From thefarsky at hotmail.com Mon Sep 1 07:20:21 2008 From: thefarsky at hotmail.com (wangqiang) Date: Mon, 1 Sep 2008 13:20:21 +0800 Subject: [Pw_forum] I cannot compile resp_mat.f90. In-Reply-To: References: Message-ID: Dear everyone, i cannot compile resp_mat.f90 file. i cannot find the llmk_ipf library. i am using intel fortran compiler 9.1. and mkl libary 10.1. what can i do now. Lee chungnam univeristy. _________________________________________________________________ ???MSN??????????????????? http://mobile.msn.com.cn/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080901/dbc44c70/attachment.htm From giannozz at democritos.it Mon Sep 1 09:15:49 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 01 Sep 2008 09:15:49 +0200 Subject: [Pw_forum] I cannot compile resp_mat.f90. In-Reply-To: References: Message-ID: <48BB96A5.7070004@democritos.it> wangqiang wrote: > i cannot compile resp_mat.f90 file. there is no resp_mat.f90 file in QE -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Mon Sep 1 09:19:02 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 01 Sep 2008 09:19:02 +0200 Subject: [Pw_forum] The scf calculation is not converged In-Reply-To: References: Message-ID: <48BB9766.30209@democritos.it> vega wrote: >[...] some scf calculations can't be converged even in 1000 cycles http://www.quantum-espresso.org/wiki/index.php/Troubleshooting_(PWscf)#self-consistency_is_slow_or_does_not_converge -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Mon Sep 1 10:03:32 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 01 Sep 2008 10:03:32 +0200 Subject: [Pw_forum] Unit of charge density In-Reply-To: <20080831001728.21066rzbnuvn37cw@web.mail.umich.edu> References: <20080831001728.21066rzbnuvn37cw@web.mail.umich.edu> Message-ID: <48BBA1D4.2050502@democritos.it> Dong-Hee Lim wrote: > I was just wondering what unit of charge density is used in Quantum > Espresso and Xcrysden viewer. the viewer just visualizes. Quantum-Espresso uses atomic units (Rydberg for PWscf, Hartree for CP, unfortunately). The charge density is not multiplied by "e" if I remember correctly, so it has the dimensions of 1/V, i.e. 1/(bohr radii)^3 Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From virginie.trinite at thalesgroup.com Mon Sep 1 11:18:43 2008 From: virginie.trinite at thalesgroup.com (Virginie Trinite) Date: Mon, 01 Sep 2008 11:18:43 +0200 Subject: [Pw_forum] Re : something strange in rutile and anatase calculation Message-ID: <48BBB373.4040905@thalesgroup.com> Hi, We have observed that including or not semi-core states in the pseudopotential af Ti, can make great difference in the stabilisation of \alpha versus \omega phases of Ti and also for anatase and rutile phase ( For Ti, you can look at Journal of Physics C, vol 20, 235239 (2008)). I am very interest of the reference of the free energy calculations showing the stabilization of \alpha-Ti by phonons (We have made a approximate evaluation of this contribution and our conclusion was that it was not sufficient). Bests ---------------------------------------- PhD Trinite Virginie Physics Research Group of Thales Research & Technology 2008/8/28, Eyvaz Isaev : >Hi, > >One more suggestion. >As far as I know the same situation is for \alpha and \omega phases of Ti, as well as HCP and DHCP phases of FeH. But free energy calculations have shown than \alpha-Ti and DHCP FeH are >stabilized due to the phonon contribution. >For FeH you can see: Eyvaz I. Isaev et al., Dynamical stability of Fe-H in the Earth's mantle and core regions, Proceedings of the National Academy of Sciences of the USA (PNAS), 104 (22), 29 May >2007, 9168. PNAS Special Issue "Geophysics". > >So, you can try follow this way, if you like. > >Bests, >Eyvaz. From giannozz at democritos.it Mon Sep 1 11:29:23 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 01 Sep 2008 11:29:23 +0200 Subject: [Pw_forum] (no subject) In-Reply-To: <10064.87.14.218.240.1220174809.squirrel@mail.unito.it> References: <10064.87.14.218.240.1220174809.squirrel@mail.unito.it> Message-ID: <48BBB5F3.9070708@democritos.it> anna.ferrari at unito.it wrote: > 1)in restart_mode > restart is done from wavefunction or from density? typically from both > 2) what is ecutrho exactly and what happens if a value is > not provided? ecutrho is the cutoff for the charge density: ecutrho=4*ecutwfc for norm-conserving PP (this is the default), ecutrho=N*ecutwfc with N something between 6 to 12 for ultrasoft PP Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From thefarsky at hotmail.com Mon Sep 1 13:30:05 2008 From: thefarsky at hotmail.com (wangqiang) Date: Mon, 1 Sep 2008 19:30:05 +0800 Subject: [Pw_forum] Did anyone try fcc-iron calculation. In-Reply-To: References: Message-ID: Dear everyone: how can i get a antiferromagnetic structure of fcc-Fe through the pwscf calculation?? Lee chungnam university. _________________________________________________________________ ?????MSN?????????? http://im.live.cn/click/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080901/296b4deb/attachment.htm From eyvaz_isaev at yahoo.com Mon Sep 1 14:10:44 2008 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 1 Sep 2008 05:10:44 -0700 (PDT) Subject: [Pw_forum] Did anyone try fcc-iron calculation. In-Reply-To: Message-ID: <640330.96779.qm@web65715.mail.ac4.yahoo.com> Hi, I did. It can be done quite easily, just take a bct unit cell (c/a=\sqrt(2)) with two Fe atoms and consider them as two different atoms with antiparallel magnetization. Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com --- On Mon, 9/1/08, wangqiang wrote: > From: wangqiang > Subject: [Pw_forum] Did anyone try fcc-iron calculation. > To: pw_forum at pwscf.org > Date: Monday, September 1, 2008, 3:30 PM > Dear everyone: > how can i get a antiferromagnetic structure of fcc-Fe > through the pwscf calculation?? > Lee > chungnam university. > _________________________________________________________________ > ?????MSN?????????? > http://im.live.cn/click/_______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From iphyboy at hotmail.com Mon Sep 1 18:30:28 2008 From: iphyboy at hotmail.com (=?gb2312?B?w/fOxMPA?=) Date: Tue, 2 Sep 2008 00:30:28 +0800 Subject: [Pw_forum] OPIUM pp *.ncpp files error in pw.x Message-ID: Dear all : I am using OPIUM to generate norm conversing pseudopotential for Ga,Ni and Zr, all the input parameters files are download from the website of OPIUM, and I choose to generate the *.ncpp (for PWSCF format) format OPIUM files, however, when I want to do self-consistent calculation with these OPIUM pps by pw.x, the calculate crashed and complained : from read_ncpp : error # 1 order of wavefunctions I find the complains in /PW/read_ncpp.f90 Can anyone tell me how to resolve this trouble ? Thanks very much . Wenmei Ming !!!!!!!!!!!!!!! my input file is : &CONTROL title = ZrNi2Ga , calculation = 'scf' , restart_mode = 'from_scratch' , pseudo_dir = '/home/phymwm/PWSCF/espresso-3.2.3/pseudo/' , / &SYSTEM ibrav = 2, celldm(1) = 11.90528, nat = 4, ntyp = 3, ecutwfc = 40 , occupations = 'smearing' , degauss = 0.02 , smearing = 'methfessel-paxton' , / &ELECTRONS conv_thr = 1.0D-9 , /ATOMIC_SPECIES Zr 91.22000 zr.ncpp Ni 58.69000 ni.ncpp Ga 69.72000 ga.ncppATOMIC_POSITIONS alat Zr 0.000000000 0.000000000 0.000000000 Ni -0.250000000 -0.250000000 -0.250000000 Ni 0.250000000 0.250000000 0.250000000 Ga 0.500000000 0.000000000 0.000000000K_POINTS automatic 8 8 8 1 1 1 _________________________________________________________________ ?????????????????????????? http://im.live.cn/Share/18.htm -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080902/f659f344/attachment.htm From iphyboy at hotmail.com Mon Sep 1 18:30:38 2008 From: iphyboy at hotmail.com (=?gb2312?B?w/fOxMPA?=) Date: Tue, 2 Sep 2008 00:30:38 +0800 Subject: [Pw_forum] OPIUM pp *.ncpp files error in pw.x Message-ID: Dear all : I am using OPIUM to generate norm conversing pseudopotential for Ga,Ni and Zr, all the input parameters files are download from the website of OPIUM, and I choose to generate the *.ncpp (for PWSCF format) format OPIUM files, however, when I want to do self-consistent calculation with these OPIUM pps by pw.x, the calculate crashed and complained : from read_ncpp : error # 1 order of wavefunctions I find the complains in /PW/read_ncpp.f90 Can anyone tell me how to resolve this trouble ? Thanks very much . Wenmei Ming !!!!!!!!!!!!!!! my input file is : &CONTROL title = ZrNi2Ga , calculation = 'scf' , restart_mode = 'from_scratch' , pseudo_dir = '/home/phymwm/PWSCF/espresso-3.2.3/pseudo/' , / &SYSTEM ibrav = 2, celldm(1) = 11.90528, nat = 4, ntyp = 3, ecutwfc = 40 , occupations = 'smearing' , degauss = 0.02 , smearing = 'methfessel-paxton' , / &ELECTRONS conv_thr = 1.0D-9 , /ATOMIC_SPECIES Zr 91.22000 zr.ncpp Ni 58.69000 ni.ncpp Ga 69.72000 ga.ncppATOMIC_POSITIONS alat Zr 0.000000000 0.000000000 0.000000000 Ni -0.250000000 -0.250000000 -0.250000000 Ni 0.250000000 0.250000000 0.250000000 Ga 0.500000000 0.000000000 0.000000000K_POINTS automatic 8 8 8 1 1 1 _________________________________________________________________ ?????????????????????????? http://im.live.cn/Share/18.htm -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080902/c2c9f63c/attachment-0001.htm From giannozz at democritos.it Mon Sep 1 18:42:35 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 01 Sep 2008 18:42:35 +0200 Subject: [Pw_forum] OPIUM pp *.ncpp files error in pw.x In-Reply-To: References: Message-ID: <48BC1B7B.5090801@democritos.it> ??? wrote: > I am using OPIUM to generate norm conversing pseudopotential for > Ga,Ni and Zr, all the input parameters files are download from the > website of OPIUM, and I choose to generate the *.ncpp (for PWSCF format) > format OPIUM files, however, when I want to do self-consistent > calculation with these OPIUM pps by pw.x, the calculate crashed and > complained : > from read_ncpp : error # 1 > order of wavefunctions > I find the complains in /PW/read_ncpp.f90 you also find the explanation there: the code expects atomic wavefunctions ordered as s,p,d, .. not sure whether it is really needed any longer. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From ejwalt at wm.edu Mon Sep 1 20:06:18 2008 From: ejwalt at wm.edu (Eric J. Walter) Date: Mon, 01 Sep 2008 14:06:18 -0400 Subject: [Pw_forum] OPIUM pp *.ncpp files error in pw.x In-Reply-To: References: Message-ID: <48BC2F1A.601@wm.edu> Dear Wenmei Ming, Please forward this post to the OPIUM mailing list: opium-talk at lists.sourceforge.net Also, please tell the OPIUM list what version of OPIUM you are using. I am sure that someone there would be happy to help you. Sincerely, Eric J. Walter ??? wrote: > Dear all : > > I am using OPIUM to generate norm conversing pseudopotential for Ga,Ni > and Zr, all the input parameters files are download from the website > of OPIUM, and I choose to generate the *.ncpp (for PWSCF format) > format OPIUM files, however, when I want to do self-consistent > calculation with these OPIUM pps by pw.x, the calculate crashed and > complained : > from read_ncpp : error # 1 > order of wavefunctions > I find the complains in /PW/read_ncpp.f90 > > Can anyone tell me how to resolve this trouble ? > > > Thanks very much . > Wenmei Ming > !!!!!!!!!!!!!!! > > my input file is : > &CONTROL > title = ZrNi2Ga , > calculation = 'scf' , > restart_mode = 'from_scratch' , > pseudo_dir = '/home/phymwm/PWSCF/espresso-3.2.3/pseudo/' , > / > &SYSTEM > ibrav = 2, > celldm(1) = 11.90528, > ! ; nat = 4, > ntyp = 3, > ecutwfc = 40 , > occupations = 'smearing' , > degauss = 0.02 , > smearing = 'methfessel-paxton' , > / > &ELECTRONS > conv_thr = 1.0D-9 , > / > ATOMIC_SPECIES > Zr&n! bsp; 91.22000 zr.ncpp > Ni 58.69000 ni.ncpp > Ga 69.72000 ga.ncpp > ATOMIC_POSITIONS alat > Zr 0.000000000 0.000000000 0.000000000 > Ni -0.250000000 -0.250000000 -0.250000000 > Ni 0.250000000 0.250000000 0.250000000 > Ga 0.500000000 0.000000000 0.000000000 > K_POINTS automatic > 8 8 8 1 1 1 > > > > > > > ------------------------------------------------------------------------ > ?????MSN?????????? ????? > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From asafis at yahoo.com.br Tue Sep 2 00:03:39 2008 From: asafis at yahoo.com.br (=?iso-8859-1?Q?=C1lvaro_Alves?=) Date: Mon, 1 Sep 2008 15:03:39 -0700 (PDT) Subject: [Pw_forum] Pseudopotential Co Message-ID: <249668.54336.qm@web52303.mail.re2.yahoo.com> Hello everyone. I am trying to do a pseudopotencial PBE fully relativistic for Co, but I'm having some difficulties.? Could someone check my input to detect the error? ?&input ??? title='Co', ??? zed=27., ??? rel=2, ??? config='[Ar] 3d7.0 4s2.0 4p0.0', ??? iswitch=3, ??? dft='PBE' ?/ ?&inputp ?? lloc=0, ?? pseudotype=3, ?? file_pseudopw='Co.rel-pbe-rrkjus.UPF', ?? nlcc=.true., ?? rcore=0.9, ?? author='ASA', ?/ 7 4P? 2? 1? 0.00 -0.10? 2.40? 2.40? 0.50 4P? 2? 1? 0.00 -0.10? 2.40? 2.40? 1.50 3D? 3? 2? 4.00? 0.00? 1.50? 2.10? 1.50 3D? 3? 2? 0.00 -0.20? 1.50? 2.10? 1.50 3D? 3? 2? 3.00? 0.00? 1.50? 2.10? 2.50 3D? 3? 2? 0.00 -0.20? 1.50? 2.10? 2.50 4S? 1? 0? 2.00? 0.00? 2.20? 2.20? 0.50 ???????????????? Sincerely A. S. Santos - Universidade Federal Fluminense Rio de Janeiro - Brasil Novos endere?os, o Yahoo! que voc? conhece. Crie um email novo com a sua cara @ymail.com ou @rocketmail.com. http://br.new.mail.yahoo.com/addresses -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080901/9ae07328/attachment.htm From zdw2000 at gmail.com Tue Sep 2 01:53:20 2008 From: zdw2000 at gmail.com (Wei Zhou) Date: Mon, 1 Sep 2008 16:53:20 -0700 Subject: [Pw_forum] how to build a I43-d structure in pwscf? Message-ID: <3e36a8fc0809011653j7c9322e6p4b40b446afd1f72b@mail.gmail.com> hello , I want to build a body center unit cell with 16 atoms in convention cell , 8 atoms in primitive cell, it has the I-43d group space with wyckoff 16c,and have a fractional coordinates( 0.044 0.044 0.044),however ,I can not build it with pwscf correctly .would some one help me? any help will be appreciated best wishes the basic vector and crystal coordination of atoms are writed as follow primitive cell, -4.999999999999993 5.000000000000001 5.000000000000004 5.000000000000001 -5.000000000000002 4.999999999999996 5.000000000000000 4.999999999999999 -4.999999999999999 coordination 0.0880000293254852 0.0880000293254852 0.0880000293254852 0.4999999999999999 0.9999999999999998 0.4119999706745147 1.0000000000000000 0.4119999706745149 0.4999999999999999 0.4119999706745147 0.4999999999999998 0.9999999999999998 0.5880000293254851 0.5880000293254851 0.5880000293254851 1.5000000000000000 1.0000000000000000 0.9119999706745148 0.9119999706745149 1.5000000000000000 0.9999999999999999 0.9999999999999999 0.9119999706745147 1.4999999999999998 . -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080901/d1f71a9e/attachment.htm From iphyboy at hotmail.com Tue Sep 2 02:28:44 2008 From: iphyboy at hotmail.com (=?gb2312?B?w/fOxMPA?=) Date: Tue, 2 Sep 2008 08:28:44 +0800 Subject: [Pw_forum] OPIUM pp *.ncpp files error in pw.x to Prof. Paolo Giannozzi In-Reply-To: <48BC1B7B.5090801@democritos.it> References: <48BC1B7B.5090801@democritos.it> Message-ID: Dear Prof. Paolo Giannozzi: thanks for you kind reply, actually I want to calculate electron-interaction in metallic ZrNi2Ga, but ultrasoft pesudopotential yields somewhat bad reuslts, so I want to try the Norm Conversing pp ,and I try to use /atomic/ld1.x to generate the TM based NCPP using the input parameters from Abinit-TMfhi pp libary. However,it is terrible from energy-volume curve. So I switch to OPIUM to generate the NCPP in .ncpp format, and the trouble "order of wavefunctions" appears. do you have the NCPP of Zr Ni and Ga atoms ? and could you send me these NCPP if possible ? Thanks very much . Wenmei Ming > Date: Mon, 1 Sep 2008 18:42:35 +0200> From: giannozz at democritos.it> To: pw_forum at pwscf.org> Subject: Re: [Pw_forum] OPIUM pp *.ncpp files error in pw.x> > ??? wrote:> > > I am using OPIUM to generate norm conversing pseudopotential for > > Ga,Ni and Zr, all the input parameters files are download from the > > website of OPIUM, and I choose to generate the *.ncpp (for PWSCF format) > > format OPIUM files, however, when I want to do self-consistent > > calculation with these OPIUM pps by pw.x, the calculate crashed and > > complained :> > from read_ncpp : error # 1> > order of wavefunctions> > I find the complains in /PW/read_ncpp.f90> > you also find the explanation there: the code expects atomic> wavefunctions ordered as s,p,d, .. not sure whether it is really> needed any longer.> > 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 _________________________________________________________________ ?????MSN?????????? http://im.live.cn/click/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080902/6c339166/attachment.htm From vegalew at hotmail.com Tue Sep 2 05:16:07 2008 From: vegalew at hotmail.com (vega) Date: Tue, 2 Sep 2008 11:16:07 +0800 Subject: [Pw_forum] where is the exact benchmark for k point sampling Message-ID: Dear all I want to calculate the atomic structure, electronic structure and the energy of a surface using slab model. I calculate the same system with 2X3X1 corresponding 3 k irregular points and gamma point only respectively to test the K point sampling. Then I compared the two case. I found that the relaxed atomic coordinates and the energy for two cases were nearly the same with ~1% errors. Do you think I could using gamma point instead for saving the computional cost? thank you for reading. vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080902/be79c125/attachment.htm From bipulrr at gmail.com Tue Sep 2 07:21:15 2008 From: bipulrr at gmail.com (Bipul Rakshit) Date: Tue, 2 Sep 2008 10:51:15 +0530 Subject: [Pw_forum] enthlapy of a system Message-ID: <3a749910809012221r1e009101naa4935d675a22df8@mail.gmail.com> Hello Everyone, I want to calculate the enthlapy at different pressures of ScAs in B1 and B2 phase, so that i able to find the transition pressure. Can anybody tell me the correct method to do that? -- Bipul Rakshit PhD Student, Barkatullah University, Bhopal 462026, MP, India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080902/32b3b215/attachment.htm From matteo at umn.edu Tue Sep 2 07:32:35 2008 From: matteo at umn.edu (Matteo Cococcioni) Date: Tue, 02 Sep 2008 00:32:35 -0500 Subject: [Pw_forum] enthlapy of a system In-Reply-To: <3a749910809012221r1e009101naa4935d675a22df8@mail.gmail.com> References: <3a749910809012221r1e009101naa4935d675a22df8@mail.gmail.com> Message-ID: <48BCCFF3.8030007@umn.edu> Dear Bipul I think all you need to do is computing the equation of state for both phases. In other words you choose a series of lattice parameters (that may be different for the two phases) and compute the total energy for each of them. In this case for B1 (NaCl) and B2 (CsCl) you don't need to relax the atomic positions for each lattice parameter as you would do in general. Once you get E(V) for each of the phase, you can use the program ev.x to fit a Murnaghan or a Birch eq. of state in your data. From the equation of state you have chosen you can compute the derivative P=-dE/dV at any V within the interval of your fit. This will be the pressure on your system. for every value of P you will have V and the corresponding E so you will be able to construct E+PV (enthalpy) for the values of P you have. hope this helps, Matteo Bipul Rakshit wrote: > Hello Everyone, > I want to calculate the enthlapy at different pressures of ScAs in B1 > and B2 phase, so that i able to find the transition pressure. Can > anybody tell me the correct method to do that? > > > -- > Bipul Rakshit > PhD Student, > Barkatullah University, > Bhopal 462026, > MP, India > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From bipulrr at gmail.com Tue Sep 2 08:14:03 2008 From: bipulrr at gmail.com (Bipul Rakshit) Date: Tue, 2 Sep 2008 11:44:03 +0530 Subject: [Pw_forum] enthlapy of a system In-Reply-To: <48BCCFF3.8030007@umn.edu> References: <3a749910809012221r1e009101naa4935d675a22df8@mail.gmail.com> <48BCCFF3.8030007@umn.edu> Message-ID: <3a749910809012314y55919f3ey16120b33104f0e1@mail.gmail.com> Dear Matteo, Got your point and really thanks for the help. Just one more question.... In scf run if we set tstress = .true., then we also get the pressure. And also we have Etot and V, so can we calculate H taking those Etot and V ?? On Tue, Sep 2, 2008 at 11:02 AM, Matteo Cococcioni wrote: > > > Dear Bipul > > I think all you need to do is computing the equation of state for both > phases. In other words you choose a series of lattice parameters (that > may be different for the two phases) and compute the total energy for > each of them. In this case for B1 (NaCl) and B2 (CsCl) you don't need to > relax the atomic positions for each lattice parameter as you would do in > general. Once you get E(V) for each of the phase, you can use the > program ev.x to fit a Murnaghan or a Birch eq. of state in your data. > From the equation of state you have chosen you can compute the > derivative P=-dE/dV at any V within the interval of your fit. This will > be the pressure on your system. for every value of P > you will have V and the corresponding E so you will be able to construct > E+PV (enthalpy) for the values of P you have. > > hope this helps, > > Matteo > > > > > Bipul Rakshit wrote: > > Hello Everyone, > > I want to calculate the enthlapy at different pressures of ScAs in B1 > > and B2 phase, so that i able to find the transition pressure. Can > > anybody tell me the correct method to do that? > > > > > > -- > > Bipul Rakshit > > PhD Student, > > Barkatullah University, > > Bhopal 462026, > > MP, India > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > 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 > -- Bipul Rakshit PhD Student, Barkatullah University, Bhopal 462026, MP, India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080902/ba7be3ac/attachment.htm From giannozz at democritos.it Tue Sep 2 10:55:27 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 02 Sep 2008 10:55:27 +0200 Subject: [Pw_forum] OPIUM pp *.ncpp files error in pw.x to Prof. Paolo Giannozzi In-Reply-To: References: <48BC1B7B.5090801@democritos.it> Message-ID: <48BCFF7F.5060902@democritos.it> ??? wrote: > thanks for you kind reply, actually I want to calculate > electron-interaction in metallic ZrNi2Ga, > but ultrasoft pseudopotential yields somewhat bad results any evidence that the "somewhat bad results" are due to the pseudopotentials? -- Paolo Giannozzi, Democritos and University of Udine, Italy From eyvaz_isaev at yahoo.com Tue Sep 2 12:31:48 2008 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 2 Sep 2008 03:31:48 -0700 (PDT) Subject: [Pw_forum] how to build a I43-d structure in pwscf? In-Reply-To: <3e36a8fc0809011653j7c9322e6p4b40b446afd1f72b@mail.gmail.com> Message-ID: <182393.20808.qm@web65701.mail.ac4.yahoo.com> Hi, See http://cst-www.nrl.navy.mil/lattice/struk/Li_cI16.html or http://www.cryst.ehu.es/cgi-bin/cryst/programs/nph-wp-list?gnum=220 > fractional coordinates (0.044 0.044 0.044) Most likely they are internal parameters. Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com --- On Tue, 9/2/08, Wei Zhou wrote: > From: Wei Zhou > Subject: [Pw_forum] how to build a I43-d structure in pwscf? > To: "PWSCF Forum" > Date: Tuesday, September 2, 2008, 3:53 AM > hello , > I want to build a body center unit cell with 16 > atoms in convention > cell , 8 atoms in primitive cell, it has the I-43d group > space with wyckoff > 16c,and have a fractional coordinates( 0.044 0.044 > 0.044),however ,I can not > build it with pwscf correctly .would some one help me? any > help will be > appreciated > best wishes > the basic vector and crystal coordination of atoms are > writed as follow > primitive cell, > -4.999999999999993 5.000000000000001 > 5.000000000000004 > 5.000000000000001 -5.000000000000002 > 4.999999999999996 > 5.000000000000000 4.999999999999999 > -4.999999999999999 > > > coordination > 0.0880000293254852 0.0880000293254852 > 0.0880000293254852 > 0.4999999999999999 0.9999999999999998 > 0.4119999706745147 > 1.0000000000000000 0.4119999706745149 > 0.4999999999999999 > 0.4119999706745147 0.4999999999999998 > 0.9999999999999998 > 0.5880000293254851 0.5880000293254851 > 0.5880000293254851 > 1.5000000000000000 1.0000000000000000 > 0.9119999706745148 > 0.9119999706745149 1.5000000000000000 > 0.9999999999999999 > 0.9999999999999999 0.9119999706745147 > 1.4999999999999998 > . > -- > ZhouDawei > JiLin Universiyt ,ChangChun ,China > zdw2000 at gmail.com > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From shrek_826 at yahoo.com.cn Tue Sep 2 14:41:20 2008 From: shrek_826 at yahoo.com.cn (Fan Yang) Date: Tue, 2 Sep 2008 20:41:20 +0800 (CST) Subject: [Pw_forum] confused about nonzero total charge Message-ID: <60872.37805.qm@web15302.mail.cnb.yahoo.com> Dear everyone, ?I?want to?calculate the charge of atoms in cell, and I'm not?sure whether the?process is correct. First,?the system?is relaxed to get the equilibrium structure. And then I run the scf and nscf calculation with?the final configuration. Then the pdos calculation is performed and the output would give the total?electrons and?also s,p,d,f electrons. Then?the total?electrons is substracted by the atom's?electrons without relaxation, for example, if the total electron of H is 0.6734,?and the charge of H is 1-0.6734=0.3266. Am I right? Once I obtain the charge of every atom in unit cell, the total charge of each atom does not equal 0, which confused me because the system of total charge is set as 0.?I also noticed?one?line in relaxation output, which said?'starting charge? 464.37813, renormalised to? 469.00000'. I?think this?may be the answer for the nonzero total charge. Best Regards, ******************************************* Fan Yang PH.D Candidate?in Electrochemistry College of Chemistry and Molecular Science Wuhan University, 430072, Hubei Province, China E-mail:shrek_826 at yahoo.com.cn ******************************************* -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080902/ecccaf03/attachment.htm From giannozz at democritos.it Tue Sep 2 16:10:26 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 02 Sep 2008 16:10:26 +0200 Subject: [Pw_forum] confused about nonzero total charge In-Reply-To: <60872.37805.qm@web15302.mail.cnb.yahoo.com> References: <60872.37805.qm@web15302.mail.cnb.yahoo.com> Message-ID: <48BD4952.2090600@democritos.it> Fan Yang wrote: > I want to calculate the charge of atoms in cell, and I'm not sure > whether the process is correct. [...] > Once I obtain the charge of every atom in unit cell, the total charge of > each atom does not equal 0, which confused me because the system of > total charge is set as 0. As explained approximately 1001 times in this mailing list, "the charge of an atom in a molecule/solid" typically depends on a particular choice of atomic orbitals over which wavefunctions are projected. The true total charge is zero, the charge you are talking about is not. From code projwfc.f90 : psum = SUM(charges(:,:,:)) / nelec WRITE( stdout, '(5x,"Spilling Parameter: ",f8.4)') 1.0d0 - psum ! ! Sanchez-Portal et al., Sol. State Commun. 95, 685 (1995). ! The spilling parameter measures the ability of the basis provided ! by the pseudo-atomic wfc to represent the PW eigenstates, ! by measuring how much of the subspace of the Hamiltonian ! eigenstates falls outside the subspace spanned by the atomic basis Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From iphyboy at hotmail.com Tue Sep 2 16:13:16 2008 From: iphyboy at hotmail.com (=?gb2312?B?w/fOxMPA?=) Date: Tue, 2 Sep 2008 22:13:16 +0800 Subject: [Pw_forum] OPIUM pp *.ncpp files error in pw.x to Prof. Paolo Giannozzi In-Reply-To: <48BCFF7F.5060902@democritos.it> References: <48BC1B7B.5090801@democritos.it> <48BCFF7F.5060902@democritos.it> Message-ID: > Date: Tue, 2 Sep 2008 10:55:27 +0200> From: giannozz at democritos.it> To: pw_forum at pwscf.org> Subject: Re: [Pw_forum] OPIUM pp *.ncpp files error in pw.x to Prof. Paolo Giannozzi> > ??? wrote:> > > thanks for you kind reply, actually I want to calculate > > electron-interaction in metallic ZrNi2Ga,> > but ultrasoft pseudopotential yields somewhat bad results> > any evidence that the "somewhat bad results" are due to the> pseudopotentials? Still some imaginary frequencies as I encountered before, which disappear after I switch to NCPP > -- > 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 _________________________________________________________________ ????????MSN????????MSN? http://im.live.cn/emoticons/?ID=6 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080902/68cf33b1/attachment.htm From akohlmey at cmm.chem.upenn.edu Tue Sep 2 16:24:40 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Tue, 2 Sep 2008 10:24:40 -0400 (EDT) Subject: [Pw_forum] confused about nonzero total charge In-Reply-To: <48BD4952.2090600@democritos.it> References: <60872.37805.qm@web15302.mail.cnb.yahoo.com> <48BD4952.2090600@democritos.it> Message-ID: On Tue, 2 Sep 2008, Paolo Giannozzi wrote: PG> Fan Yang wrote: PG> PG> > I want to calculate the charge of atoms in cell, and I'm not sure PG> > whether the process is correct. [...] PG> > Once I obtain the charge of every atom in unit cell, the total charge of PG> > each atom does not equal 0, which confused me because the system of PG> > total charge is set as 0. PG> PG> As explained approximately 1001 times in this mailing list, paolo, does that mean, we have to start calling you "scheherazade" now? , axel. PG> "the charge of an atom in a molecule/solid" typically depends on a PG> particular choice of atomic orbitals over which wavefunctions are PG> projected. The true total charge is zero, the charge you are PG> talking about is not. From code projwfc.f90 : PG> PG> psum = SUM(charges(:,:,:)) / nelec PG> WRITE( stdout, '(5x,"Spilling Parameter: ",f8.4)') 1.0d0 - psum PG> ! PG> ! Sanchez-Portal et al., Sol. State Commun. 95, 685 (1995). PG> ! The spilling parameter measures the ability of the basis provided PG> ! by the pseudo-atomic wfc to represent the PW eigenstates, PG> ! by measuring how much of the subspace of the Hamiltonian PG> ! eigenstates falls outside the subspace spanned by the atomic basis PG> PG> Paolo 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 matteo at umn.edu Tue Sep 2 18:25:53 2008 From: matteo at umn.edu (Matteo Cococcioni) Date: Tue, 02 Sep 2008 11:25:53 -0500 Subject: [Pw_forum] enthlapy of a system In-Reply-To: <3a749910809012314y55919f3ey16120b33104f0e1@mail.gmail.com> References: <3a749910809012221r1e009101naa4935d675a22df8@mail.gmail.com> <48BCCFF3.8030007@umn.edu> <3a749910809012314y55919f3ey16120b33104f0e1@mail.gmail.com> Message-ID: <48BD6911.7020103@umn.edu> Dear Bipul, the pressure you get as the Tr of the stress tensor can be a little inconsistent with the one you would get from the E(V) curve unless you keep the number of plane waves constant for all the lattice spacing (i.e. you change the cutoff with the volume of the unit cell). The reason is the variational nature of E: for larger cells you get smaller Brillouin zones and a larger number of plane waves in the calculation that makes E decrease more than "expected" just from the increase of lattice spacing. The stress instead is computed for a given unit cell and a fixed number of plane waves. Probably you can find other messages about this in the forum archive. Regards, Matteo Bipul Rakshit wrote: > Dear Matteo, > Got your point and really thanks for the help. > Just one more question.... > In scf run if we set tstress = .true., then we also get the pressure. > And also we have Etot and V, so can we calculate H taking those Etot > and V ?? > > On Tue, Sep 2, 2008 at 11:02 AM, Matteo Cococcioni > wrote: > > > > Dear Bipul > > I think all you need to do is computing the equation of state for both > phases. In other words you choose a series of lattice parameters (that > may be different for the two phases) and compute the total energy for > each of them. In this case for B1 (NaCl) and B2 (CsCl) you don't > need to > relax the atomic positions for each lattice parameter as you would > do in > general. Once you get E(V) for each of the phase, you can use the > program ev.x to fit a Murnaghan or a Birch eq. of state in your data. > From the equation of state you have chosen you can compute the > derivative P=-dE/dV at any V within the interval of your fit. This > will > be the pressure on your system. for every value of P > you will have V and the corresponding E so you will be able to > construct > E+PV (enthalpy) for the values of P you have. > > hope this helps, > > Matteo > > > > > Bipul Rakshit wrote: > > Hello Everyone, > > I want to calculate the enthlapy at different pressures of ScAs > in B1 > > and B2 phase, so that i able to find the transition pressure. Can > > anybody tell me the correct method to do that? > > > > > > -- > > Bipul Rakshit > > PhD Student, > > Barkatullah University, > > Bhopal 462026, > > MP, India > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > 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 > > > > > -- > Bipul Rakshit > PhD Student, > Barkatullah University, > Bhopal 462026, > MP, India > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- A non-text attachment was scrubbed... Name: matteo.vcf Type: text/x-vcard Size: 294 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20080902/650a1012/attachment.vcf From thefarsky at hotmail.com Wed Sep 3 09:29:35 2008 From: thefarsky at hotmail.com (wangqiang) Date: Wed, 3 Sep 2008 15:29:35 +0800 Subject: [Pw_forum] I cannot compile resp_mat.f90. In-Reply-To: References: Message-ID: Dear everyone: is there anybody who has the experience compiling the code resp_mat.f90. actually it is not a part of QE. however i got it from the tutorial lda+U from Vlab lectures which is included by QE website. i got an error when i compiled this file. it seems the computer cannot find lmkl_ipf. Lee chungnam university. _________________________________________________________________ ?????????????????????????? http://im.live.cn/Share/18.htm -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080903/35b9a6ab/attachment.htm From yang at sri.jussieu.fr Wed Sep 3 10:01:31 2008 From: yang at sri.jussieu.fr (yang) Date: Wed, 3 Sep 2008 10:01:31 +0200 Subject: [Pw_forum] question about example16 (pengyang) Message-ID: Dear All, I am a beginner of PWscf and want to calculate some STM images. There is a problem for me about the example16 in the version 3.2.3. The 'make.power5-aix-serial' file in the 'install' directory does not work for the compilation in our server. I just change the 'MPI90 = mpixlf90_r' to 'MPI90 = xlf90_r' in the defaut file 'make.sys' obtained by the command 'configure', then the program was sucessfully compiled on the AIX 5 system. The program was stopped during running the example16, and i cann't get the 'AlAs110-1.0eV.ps' file. The erro message was attached below. I'm afraid that it is due to the incorrect compilation. Anyone can help me? Thanks in advance! Best regards, Peng ------------------------------ ---------------------------------------------------------------- /san/home/yangpeng/pwscf/examples/examples/example16 : starting This example shows a calculation of STM maps. executables directory: /san/home/yangpeng/pwscf/espresso/bin pseudo directory: /san/home/yangpeng/pwscf/espresso/pseudo temporary directory: /san/home/yangpeng/pwscf/tmp checking that needed directories and files exist... done running pw.x as: /san/home/yangpeng/pwscf/espresso/bin/pw.x running pp.x as: /san/home/yangpeng/pwscf/espresso/bin/pp.x running plotrho.x as: /san/home/yangpeng/pwscf/espresso/bin/plotrho.x cleaning /san/home/yangpeng/pwscf/tmp... done running the scf calculation... done running the non-scf calculation... done running the post-processing phase, negative bias... done running plotrho on negative bias data..."plotrho.f90", line 88: 1525-097 A READ statement using decimal b ase input found the invalid digit 'n' in the input file. The program will recover by assuming a zero in it s place. "plotrho.f90", line 88: 1525-097 A READ statement using decimal base input found the invalid digit '.' in t he input file. The program will recover by assuming a zero in its place. "plotrho.f90", line 114: 1525-001 The READ statement on the file (standard input) cannot be completed becau se the end of the file was reached. The program will stop. done running the post-processing phase, positive bias... done running plotrho on positive bias data..."plotrho.f90", line 88: 1525-097 A READ statement using decimal b ase input found the invalid digit 'n' in the input file. The program will recover by assuming a zero in it s place. "plotrho.f90", line 88: 1525-097 A READ statement using decimal base input found the invalid digit '.' in t he input file. The program will recover by assuming a zero in its place. "plotrho.f90", line 114: 1525-001 The READ statement on the file (standard input) cannot be completed becau se the end of the file was reached. The program will stop. done ----------------------------------------------------------- Peng YANG Doctorant Laboratoire LM2N Directeur: Pr M.-P. Pileni Universit? Pierre et Marie Curie 4 place Jussieu 75252 Paris Cedex 05 tel: 01 44 27 26 96 fax: 01 44 27 25 15 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080903/19667d80/attachment.htm From giannozz at democritos.it Wed Sep 3 10:19:02 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 3 Sep 2008 10:19:02 +0200 Subject: [Pw_forum] I cannot compile resp_mat.f90. In-Reply-To: References: Message-ID: On Sep 3, 2008, at 9:29 , wangqiang wrote: > it seems the computer cannot find lmkl_ipf. this is a linking problem, not a compilation problem (if you do not know the difference between the two, please learn it). The correct way of linking mkl libraries depends upon the version of mkl installed on your machine (if any) Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at democritos.it Wed Sep 3 10:35:12 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 3 Sep 2008 10:35:12 +0200 Subject: [Pw_forum] question about example16 (pengyang) In-Reply-To: References: Message-ID: <3FAA6606-5182-4496-8D71-44F17EDD0340@democritos.it> On Sep 3, 2008, at 10:01 , yang wrote: > I am a beginner of PWscf and want to calculate some STM images. > There is a problem for me about the example16 in the version 3.2.3. if you are a beginner, why are you beginning with an old version of the code? > The 'make.power5-aix-serial' file in the 'install' directory does > not work it is basically an example file, and I am not sure it has ever updated to reflect changes in the installation mechanism. Anyway the sentence "it doesn't work" contains 0 bits of information. > for the compilation in our server. I just change the 'MPI90 = > mpixlf90_r' to > 'MPI90 = xlf90_r' in the defaut file 'make.sys' obtained by the > command 'configure' this means that you have a parallel machine, but that you are compiling for serial execution. Are you sure this is what you want? > [...] The error message was attached below. more 0 bits of information > I'm afraid that it is due to the incorrect compilation. unlikely. Run 'plotrho' interactively and see what happens P. --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From munima at iitg.ernet.in Wed Sep 3 10:48:32 2008 From: munima at iitg.ernet.in (Munima Bora Sahariah) Date: Wed, 3 Sep 2008 14:18:32 +0530 (IST) Subject: [Pw_forum] wrong order of k points Message-ID: <55767.172.16.48.135.1220431712.squirrel@webmail.iitg.ac.in> Dear Miguel, Yesterday I came across a strange problem that you faced some 3 years back. I tracked your mail in pwforum while trying to dig out the problem. In order to remind you of the problem I am including your old mail here describing the error. I am having exactly the same error message while trying to compute phonons at X-point (1 0 0) of a fcc system. The gamma point did not create any problem. I am very anxious to know how you resolved the case. Please help me out. regards munima BELOW IS YOUR OLD MAIL: >Miguel Mart?nez Canales escribi?: > Dear Dr. Gianozzi > > First, thanks to you and Dr.Isaev for your answers. > >> >> IF ( ABS( xq(1) - ( xk(1,ikq) - xk(1,ikk) ) ) > eps8 .OR. & >> ABS( xq(2) - ( xk(2,ikq) - xk(2,ikk) ) ) > eps8 .OR. & >> ABS( xq(3) - ( xk(3,ikq) - xk(3,ikk) ) ) > eps8 ) THEN >> WRITE( stdout, * ) ikk, ikq, nksq >> WRITE( stdout, * ) ( xq(ipol), ipol = 1, 3) >> WRITE( stdout, * ) ( xk(ipol,ikq), ipol = 1, 3) >> WRITE( stdout, * ) ( xk(ipol,ikk), ipol = 1, 3) >> CALL errore( 'phq_init', 'wrong order of k points', 1 ) >> END IF >> >> what does it print immediately before that? >> > > The outpur (from the last k point to the error is): > > > k(16250) = ( 0.4200000 1.4200000 -0.2500000), wk = > 0.0000000 > > pseudo 1 is Al (US) zval = 3.0 lmax= 1 lloc= 0 > Version 0 0 0 of US pseudo code > > Using log mesh of 893 points > > The pseudopotential has 2 beta functions with: > > l(1) = 0 > l(2) = 1 > > Q(r) pseudized with 8 coefficients, rinner = 0.900 0.900 > 0.900 > > > > Atomic displacements: > There are 3 irreducible representations > > Representation 1 1 modes - To be done > > Representation 2 1 modes - To be done > > Representation 3 1 modes - To be done > 1 2 8125 > -0.125000000000000 -0.125000000000000 0.375000000000000 > -2.00000000000000D-002 2.00000000000000D-002 0.270000000000000 > -2.00000000000000D-002 2.00000000000000D-002 2.00000000000000D-002 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > from phq_init : error # 1 > wrong order of k points > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > > stopping ... > > > If it helps, this tipically occurs in low symmetry points (i.e. > outside 00x or xxx lines). Thanks again, and sorry for the late answer, > > Miguel > Dr. Munima B. Sahariah PI, DST Fast Track Project IASST, Guwahati Assam, INDIA ----------------------------------------------------------------------------------- This email was sent from IIT Guwahati Webmail. