at sissa.it Mon Apr 30 21:39:03 2007 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 30 Apr 2007 20:39:03 +0100 Subject: [Pw_forum] transition state In-Reply-To: <81867.45049.qm@web52306.mail.re2.yahoo.com> References: <81867.45049.qm@web52306.mail.re2.yahoo.com> Message-ID: <95094276-441A-478C-BA35-816D532C08DC@sissa.it> On Apr 30, 2007, at 6:16 PM, ????? ?????? wrote: > Thank you for replaying ,NEB take much time I`m trying > to built potential surface by taking many scf > calculations for different position for reactant by > changing the bond length ,angles and distance ,and > then I`ll try to find saddle point .Is this way is > correct ,how could I know that the atomic position are > wrong . if there was any other (much) simpler way of locating the saddle point, noone would be so silly to do NEB ... I am afraid, Ihsan, that there are very few other options ... > Thanks > Ihsan Ihsan who? from where? Stefano Baroni SISSA - Trieste > > __________________________________________________ > 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 --- 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: /pipermail/attachments/20070430/2893c597/attachment.htm From konstantin_kudin at yahoo.com Mon Apr 30 22:23:16 2007 From: konstantin_kudin at yahoo.com (Konstantin Kudin) Date: Mon, 30 Apr 2007 13:23:16 -0700 (PDT) Subject: [Pw_forum] spilling parameter in Lowdin - how to assign the charge that went into Nirvana ? In-Reply-To: <2A0AF24D-13E9-4EC1-8CD6-EA9763C6666E@nest.sns.it> Message-ID: <540288.28763.qm@web57708.mail.re3.yahoo.com> Hi all, Revisiting this month old topic, I'd like to ask how easy/difficult would it be to add the optimization of the atomic basis for the Lowdin charge evaluation? The optimization is described here: http://dx.doi.org/10.1016/0038-1098(95)00341-X With that, one could perhaps even calculate the spread of the atomic density, and figure out if the atom became larger/smaller as a result of chemical interaction. Thanks! Kostya --- Paolo Giannozzi wrote: > > On Mar 24, 2007, at 15:54 , Andrea Ferretti wrote: > > > Lowding orbitals in Quantum-Espresso (at least as far as I know) > are > > computed from the atomic reference states from pseudopotential data > > according to: > > > > S_ij = < i | j > are the overlaps of atomic orbitals i and j > > > > lowding orbitals are then obtained as > > | low_i > = R_ij | j > > > > > where R = S^{-1/2} > > > > so I think there is no optimization (in the sense of the paper > above) > > but maybe I am missing something > > no, you aren't missing anything: the atomic basis is not optimized > > Paolo > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From degironc at sissa.it Mon Apr 30 23:18:58 2007 From: degironc at sissa.it (degironc) Date: Mon, 30 Apr 2007 23:18:58 +0200 Subject: [Pw_forum] spilling parameter in Lowdin - how to assign the charge that went into Nirvana ? In-Reply-To: <540288.28763.qm@web57708.mail.re3.yahoo.com> References: <540288.28763.qm@web57708.mail.re3.yahoo.com> Message-ID: <46365D42.5020907@sissa.it> In order to calculate the spilling parameter with scaled atomic wfc's the easiest way is to modify the atomic_wfc.f90 routine scaling the k+g vectors [arrays gk and qg at the beginning of it] ... this is rather easy I think. In order to optimize the scaling one needs to implement some minimization startegy... this implies repeatedly calculate the spilling parameter (if using a method not requiring gradients such as the simplex algorithm) or code the calculation of its derivatives if one wants to use CG or similar algorithms. stefano Konstantin Kudin wrote: > Hi all, > > Revisiting this month old topic, I'd like to ask how easy/difficult >would it be to add the optimization of the atomic basis for the Lowdin >charge evaluation? > > The optimization is described here: >http://dx.doi.org/10.1016/0038-1098(95)00341-X > > With that, one could perhaps even calculate the spread of the atomic >density, and figure out if the atom became larger/smaller as a result >of chemical interaction. > > Thanks! > Kostya > > >--- Paolo Giannozzi wrote: > > > >>On Mar 24, 2007, at 15:54 , Andrea Ferretti wrote: >> >> >> >>>Lowding orbitals in Quantum-Espresso (at least as far as I know) >>> >>> >>are >> >> >>>computed from the atomic reference states from pseudopotential data >>>according to: >>> >>>S_ij = < i | j > are the overlaps of atomic orbitals i and j >>> >>>lowding orbitals are then obtained as >>>| low_i > = R_ij | j > >>> >>>where R = S^{-1/2} >>> >>>so I think there is no optimization (in the sense of the paper >>> >>> >>above) >> >> >>>but maybe I am missing something >>> >>> >>no, you aren't missing anything: the atomic basis is not optimized >> >>Paolo >>--- >>Paolo Giannozzi, Democritos and University of Udine, Italy >> >> >>