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. http://www.iitg.ac.in From giannozz at democritos.it Wed Sep 3 12:30:36 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 3 Sep 2008 12:30:36 +0200 Subject: [Pw_forum] wrong order of k points In-Reply-To: <55767.172.16.48.135.1220431712.squirrel@webmail.iitg.ac.in> References: <55767.172.16.48.135.1220431712.squirrel@webmail.iitg.ac.in> Message-ID: On Sep 3, 2008, at 10:48 , Munima Bora Sahariah wrote: >> from phq_init : error # 1 >> wrong order of k points you either did not perform the intermediate non-scf calculation needed for q/=0 (old QE versions), or you did not specify the option lnscf=.true. to phonon (new QE versions) Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From gaoguoying at gmail.com Wed Sep 3 12:44:09 2008 From: gaoguoying at gmail.com (Guoying Gao) Date: Wed, 3 Sep 2008 18:44:09 +0800 Subject: [Pw_forum] error in the phonon calculation at Gamma point Message-ID: <6234c0c60809030344g9332fbai9fadf4d341352da1@mail.gmail.com> Dear everyone, When I calculate the phonon frequency at Gamma point, the following errors are encountered in XX.ph.out file Program PHONON v.3.2 starts ... Today is 3Sep2008 at 18:29:37 Ultrasoft (Vanderbilt) Pseudopotentials nbndx = 20 nbnd = 20 natomwfc = 128 npwx = 3784 nelec = 40.00 nkb = 88 ngl = 5829 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from divide_class : error # 1 wrong axis %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... Any suggestions will be appreciated. Thanks in advance! Best wishes! Guoying Jilin Univ. P.R.China From anna.ferrari at unito.it Wed Sep 3 12:52:27 2008 From: anna.ferrari at unito.it (anna.ferrari at unito.it) Date: Wed, 3 Sep 2008 12:52:27 +0200 (CEST) Subject: [Pw_forum] files for PP and guess Message-ID: <10112.87.21.21.158.1220439147.squirrel@mail.unito.it> dear all i'm confused about files produced by pw, in particular 1) for post-processing which files i need (that i should copy back from the scratch)? (i'm used to other codes where a unique file is produced at the end of the run useful for properties and restart) 2) if i want for instance a new scf calculation with a better ecutrho, can i read a guess form a previuos (worse) calculation? In this case which file(s) i need (the minimal information of course) to provide? thanks again anna From hviet.nguyen at gmail.com Wed Sep 3 13:24:19 2008 From: hviet.nguyen at gmail.com (Nguyen Huy Viet) Date: Wed, 3 Sep 2008 13:24:19 +0200 Subject: [Pw_forum] wrong order of k points In-Reply-To: References: <55767.172.16.48.135.1220431712.squirrel@webmail.iitg.ac.in> Message-ID: On Wed, Sep 3, 2008 at 12:30 PM, Paolo Giannozzi wrote: > > On Sep 3, 2008, at 10:48 , Munima Bora Sahariah wrote: > > >> from phq_init : error # 1 > >> wrong order of k points > > you either did not perform the intermediate non-scf calculation > needed for q/=0 (old QE versions), or you did not specify the > option lnscf=.true. to phonon (new QE versions) > q-vector in non-scf and phonon calculations are not the same might be another possibility. Viet -- Mr. Viet Huy Nguyen PhD student, Condensed Matter Theory Sector SISSA/ISAS Via Beirut 2-4, 34014, Trieste, Italy Tel: +39 040 3787 492 Fax: +39 040 3787 528 Email: nhviet at sissa.it hviet.nguyen at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080903/eb1fc6b2/attachment.htm From dalcorso at sissa.it Wed Sep 3 13:55:59 2008 From: dalcorso at sissa.it (Dal Corso Andrea) Date: Wed, 03 Sep 2008 13:55:59 +0200 Subject: [Pw_forum] error in the phonon calculation at Gamma point In-Reply-To: <6234c0c60809030344g9332fbai9fadf4d341352da1@mail.gmail.com> References: <6234c0c60809030344g9332fbai9fadf4d341352da1@mail.gmail.com> Message-ID: <1220442959.3338.18.camel@dhpc-5-53.sissa.it> On Wed, 2008-09-03 at 18:44 +0800, Guoying Gao wrote: > Dear everyone, > > When I calculate the phonon frequency at Gamma point, the > following errors are encountered in XX.ph.out file > > > Program PHONON v.3.2 starts ... > Today is 3Sep2008 at 18:29:37 > > Ultrasoft (Vanderbilt) Pseudopotentials > > nbndx = 20 nbnd = 20 natomwfc = 128 npwx = 3784 > nelec = 40.00 nkb = 88 ngl = 5829 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from divide_class : error # 1 > wrong axis > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > Please check if this error is still present with the 4.0.1 version of the code. If it is, please provide the input file. Andrea > Any suggestions will be appreciated. Thanks in advance! > > Best wishes! > > Guoying > Jilin Univ. > P.R.China > _______________________________________________ > 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 gaoguoying at gmail.com Wed Sep 3 15:42:04 2008 From: gaoguoying at gmail.com (Guoying Gao) Date: Wed, 3 Sep 2008 21:42:04 +0800 Subject: [Pw_forum] error in the phonon calculation at Gamma point In-Reply-To: <1220442959.3338.18.camel@dhpc-5-53.sissa.it> References: <6234c0c60809030344g9332fbai9fadf4d341352da1@mail.gmail.com> <1220442959.3338.18.camel@dhpc-5-53.sissa.it> Message-ID: <6234c0c60809030642s686ca582lca173144a901851b@mail.gmail.com> Dear Andrea, Thanks for your kindly help. The version of 4.0.1 works. Regards! Guoying From kazempoor2000 at yahoo.com Wed Sep 3 16:27:05 2008 From: kazempoor2000 at yahoo.com (ali kazempoor) Date: Wed, 3 Sep 2008 07:27:05 -0700 (PDT) Subject: [Pw_forum] spin-orbit2 Message-ID: <932611.34002.qm@web33102.mail.mud.yahoo.com> Dear all I want to calculate the contribution of two terms in magnetic energy. the first , is crystallin anisotropy energy and the second is shape anisotropy. for crystalline anisotropy I run relax calculation for a Ni?[001]?nanowire with 1 1 40 kpoint and smearing =.002 and ecut =40 and ultrasoft full relativistic (includ the spin-orbit term) and true switch for noncoliin and spinorb .?but the message is convergency not achieved after 100 iteration. where is the error? Is there?any way to determine the contribution of shape anisotrpy in total energy? thanks ?Ali Kazempour, Isfahan Univ. of .Tech, Isfahan, Iran ? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080903/35a0c4a9/attachment.htm From giannozz at democritos.it Wed Sep 3 18:30:30 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 3 Sep 2008 18:30:30 +0200 Subject: [Pw_forum] files for PP and guess In-Reply-To: <10112.87.21.21.158.1220439147.squirrel@mail.unito.it> References: <10112.87.21.21.158.1220439147.squirrel@mail.unito.it> Message-ID: <28CA7AC9-A940-48CC-BF2C-0039B72D700A@democritos.it> On Sep 3, 2008, at 12:52 , anna.ferrari at unito.it wrote: > 1) for post-processing which files i need (that i should > copy back from the scratch)? the directory $prefix.save/ in $outdir, plus, if variable $wf_collect is not set to true, wavefunction files $prefix.wfc*. > (i'm used to other codes where a unique file is > produced at the end of the run useful for properties > and restart) some like it that way. By using tar or zip, you can always reduce to a single file. > 2) if i want for instance a new scf calculation with a > better ecutrho, can i read a guess form a previuos (worse) > calculation? no, this is currently not implemented (I think) Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From ravitri2k at gmail.com Wed Sep 3 20:05:57 2008 From: ravitri2k at gmail.com (Ravi Maranganti) Date: Wed, 3 Sep 2008 13:05:57 -0500 Subject: [Pw_forum] How to calculate real space inter-atomic force constants excluding the contribution due to macroscopic electric field Message-ID: <59b3d4f30809031105mfbedea3he498815db9251f22@mail.gmail.com> Hello All How should i calculate the real space interatomic force constants with the contribution due to the macroscopic electric field being excluded. According to Giannozzi et al (Ab initio calculation of phonons in semiconductors, Phys Rev B, v. 43, p.7231), the contribution to the dynamical matrix in polar materials can be written as a combination of an analytical term and a non-analytical term (at low k-vectors). The non-analytical term involves the Born effective charges and the dielectric constant. so, should i subtract this term from the dynamical matrix at all k-vectors (or only at certain k-vectors?) and then use q2r.x to obtain the real space force constants? Also, should i *not* specify .epsil. = true? I am new to pwscf so please bear with me...thanks! Ravi Maranganti, PhD Candidate, Mechanical Engg Dept, Univ of Houston, TX -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080903/285cf5a5/attachment.htm From anna.ferrari at unito.it Wed Sep 3 20:54:50 2008 From: anna.ferrari at unito.it (anna.ferrari at unito.it) Date: Wed, 3 Sep 2008 20:54:50 +0200 (CEST) Subject: [Pw_forum] files for PP and guess In-Reply-To: <28CA7AC9-A940-48CC-BF2C-0039B72D700A@democritos.it> References: <10112.87.21.21.158.1220439147.squirrel@mail.unito.it> <28CA7AC9-A940-48CC-BF2C-0039B72D700A@democritos.it> Message-ID: <10381.79.17.215.128.1220468090.squirrel@mail.unito.it> reading the pw_forum i'm learning about a lot of keyword (as $wf_collect ) i didn't know (are not in the PW doc files) where i can find a complete list of keywords and their settings and i see i have a $prefix.wfc* file for each processor, what i can do if i need to change the number of processors? thanks again anna > > > ------------------ Messaggio originale ------------------- > Oggetto: Re: [Pw_forum] files for PP and guess > Da: "Paolo Giannozzi" > Data: Mer, 3 Settembre 2008, 6:30 pm > A: "PWSCF Forum" > ---------------------------------------------------------- > > > On Sep 3, 2008, at 12:52 , anna.ferrari at unito.it wrote: > >> 1) for post-processing which files i need (that i should >> copy back from the scratch)? > > the directory $prefix.save/ in $outdir, plus, if variable > $wf_collect is not set to true, wavefunction files > $prefix.wfc*. > >> (i'm used to other codes where a unique file is >> produced at the end of the run useful for properties >> and restart) > > some like it that way. By using tar or zip, you can always > reduce to a single file. > >> 2) if i want for instance a new scf calculation with a >> better ecutrho, can i read a guess form a previuos >> (worse) >> calculation? > > no, this is currently not implemented (I think) > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > From ezadshojaee at hotmail.com Wed Sep 3 21:35:13 2008 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Wed, 3 Sep 2008 19:35:13 +0000 Subject: [Pw_forum] Symdynph_gq Message-ID: Hi May i know the reference that the routine "Symdynphq_gq" is based on? I want to know that how does the code apply a rotation of the small groupof q to the dynamical matrix. Thank You _________________________________________________________________ News, entertainment and everything you care about at Live.com. Get it now! http://www.live.com/getstarted.aspx -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080903/cc85ce6f/attachment.htm From ski2 at mail.uh.edu Thu Sep 4 00:14:41 2008 From: ski2 at mail.uh.edu (ski2 at mail.uh.edu) Date: Wed, 03 Sep 2008 17:14:41 -0500 Subject: [Pw_forum] question related to ATOMIC POSITIONS Message-ID: Dear All, I am a new user. I have a quick question in terms of structure input preparation. There is no choice to specify a space group of a system. So, I assume that, in ATOMIC_POSITIONS, all atomic positions in a primitive cell must be included in the script file, although there are symmetrically equivalent atoms. Am I right? Best Wishes, Sang-Hwan Kim Department of Chemistry University of Houston 136 Fleming Building Houston, TX 77204-5003 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080903/a789d035/attachment-0001.htm From eyvaz_isaev at yahoo.com Thu Sep 4 00:43:14 2008 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 3 Sep 2008 15:43:14 -0700 (PDT) Subject: [Pw_forum] question related to ATOMIC POSITIONS In-Reply-To: Message-ID: <850144.31670.qm@web65701.mail.ac4.yahoo.com> Dear Kim, --- On Thu, 9/4/08, ski2 at mail.uh.edu wrote: > I am a new user. Welcome! > I have a quick question in terms of structure input > preparation. There is no choice to specify a space > group of a system. So, I assume that, in > ATOMIC_POSITIONS, all atomic positions in a primitive > cell must be included in the script file, although > there are symmetrically equivalent atoms. > Am I right? A quick answer: yes. All atoms in the unit cell you have chosen should be specified. Even they are equivalent by symmetry. Bests, Eyvaz. > Best Wishes, > Sang-Hwan Kim > > > > > > Department of Chemistry > University of Houston > 136 Fleming Building > Houston, TX 77204-5003 > _______________________________________________ ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From wangqj1 at 126.com Thu Sep 4 05:31:14 2008 From: wangqj1 at 126.com (wangqj1) Date: Thu, 4 Sep 2008 11:31:14 +0800 (CST) Subject: [Pw_forum] How to analyze the result of pdos Message-ID: <7169925.423311220499074710.JavaMail.coremail@bj126app104.126.com> Dear all After I used projwfc.x ,I got the pdos of every atom ,but I don't know the signification of the following characters of pdos_atm#1(Zn)_wfc#1(d) and pdos_atm#1(Zn)_wfc#2(s). pdos_atm#1(Zn)_wfc#1(d): # E (eV) ldosup(E) ldosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdos_atm#1(Zn)_wfc#2(ds): # E (eV) ldosup(E) ldosdw(E) pdosup(E) pdosdw(E) Besides,I want to draw the figure of pdos of every atom on the same figure .for Zn3p,Zn4s,and so on .So what data should I extract from the pdos_atm#1(Zn)_wfc#1(d)and pdos_atm#1(Zn)_wfc#2(ds),for there are so many information . Any advice will be appreciated ! sincerely XiangTan University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080904/c7d1006f/attachment.htm From szs_naghavi at yahoo.com Thu Sep 4 08:12:03 2008 From: szs_naghavi at yahoo.com (zahra sadat naghavi) Date: Wed, 3 Sep 2008 23:12:03 -0700 (PDT) Subject: [Pw_forum] old error Message-ID: <22569.57454.qm@web63003.mail.re1.yahoo.com> Dear all I have my old problem with ph.x:? ?? from phq_readin : error #???????? 1 ???? cannot start from pw.x data file using Gamma-point?tricks my ph.x is cut very soon. I test diffrent automatics in?pw.x run but nothing happens! ? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080903/0e63ae4a/attachment.htm From giannozz at democritos.it Thu Sep 4 08:42:18 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 4 Sep 2008 08:42:18 +0200 Subject: [Pw_forum] files for PP and guess In-Reply-To: <10381.79.17.215.128.1220468090.squirrel@mail.unito.it> References: <10112.87.21.21.158.1220439147.squirrel@mail.unito.it> <28CA7AC9-A940-48CC-BF2C-0039B72D700A@democritos.it> <10381.79.17.215.128.1220468090.squirrel@mail.unito.it> Message-ID: On Sep 3, 2008, at 20:54 , anna.ferrari at unito.it wrote: > reading the pw_forum i'm learning about a lot of keyword > (as $wf_collect ) i didn't know (are not in the PW doc files) they are, they are... > where i can find a complete list of keywords and their settings Doc/INPUT_PW.* > and i see i have a $prefix.wfc* file for each processor, > what i can do if i need to change the number of processors? this is exacty what the wf_collect option is good for Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at democritos.it Thu Sep 4 08:45:21 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 4 Sep 2008 08:45:21 +0200 Subject: [Pw_forum] old error In-Reply-To: <22569.57454.qm@web63003.mail.re1.yahoo.com> References: <22569.57454.qm@web63003.mail.re1.yahoo.com> Message-ID: On Sep 4, 2008, at 8:12 , zahra sadat naghavi wrote: > from phq_readin : error # 1 > cannot start from pw.x data file using Gamma-point tricks you need to execute the self-consistent calculation without the Gamma-point specific tricks. Do not specify K_POINTS gamma --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From anna.ferrari at unito.it Thu Sep 4 12:04:41 2008 From: anna.ferrari at unito.it (Anna Maria Ferrari) Date: Thu, 4 Sep 2008 11:04:41 +0100 Subject: [Pw_forum] files for PP and guess References: <10112.87.21.21.158.1220439147.squirrel@mail.unito.it><28CA7AC9-A940-48CC-BF2C-0039B72D700A@democritos.it><10381.79.17.215.128.1220468090.squirrel@mail.unito.it> Message-ID: <00f601c90e75$b0ce1c70$2376c082@desdemona> YES yes you are right two more questions: 1) i need to calculate properties, DOS and STM images i need a good charge density... I suppose i need a better value of ecutrho than default but i see in the manual i can also change nr* and nr*s. What are the relevant keywords to improve accuracy for STM? and how large they should be with respect to default values? 2) in plot x0* ed e1* coordinates are refered to the crystal coordinates? (or the a_0 ones) thanks again anna > > On Sep 3, 2008, at 20:54 , anna.ferrari at unito.it wrote: > >> reading the pw_forum i'm learning about a lot of keyword >> (as $wf_collect ) i didn't know (are not in the PW doc files) > > they are, they are... > >> where i can find a complete list of keywords and their settings > > Doc/INPUT_PW.* > >> and i see i have a $prefix.wfc* file for each processor, >> what i can do if i need to change the number of processors? > > this is exacty what the wf_collect option is good for > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > From giannozz at democritos.it Thu Sep 4 11:11:43 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 04 Sep 2008 11:11:43 +0200 Subject: [Pw_forum] files for PP and guess In-Reply-To: <00f601c90e75$b0ce1c70$2376c082@desdemona> References: <10112.87.21.21.158.1220439147.squirrel@mail.unito.it><28CA7AC9-A940-48CC-BF2C-0039B72D700A@democritos.it><10381.79.17.215.128.1220468090.squirrel@mail.unito.it> <00f601c90e75$b0ce1c70$2376c082@desdemona> Message-ID: <48BFA64F.6080907@democritos.it> Anna Maria Ferrari wrote: > I suppose i need a better value of ecutrho than default > but i see in the manual i can also change nr* and nr*s. but you shouldn't: they are automatically calculated > 2) in plot x0* ed e1* coordinates are refered to the crystal coordinates? > (or the a_0 ones) a_0 units -- Paolo Giannozzi, Democritos and University of Udine, Italy From prasenjit.jnc at gmail.com Thu Sep 4 11:26:22 2008 From: prasenjit.jnc at gmail.com (Prasenjit Ghosh) Date: Thu, 4 Sep 2008 11:26:22 +0200 Subject: [Pw_forum] files for PP and guess In-Reply-To: <00f601c90e75$b0ce1c70$2376c082@desdemona> References: <10112.87.21.21.158.1220439147.squirrel@mail.unito.it> <28CA7AC9-A940-48CC-BF2C-0039B72D700A@democritos.it> <10381.79.17.215.128.1220468090.squirrel@mail.unito.it> <00f601c90e75$b0ce1c70$2376c082@desdemona> Message-ID: <627e0ffa0809040226w55d86b8exa20327911b96688@mail.gmail.com> Dear Anna, For any calculation you need to choose a well converged value of ecutrho. However, for DOS and STM, since one needs to do integration over the Brillouin zone, one should use large no. of k-points, much more than one uses for a typical scf or relaxation cal. Prasenjit. On 04/09/2008, Anna Maria Ferrari wrote: > > YES yes you are right > > two more questions: > > 1) i need to calculate properties, DOS and STM images > i need a good charge density... > > I suppose i need a better value of ecutrho than default > but i see in the manual i can also change nr* and nr*s. > What are the relevant keywords to improve accuracy for STM? and how large > they should be with respect to default values? > > 2) in plot x0* ed e1* coordinates are refered to the crystal coordinates? > (or the a_0 ones) > > thanks again > anna > > > > > > > On Sep 3, 2008, at 20:54 , anna.ferrari at unito.it wrote: > > > >> reading the pw_forum i'm learning about a lot of keyword > >> (as $wf_collect ) i didn't know (are not in the PW doc files) > > > > they are, they are... > > > >> where i can find a complete list of keywords and their settings > > > > Doc/INPUT_PW.* > > > >> and i see i have a $prefix.wfc* file for each processor, > >> what i can do if i need to change the number of processors? > > > > this is exacty what the wf_collect option is good for > > > > Paolo > > --- > > Paolo Giannozzi, Dept of Physics, University of Udine > > via delle Scienze 208, 33100 Udine, Italy > > Phone +39-0432-558216, fax +39-0432-558222 > > > > > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- PRASENJIT GHOSH, POST-DOC, ROOM NO: 265, MAIN BUILDING, CM SECTION, ICTP, STRADA COSTERIA 11, TRIESTE, 34104, ITALY PHONE: +39 040 2240 369 (O) +39 3807528672 (M) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080904/5e882bf6/attachment.htm From paulatto at sissa.it Thu Sep 4 11:42:03 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Thu, 4 Sep 2008 11:42:03 +0200 (CEST) Subject: [Pw_forum] How to analyze the result of pdos In-Reply-To: <7169925.423311220499074710.JavaMail.coremail@bj126app104.126.com> References: <7169925.423311220499074710.JavaMail.coremail@bj126app104.126.com> Message-ID: <14079.147.122.5.216.1220521323.squirrel@webmail.sissa.it> On Gio, Settembre 4, 2008 05:31, wangqj1 wrote: > > Dear all > After I used projwfc.x ,I got the pdos of every atom ,but I don't > know the signification of the following characters of > pdos_atm#1(Zn)_wfc#1(d) and pdos_atm#1(Zn)_wfc#2(s). > pdos_atm#1(Zn)_wfc#1(d) projected dos of atom number 1 (Zinc) wavefunction number 1 (d-orbital) > pdos_atm#1(Zn)_wfc#2(s) projected dos of atom number 1 (Zinc) wavefunction number 2 (s-orbital) Each file contains 3 + (2l+1) columns, where l is the angular momentum of its specific wavefunction. The first column is the energy. The second and third are the total spin up and spin down dos; where "total" means its summed over all values of m; than you find the projected dos for each value of m, up and down, one at a time. e.g. for a p orbital you have p_x, p_y and p_z orbitals: # E(eV) ldosup(E) ldosdw(E) pdosup(E) pdosdw(E) pdosup(E)... energy [sum_xyz p up] [sum_xyz p down] [p_z up] [p_z down] [p_x up]... I hope I was clear, regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From giannozz at democritos.it Thu Sep 4 11:51:17 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 04 Sep 2008 11:51:17 +0200 Subject: [Pw_forum] Symdynph_gq In-Reply-To: References: Message-ID: <48BFAF95.8020205@democritos.it> Ezad Shojaee wrote: > May i know the reference that the routine "Symdynphq_gq" is based on? there is unfortunately no reference, apart from the code itself and from some "oral tradition" P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From munima at alumni.iitg.ernet.in Thu Sep 4 12:56:03 2008 From: munima at alumni.iitg.ernet.in (munima at alumni.iitg.ernet.in) Date: Thu, 4 Sep 2008 16:26:03 +0530 Subject: [Pw_forum] wrong order of k points In-Reply-To: References: <51367.125.20.8.170.1220500283.squirrel@webmail.iitg.ac.in> Message-ID: Dear Paolo and Viet Thanks for your replies. I did do nscf calculations and there was no mismatch of q-vector in nscf and phonon. I rechecked outdir's also. But still I was getting the error. However today I tried lnscf=.true. (thanks to paolo, I did not try this option before) in phonon instead of doing nscf calculation separately. Surprizingly, it worked. I do not know why the phonon code could not read nscf outputs properly in earlier case. I have done such calculations several times before and never encountered this problem. with regards munima ________________________________ > > > ---------------------------- Original Message ---------------------------- > Subject: Re: [Pw_forum] wrong order of k points > From: "Nguyen Huy Viet" > Date: Wed, September 3, 2008 4:54 pm > To: "PWSCF Forum" > -------------------------------------------------------------------------- > > On Wed, Sep 3, 2008 at 12:30 PM, Paolo Giannozzi > wrote: > >> >> On Sep 3, 2008, at 10:48 , Munima Bora Sahariah wrote: >> >>>> from phq_init : error # 1 >>>> wrong order of k points >> >> you either did not perform the intermediate non-scf calculation >> needed for q/=0 (old QE versions), or you did not specify the >> option lnscf=.true. to phonon (new QE versions) >> > > q-vector in non-scf and phonon calculations are not the same might be > another possibility. > Viet > -- > Mr. Viet Huy Nguyen > PhD student, Condensed Matter Theory Sector > SISSA/ISAS Via Beirut 2-4, 34014, Trieste, Italy > Tel: +39 040 3787 492 Fax: +39 040 3787 528 > Email: nhviet at sissa.it hviet.nguyen at gmail.com > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > ----------------------------------------------------------------------------------- > This email was sent from IIT Guwahati Webmail. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. > http://www.iitg.ac.in From jtmullen at ncsu.edu Thu Sep 4 15:55:13 2008 From: jtmullen at ncsu.edu (jtmullen at ncsu.edu) Date: Thu, 4 Sep 2008 09:55:13 -0400 (EDT) Subject: [Pw_forum] Writing output files in parallel Message-ID: <59725.75.182.109.108.1220536513.squirrel@webmail.ncsu.edu> Greetings all, I am interested in modifying my local copy of espresso to write some data being calculated across multiple processors in much the same manner as the output wfc* (* = 1,2,3,4,..., N) files from an SCF calculation. Can someone point me to the relevant code in v4.0.1 which writes the wavefunction information to the "prefix.wfc*" file? Many hours worth of egrep have failed to point me in the correct direction. Thanks in advance Jeff Mullen Physics North Carolina State University From stewart at cnf.cornell.edu Thu Sep 4 16:28:41 2008 From: stewart at cnf.cornell.edu (Derek Stewart) Date: Thu, 04 Sep 2008 10:28:41 -0400 Subject: [Pw_forum] How to analyze the result of pdos In-Reply-To: <7169925.423311220499074710.JavaMail.coremail@bj126app104.126.com> References: <7169925.423311220499074710.JavaMail.coremail@bj126app104.126.com> Message-ID: <20080904142841.77913.qmail@mail.spidergraphics.com> Just to follow up on Lorenzo's comments, you can find more information on the partial DOS order in the archives of the pwscf forum. This is always good place to start with your questions. In particular, see the following comment by Stefano de Gironcoli (see below). Best regards, Derek http://www.democritos.it/pipermail/pw_forum/2006-November/005393.html Stefano de Gironcoli degironc at sissa.it Sat Nov 25 08:45:04 CET 2006 * Previous message: [Pw_forum] help needed with projwfc.x * Next message: [Pw_forum] help needed with projwfc.x * Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] the sequence is: 1 dz2 (m=0) 2 dzx (real combination of m=+/-1 with cosine) 3 dzy (real combination of m=+/-1 with sine) 4 dx2-y2 (real combination of m=+/-2 with cosine) 5 dxy (real combination of m=+/-1 with sine) similarly for p orbitals 1 pz (m=0) 2 px (real combination of m=+/-1 with cosine) 3 py (real combination of m=+/-1 with sine) stefano On Sat, 25 Nov 2006, Clark Lee wrote: > The same question. > Anyone else knows what's the sequence of the d-orbitals > (dxy,dxz,dyz,dx2-y2,dz2) are assigned in results from projwfc.x > wangqj1 writes: > > Dear all > After I used projwfc.x ,I got the pdos of every atom ,but I don't know the signification of the following characters of pdos_atm#1(Zn)_wfc#1(d) and pdos_atm#1(Zn)_wfc#2(s). > > pdos_atm#1(Zn)_wfc#1(d): > # E (eV) ldosup(E) ldosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) pdosup(E) pdosdw(E) > > pdos_atm#1(Zn)_wfc#2(ds): > # E (eV) ldosup(E) ldosdw(E) pdosup(E) pdosdw(E) > > Besides,I want to draw the figure of pdos of every atom on the same figure .for Zn3p,Zn4s,and so on .So what data should I extract from the pdos_atm#1(Zn)_wfc#1(d)and pdos_atm#1(Zn)_wfc#2(ds),for there are so many information . > Any advice will be appreciated ! > sincerely > > XiangTan University ################################ Derek Stewart, Ph. D. Scientific Computation Associate http://www.people.cornell.edu/pages/das248/ 250 Duffield Hall Cornell Nanoscale Facility (CNF) Ithaca, NY 14853 stewart (at) cnf.cornell.edu (607) 255-2856 From eyvaz_isaev at yahoo.com Thu Sep 4 17:53:52 2008 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 4 Sep 2008 08:53:52 -0700 (PDT) Subject: [Pw_forum] How to calculate real space inter-atomic force constants excluding the contribution due to macroscopic electric field In-Reply-To: <59b3d4f30809031105mfbedea3he498815db9251f22@mail.gmail.com> Message-ID: <402921.77802.qm@web65711.mail.ac4.yahoo.com> Dear Ravi, From conor.hogan at roma2.infn.it Thu Sep 4 17:54:18 2008 From: conor.hogan at roma2.infn.it (Conor Hogan) Date: Thu, 4 Sep 2008 17:54:18 +0200 (CEST) Subject: [Pw_forum] Yambo code released under GPL license! Message-ID: <44487.141.108.248.231.1220543658.squirrel@141.108.248.231> Dear PWscf/Quantum-Expresso community, We would like to announce the release, under the GPL license, of the "Yambo" code. Yambo is a multi-purpose C/Fortran code which can compute various excited state properties (GW corrections, absorption spectra including excitonic effects, etc) using many body perturbation theory. We invite you to download the code freely at http://www.yambo-code.org. Yambo is interfaced with the PWscf code, following a standard non-self-consistent run. The interface makes use of the IOTK library of Giovanni Bussi, now part of the standard QE distribution, and the QEXML library routies of Andrea Ferretti. See the webpage for instructions for use, tutorials, and so on. We hope you enjoy using Yambo! Best regards, The Yambo development team - Andrea Marini, Conor Hogan, Myrta Gruning, Daniele Varsano (yambo at yambo-code.org) NB: Please note that Yambo has no affiliation with PWscf, so do NOT post issues regarding Yambo to this mailing list! ---- Dr. Conor Hogan --- ----- --- --- Dipartimento di Fisica e CNR-INFM | : | : Universita' di Roma "Tor Vergata" |-- : --- |--- Tel: +39 06 7259 4121 *NEW* | : | : Fax: +39 06 2023507 --- : --- : http://www.fisica.uniroma2.it/~cmtheo-group/ European Theoretical Spectroscopy Facility http://www.etsf.eu Q. What do you call a fish with no eyes? A. A fsh. (Tommy Cooper) The early bird gets the worm, but the second mouse gets the cheese. - S. Wright If you go through a lot of hammers each month, I don't think it necessarily means you're a hard worker. It may just mean that you have a lot to learn about proper hammer maintenance - J. Handey From giannozz at democritos.it Fri Sep 5 09:52:56 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 05 Sep 2008 09:52:56 +0200 Subject: [Pw_forum] Writing output files in parallel In-Reply-To: <59725.75.182.109.108.1220536513.squirrel@webmail.ncsu.edu> References: <59725.75.182.109.108.1220536513.squirrel@webmail.ncsu.edu> Message-ID: <48C0E558.5030506@democritos.it> jtmullen at ncsu.edu wrote: > I am interested in modifying my local copy of espresso to write some data > being calculated across multiple processors in much the same manner as the > output wfc* (* = 1,2,3,4,..., N) files from an SCF calculation. Can > someone point me to the relevant code in v4.0.1 which writes the > wavefunction information to the "prefix.wfc*" file? At the end of the calculation? look for "pw_writefile", in PW/pw_restart.f90 . During the calculation, wavefunctions are accessed/read/written all the time. Look for "iunwfc", "davcio", "get_buffer", "save_buffer". The latter routines read/write to file or to memory, depending on the options Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From paulatto at sissa.it Fri Sep 5 10:25:32 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Fri, 5 Sep 2008 10:25:32 +0200 (CEST) Subject: [Pw_forum] Writing output files in parallel In-Reply-To: <59725.75.182.109.108.1220536513.squirrel@webmail.ncsu.edu> References: <59725.75.182.109.108.1220536513.squirrel@webmail.ncsu.edu> Message-ID: <16367.82.50.174.28.1220603132.squirrel@webmail.sissa.it> On Gio, Settembre 4, 2008 15:55, jtmullen at ncsu.edu wrote: > Can > someone point me to the relevant code in v4.0.1 which writes the > wavefunction information to the "prefix.wfc*" file? Dear Jeff, it may be easier than you think: most code is executed by all processors, so you only need to create a unique filename for each processor. It can be done using the global variables mpime (absolute processor index) or me_image and my_image_id (processor index within an image and index of the image); all of them are in module mp_global. The QE approach is to create a different directory for each image. Images are used only for NEB calculations and similar stuff, so I will ignore them from here on. The following is a code example that should work: USE mp_global, ONLY : me_image, my_image_id, nimage USE io_files, ONLY : tmp_dir, prefix, iunmystuff !<- you should add it USE io_global, ONLY : ionode CHARACTER(len=256) :: filename, basename basename = tmp_dir // '/' // prefix // '_' // "my_stuff" // '.' IF (nimage>1) STOP 1 IF(ionode) WRITE(stdout, '(5x,a)') "I'm going to write my stuff..." WRITE(filename, '(a,I5.5)' ) basename, me_image OPEN(filename, unit=iunmystuff ...) ... If you really want to see how it is done in pw, you'll find it in PW/pw_restart.f90 and Modules/xml_io_base.f90, but that part of code is a bit difficult to read, as it takes in account many different cases. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From ierrea001 at ikasle.ehu.es Fri Sep 5 15:11:08 2008 From: ierrea001 at ikasle.ehu.es (Ion Errea Lope) Date: Fri, 05 Sep 2008 15:11:08 +0200 Subject: [Pw_forum] original symmetry operations not satisfied Message-ID: <1220620268.7011.57.camel@lcpybm> Dear all, Lately, I am trying to relax a body centered tetragonal structure with 16 atoms per cell using a 'relax' calculation in order to minimize the forces for future phonon calculations. Here is my input: &control calculation='relax' restart_mode='from_scratch', prefix='ca', dt = 40.0, etot_conv_thr = 1.0D-5 , forc_conv_thr = 1.0D-4 , tstress = .true. tprnfor = .true. ... / &ions ion_dynamics = 'damp' pot_extrapolation = 'first_order' , wfc_extrapolation = 'first_order' , ... ATOMIC_POSITIONS {crystal} ... Ca 0.50000 0.81573 0.01829 0 1 1 Ca 0.50000 0.51829 0.31573 0 1 1 Ca 0.50000 0.48171 0.68427 0 1 1 Ca 0.50000 0.18427 0.98171 0 1 1 ... Once the first scf calculation is done and the new coordinates calculated, I get the following error message in the output file: ATOMIC_POSITIONS (crystal) ... Ca 0.500116912 0.815774675 0.018451586 0 1 1 Ca 0.499883088 0.518568498 0.315891586 0 1 1 Ca 0.500116912 0.481431502 0.684108414 0 1 1 Ca 0.499883088 0.184225325 0.981548414 0 1 1 SYMMETRIZED ATOMIC COORDINATES) ATOMIC_POSITIONS (crystal) ... Ca 0.500000000 0.815803903 0.018539270 0 1 1 Ca 0.500000000 0.518539270 0.315803903 0 1 1 Ca 0.500000000 0.481460730 0.684196097 0 1 1 Ca 0.500000000 0.184196097 0.981460730 0 1 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%% from checkallsym : error # 1 some of the original symmetry operations not satisfied %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%% stopping ... Apparently, after the symmetrized atomic coordinates are written the calculation stops without proceeding with the next scf calculation. I have checked the mail list and realised more people encountered this problem. Some of the proposals I found mentioned trying with different ion_dynamics (bfgs and damp), changing first order extrapolation to second order and finally, using different number of processors. I have tried both bfgs and damp calculations, fisrt order and second order extrapolation, and also calculations with different numbers of processors. However neither of them worked in my case and the same error was found. Another option is not to consider symmetry operations including the flag no_sym = .true. Although this works correctly and relaxes the structure (taking much longer though due to the larger number of kpoints) does not keep the atoms at the symmetry point 0.5. I wonder if there is any option to make the relaxation in y and z coordinates without relaxing in x, that is, keeping symmetry operations but without having this error. Thanks in advance and kind regards, Ion -- ?======================================== Ion Errea Lope Materia Kondentsatuaren Fisika Saila Zientzia eta Teknologia Fakultatea Euskal Herriko Unibertsitatea 644 Posta Kutxatila, 48080 Bilbo, Spain Telefonoa: +34 946015322 Mail: ierrea001 at ikasle.ehu.es ======================================== From calandra at impmc.jussieu.fr Fri Sep 5 15:22:28 2008 From: calandra at impmc.jussieu.fr (matteo calandra) Date: Fri, 5 Sep 2008 15:22:28 +0200 Subject: [Pw_forum] XSpectra: x-ray absorption spectra in Quantom Expresso Message-ID: <20080905152228.q7tml1tkf4os44w4@www-int.impmc.jussieu.fr> > Dear all, > Can Quantum Espresso calculate x-ray absorption spectra? or you can > recommend me some > good software to calculate XAS? > Regards > Lily Dear Lily, a new package has been included in the current CVS of the quantum espresso distribution calculating K-edge X-ray absorption spectra. The calculation uses gipaw and the lanczos method. Norm conserving pseudopotentials are allowed and in september we will also upgrade to ultrasoft pseudopotential. Spin polarised and DFT+U calculations are allowed. The main advantage of this code is that, using the continous fraction, it only requires the calculation of the occupied states and thus no calculation of empty bands is necessary. Core-hole effects are treated generating a pseudopotential with a core-hole in the 1s state and then performing the calculation on a supercell with a single absorbing atom. The code is available under the directory XSpectra of the current CVS. There is a short manual and an example is available under the examples directory (NiO). It requires the generation of a GIPAW pseudopotential with a core-hole in the 1s state for the absorbing atom. We have pseudopotentials of this kind for several elements and they will be made public very soon. More examples will be soon included. The generalisation to ultrasoft pseudopotential has been terminated and will be uploaded in september. Please contact me if you need more informations on the code. Matteo Calandra ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From giannozz at democritos.it Fri Sep 5 15:59:13 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 05 Sep 2008 15:59:13 +0200 Subject: [Pw_forum] original symmetry operations not satisfied In-Reply-To: <1220620268.7011.57.camel@lcpybm> References: <1220620268.7011.57.camel@lcpybm> Message-ID: <48C13B31.7010904@democritos.it> Ion Errea Lope wrote: > Some of the proposals I found mentioned trying with different > ion_dynamics (bfgs and damp), changing first order extrapolation to > second order and finally, using different number of processors none of which is good (actually, none of which make any sense at all). You should specify which version of the code you used and report the complete input data. > I wonder if there is any option to make the relaxation in y and z > coordinates without relaxing in x, that is, keeping symmetry operations > but without having this error. it depends. If you have symmetry operations sending x into either y or z, you cannot. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From lfhuang at theory.issp.ac.cn Fri Sep 5 15:53:29 2008 From: lfhuang at theory.issp.ac.cn (L.F.Huang) Date: Fri, 5 Sep 2008 21:53:29 +0800 (CST) Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell Message-ID: <32973.210.73.17.131.1220622809.squirrel@210.73.16.4> Dear PWSCF developers and users: I am calculating a graphene supercell with one H atom one it, and I take 4*4*1 size of supercell now, and there are 33 atoms with 32 C atoms and 1 H atom. I calculate its phonon dispersions with 13 nodes with 2G random memory for each node. When frequencies of one phonon mode are calculated, the program enter into calculation of electron-phonon interactions. But it has passed nearly 30 hours, and the calculation of el-ph interaction still not completed! And for graphene unit cell with 2 C atoms, this process just costs about 10 seconds. I don't know it is normal that the calculation of el-ph interaction of my graphene supercell costs so much or so much more time? Any one has the experiences in the el-ph calculation of supercells? Could anyone tell me how can I do or how much time I would expect to wait? Thanks very much! Best Wishes! Yours Sincerely L.F.Huang ====================================================================== L.F.Huang(???) lfhuang at theory.issp.ac.cn ====================================================================== Add: Research Laboratory for Computational Materials Sciences, Instutue of Solid State Physics,the Chinese Academy of Sciences, P.O.Box 1129, Hefei 230031, P.R.China Tel: 86-551-5591464-328(office) Fax: 86-551-5591434 Web: http://theory.issp.ac.cn (website of our theory group) http://www.issp.ac.cn (website of our institute) ====================================================================== From akohlmey at cmm.chem.upenn.edu Fri Sep 5 16:33:27 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 5 Sep 2008 10:33:27 -0400 (EDT) Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: <32973.210.73.17.131.1220622809.squirrel@210.73.16.4> References: <32973.210.73.17.131.1220622809.squirrel@210.73.16.4> Message-ID: On Fri, 5 Sep 2008, L.F.Huang wrote: LFH> Dear PWSCF developers and users: LFH> I am calculating a graphene supercell with one H atom one it, and I LFH> take 4*4*1 size of supercell now, and there are 33 atoms with 32 C atoms LFH> and 1 H atom. I calculate its phonon dispersions with 13 nodes with 2G LFH> random memory for each node. When frequencies of one phonon mode are dear l.f. huang, more important that the amount of memory in your nodes would be know what kind of interconnect (gigabit, infiniband, myrinet, ...) you have and how you run in parallel (g-space parallelization only, g-space+k-points, ...) and what type and how many cpus/cores per node you have and even what kind of parallel library and math library you use. all of that can have a serious impact on performance. have you done a scaling test for the first part of the calculation? unless you are smart about it, adding more nodes does not necessarily speed up your calculations, specifically, if you have only a gigabit ethernet interconnect. please see, e.g., http://www.democritos.it/pipermail/pw_forum/2008-June/009293.html for a demonstration of how one can affect performance through the way the QE codes are run. LFH> calculated, the program enter into calculation of electron-phonon LFH> interactions. But it has passed nearly 30 hours, and the calculation of LFH> el-ph interaction still not completed! And for graphene unit cell with LFH> 2 C atoms, this process just costs about 10 seconds. you should be aware of the fact, that this kind of calculation does not scale linearly; the cost of an scf caclculation alone generally goes up by about an order magnitude when doubling the system size. cheers, axel. LFH> I don't know it is normal that the calculation of el-ph interaction LFH> of my graphene supercell costs so much or so much more time? Any one LFH> has the experiences in the el-ph calculation of supercells? Could anyone LFH> tell me how can I do or how much time I would expect to wait? LFH> LFH> Thanks very much! LFH> Best Wishes! LFH> LFH> Yours Sincerely LFH> L.F.Huang LFH> LFH> ====================================================================== LFH> L.F.Huang(???) lfhuang at theory.issp.ac.cn LFH> ====================================================================== LFH> Add: Research Laboratory for Computational Materials Sciences, LFH> Instutue of Solid State Physics,the Chinese Academy of Sciences, LFH> P.O.Box 1129, Hefei 230031, P.R.China LFH> Tel: 86-551-5591464-328(office) LFH> Fax: 86-551-5591434 LFH> Web: http://theory.issp.ac.cn (website of our theory group) LFH> http://www.issp.ac.cn (website of our institute) LFH> ====================================================================== LFH> LFH> LFH> _______________________________________________ LFH> Pw_forum mailing list LFH> Pw_forum at pwscf.org LFH> http://www.democritos.it/mailman/listinfo/pw_forum LFH> -- ======================================================================= 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 democritos.it Fri Sep 5 16:44:09 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 05 Sep 2008 16:44:09 +0200 Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: <32973.210.73.17.131.1220622809.squirrel@210.73.16.4> References: <32973.210.73.17.131.1220622809.squirrel@210.73.16.4> Message-ID: <48C145B9.9080705@democritos.it> L.F.Huang wrote: > Could anyone tell me how can I do or how much time I would expect to wait? nobody can tell you anything like this. You should try to have an idea of the needed time by trying first with the smaller cell, then with a small supercell, etc.. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From lfhuang at theory.issp.ac.cn Sat Sep 6 14:26:07 2008 From: lfhuang at theory.issp.ac.cn (L.F.Huang) Date: Sat, 6 Sep 2008 20:26:07 +0800 (CST) Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: References: Message-ID: <32820.210.73.17.131.1220703967.squirrel@210.73.16.4> Dear prof. Kohlmeyer and Giannozzi: Thank both of you for your kindly help! Frankly, I have not taken serious tests. When I found the calculations of frequencies(99 modes) for one wave vector of phonon in the 4*4*1 graphene supercell with one H atom on it took about 20 days, then I took it for granted that the calculation of el-ph interaction would accomplish for several hours or a little more. But nearly 3 days having passed, it is still doing the calculation......I realy need some serious tests to let it clear in my mind. Thank you for your advices for testing! Thank you very much! Best Wishes! Yours Sincerely! L.F.Huang PostScript: Amount of memory: 2GB per node kind of interconnect: myrinet 2000 No. of reduced k points: 52 No. of nodes: 13 (with npool by default 1) Parallel environment: MPI 1.2 Math library: NAG ACML 1.5?ATLAS 3.6 No. of cpus per node: 4 > Date: Fri, 5 Sep 2008 10:33:27 -0400 (EDT) > From: Axel Kohlmeyer > Subject: Re: [Pw_forum] time for calculation of el-ph interactions of > graphene supercell > To: "L.F.Huang" > Cc: pw_forum at pwscf.org > Message-ID: > Content-Type: text/plain; charset="gb2312" > > On Fri, 5 Sep 2008, L.F.Huang wrote: > > LFH> Dear PWSCF developers and users: > LFH> I am calculating a graphene supercell with one H atom one it, and > I LFH> take 4*4*1 size of supercell now, and there are 33 atoms with 32 > C atoms LFH> and 1 H atom. I calculate its phonon dispersions with 13 > nodes with 2G LFH> random memory for each node. When frequencies of one > phonon mode are > > dear l.f. huang, > > more important that the amount of memory in your nodes would > be know what kind of interconnect (gigabit, infiniband, myrinet, ...) > you have and how you run in parallel (g-space parallelization only, > g-space+k-points, ...) and what type and how many cpus/cores per node > you have and even what kind of parallel library and math library you > use. all of that can have a serious impact on performance. > > have you done a scaling test for the first part of the calculation? > unless you are smart about it, adding more nodes does not necessarily > speed up your calculations, specifically, if you have only a gigabit > ethernet interconnect. > > please see, e.g., > http://www.democritos.it/pipermail/pw_forum/2008-June/009293.html > for a demonstration of how one can affect performance through the > way the QE codes are run. > > LFH> calculated, the program enter into calculation of electron-phonon > LFH> interactions. But it has passed nearly 30 hours, and the > calculation of LFH> el-ph interaction still not completed! And for > graphene unit cell with LFH> 2 C atoms, this process just costs about 10 > seconds. > > you should be aware of the fact, that this kind of calculation does not > scale linearly; the cost of an scf caclculation alone generally goes up > by about an order magnitude when doubling the system size. > > cheers, > axel. > > LFH> I don't know it is normal that the calculation of el-ph > interaction LFH> of my graphene supercell costs so much or so much more > time? Any one LFH> has the experiences in the el-ph calculation of > supercells? Could anyone LFH> tell me how can I do or how much time I > would expect to wait? LFH> > LFH> Thanks very much! > LFH> Best Wishes! > LFH> > LFH> Yours Sincerely > LFH> L.F.Huang > LFH> > LFH> > ====================================================================== > LFH> L.F.Huang(???) lfhuang at theory.issp.ac.cn > LFH> > ====================================================================== > LFH> Add: Research Laboratory for Computational Materials Sciences, > LFH> Instutue of Solid State Physics,the Chinese Academy of > Sciences, LFH> P.O.Box 1129, Hefei 230031, P.R.China > LFH> Tel: 86-551-5591464-328(office) > LFH> Fax: 86-551-5591434 > LFH> Web: http://theory.issp.ac.cn (website of our theory group) > LFH> http://www.issp.ac.cn (website of our institute) > LFH> > ====================================================================== > LFH> > LFH> > LFH> _______________________________________________ > LFH> Pw_forum mailing list > LFH> Pw_forum at pwscf.org > LFH> http://www.democritos.it/mailman/listinfo/pw_forum > LFH> > > -- > ======================================================================= > 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. > > ------------------------------ > > Date: Fri, 05 Sep 2008 16:44:09 +0200 > From: Paolo Giannozzi > Subject: Re: [Pw_forum] time for calculation of el-ph interactions of > graphene supercell > To: PWSCF Forum > Message-ID: <48C145B9.9080705 at democritos.it> > Content-Type: text/plain; charset=GB2312 > > L.F.Huang wrote: > >> Could anyone tell me how can I do or how much time I would expect to >> wait? > > nobody can tell you anything like this. You should try to have an > idea of the needed time by trying first with the smaller cell, then with > a small supercell, etc.. > > 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 lfhuang at theory.issp.ac.cn Sat Sep 6 14:32:24 2008 From: lfhuang at theory.issp.ac.cn (L.F.Huang) Date: Sat, 6 Sep 2008 20:32:24 +0800 (CST) Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell Message-ID: <32835.210.73.17.131.1220704344.squirrel@210.73.16.4> Sorry! the amount of memory is 8GB per node, not 2GB. Yours Sincerely L.F.Huang From wangqj1 at 126.com Sun Sep 7 03:49:08 2008 From: wangqj1 at 126.com (wangqj1) Date: Sun, 7 Sep 2008 09:49:08 +0800 (CST) Subject: [Pw_forum] How to computer on parallel on machines Message-ID: <30568826.20971220752148791.JavaMail.coremail@bj126app75.126.com> Dear pwscf users I want to computer on two computers,but I don't know how to connect the two computers, I want to know the specific approaches.Could someone give me some advice ? sincerely XiangTan University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080907/41980fd8/attachment.htm From clie at aphy.iphy.ac.cn Sun Sep 7 05:11:12 2008 From: clie at aphy.iphy.ac.cn (clie) Date: Sun, 07 Sep 2008 03:11:12 GMT Subject: [Pw_forum] Pw_forum Digest, Vol 15, Issue 11 In-Reply-To: References: Message-ID: <20080907031112.27825.eqmail@aphy.iphy.ac.cn> dear, L.F.Huang, you will be waiting for several weeks or one month until the calculation finished, i guess. > Message: 7 > Date: Fri, 5 Sep 2008 10:33:27 -0400 (EDT) > From: Axel Kohlmeyer > Subject: Re: [Pw_forum] time for calculation of el-ph interactions of > graphene supercell > To: "L.F.Huang" > Cc: pw_forum at pwscf.org > Message-ID: > Content-Type: text/plain; charset="gb2312" > > On Fri, 5 Sep 2008, L.F.Huang wrote: > > LFH> Dear PWSCF developers and users: > LFH> I am calculating a graphene supercell with one H atom one it, and I > LFH> take 4*4*1 size of supercell now, and there are 33 atoms with 32 C atoms > LFH> and 1 H atom. I calculate its phonon dispersions with 13 nodes with 2G > LFH> random memory for each node. When frequencies of one phonon mode are > > dear l.f. huang, > > more important that the amount of memory in your nodes would > be know what kind of interconnect (gigabit, infiniband, myrinet, ...) > you have and how you run in parallel (g-space parallelization only, > g-space+k-points, ...) and what type and how many cpus/cores per node > you have and even what kind of parallel library and math library you use. > all of that can have a serious impact on performance. > > have you done a scaling test for the first part of the calculation? > unless you are smart about it, adding more nodes does not necessarily > speed up your calculations, specifically, if you have only a gigabit > ethernet interconnect. > > please see, e.g., > http://www.democritos.it/pipermail/pw_forum/2008-June/009293.html > for a demonstration of how one can affect performance through the > way the QE codes are run. > > LFH> calculated, the program enter into calculation of electron-phonon > LFH> interactions. But it has passed nearly 30 hours, and the calculation of > LFH> el-ph interaction still not completed! And for graphene unit cell with > LFH> 2 C atoms, this process just costs about 10 seconds. > > you should be aware of the fact, that this kind of calculation does not > scale linearly; the cost of an scf caclculation alone generally goes > up by about an order magnitude when doubling the system size. > > cheers, > axel. > > LFH> I don't know it is normal that the calculation of el-ph interaction > LFH> of my graphene supercell costs so much or so much more time? Any one > LFH> has the experiences in the el-ph calculation of supercells? Could anyone > LFH> tell me how can I do or how much time I would expect to wait? > LFH> > LFH> Thanks very much! > LFH> Best Wishes! > LFH> > LFH> Yours Sincerely > LFH> L.F.Huang > LFH> > LFH> ====================================================================== > LFH> L.F.Huang(???) lfhuang at theory.issp.ac.cn > LFH> ====================================================================== > LFH> Add: Research Laboratory for Computational Materials Sciences, > LFH> Instutue of Solid State Physics,the Chinese Academy of Sciences, > LFH> P.O.Box 1129, Hefei 230031, P.R.China > LFH> Tel: 86-551-5591464-328(office) > LFH> Fax: 86-551-5591434 > LFH> Web: http://theory.issp.ac.cn (website of our theory group) > LFH> http://www.issp.ac.cn (website of our institute) > LFH> ====================================================================== > LFH> > LFH> > LFH> _______________________________________________ > LFH> Pw_forum mailing list > LFH> Pw_forum at pwscf.org > LFH> http://www.democritos.it/mailman/listinfo/pw_forum > LFH> > > -- > ======================================================================= > 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. > > ------------------------------ > > Message: 8 > Date: Fri, 05 Sep 2008 16:44:09 +0200 > From: Paolo Giannozzi > Subject: Re: [Pw_forum] time for calculation of el-ph interactions of > graphene supercell > To: PWSCF Forum > Message-ID: <48C145B9.9080705 at democritos.it> > Content-Type: text/plain; charset=GB2312 > > L.F.Huang wrote: > >> Could anyone tell me how can I do or how much time I would expect to wait? > > nobody can tell you anything like this. You should try to have an > idea of the needed time by trying first with the smaller cell, then > with a small supercell, etc.. > > 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 > > > End of Pw_forum Digest, Vol 15, Issue 11 > **************************************** From lfhuang at theory.issp.ac.cn Sun Sep 7 10:10:17 2008 From: lfhuang at theory.issp.ac.cn (L.F.Huang) Date: Sun, 7 Sep 2008 16:10:17 +0800 (CST) Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: References: Message-ID: <32947.210.73.17.131.1220775017.squirrel@210.73.16.4> Dear clie: Thank you for your kindly attention! Would you please tell me how do you make your guess? Or could you give me some examples in your own experience? Best Wishes! Yours Sincerely L.F.Huang > Date: Sun, 07 Sep 2008 03:11:12 GMT > From: "clie" > Subject: Re: [Pw_forum] Pw_forum Digest, Vol 15, Issue 11 > To: pw_forum at pwscf.org > Message-ID: <20080907031112.27825.eqmail at aphy.iphy.ac.cn> > Content-Type: text/plain; format=flowed; charset="gb2312" > > dear, L.F.Huang, you will be waiting for several weeks or one month > until the calculation finished, i guess. ====================================================================== L.F.Huang(???) lfhuang at theory.issp.ac.cn ====================================================================== Add: Research Laboratory for Computational Materials Sciences, Instutue of Solid State Physics,the Chinese Academy of Sciences, P.O.Box 1129, Hefei 230031, P.R.China Tel: 86-551-5591464-328(office) Fax: 86-551-5591434 Web: http://theory.issp.ac.cn (website of our theory group) http://www.issp.ac.cn (website of our institute) ====================================================================== From akohlmey at cmm.chem.upenn.edu Sun Sep 7 20:29:11 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sun, 7 Sep 2008 14:29:11 -0400 (EDT) Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: <32820.210.73.17.131.1220703967.squirrel@210.73.16.4> References: <32820.210.73.17.131.1220703967.squirrel@210.73.16.4> Message-ID: On Sat, 6 Sep 2008, L.F.Huang wrote: LFH> PostScript: LFH> Amount of memory: 2GB per node LFH> kind of interconnect: myrinet 2000 LFH> No. of reduced k points: 52 LFH> No. of nodes: 13 (with npool by default 1) parallelization over k-points is always the first item to try. not all components in Q-E support it fully, but when it is supported, it is superior over the g-space parallelization, which in itself has scaling limitationss. please see how in my previous demonstration to vega, the execution time goes up significantly, when using too many nodes and the wrong parallelization scheme. i would also check, whether your machine is communicating correctly. i've seen myrinet machines just stall occasionally, due to firmware crashes. LFH> Parallel environment: MPI 1.2 i assume it is MPICH-GM. mpi 1.2 does not make that much sense (there is an MPI 1.2 standard but that numbering is usually independent of the MPI standard version. e.g., i'm using OpenMPI 1.2.7 and it implements all of MPI 2.0 ... ). LFH> Math library: NAG ACML 1.5?ATLAS 3.6 ACML is from AMD not NAG, and that would conflict with ATLAS as both implement BLAS/LAPACK. LFH> No. of cpus per node: 4 which type? looks like opteron dual-core? 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 akohlmey at cmm.chem.upenn.edu Sun Sep 7 20:31:35 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sun, 7 Sep 2008 14:31:35 -0400 (EDT) Subject: [Pw_forum] How to computer on parallel on machines In-Reply-To: <30568826.20971220752148791.JavaMail.coremail@bj126app75.126.com> References: <30568826.20971220752148791.JavaMail.coremail@bj126app75.126.com> Message-ID: On Sun, 7 Sep 2008, wangqj1 wrote: WQ> WQ> WQ> Dear pwscf users WQ> I want to computer on two computers,but I don't know how to WQ> connect the two computers, I want to know the specific WQ> approaches.Could someone give me some advice ? you just set up a network between them. as to how this can be done, please contact a linux (i assume you are running linux) beginners forum and search the web for tutorials (there have to be (too) many). for this mailing list this would be off-topic. cheers, axel. WQ> WQ> sincerely WQ> WQ> XiangTan University -- ======================================================================= 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 zdw2000 at gmail.com Mon Sep 8 03:33:35 2008 From: zdw2000 at gmail.com (Wei Zhou) Date: Mon, 8 Sep 2008 09:33:35 +0800 Subject: [Pw_forum] how to build a I43-d structure in pwscf? In-Reply-To: <182393.20808.qm@web65701.mail.ac4.yahoo.com> References: <3e36a8fc0809011653j7c9322e6p4b40b446afd1f72b@mail.gmail.com> <182393.20808.qm@web65701.mail.ac4.yahoo.com> Message-ID: <3e36a8fc0809071833r545f2701x9c4b509fc82377f3@mail.gmail.com> thank you very much ,it is very kind of you 2008/9/2 Eyvaz Isaev > Hi, > > See http://cst-www.nrl.navy.mil/lattice/struk/Li_cI16.html > or > http://www.cryst.ehu.es/cgi-bin/cryst/programs/nph-wp-list?gnum=220 > > > fractional coordinates (0.044 0.044 0.044) > Most likely they are internal parameters. > > Bests, > Eyvaz. > > ------------------------------------------------------------------- > Prof. Eyvaz Isaev, > Theoretical Physics Department, Moscow State Institute of Steel & Alloys, > Russia, > Department of Physics, Chemistry, and Biology (IFM), Linkoping University, > Sweden > Condensed Matter Theory Group, Uppsala University, Sweden > Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com > > > --- On Tue, 9/2/08, Wei Zhou wrote: > > > From: Wei Zhou > > Subject: [Pw_forum] how to build a I43-d structure in pwscf? > > To: "PWSCF Forum" > > Date: Tuesday, September 2, 2008, 3:53 AM > > hello , > > I want to build a body center unit cell with 16 > > atoms in convention > > cell , 8 atoms in primitive cell, it has the I-43d group > > space with wyckoff > > 16c,and have a fractional coordinates( 0.044 0.044 > > 0.044),however ,I can not > > build it with pwscf correctly .would some one help me? any > > help will be > > appreciated > > best wishes > > the basic vector and crystal coordination of atoms are > > writed as follow > > primitive cell, > > -4.999999999999993 5.000000000000001 > > 5.000000000000004 > > 5.000000000000001 -5.000000000000002 > > 4.999999999999996 > > 5.000000000000000 4.999999999999999 > > -4.999999999999999 > > > > > > coordination > > 0.0880000293254852 0.0880000293254852 > > 0.0880000293254852 > > 0.4999999999999999 0.9999999999999998 > > 0.4119999706745147 > > 1.0000000000000000 0.4119999706745149 > > 0.4999999999999999 > > 0.4119999706745147 0.4999999999999998 > > 0.9999999999999998 > > 0.5880000293254851 0.5880000293254851 > > 0.5880000293254851 > > 1.5000000000000000 1.0000000000000000 > > 0.9119999706745148 > > 0.9119999706745149 1.5000000000000000 > > 0.9999999999999999 > > 0.9999999999999999 0.9119999706745147 > > 1.4999999999999998 > > . > > -- > > ZhouDawei > > JiLin Universiyt ,ChangChun ,China > > zdw2000 at gmail.com > > _______________________________________________ > > 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 > -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/3c683623/attachment.htm From lfhuang at theory.issp.ac.cn Mon Sep 8 03:40:59 2008 From: lfhuang at theory.issp.ac.cn (L.F.Huang) Date: Mon, 8 Sep 2008 09:40:59 +0800 (CST) Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: References: <32820.210.73.17.131.1220703967.squirrel@210.73.16.4> Message-ID: <32831.210.73.17.131.1220838059.squirrel@210.73.16.4> Dear prof. Kohlmeyer: Thank you for your kind help! I have read your previous demonstration to vega. I am also doing some tests now. I am sorry for my limitation in computer knowledge! And the computer I am using is thousands miles away from me,so I just find information from its webpage, which I list below and try to make it as complete as possible: Peak Performance: 10.2Tflops Computing Nodes: 512 4-way nodes,each composed of 4 CPUs Storage Nodes: 16 4-way nodes,each composed of 4 CPUs Accessing Nodes: 4 4-way nodes,each composed of 4 CPUs CPU: AMD OPTERON 850, 2.4GHz?2128 CPU tatally System Memory: 4256GB System Storage: 20TB Architecture: Cluster?Myrinet 2000 Operating System: Turbo Linux 8.0 Language: C, C++, Fortran77/90/95, Java Compliers: gcc3.2.2-4, Java1.2 , tcl/tk 8.0, Perl 5.0 Mathematic Libraries: NAG ACML 1.5, ATLAS 3.6 Parallel Environment: BCL4?Myricom GM, MPI 1.2, PVM 3.3.6, JobManager Tools: gdb5.3, OpenLDAP2.2, Python2.3, X-window Sysytem Administration: LSF, Dawning Cluster Operating System, DCOS, including DCMS(Dawning Cluster Management System ), DCIS(Dawning Cluster Installation System), DCMM(Dawning Cluster Monitor and Management System), Mterm(Multi-terminal software), Dawing Large-Scale KVM System(SKVM) Another question is: Prof. Kohlmeyer, you have mentioned that not all components in Q-E support k-points parallelization fully. I wonder are there still any other components like that besides postprocessing components? Are elphon.f90 and elph.f90? Thanks! Yours Sincerely L.F.Huang > On Sat, 6 Sep 2008, L.F.Huang wrote: > > LFH> PostScript: > LFH> Amount of memory: 2GB per node > LFH> kind of interconnect: myrinet 2000 > LFH> No. of reduced k points: 52 > LFH> No. of nodes: 13 (with npool by default 1) > > parallelization over k-points is always the > first item to try. not all components in Q-E > support it fully, but when it is supported, > it is superior over the g-space parallelization, > which in itself has scaling limitationss. please > see how in my previous demonstration to vega, > the execution time goes up significantly, when > using too many nodes and the wrong parallelization > scheme. i would also check, whether your machine > is communicating correctly. i've seen myrinet machines > just stall occasionally, due to firmware crashes. > > LFH> Parallel environment: MPI 1.2 > > i assume it is MPICH-GM. mpi 1.2 does not make that > much sense (there is an MPI 1.2 standard but that > numbering is usually independent of the MPI standard > version. e.g., i'm using OpenMPI 1.2.7 and it implements > all of MPI 2.0 ... ). > > LFH> Math library: NAG ACML 1.5?ATLAS 3.6 > > ACML is from AMD not NAG, and that would conflict > with ATLAS as both implement BLAS/LAPACK. > > LFH> No. of cpus per node: 4 > > which type? looks like opteron dual-core? > > 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 sauluck at iitk.ac.in Mon Sep 8 06:42:18 2008 From: sauluck at iitk.ac.in (Sushil Auluck) Date: Mon, 08 Sep 2008 10:12:18 +0530 Subject: [Pw_forum] How to computer on parallel on machines In-Reply-To: References: <30568826.20971220752148791.JavaMail.coremail@bj126app75.126.com> Message-ID: <48C4AD2A.5090009@iitk.ac.in> hi, for connecting two computers you can use a cheap ( maybe not efficient option) option. use a cross/twisted cable. both machines must have ethernet cards. i have wanted to do this for a long time. let me know your experiences. for more than 2 computers you will need a fast switch....... s.auluck -- ....................................................................... Prof. Sushil Auluck Phone:+91-512-6797092/6148 Department of Physics +91-512-6798177(Home) Indian Institute of Technology Cell :+91-9305548667 Kanpur 208016 (UP) Fax :+91-512-6790914 India E-mail:sauluck at iitk.ac.in ...............................................:sauluck at gmail.com http://www.iitk.ac.in/phy/People/phy_facvis.html http://www.iitk.ac.in/phy/New01/profile_SA.html ....................................................................... From kazempoor2000 at yahoo.com Mon Sep 8 10:11:58 2008 From: kazempoor2000 at yahoo.com (ali kazempoor) Date: Mon, 8 Sep 2008 01:11:58 -0700 (PDT) Subject: [Pw_forum] relax-spinorbit Message-ID: <844011.18596.qm@web33101.mail.mud.yahoo.com> Dear all For test ,I constructed a simple Ni nanowire along [001] and I want to calculate the magnetic anisotropy energy. I get the full relativistic pseudopotential and lattice constant from example 26 of espresso.But for relaxation with nanocoliin and lspinorbit switch it does'nt converge . where is the error or which paprameter must be modified? thanks Ali Kazempour,Isfahan univ .Of .Tech,Iran &control title = ni , calculation = 'relax' , restart_mode = 'from_scratch' , tstress = .true. tprnfor = .true. outdir = '/scratch19/user14/' , pseudo_dir = './' , prefix = ni111 , / &system ibrav= 6, celldm(1)=19.95, celldm(3)=0.3333333,nat=5, ntyp=1, ecutwfc = 40 , ecutrho =400, lspinorb=.true., noncolin=.true., starting_magnetization(1)=0.8, degauss=0.001, smearing='mp', angle1(1)=60 angle2(1)=90 occupations='smearing', ecutwfc =40 ecutrho =500.0 / &electrons conv_thr = 1.0d-4 mixing_beta = 0.8 / &IONS ion_dynamics='bfgs' / ATOMIC_SPECIES Ni 28.0 NirelPBE.RRKJ3.UPF ATOMIC_POSITIONS crystal Ni 0.5 0.5 0.0 Ni 0.3333333 0.5 0.5 Ni 0.5 0.66666666 0.5 Ni 0.6666666 0.5 0.5 Ni 0.5 0.33333333 0.5 K_POINTS automatic 1 1 50 0 0 1 From mansourehp at gmail.com Mon Sep 8 10:51:13 2008 From: mansourehp at gmail.com (Mansoureh Pashangpour) Date: Mon, 8 Sep 2008 12:21:13 +0330 Subject: [Pw_forum] relax-spinorbit In-Reply-To: <844011.18596.qm@web33101.mail.mud.yahoo.com> References: <844011.18596.qm@web33101.mail.mud.yahoo.com> Message-ID: Dear Ali First check your input file with xcrysden . then check ecut and ecutrho(8-10 times of ecut is good).then use &ions pot_extrapolation='second_order', wfc_extrapolation='second_order', good luck Mansoureh On Mon, Sep 8, 2008 at 11:41 AM, ali kazempoor wrote: > Dear all > For test ,I constructed a simple Ni nanowire along [001] and I want to > calculate the magnetic anisotropy energy. I get the full relativistic > pseudopotential and lattice constant from example 26 of espresso.But for > relaxation with nanocoliin and lspinorbit switch it does'nt converge . where > is the error or which paprameter must be modified? > thanks > Ali Kazempour,Isfahan univ .Of .Tech,Iran > &control > title = ni , > calculation = 'relax' , > restart_mode = 'from_scratch' , > tstress = .true. > tprnfor = .true. > outdir = '/scratch19/user14/' , > pseudo_dir = './' , > prefix = ni111 , > / > &system > ibrav= 6, celldm(1)=19.95, celldm(3)=0.3333333,nat=5, ntyp=1, > ecutwfc = 40 , ecutrho =400, > lspinorb=.true., > noncolin=.true., > starting_magnetization(1)=0.8, > degauss=0.001, > smearing='mp', > angle1(1)=60 > angle2(1)=90 > occupations='smearing', > ecutwfc =40 > ecutrho =500.0 > > / > &electrons > conv_thr = 1.0d-4 > mixing_beta = 0.8 > / > &IONS > ion_dynamics='bfgs' > / > ATOMIC_SPECIES > Ni 28.0 NirelPBE.RRKJ3.UPF > ATOMIC_POSITIONS crystal > Ni 0.5 0.5 0.0 > Ni 0.3333333 0.5 0.5 > Ni 0.5 0.66666666 0.5 > Ni 0.6666666 0.5 0.5 > Ni 0.5 0.33333333 0.5 > K_POINTS automatic > 1 1 50 0 0 1 > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/07983797/attachment.htm From mansourehp at gmail.com Mon Sep 8 11:03:36 2008 From: mansourehp at gmail.com (Mansoureh Pashangpour) Date: Mon, 8 Sep 2008 12:33:36 +0330 Subject: [Pw_forum] vc-cp Message-ID: Dear all I used CP molecular dynamics for a metalic system.Now I want to use external pressure on this system.I have used vc-cp instead of cp and &cell cell_dynamics='pr', press=10, but I think its result isn't acceptable.cell temprature increased fast. I must say that I didn't use any thermostats in my simulation. how can I add them in &cell , &ions and &electron parts? Thanks in advance for your help and attention good luck Mansoureh IPM,Tehran,Iran -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/dc743f06/attachment.htm From giannozz at democritos.it Mon Sep 8 11:24:16 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 08 Sep 2008 11:24:16 +0200 Subject: [Pw_forum] relax-spinorbit In-Reply-To: References: <844011.18596.qm@web33101.mail.mud.yahoo.com> Message-ID: <48C4EF40.6030106@democritos.it> Mansoureh Pashangpour wrote: > use > pot_extrapolation='second_order', > wfc_extrapolation='second_order', first- and second-extrapolations should be used for MD calculations only. There one has a sequence of wavefunctions depending upon a single variable (time) and can perform extrapolation wrt that single variable. In a structural optimization run using BFGS, one performs a line minimization along a given directions, finds a new minimization direction and so on. I don't think first- or second-extrapolation will give any advantage over the default choice Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From zhaohscas at yahoo.com.cn Mon Sep 8 11:28:15 2008 From: zhaohscas at yahoo.com.cn (Hongsheng Zhao) Date: Mon, 8 Sep 2008 17:28:15 +0800 Subject: [Pw_forum] About the atomic radii used for visuliztion of a structre. Message-ID: <1039503785.20080908172815@yahoo.com.cn> Dear all, This issue is not so pertinent to the usage of PWSCF itself, but to the visuliztion of a structure. If my question bring inconvenience to this list, I really feel sorry for this! My issue is as follows: What's relationship between the atomic radii used for visuliztion of a structre and the atomic real radii of the atoms/ions? This issue arise from the fact that I find many crystal visuliztion tools, i.e., diamond, rasmol, and etc., don't treat the atomic radii according to the real radii of the atoms/ions belong to the structure, never so much as proportionate to the counterpart real radii. I just cann't understand this thing. In my mind, the atoms/ions should be represent proportional to its real magnitude. Who can give me some hints? Sincerely yours, -- Hongsheng Zhao Xinjiang Technical Institute of Physics and Chemistry Chinese Academy of Sciences GnuPG DSA: 0xD108493 2008-9-8 From vegalew at hotmail.com Mon Sep 8 13:39:52 2008 From: vegalew at hotmail.com (vega) Date: Mon, 8 Sep 2008 19:39:52 +0800 Subject: [Pw_forum] How to plot the 2D-charge density for a triclinic or monoclinic box Message-ID: Dear all, I need to plot 2D charge density for a triclinic or monoclinic box. while the section plane along the c-axis is not a rectangle but a parallelogram. so the vector of the section plane e1 and e2 is not orthogonal. how could I specify the e1(1)~e1(3) and e2(1)~e(3) in this case. thank you for reading vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/a5f2a79f/attachment.htm From kada_yesudas at yahoo.co.in Mon Sep 8 14:31:09 2008 From: kada_yesudas at yahoo.co.in (kada yesudas) Date: Mon, 8 Sep 2008 18:01:09 +0530 (IST) Subject: [Pw_forum] parallel processing Message-ID: <763116.96452.qm@web8407.mail.in.yahoo.com> Hai I have a problem in running parallel pw.x my hardware is amd64 bit dual core with 8 node and master. my job is running in serial but not running in parallel. i tried several options as described in user manual. can any help me in this regard. my input is as follows. &CONTROL calculation = "relax", pseudo_dir = "/opt/Apps/PWscf-4.0.1/pseudo", title = "TiO2-supercell at 500ev", verbosity = "high", outdir = "/tmp", wfcdir = "/tmp", nstep = 300, restart_mode = "from_scratch", disk_io = "high", etot_conv_thr = 1.0D-5, forc_conv_thr = 1.0D-4, / &SYSTEM ibrav = 14, A = 10.2395, B = 11.3526, C = 20.8588, cosAB = 0, cosAC = 0, cosBC = 0, nat = 72, ntyp = 2, ecutwfc = 25, / &ELECTRONS electron_maxstep = 500, diago_david_ndim = 6, conv_thr = 1.0D-7, / &IONS ion_dynamics = 'bfgs', / ATOMIC_SPECIES O 16.000 O.pw91-van_ak.UPF Ti 47.862 Ti.pw91-nsp-van.UPF ATOMIC_POSITIONS angstrom Ti 7.80260139 0.00000000 0.73172670 0 0 0 Ti 7.80260139 3.78420000 0.73172670 0 0 0 Ti 7.80260139 7.56840000 0.73172670 0 0 0 Ti 9.20111230 0.00000000 4.24810321 1 1 1 Ti 9.20111230 3.78420000 4.24810321 1 1 1 Ti 9.20111230 7.56840000 4.24810321 1 1 1 Ti 2.68285140 1.89210000 0.73172670 0 0 0 Ti 2.68285140 5.67630000 0.73172670 0 0 0 Ti 2.68285140 9.46050000 0.73172670 0 0 0 Ti 4.08136231 1.89210000 4.24810321 1 1 1 Ti 4.08136231 5.67630000 4.24810321 1 1 1 Ti 4.08136231 9.46050000 4.24810321 1 1 1 Ti 5.59240532 1.89210000 1.61071654 0 0 0 Ti 5.59240532 5.67630000 1.61071654 0 0 0 Ti 5.59240532 9.46050000 1.61071654 0 0 0 Ti 6.99091623 1.89210000 5.12709304 1 1 1 Ti 6.99091623 5.67630000 5.12709304 1 1 1 Ti 6.99091623 9.46050000 5.12709304 1 1 1 Ti 0.47265532 0.00000000 1.61071654 0 0 0 Ti 0.47265532 3.78420000 1.61071654 0 0 0 Ti 0.47265532 7.56840000 1.61071654 0 0 0 Ti 1.87116623 0.00000000 5.12709304 1 1 1 Ti 1.87116623 3.78420000 5.12709304 1 1 1 Ti 1.87116623 7.56840000 5.12709304 1 1 1 O 5.96276803 0.00000000 1.46345341 0 0 0 O 5.96276803 3.78420000 1.46345341 0 0 0 O 5.96276803 7.56840000 1.46345341 0 0 0 O 7.36127895 0.00000000 4.97982991 1 1 1 O 7.36127895 3.78420000 4.97982991 1 1 1 O 7.36127895 7.56840000 4.97982991 1 1 1 O 0.84301804 1.89210000 1.46345341 0 0 0 O 0.84301804 5.67630000 1.46345341 0 0 0 O 0.84301804 9.46050000 1.46345341 0 0 0 O 2.24152895 1.89210000 4.97982991 1 1 1 O 2.24152895 5.67630000 4.97982991 1 1 1 O 2.24152895 9.46050000 4.97982991 1 1 1 O 3.75257196 1.89210000 2.34265183 0 0 0 O 3.75257196 5.67630000 2.34265183 0 0 0 O 3.75257196 9.46050000 2.34265183 0 0 0 O 5.15108287 1.89210000 5.85881974 1 1 1 O 5.15108287 5.67630000 5.85881974 1 1 1 O 5.15108287 9.46050000 5.85881974 1 1 1 O 8.87232196 0.00000000 2.34265183 0 0 0 O 8.87232196 3.78420000 2.34265183 0 0 0 O 8.87232196 7.56840000 2.34265183 0 0 0 O 0.03133287 0.00000000 5.85881974 1 1 1 O 0.03133287 3.78420000 5.85881974 1 1 1 O 0.03133287 7.56840000 5.85881974 1 1 1 O 2.31238629 0.00000000 0.87898983 0 0 0 O 2.31238629 3.78420000 0.87898983 0 0 0 O 2.31238629 7.56840000 0.87898983 0 0 0 O 3.71099959 0.00000000 4.39536634 1 1 1 O 3.71099959 3.78420000 4.39536634 1 1 1 O 3.71099959 7.56840000 4.39536634 1 1 1 O 7.43213628 1.89210000 0.87898983 0 0 0 O 7.43213628 5.67630000 0.87898983 0 0 0 O 7.43213628 9.46050000 0.87898983 0 0 0 O 8.83074959 1.89210000 4.39536634 1 1 1 O 8.83074959 5.67630000 4.39536634 1 1 1 O 8.83074959 9.46050000 4.39536634 1 1 1 O 4.52268476 1.89210000 0.00000000 0 0 0 O 4.52268476 5.67630000 0.00000000 0 0 0 O 4.52268476 9.46050000 0.00000000 0 0 0 O 5.92119566 1.89210000 3.51637650 1 1 1 O 5.92119566 5.67630000 3.51637650 1 1 1 O 5.92119566 9.46050000 3.51637650 1 1 1 O 9.64243475 0.00000000 0.00000000 0 0 0 O 9.64243475 3.78420000 0.00000000 0 0 0 O 9.64243475 7.56840000 0.00000000 0 0 0 O 0.80144567 0.00000000 3.51637650 1 1 1 O 0.80144567 3.78420000 3.51637650 1 1 1 O 0.80144567 7.56840000 3.51637650 1 1 1 K_POINTS automatic 1 2 2 0 0 0 Kada Yesudas C/O: Dr. K. Bhanuprakash, Indian Institute of Chemical Technology(IICT), Habsiguda, Hyderabad-500007 Andhra Pradesh, INDIA Tel:27160123 Ext:2430 Connect with friends all over the world. Get Yahoo! India Messenger at http://in.messenger.yahoo.com/?wm=n/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/4dab77b8/attachment-0001.htm From kada_yesudas at yahoo.co.in Mon Sep 8 14:31:09 2008 From: kada_yesudas at yahoo.co.in (kada yesudas) Date: Mon, 8 Sep 2008 18:01:09 +0530 (IST) Subject: [Pw_forum] parallel processing Message-ID: <763116.96452.qm@web8407.mail.in.yahoo.com> Hai I have a problem in running parallel pw.x my hardware is amd64 bit dual core with 8 node and master. my job is running in serial but not running in parallel. i tried several options as described in user manual. can any help me in this regard. my input is as follows. &CONTROL calculation = "relax", pseudo_dir = "/opt/Apps/PWscf-4.0.1/pseudo", title = "TiO2-supercell at 500ev", verbosity = "high", outdir = "/tmp", wfcdir = "/tmp", nstep = 300, restart_mode = "from_scratch", disk_io = "high", etot_conv_thr = 1.0D-5, forc_conv_thr = 1.0D-4, / &SYSTEM ibrav = 14, A = 10.2395, B = 11.3526, C = 20.8588, cosAB = 0, cosAC = 0, cosBC = 0, nat = 72, ntyp = 2, ecutwfc = 25, / &ELECTRONS electron_maxstep = 500, diago_david_ndim = 6, conv_thr = 1.0D-7, / &IONS ion_dynamics = 'bfgs', / ATOMIC_SPECIES O 16.000 O.pw91-van_ak.UPF Ti 47.862 Ti.pw91-nsp-van.UPF ATOMIC_POSITIONS angstrom Ti 7.80260139 0.00000000 0.73172670 0 0 0 Ti 7.80260139 3.78420000 0.73172670 0 0 0 Ti 7.80260139 7.56840000 0.73172670 0 0 0 Ti 9.20111230 0.00000000 4.24810321 1 1 1 Ti 9.20111230 3.78420000 4.24810321 1 1 1 Ti 9.20111230 7.56840000 4.24810321 1 1 1 Ti 2.68285140 1.89210000 0.73172670 0 0 0 Ti 2.68285140 5.67630000 0.73172670 0 0 0 Ti 2.68285140 9.46050000 0.73172670 0 0 0 Ti 4.08136231 1.89210000 4.24810321 1 1 1 Ti 4.08136231 5.67630000 4.24810321 1 1 1 Ti 4.08136231 9.46050000 4.24810321 1 1 1 Ti 5.59240532 1.89210000 1.61071654 0 0 0 Ti 5.59240532 5.67630000 1.61071654 0 0 0 Ti 5.59240532 9.46050000 1.61071654 0 0 0 Ti 6.99091623 1.89210000 5.12709304 1 1 1 Ti 6.99091623 5.67630000 5.12709304 1 1 1 Ti 6.99091623 9.46050000 5.12709304 1 1 1 Ti 0.47265532 0.00000000 1.61071654 0 0 0 Ti 0.47265532 3.78420000 1.61071654 0 0 0 Ti 0.47265532 7.56840000 1.61071654 0 0 0 Ti 1.87116623 0.00000000 5.12709304 1 1 1 Ti 1.87116623 3.78420000 5.12709304 1 1 1 Ti 1.87116623 7.56840000 5.12709304 1 1 1 O 5.96276803 0.00000000 1.46345341 0 0 0 O 5.96276803 3.78420000 1.46345341 0 0 0 O 5.96276803 7.56840000 1.46345341 0 0 0 O 7.36127895 0.00000000 4.97982991 1 1 1 O 7.36127895 3.78420000 4.97982991 1 1 1 O 7.36127895 7.56840000 4.97982991 1 1 1 O 0.84301804 1.89210000 1.46345341 0 0 0 O 0.84301804 5.67630000 1.46345341 0 0 0 O 0.84301804 9.46050000 1.46345341 0 0 0 O 2.24152895 1.89210000 4.97982991 1 1 1 O 2.24152895 5.67630000 4.97982991 1 1 1 O 2.24152895 9.46050000 4.97982991 1 1 1 O 3.75257196 1.89210000 2.34265183 0 0 0 O 3.75257196 5.67630000 2.34265183 0 0 0 O 3.75257196 9.46050000 2.34265183 0 0 0 O 5.15108287 1.89210000 5.85881974 1 1 1 O 5.15108287 5.67630000 5.85881974 1 1 1 O 5.15108287 9.46050000 5.85881974 1 1 1 O 8.87232196 0.00000000 2.34265183 0 0 0 O 8.87232196 3.78420000 2.34265183 0 0 0 O 8.87232196 7.56840000 2.34265183 0 0 0 O 0.03133287 0.00000000 5.85881974 1 1 1 O 0.03133287 3.78420000 5.85881974 1 1 1 O 0.03133287 7.56840000 5.85881974 1 1 1 O 2.31238629 0.00000000 0.87898983 0 0 0 O 2.31238629 3.78420000 0.87898983 0 0 0 O 2.31238629 7.56840000 0.87898983 0 0 0 O 3.71099959 0.00000000 4.39536634 1 1 1 O 3.71099959 3.78420000 4.39536634 1 1 1 O 3.71099959 7.56840000 4.39536634 1 1 1 O 7.43213628 1.89210000 0.87898983 0 0 0 O 7.43213628 5.67630000 0.87898983 0 0 0 O 7.43213628 9.46050000 0.87898983 0 0 0 O 8.83074959 1.89210000 4.39536634 1 1 1 O 8.83074959 5.67630000 4.39536634 1 1 1 O 8.83074959 9.46050000 4.39536634 1 1 1 O 4.52268476 1.89210000 0.00000000 0 0 0 O 4.52268476 5.67630000 0.00000000 0 0 0 O 4.52268476 9.46050000 0.00000000 0 0 0 O 5.92119566 1.89210000 3.51637650 1 1 1 O 5.92119566 5.67630000 3.51637650 1 1 1 O 5.92119566 9.46050000 3.51637650 1 1 1 O 9.64243475 0.00000000 0.00000000 0 0 0 O 9.64243475 3.78420000 0.00000000 0 0 0 O 9.64243475 7.56840000 0.00000000 0 0 0 O 0.80144567 0.00000000 3.51637650 1 1 1 O 0.80144567 3.78420000 3.51637650 1 1 1 O 0.80144567 7.56840000 3.51637650 1 1 1 K_POINTS automatic 1 2 2 0 0 0 Kada Yesudas C/O: Dr. K. Bhanuprakash, Indian Institute of Chemical Technology(IICT), Habsiguda, Hyderabad-500007 Andhra Pradesh, INDIA Tel:27160123 Ext:2430 Connect with friends all over the world. Get Yahoo! India Messenger at http://in.messenger.yahoo.com/?wm=n/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/4dab77b8/attachment-0002.htm From kada_yesudas at yahoo.co.in Mon Sep 8 14:41:20 2008 From: kada_yesudas at yahoo.co.in (kada yesudas) Date: Mon, 8 Sep 2008 18:11:20 +0530 (IST) Subject: [Pw_forum] parallel processing Message-ID: <678358.76149.qm@web8408.mail.in.yahoo.com> Hai I have a problem in running parallel pw.x my hardware is amd64 bit dual core with 8 node and master. my job is running in serial but not running in parallel. i tried several options as described in user manual. can any help me in this regard. my input is as follows. &CONTROL calculation = "relax", pseudo_dir = "/opt/Apps/PWscf-4.0.1/pseudo", title = "TiO2-supercell at 500ev", verbosity = "high", outdir = "/tmp", wfcdir = "/tmp", nstep = 300, restart_mode = "from_scratch", disk_io = "high", etot_conv_thr = 1.0D-5, forc_conv_thr = 1.0D-4, / &SYSTEM ibrav = 14, A = 10.2395, B = 11.3526, C = 20.8588, cosAB = 0, cosAC = 0, cosBC = 0, nat = 72, ntyp = 2, ecutwfc = 25, / &ELECTRONS electron_maxstep = 500, diago_david_ndim = 6, conv_thr = 1.0D-7, / &IONS ion_dynamics = 'bfgs', / ATOMIC_SPECIES O 16.000 O.pw91-van_ak.UPF Ti 47.862 Ti.pw91-nsp-van.UPF ATOMIC_POSITIONS angstrom Ti 7.80260139 0.00000000 0.73172670 0 0 0 Ti 7.80260139 3.78420000 0.73172670 0 0 0 Ti 7.80260139 7.56840000 0.73172670 0 0 0 Ti 9.20111230 0.00000000 4.24810321 1 1 1 Ti 9.20111230 3.78420000 4.24810321 1 1 1 Ti 9.20111230 7.56840000 4.24810321 1 1 1 Ti 2.68285140 1.89210000 0.73172670 0 0 0 Ti 2.68285140 5.67630000 0.73172670 0 0 0 Ti 2.68285140 9.46050000 0.73172670 0 0 0 Ti 4.08136231 1.89210000 4.24810321 1 1 1 Ti 4.08136231 5.67630000 4.24810321 1 1 1 Ti 4.08136231 9.46050000 4.24810321 1 1 1 Ti 5.59240532 1.89210000 1.61071654 0 0 0 Ti 5.59240532 5.67630000 1.61071654 0 0 0 Ti 5.59240532 9.46050000 1.61071654 0 0 0 Ti 6.99091623 1.89210000 5.12709304 1 1 1 Ti 6.99091623 5.67630000 5.12709304 1 1 1 Ti 6.99091623 9.46050000 5.12709304 1 1 1 Ti 0.47265532 0.00000000 1.61071654 0 0 0 Ti 0.47265532 3.78420000 1.61071654 0 0 0 Ti 0.47265532 7.56840000 1.61071654 0 0 0 Ti 1.87116623 0.00000000 5.12709304 1 1 1 Ti 1.87116623 3.78420000 5.12709304 1 1 1 Ti 1.87116623 7.56840000 5.12709304 1 1 1 O 5.96276803 0.00000000 1.46345341 0 0 0 O 5.96276803 3.78420000 1.46345341 0 0 0 O 5.96276803 7.56840000 1.46345341 0 0 0 O 7.36127895 0.00000000 4.97982991 1 1 1 O 7.36127895 3.78420000 4.97982991 1 1 1 O 7.36127895 7.56840000 4.97982991 1 1 1 O 0.84301804 1.89210000 1.46345341 0 0 0 O 0.84301804 5.67630000 1.46345341 0 0 0 O 0.84301804 9.46050000 1.46345341 0 0 0 O 2.24152895 1.89210000 4.97982991 1 1 1 O 2.24152895 5.67630000 4.97982991 1 1 1 O 2.24152895 9.46050000 4.97982991 1 1 1 O 3.75257196 1.89210000 2.34265183 0 0 0 O 3.75257196 5.67630000 2.34265183 0 0 0 O 3.75257196 9.46050000 2.34265183 0 0 0 O 5.15108287 1.89210000 5.85881974 1 1 1 O 5.15108287 5.67630000 5.85881974 1 1 1 O 5.15108287 9.46050000 5.85881974 1 1 1 O 8.87232196 0.00000000 2.34265183 0 0 0 O 8.87232196 3.78420000 2.34265183 0 0 0 O 8.87232196 7.56840000 2.34265183 0 0 0 O 0.03133287 0.00000000 5.85881974 1 1 1 O 0.03133287 3.78420000 5.85881974 1 1 1 O 0.03133287 7.56840000 5.85881974 1 1 1 O 2.31238629 0.00000000 0.87898983 0 0 0 O 2.31238629 3.78420000 0.87898983 0 0 0 O 2.31238629 7.56840000 0.87898983 0 0 0 O 3.71099959 0.00000000 4.39536634 1 1 1 O 3.71099959 3.78420000 4.39536634 1 1 1 O 3.71099959 7.56840000 4.39536634 1 1 1 O 7.43213628 1.89210000 0.87898983 0 0 0 O 7.43213628 5.67630000 0.87898983 0 0 0 O 7.43213628 9.46050000 0.87898983 0 0 0 O 8.83074959 1.89210000 4.39536634 1 1 1 O 8.83074959 5.67630000 4.39536634 1 1 1 O 8.83074959 9.46050000 4.39536634 1 1 1 O 4.52268476 1.89210000 0.00000000 0 0 0 O 4.52268476 5.67630000 0.00000000 0 0 0 O 4.52268476 9.46050000 0.00000000 0 0 0 O 5.92119566 1.89210000 3.51637650 1 1 1 O 5.92119566 5.67630000 3.51637650 1 1 1 O 5.92119566 9.46050000 3.51637650 1 1 1 O 9.64243475 0.00000000 0.00000000 0 0 0 O 9.64243475 3.78420000 0.00000000 0 0 0 O 9.64243475 7.56840000 0.00000000 0 0 0 O 0.80144567 0.00000000 3.51637650 1 1 1 O 0.80144567 3.78420000 3.51637650 1 1 1 O 0.80144567 7.56840000 3.51637650 1 1 1 K_POINTS automatic 1 2 2 0 0 0 Kada Yesudas C/O: Dr. K. Bhanuprakash, Indian Institute of Chemical Technology(IICT), Habsiguda, Hyderabad-500007 Andhra Pradesh, INDIA Tel:27160123 Ext:2430 Unlimited freedom, unlimited storage. Get it now, on http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/7be1618f/attachment-0001.htm From paulatto at sissa.it Mon Sep 8 14:41:56 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Mon, 08 Sep 2008 14:41:56 +0200 Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: <32831.210.73.17.131.1220838059.squirrel@210.73.16.4> References: <32820.210.73.17.131.1220703967.squirrel@210.73.16.4> <32831.210.73.17.131.1220838059.squirrel@210.73.16.4> Message-ID: <48C51D94.5030107@sissa.it> L.F.Huang ha scritto: > Prof. Kohlmeyer, you have mentioned that not all components in > Q-E support k-points parallelization fully. I wonder are there > still any other components like that besides postprocessing components? > Are elphon.f90 and elph.f90? > Thanks! > Dear Huang, I don't know if el-ph coupling supports k-points parallelizations; but if you do your scf calculation using the wf_collect=.true. option you can (usually) post-process the data with a different parallel configuration. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From giannozz at democritos.it Mon Sep 8 14:46:20 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 08 Sep 2008 14:46:20 +0200 Subject: [Pw_forum] parallel processing In-Reply-To: <763116.96452.qm@web8407.mail.in.yahoo.com> References: <763116.96452.qm@web8407.mail.in.yahoo.com> Message-ID: <48C51E9C.3090309@democritos.it> kada yesudas wrote: > my hardware is amd64 bit dual core with 8 node and master. > my job is running in serial but not running in parallel. unlike your mailer that seems to run in parallel, since we received three copies of your message ... Seriously: there is a long section in the user guide about how what is needed to compile for parallel execution and how to run in parallel. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From kondor.jess at gmail.com Mon Sep 8 14:52:05 2008 From: kondor.jess at gmail.com (Jess Kondor) Date: Mon, 8 Sep 2008 07:52:05 -0500 Subject: [Pw_forum] relax-spinorbit In-Reply-To: <844011.18596.qm@web33101.mail.mud.yahoo.com> References: <844011.18596.qm@web33101.mail.mud.yahoo.com> Message-ID: <1d9d5d9d0809080552p16bf06cel93c3a54f04a84ff@mail.gmail.com> On Mon, Sep 8, 2008 at 3:11 AM, ali kazempoor wrote: > &system > ecutwfc = 40 , ecutrho =400, ... > ecutwfc =40 > ecutrho =500.0 > Why do you use twice (and different) ecutwfc/ecutrho values? jess > / > &electrons > conv_thr = 1.0d-4 > mixing_beta = 0.8 > / > &IONS > ion_dynamics='bfgs' > / > ATOMIC_SPECIES > Ni 28.0 NirelPBE.RRKJ3.UPF > ATOMIC_POSITIONS crystal > Ni 0.5 0.5 0.0 > Ni 0.3333333 0.5 0.5 > Ni 0.5 0.66666666 0.5 > Ni 0.6666666 0.5 0.5 > Ni 0.5 0.33333333 0.5 > K_POINTS automatic > 1 1 50 0 0 1 > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- ===================================== Jess Kondor, PhD PICYT, ===================================== From fratesi at mater.unimib.it Mon Sep 8 15:18:17 2008 From: fratesi at mater.unimib.it (Guido Fratesi) Date: Mon, 8 Sep 2008 15:18:17 +0200 (CEST) Subject: [Pw_forum] STS spectra In-Reply-To: <4873AAB2.11F0.001D.0@empa.ch> References: <4873AAB2.11F0.001D.0@empa.ch> Message-ID: Dear Daniele, I'm sorry to reply only now to your message which dates two months ago, and I guess you could have probably found a convenient solution to get the LDOS by now. Sometime ago I was also interested in computing the LDOS at fixed coordinate, r, and at variance with the energy, E. I added a few bits to the code (3.1.1 at the time) which might be useful to you as well, but this never come to a stage sufficiently refined to be distributed and eventually got obsolete. I attach here a few lines of the algorithm which I was adding into projwfc.f90, as you might be curious about that. Only the simplest, norm-conserving, collinear case was implemented. The basic idea is the following (possibly not the most efficient one), and closely follows the one for computing the projected DOS: * given regions where to integrate the LDOS * define "theta" arrays (dimension nrxx) which evaluate 1 inside and 0 outside these regions --> thetathisproc(1:nrxx,ibox) * for each band, each kpoint --> proj(ibox,ibnd,ik) - get the |wfc^2| (array of dimension nrxx) - multiply times the "theta" array and sum over space (this step is analogous to performing the projection over the atomic orbitals) * sum over kpoints and bands --> ldosbox(ie,ibox,current_spin) I would be interested to know if you find this way of proceeding convenient to your case (and in case I can send a sort-of-working example) or if you have spotted a better way to proceed, in the while. Regards, Guido ---------------------- compunting proj k_loop: DO ik = 1, nks ! IF ( lsda ) current_spin = isk(ik) CALL gk_sort (xk (1, ik), ngm, g, ecutwfc / tpiba2, npw, igk, g2kin) CALL davcio (evc, nwordwfc, iunwfc, ik, - 1) ! bnd_loop: DO ibnd = 1, nbnd ! IF (noncolin) THEN ! CALL errore( 'projwfc', 'boxes and noncolin not implemented', 1 ) ! ELSE ! caux(1:nrxx) = (0._DP,0._DP) DO ig = 1, npw caux (nl (igk (ig) ) ) = evc (ig, ibnd) ENDDO IF (gamma_only) THEN DO ig = 1, npw caux (nlm(igk (ig) ) ) = CONJG(evc (ig, ibnd)) ENDDO END IF CALL cft3 (caux, nr1, nr2, nr3, nrx1, nrx2, nrx3, 1) ! raux(:) = DBLE( caux(:) )**2 + AIMAG( caux(:) )**2 ! END IF ! ! The contribution of this wavefunction to the LDOS ! integrated in the volume is the projection of the ! squared wfc on a function =1 in the volume itself: ! DO ibox = iboxmin, n_proj_boxes proj(ibox,ibnd,ik) = DDOT(nrxx,raux(:),1,thetathisproc(:,ibox),1) & & / (nrx1*nrx2*nrx3) END DO ! END DO bnd_loop ! CALL reduce ((n_proj_boxes-iboxmin+1)*nbnd , proj(iboxmin,1,ik)) ! END DO k_loop ---------------------- summing over kp and bands DO ik = 1,nkstot IF ( nspin == 2 ) current_spin = isk ( ik ) DO ibnd = 1, nbnd etev = et(ibnd,ik) ie_mid = NINT( (etev-Emin)/DeltaE ) DO ie = MAX(ie_mid-ie_delta, 0), MIN(ie_mid+ie_delta, ne) delta = w0gauss((Emin+DeltaE*ie-etev)/degauss,ngauss) & / degauss / rytoev ! DO ibox = iboxmin, n_proj_boxes ldosbox(ie,ibox,current_spin) = & ldosbox(ie,ibox,current_spin) + & wk(ik) * delta * proj (ibox, ibnd, ik) END DO ! ! dostot(:,ns) = total DOS (states/eV) for spin "ns" ! dostot(ie,current_spin) = dostot(ie,current_spin) + & wk(ik) * delta END DO END DO END DO ---------------------- On Tue, 8 Jul 2008, Daniele Passerone wrote: | Dear Pwscf users and developers, The postprocessing facilities of espresso allow to extract stm images, that correspond in Tersoff-Hamann approximation to the localized density of states integrated between Efermi and a given (positive or negative) bias. The more convenient way to analyze this quantity is in my opinion by generating a 3d grid and visualizing it with xcrysden, which incidentally can also show planes parallel to the cell sides. The ILDOS option of pp.x (plot_num=10) somewhat extends this, allowing to compute the LDOS integrated between emin and emax. STS (scanning tunneling spectroscopy) experiments sample the local density of states at a surface: one obtains a profile of dI/dV at a certain point on the surface, and this gives valuable information about the local conductivity and presence of peaks in the DOS etc. The question is: what is the more convenient and fast way to get the LDOS profile from Emin to Emax, on a certain plane (let us say, 2 Angstrom above the surface)? I thought one could make several jobs of pp.x : from Emin to Emin+DeltaE, from Emin + Delta E to Emin + 2*Delta E.... and so on... but maybe there are much better ways to do that. Thank you in advance for any advice, Daniele -- Guido Fratesi Dipartimento di Scienza dei Materiali Universita` degli Studi di Milano-Bicocca From ihsanas at yahoo.com Mon Sep 8 16:07:39 2008 From: ihsanas at yahoo.com (=?iso-8859-1?Q?=C7=CD=D3=C7=E4_=DA=D1=ED=DE=C7=CA?=) Date: Mon, 8 Sep 2008 07:07:39 -0700 (PDT) Subject: [Pw_forum] smearing Message-ID: <268611.37400.qm@web52305.mail.re2.yahoo.com> Dear pwscf user How can one be shure that he is using the best smearing and best degauss for certain smearing .could convergance test for k-point and different degauss be the best way . best regards Ihsan physics department -Jordan university From ski2 at mail.uh.edu Mon Sep 8 16:26:27 2008 From: ski2 at mail.uh.edu (ski2 at mail.uh.edu) Date: Mon, 08 Sep 2008 09:26:27 -0500 Subject: [Pw_forum] nbnd and semicore in PP Message-ID: Dear All, I am wondering if I use PP containing semicore states in valence state, do I need to count the bands of the semicore states and add into "nbnd"? Best Wishes, Sang-Hwan Kim (Postdoc) University of Houston Department of Chemistry University of Houston 136 Fleming Building Houston, TX 77204-5003 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/641f1c19/attachment.htm From paulatto at sissa.it Mon Sep 8 16:28:19 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Mon, 08 Sep 2008 16:28:19 +0200 Subject: [Pw_forum] nbnd and semicore in PP In-Reply-To: References: Message-ID: <48C53683.5050506@sissa.it> ski2 at mail.uh.edu ha scritto: > Dear All, > > I am wondering if I use PP containing semicore states in valence > state, do I need to count the bands of the semicore states and add > into "nbnd"? Yes, you have to. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From giannozz at democritos.it Mon Sep 8 16:29:43 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 08 Sep 2008 16:29:43 +0200 Subject: [Pw_forum] nbnd and semicore in PP In-Reply-To: <48C53683.5050506@sissa.it> References: <48C53683.5050506@sissa.it> Message-ID: <48C536D7.1030907@democritos.it> Lorenzo Paulatto wrote: > ski2 at mail.uh.edu ha scritto: >> Dear All, >> >> I am wondering if I use PP containing semicore states in valence >> state, do I need to count the bands of the semicore states and add >> into "nbnd"? > Yes, you have to. unless you are happy with the default value calculated by the code Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From eariel99 at gmail.com Mon Sep 8 18:06:55 2008 From: eariel99 at gmail.com (Eduardo Ariel Menendez Proupin) Date: Mon, 8 Sep 2008 12:06:55 -0400 Subject: [Pw_forum] running in paralell Message-ID: Hi, I have problem running pw using OpenMPI. It runs terribly slow if run in different nodes (2 in this case). I think that the problem is in the OpenMP parallelism included in the Intel Compiler and the MKL libraries. The configure program sets the flag -openmp. If delete the flag, get unresolved libraries. To run using MPI, the MKL need the shell variable OMP_NUMTHREADS=1 and MKL_NUM_THREADS=1. With this, MPI works fine in one node. The problem is that running in two nodes, the shell variables are not passed to the remote machine, and I have not found how to do it, but there shluld be an easy way, or maybe in the configuration of MPI. Thanks -- Eduardo Menendez University of Chile -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/a6de4011/attachment.htm From paulatto at sissa.it Mon Sep 8 18:27:10 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Mon, 08 Sep 2008 18:27:10 +0200 Subject: [Pw_forum] running in paralell In-Reply-To: References: Message-ID: <48C5525E.2070307@sissa.it> Eduardo Ariel Menendez Proupin ha scritto: > With this, MPI works fine in one node. The problem is that running in > two nodes, the shell variables are not passed to the remote machine, > and I have not found how to do it, but there shluld be an easy way, or > maybe in the configuration of MPI. Dear Eduardo, yours is a simple openmpi question, it has nothing to do with QE. Anyway the solution is to use the -x option of mpirun (check the manual), or to set the variable in your ~/.bashrc file. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From akohlmey at cmm.chem.upenn.edu Mon Sep 8 19:18:01 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 8 Sep 2008 13:18:01 -0400 (EDT) Subject: [Pw_forum] running in paralell In-Reply-To: References: Message-ID: On Mon, 8 Sep 2008, Eduardo Ariel Menendez Proupin wrote: EAMP> Hi, EAMP> I have problem running pw using OpenMPI. It runs terribly slow if run in EAMP> different nodes (2 in this case). I think that the problem is in the OpenMP EAMP> parallelism included in the Intel Compiler and the MKL libraries. EAMP> The configure program sets the flag -openmp. If delete the flag, get EAMP> unresolved libraries. To run using MPI, the MKL need the shell variable EAMP> OMP_NUMTHREADS=1 and MKL_NUM_THREADS=1. apart from setting environment variables, you can also link with the serial version of mkl. in this case neither -openmp nor -lguide and -lpthread are needed. ...and while at it, you can also hardcode the search path to MKL into the binary. the BLAS_LIBS entry in make.sys becomes then: BLAS_LIBS = -L/opt/intel/mkl/10.0.1.014/lib/em64t \ -Wl,-rpath,/opt/intel/mkl/10.0.1.014/lib/em64t \ -lmkl_intel_lp64 -lmkl_sequential -lmkl_core and LAPACK_LIBS = can be left empty. the usual -lmkl is just a wrapper to link -lmkl_intel_lp64 -lmkl_thread_intel -lmkl_core and -lmkl_em64t will link the equivalent static libraries. if you want to make your life (and that of other users) _really_ easy, you can edit both libmkl.so and libmkl_em64t.a in a text editor and just change libmkl_intel_thread.so/.a to libmkl_sequential.so/.a and you won't have to worry about this issue again (until you install a new version of MKL). cheers, axel. EAMP> With this, MPI works fine in one node. The problem is that running in two EAMP> nodes, the shell variables are not passed to the remote machine, and I have EAMP> not found how to do it, but there shluld be an easy way, or maybe in the EAMP> configuration of MPI. EAMP> Thanks EAMP> -- ======================================================================= 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 kazempoor2000 at yahoo.com Mon Sep 8 20:08:52 2008 From: kazempoor2000 at yahoo.com (ali kazempoor) Date: Mon, 8 Sep 2008 11:08:52 -0700 (PDT) Subject: [Pw_forum] nancollin Message-ID: <78556.43268.qm@web33105.mail.mud.yahoo.com> Dear all For calculating MAE energy I run relax calculation according to below with Nirelpbe.rrkg3.UPF pseudopotential but one error appears like ?task #???????? 6 ???? from stres_gradcorr : error #???????? 1 ???? noncollinear stress + GGA not implemented which parameter mast be changed? thanks Ali Kazempour,Isfahan univ. Of Tech, Iran -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/cf4cea32/attachment.htm From ezadshojaee at hotmail.com Mon Sep 8 21:08:12 2008 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Mon, 8 Sep 2008 19:08:12 +0000 Subject: [Pw_forum] about matching grid's Message-ID: Hi i have a technical question, maybe so trivial. Imagine our unit cell consists of two different atoms which their pseudo file are constructed in different # of points as a radial mesh(coarser near their nucleus ..) Is matching the grid points for the scf calculation in the cell a problem? How does the code do this? I mean how does the code choose the grid for scf and what is it's relation to the atomic grids? Thanx _________________________________________________________________ Invite your mail contacts to join your friends list with Windows Live Spaces. It's easy! http://spaces.live.com/spacesapi.aspx?wx_action=create&wx_url=/friends.aspx&mkt=en-us -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080908/bc57803b/attachment.htm From akohlmey at cmm.chem.upenn.edu Mon Sep 8 21:04:56 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 8 Sep 2008 15:04:56 -0400 (EDT) Subject: [Pw_forum] nancollin In-Reply-To: <78556.43268.qm@web33105.mail.mud.yahoo.com> References: <78556.43268.qm@web33105.mail.mud.yahoo.com> Message-ID: On Mon, 8 Sep 2008, ali kazempoor wrote: AK> Dear all AK> AK> For calculating MAE energy I run relax calculation according to below with Nirelpbe.rrkg3.UPF AK> AK> pseudopotential but one error appears like AK> AK> ?task #???????? 6 AK> AK> ???? from stres_gradcorr : error #???????? 1 AK> AK> ???? noncollinear stress + GGA not implemented AK> AK> which parameter mast be changed? ali, please read what the code is telling you. Q-E may have cryptic error messages, but this is not one of them. this cannot be fixed by changing parameters (except for trivial ones, i.e. not using stress or GGA functionals), but by programming. the stress tensor for non-collinear with gradient corrected functionals is not supported yet. to make your calculation work, you would have to implement it. cheers, axel. AK> AK> thanks AK> AK> Ali Kazempour,Isfahan univ. Of Tech, Iran AK> AK> -- ======================================================================= 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 Mon Sep 8 21:11:04 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 8 Sep 2008 15:11:04 -0400 (EDT) Subject: [Pw_forum] about matching grid's In-Reply-To: References: Message-ID: On Mon, 8 Sep 2008, Ezad Shojaee wrote: ES> ES> Hi ES> i have a technical question, maybe so trivial. ES> Imagine our unit cell consists of two different atoms which their ES> pseudo file are constructed in different # of points as a radial there is no relation between the grid for the pseudopotentials and the selection of the grids for density and wavefunction. ES> mesh(coarser near their nucleus ..) Is matching the grid points for ES> the scf calculation in the cell a problem? How does the code do ES> this? I mean how does the code choose the grid for scf and what is ES> it's relation to the atomic grids? the grid for wavefunction is chosen based on the wavefunction cutoff, box size and the fourier transform library. for computational efficiency, you want the smallest possible number of grid points that allows to represent the highest frequency plane wave determined through the wavefunction cutoff. this grid is compared to a list of grids supported grids by FFT library and the smallest grid that has equal or larger number of grid points is chosen. similar for the density cutoff (if not predetermined by ecutwfc*4, which is always supported...) cheers, axel. ES> Thanx ES> _________________________________________________________________ ES> Invite your mail contacts to join your friends list with Windows Live Spaces. It's easy! ES> http://spaces.live.com/spacesapi.aspx?wx_action=create&wx_url=/friends.aspx&mkt=en-us -- ======================================================================= 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 lfhuang at theory.issp.ac.cn Tue Sep 9 03:44:10 2008 From: lfhuang at theory.issp.ac.cn (L.F.Huang) Date: Tue, 9 Sep 2008 09:44:10 +0800 (CST) Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: References: Message-ID: <32795.210.73.17.131.1220924650.squirrel@210.73.16.4> Dear Paulatto: I am sorry that I could not get exactly what you mean when you mentioned "if you do your scf calculation using the wf_collect=.true. option you can (usually) post-process the data with a different parallel configuration". Does "wf_collect=.true." is related with the calculation of EPC ? I have read throug some documents for inputs, but can not find that relation. Could you please teach how to use "wf_collect=.true." in the calculations of el-ph interaction? Or could please you tell me some examples? Additionally, the calculation of el-ph interaction in my case finally complete in this morning, 5 days elapsed. Thank you very much Paulatto! Also Thank you for your lecture given in Shanghai China this July! Best Wishes! Yours Sincerely L.F.Huang > Date: Mon, 08 Sep 2008 14:41:56 +0200 > From: Lorenzo Paulatto > Subject: Re: [Pw_forum] time for calculation of el-ph interactions of > graphene supercell > To: PWSCF Forum > Message-ID: <48C51D94.5030107 at sissa.it> > Content-Type: text/plain; charset=ISO-8859-15; format=flowed > > Dear Huang, > I don't know if el-ph coupling supports k-points parallelizations; but > if you do your scf calculation using the wf_collect=.true. option you > can (usually) post-process the data with a different parallel > configuration. > > regards > > -- > Lorenzo Paulatto > SISSA & DEMOCRITOS (Trieste) > phone: +39 040 3787 511 > skype: paulatz > www: http://people.sissa.it/~paulatto/ From lanhaiping at gmail.com Tue Sep 9 05:56:41 2008 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 9 Sep 2008 11:56:41 +0800 Subject: [Pw_forum] Unit of potential grids ? Message-ID: Dear All: I have a confusion about the unit of potential grids. In my test calculations, i applied a external electric field by a dipole layer at vacuum slab. But i found the output result seems not very consistent with my settings. For example, I add a 0.1 V/Ang field along Z-direction,while the result that pp.x gives to me is 0.2 V/Ang if I assume that the unit of potential grids is just atomic unit . Would you please help me clarify this point ? Best , Hai-Ping -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080909/0558a1fe/attachment.htm From mansourehp at gmail.com Tue Sep 9 06:03:06 2008 From: mansourehp at gmail.com (Mansoureh Pashangpour) Date: Tue, 9 Sep 2008 07:33:06 +0330 Subject: [Pw_forum] molecular dynamics with pressure Message-ID: Dear all I am performing a molecular dynamics with external pressure .I used calculation=vc-cp and I added &cell cell_dynamics='pr' press=200, now I've got two question: 1. econt is constant during this simulation or not? 2.how can I control the temperature of the system?how can I find correct fnosep? &electrons emass = 500.d0, emass_cutoff = 2.5d0, electron_dynamics = 'verlet', electron_velocities='zero', / &ions ion_dynamics = 'verlet', ion_velocities='zero', ion_temperature='nose', tempew=300, fnosep=9, / &cell cell_dynamics = 'none', press = 200, / THANKS ALOT Mansoureh -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080909/6f2dc659/attachment.htm From lanhaiping at gmail.com Tue Sep 9 06:06:40 2008 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 9 Sep 2008 12:06:40 +0800 Subject: [Pw_forum] Unit of potential grids ? In-Reply-To: References: Message-ID: I am really sorrry for nochecking archieve. I confirm it, the unit of potential grids is Ryd . Sorry ? On Tue, Sep 9, 2008 at 11:56 AM, lan haiping wrote: > Dear All: > > I have a confusion about the unit of potential grids. > In my test calculations, i applied a external electric field by a dipole > layer at vacuum slab. But i found the output result seems not very > consistent with my settings. For example, I add a 0.1 V/Ang field along > Z-direction,while the result that pp.x gives to me is 0.2 V/Ang if I assume > that the unit of potential grids is just atomic unit . > > Would you please help me clarify this point ? > > Best , > > Hai-Ping > > > -- > Hai-Ping Lan > Department of Electronics , > Peking University , Bejing, 100871 > lanhaiping at gmail.com, hplan at pku.edu.cn > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080909/d67ec5d0/attachment.htm From paulatto at sissa.it Tue Sep 9 11:06:41 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Tue, 09 Sep 2008 11:06:41 +0200 Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: <32795.210.73.17.131.1220924650.squirrel@210.73.16.4> References: <32795.210.73.17.131.1220924650.squirrel@210.73.16.4> Message-ID: <48C63CA1.5090403@sissa.it> L.F.Huang ha scritto: > Does "wf_collect=.true." is related with the calculation of EPC ? I have > read throug some documents for inputs, but can not find that relation. > Could you please teach how to use "wf_collect=.true." in the calculations > of el-ph interaction? Or could please you tell me some examples? The option only affect parallel runs, it instructs the code to recollect the wavefunction files from all the processor pools in a single directory, as if the calculation was run with only one pool. It has no specific relationship with el-ph calculations. I don't know if el-ph has been implemented with pools parallelizations, but if it hasn't you can still use pools in the scf calculation using the wf_collect option. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From zucco at dipteris.unige.it Tue Sep 9 11:53:51 2008 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Tue, 9 Sep 2008 11:53:51 +0200 Subject: [Pw_forum] Characters table for R-3c Message-ID: <21C411C2-DA30-474E-A9D3-D972226101E7@dipteris.unige.it> Hello to users, I'm working on a carbonate with R-3c (hexagonal axes) structure and the char table is reported as below in the output file: [...] point group S_6 (-3) there are 6 classes the character table: E C3 C3^2 i S6^5 S6 A_g 1.00 1.00 1.00 1.00 1.00 1.00 E_g 1.00 -0.50 -0.50 1.00 -0.50 -0.50 E_g* 1.00 -0.50 -0.50 1.00 -0.50 -0.50 A_u 1.00 1.00 1.00 -1.00 -1.00 -1.00 E_u 1.00 -0.50 -0.50 -1.00 0.50 0.50 E_u* 1.00 -0.50 -0.50 -1.00 0.50 0.50 imaginary part A_g 0.00 0.00 0.00 0.00 0.00 0.00 E_g 0.00 0.87 -0.87 0.00 0.87 -0.87 E_g* 0.00 -0.87 0.87 0.00 -0.87 0.87 A_u 0.00 0.00 0.00 0.00 0.00 0.00 E_u 0.00 0.87 -0.87 0.00 -0.87 0.87 E_u* 0.00 -0.87 0.87 0.00 0.87 -0.87 the symmetry operations in each class: E 1 C3 2 C3^2 3 i 4 S6^5 5 >> S6^5 6 << [...] The class it seems to be recognized correctly but the count of the sym op are not. Is it correct? note: I'm using serial espresso 4.0 cvs 2 on macosx. thanks to all marino ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From giannozz at democritos.it Tue Sep 9 12:40:31 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 09 Sep 2008 12:40:31 +0200 Subject: [Pw_forum] time for calculation of el-ph interactions of graphene supercell In-Reply-To: <48C63CA1.5090403@sissa.it> References: <32795.210.73.17.131.1220924650.squirrel@210.73.16.4> <48C63CA1.5090403@sissa.it> Message-ID: <48C6529F.7030004@democritos.it> Lorenzo Paulatto wrote: > I don't know if el-ph has been implemented with pools parallelization it hasn't Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From dalcorso at sissa.it Tue Sep 9 13:56:58 2008 From: dalcorso at sissa.it (Dal Corso Andrea) Date: Tue, 09 Sep 2008 13:56:58 +0200 Subject: [Pw_forum] Characters table for R-3c In-Reply-To: <21C411C2-DA30-474E-A9D3-D972226101E7@dipteris.unige.it> References: <21C411C2-DA30-474E-A9D3-D972226101E7@dipteris.unige.it> Message-ID: <1220961418.3339.4.camel@dhpc-5-53.sissa.it> On Tue, 2008-09-09 at 11:53 +0200, Marino Vetuschi Zuccolini wrote: > Hello to users, > > I'm working on a carbonate with R-3c (hexagonal axes) structure and > the char table is reported as below in the output file: > [...] > point group S_6 (-3) > there are 6 classes > the character table: > > E C3 C3^2 i S6^5 S6 > A_g 1.00 1.00 1.00 1.00 1.00 1.00 > E_g 1.00 -0.50 -0.50 1.00 -0.50 -0.50 > E_g* 1.00 -0.50 -0.50 1.00 -0.50 -0.50 > A_u 1.00 1.00 1.00 -1.00 -1.00 -1.00 > E_u 1.00 -0.50 -0.50 -1.00 0.50 0.50 > E_u* 1.00 -0.50 -0.50 -1.00 0.50 0.50 > imaginary part > A_g 0.00 0.00 0.00 0.00 0.00 0.00 > E_g 0.00 0.87 -0.87 0.00 0.87 -0.87 > E_g* 0.00 -0.87 0.87 0.00 -0.87 0.87 > A_u 0.00 0.00 0.00 0.00 0.00 0.00 > E_u 0.00 0.87 -0.87 0.00 -0.87 0.87 > E_u* 0.00 -0.87 0.87 0.00 0.87 -0.87 > > the symmetry operations in each class: > E 1 > C3 2 > C3^2 3 > i 4 > S6^5 5 > >> S6^5 6 << > [...] Please, provide the input file. You have probably found a bug. Andrea > > The class it seems to be recognized correctly but the count of the sym > op are not. Is it correct? > > note: I'm using serial espresso 4.0 cvs 2 on macosx. > > thanks to all > > marino > > > > ******************************************************* > Marino Vetuschi Zuccolini > zucco at dipteris.unige.it > Researcher / Geochemist > Laboratory of Geochemistry > > DIPartimento per lo studio della TErra e delle sue RISorse - > Universit? di Genova > Tel. ++39 010 3538136 Fax. ++39 010 352169 > Corso Europa 26, 16132 - Genova - Italy > > > _______________________________________________ > 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 zucco at dipteris.unige.it Tue Sep 9 14:11:01 2008 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Tue, 9 Sep 2008 14:11:01 +0200 Subject: [Pw_forum] Characters table for R-3c In-Reply-To: <1220961418.3339.4.camel@dhpc-5-53.sissa.it> References: <21C411C2-DA30-474E-A9D3-D972226101E7@dipteris.unige.it> <1220961418.3339.4.camel@dhpc-5-53.sissa.it> Message-ID: <3FAF1231-749B-441F-9D41-ABD6D46E3ECA@dipteris.unige.it> Dear Andrea, this is the input file that originates the char table in the output. m. -------------- next part -------------- A non-text attachment was scrubbed... Name: siderite_espresso.in Type: application/octet-stream Size: 2800 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20080909/89700002/attachment.obj -------------- next part -------------- On 09/set/08, at 13:56, Dal Corso Andrea wrote: > On Tue, 2008-09-09 at 11:53 +0200, Marino Vetuschi Zuccolini wrote: >> Hello to users, >> >> I'm working on a carbonate with R-3c (hexagonal axes) structure and >> the char table is reported as below in the output file: >> [...] >> point group S_6 (-3) >> there are 6 classes >> the character table: >> >> E C3 C3^2 i S6^5 S6 >> A_g 1.00 1.00 1.00 1.00 1.00 1.00 >> E_g 1.00 -0.50 -0.50 1.00 -0.50 -0.50 >> E_g* 1.00 -0.50 -0.50 1.00 -0.50 -0.50 >> A_u 1.00 1.00 1.00 -1.00 -1.00 -1.00 >> E_u 1.00 -0.50 -0.50 -1.00 0.50 0.50 >> E_u* 1.00 -0.50 -0.50 -1.00 0.50 0.50 >> imaginary part >> A_g 0.00 0.00 0.00 0.00 0.00 0.00 >> E_g 0.00 0.87 -0.87 0.00 0.87 -0.87 >> E_g* 0.00 -0.87 0.87 0.00 -0.87 0.87 >> A_u 0.00 0.00 0.00 0.00 0.00 0.00 >> E_u 0.00 0.87 -0.87 0.00 -0.87 0.87 >> E_u* 0.00 -0.87 0.87 0.00 0.87 -0.87 >> >> the symmetry operations in each class: >> E 1 >> C3 2 >> C3^2 3 >> i 4 >> S6^5 5 >>>> S6^5 6 << >> [...] > > Please, provide the input file. You have probably found a bug. > > Andrea > > >> >> The class it seems to be recognized correctly but the count of the >> sym >> op are not. Is it correct? >> >> note: I'm using serial espresso 4.0 cvs 2 on macosx. >> >> thanks to all >> >> marino >> >> >> >> ******************************************************* >> Marino Vetuschi Zuccolini >> zucco at dipteris.unige.it >> Researcher / Geochemist >> Laboratory of Geochemistry >> >> DIPartimento per lo studio della TErra e delle sue RISorse - >> Universit? di Genova >> Tel. ++39 010 3538136 Fax. ++39 010 352169 >> Corso Europa 26, 16132 - Genova - Italy >> >> >> _______________________________________________ >> 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 > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From zucco at dipteris.unige.it Tue Sep 9 14:14:14 2008 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Tue, 9 Sep 2008 14:14:14 +0200 Subject: [Pw_forum] Characters table for R-3c Message-ID: <2ACB8477-DA95-4F1F-818B-527218E63567@dipteris.unige.it> Dear Andrea, in the input I sent just few second ago, I've added the keyword nosym = .true., to avoid problem of broken symmetry during relaxation. Sorry. m. ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From dalcorso at sissa.it Tue Sep 9 17:32:48 2008 From: dalcorso at sissa.it (Dal Corso Andrea) Date: Tue, 09 Sep 2008 17:32:48 +0200 Subject: [Pw_forum] Characters table for R-3c In-Reply-To: <2ACB8477-DA95-4F1F-818B-527218E63567@dipteris.unige.it> References: <2ACB8477-DA95-4F1F-818B-527218E63567@dipteris.unige.it> Message-ID: <1220974368.3339.11.camel@dhpc-5-53.sissa.it> Thank you for reporting this. Actually there was a bug with the improper rotations. I have commited the fix on the cvs version. To correct the bug you can substitute the two files PW/divide_class.f90 PW/divide_class_so.f90 with those of the cvs version. Andrea On Tue, 2008-09-09 at 14:14 +0200, Marino Vetuschi Zuccolini wrote: > Dear Andrea, > in the input I sent just few second ago, I've added the keyword nosym > = .true., to avoid problem of broken symmetry during relaxation. > > Sorry. > > m. > > > ******************************************************* > Marino Vetuschi Zuccolini > zucco at dipteris.unige.it > Researcher / Geochemist > Laboratory of Geochemistry > > DIPartimento per lo studio della TErra e delle sue RISorse - > Universit? di Genova > Tel. ++39 010 3538136 Fax. ++39 010 352169 > Corso Europa 26, 16132 - Genova - Italy > > > _______________________________________________ > 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 lfhuang at theory.issp.ac.cn Wed Sep 10 07:28:48 2008 From: lfhuang at theory.issp.ac.cn (=?utf-8?B?6buE6Imv6ZSL?=) Date: Wed, 10 Sep 2008 13:28:48 +0800 Subject: [Pw_forum] =?utf-8?q?time_for_calculation_of_el-ph_interactions_o?= =?utf-8?q?f_graphene_supercell?= Message-ID: <20080910052848.32341.qmail@ms.hfcas.ac.cn> Dear Paulatto and prof. Giannozzi: Thank both of you for your kindly help! Best Wishes! Yours Sincerely L.F.Huang -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080910/faf3ff54/attachment.htm From chemseenu at gmail.com Wed Sep 10 08:00:35 2008 From: chemseenu at gmail.com (kola srinivas) Date: Tue, 9 Sep 2008 23:00:35 -0700 Subject: [Pw_forum] need help Message-ID: Respected Sir, I am doing some adsorption studies with anatase(101) surface with ultrasoft pseudopotentials. my surface model is 2x1x2 with 15 angstroms vaccum space. I am able to run in parallel with water adsorbed on the TiO2 surface. the moment when i replaced water molecule with 2-Cyano-3-phenyl-acrylic acid it is not doing any calculations in parallel. According to the process pid it is showing that in all nodes the job has distributed, but doing nothing. I tried with several options with npool and number of processors. please help me in this regard. yours sincerely, K.Srinivas -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080909/35becf0e/attachment-0001.htm From vegalew at hotmail.com Wed Sep 10 08:09:44 2008 From: vegalew at hotmail.com (vega) Date: Wed, 10 Sep 2008 14:09:44 +0800 Subject: [Pw_forum] need help In-Reply-To: References: Message-ID: >According to the process pid it is showing that in all nodes the job has distributed, but doing nothing. what do you mean by 'doing nothing'? I think you can try to use '-in' instead of '<' in command line. for example, if you type pw.x < inputfile > outputfile, nothing printed on the screen. Then you could try to type pw.x -in inputfile > outputfile. hope helps. by the way I also encounter another situation. If the subnet mask is not consistent among the clusters, problems may be happen. so you may check the subnet mask. ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: kola srinivas Sent: Wednesday, September 10, 2008 2:00 PM To: pw_forum at pwscf.org Subject: [Pw_forum] need help Respected Sir, I am doing some adsorption studies with anatase(101) surface with ultrasoft pseudopotentials. my surface model is 2x1x2 with 15 angstroms vaccum space. I am able to run in parallel with water adsorbed on the TiO2 surface. the moment when i replaced water molecule with 2-Cyano-3-phenyl-acrylic acid it is not doing any calculations in parallel. According to the process pid it is showing that in all nodes the job has distributed, but doing nothing. I tried with several options with npool and number of processors. please help me in this regard. yours sincerely, K.Srinivas -------------------------------------------------------------------------------- _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080910/cd3319c5/attachment.htm From giannozz at democritos.it Wed Sep 10 12:34:03 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 10 Sep 2008 12:34:03 +0200 Subject: [Pw_forum] need help In-Reply-To: References: Message-ID: <48C7A29B.20806@democritos.it> kola srinivas wrote: > I am able to run in parallel with water adsorbed on the > TiO2 surface. the moment when i replaced water molecule with > 2-Cyano-3-phenyl-acrylic acid it is not doing any calculations in > parallel. of course this happens for some other reason and not because the code doesn't like 2-Cyano-whatever instead of water on TiO2 Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From vegalew at hotmail.com Wed Sep 10 13:21:32 2008 From: vegalew at hotmail.com (vega) Date: Wed, 10 Sep 2008 19:21:32 +0800 Subject: [Pw_forum] BUG report about compatibility between he openmpi and QE Message-ID: Dear all, I found something strange when I use QE with openmpi to calculate the surface geometry of anatase 101. QE with openmpi could calculate very thin slabs of 1X3 101 anatase surface. But when I increase the thickness of the slab, the system has 96 atoms or more, QE with openmpi calculation crashed and reported that the 'Stack trace terminated abnormally', then process killed on some nodes. I tried times times again, but the job crashed again and again. I recompile the code again with the mpich2 again with the same compiler same MKL same fftw, today. Then I triggered the same job. The job can run smoothly with MPICH2. So I think openmpi may have some kind of bug. vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080910/7dacad5e/attachment.htm From Giovanni.Cantele at na.infn.it Wed Sep 10 14:35:00 2008 From: Giovanni.Cantele at na.infn.it (Giovanni Cantele) Date: Wed, 10 Sep 2008 14:35:00 +0200 Subject: [Pw_forum] BUG report about compatibility between he openmpi and QE In-Reply-To: References: Message-ID: <48C7BEF4.1070207@na.infn.it> Hi, i'm sorry I can't help you, I want just to mention that the bug, if any, should concern something else than just openmpi + QE. I compiled openmpi-1.2.5 on my cluster and then used it to compile QE (3.2.3, 4.0 and 4.1) and everything works perfectly, even for large jobs. giovanni vega wrote: > Dear all, > I found something strange when I use QE with openmpi to calculate the > surface geometry of anatase 101. > QE with openmpi could calculate very thin slabs of 1X3 101 anatase > surface. But when I increase the thickness > of the slab, the system has 96 atoms or more, QE with openmpi > calculation crashed and reported that the 'Stack trace terminated > abnormally', then process killed on some nodes. I tried times times > again, but the job crashed again and again. > I recompile the code again with the mpich2 again with the same > compiler same MKL same fftw, today. > Then I triggered the same job. The job can run smoothly with MPICH2. > So I think openmpi may have some kind of bug. > vega > ================================================================================= > Vega Lew (weijia liu) > PH.D Candidate in Chemical Engineering > State Key Laboratory of Materials-oriented Chemical Engineering > College of Chemistry and Chemical Engineering > Nanjing University of Technology, 210009, Nanjing, Jiangsu, China > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- 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 cnr.it giovanni.cantele at na.infn.it Web: http://people.na.infn.it/~cantele Research Group: http://www.nanomat.unina.it From chemseenu at gmail.com Wed Sep 10 14:45:50 2008 From: chemseenu at gmail.com (kola srinivas) Date: Wed, 10 Sep 2008 05:45:50 -0700 Subject: [Pw_forum] thank you Message-ID: dear vega lew, i tried the '-in' option instead of '<' in the command, its working nicely, thank you very much for your help, thanks to pw_forum. yours sincerely, K.Srinivas ---------- Forwarded message ---------- From: Date: Wed, Sep 10, 2008 at 12:56 AM Subject: Pw_forum Digest, Vol 15, Issue 19 To: pw_forum at pwscf.org Send Pw_forum mailing list submissions to pw_forum at pwscf.org To subscribe or unsubscribe via the World Wide Web, visit http://www.democritos.it/mailman/listinfo/pw_forum or, via email, send a message with subject or body 'help' to pw_forum-request at pwscf.org You can reach the person managing the list at pw_forum-owner at pwscf.org When replying, please edit your Subject line so it is more specific than "Re: Contents of Pw_forum digest..." Today's Topics: 1. Re: need help (vega) ---------------------------------------------------------------------- Message: 1 Date: Wed, 10 Sep 2008 14:09:44 +0800 From: "vega" Subject: Re: [Pw_forum] need help To: "PWSCF Forum" Message-ID: Content-Type: text/plain; charset="iso-8859-1" >According to the process pid it is showing that in all nodes the job has distributed, but doing nothing. what do you mean by 'doing nothing'? I think you can try to use '-in' instead of '<' in command line. for example, if you type pw.x < inputfile > outputfile, nothing printed on the screen. Then you could try to type pw.x -in inputfile > outputfile. hope helps. by the way I also encounter another situation. If the subnet mask is not consistent among the clusters, problems may be happen. so you may check the subnet mask. ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: kola srinivas Sent: Wednesday, September 10, 2008 2:00 PM To: pw_forum at pwscf.org Subject: [Pw_forum] need help Respected Sir, I am doing some adsorption studies with anatase(101) surface with ultrasoft pseudopotentials. my surface model is 2x1x2 with 15 angstroms vaccum space. I am able to run in parallel with water adsorbed on the TiO2 surface. the moment when i replaced water molecule with 2-Cyano-3-phenyl-acrylic acid it is not doing any calculations in parallel. According to the process pid it is showing that in all nodes the job has distributed, but doing nothing. I tried with several options with npool and number of processors. please help me in this regard. yours sincerely, K.Srinivas -------------------------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080910/2bd18798/attachment.htm From sagarambavale at yahoo.co.in Wed Sep 10 16:04:17 2008 From: sagarambavale at yahoo.co.in (ambavale sagar) Date: Wed, 10 Sep 2008 19:34:17 +0530 (IST) Subject: [Pw_forum] Parallel PWCOND problem Message-ID: <442651.34866.qm@web94616.mail.in2.yahoo.com> Hello all, I am getting different results for PWCOND calculations of same system if different number of processors are used. My system is of x86_64 architacture with dual cpu quad core xeon machine (total 8 cores). Total memory is 8 GB. Here LAM-MPI-7.1.4 is compiled with intel fortran-10.1.12 and intel MKL-10.0.1.013 is used. More Surprisingly, we can get same results with 2 and 8 cores while for results (value of Transmission coefficient and eigen channels) with 4 processors are very different. In some cases results are same and sometimes they differ. Where the things may going wrong? Input files for al.scf, al-c7-al.scf and al-c7-al.cond are here: ------------------------------- al.scf &CONTROL calculation = 'scf' restart_mode= 'from_scratch' prefix = 'al_nc' pseudo_dir = '/home/sagar/espresso-4.0/pseudo/' outdir = '/home/sagar/sagar_espresso_tmp/' / &SYSTEM ibrav = 6, celldm(1) = 22.9679, celldm(3) = 0.333, nat = 9, ntyp = 1, ecutwfc = 30.0, ecutrho = 180.0, occupations = "smearing", smearing = "methfessel-paxton", degauss = 0.01, / &ELECTRONS conv_thr = 1.d-12, mixing_beta = 0.5, / ATOMIC_SPECIES Al 26.98 Al.vbc.UPF ATOMIC_POSITIONS Al 0.0 0.0 0.0 Al 0.333 0.0 0.0 Al 0.0 0.333 0..0 Al 0.333 0.333 0.0 Al .1665 .1665 0.0 Al 0.0 0.1665 0.1665 Al 0.1665 0.0 0.1665 Al 0.333 0.1665 0.1665 Al 0.1665 0.333 0.1665 K_POINTS (automatic) 1 1 2 1 1 1 --------------------------------------- al-c7-al.scf &CONTROL calculation = 'scf' restart_mode= 'from_scratch' prefix = 'al_7c_al-nc' pseudo_dir = '/home/sagar/espresso-4.0/pseudo/' outdir = '/home/sagar/sagar_espresso_tmp/' / &SYSTEM ibrav = 6, celldm(1) = 22.9679, celldm(3) = 1.7884518095, nat = 38, ntyp = 2, ecutwfc = 30.0, ecutrho = 180.0, occupations = 'smearing' smearing = 'methfessel-paxton' degauss = 0.01, / &ELECTRONS conv_thr = 1.d-8, mixing_beta = 0.5, / ATOMIC_SPECIES Al 26.98 Al.vbc.UPF C 12.01 C.pz-vbc.UPF ATOMIC_POSITIONS (angstrom) Al 0.0000000 0.0000000 0.0000000 Al 4.0509582 0.0000000 0.0000000 Al 0.0000000 4.0473120 0.0000000 Al 4.0509582 4.0509582 0.0000000 Al 2.0236560 2.0236560 0.0000000 Al 0.0000000 2.0236560 2.0236560 Al 2.0236560 0.0000000 2.0236560 Al 4.0509582 2.0236560 2.0236560 Al 2.0236560 4.0513593 2.0236560 Al 0.0000000 0.0000000 4.0513593 Al 4.0509582 0.0000000 4.0513593 Al 0.0000000 4.0473120 4.0513593 Al 4.0509582 4.0509582 4.0513593 Al 2.0236560 2.0236560 4.0513593 Al 0.0000000 2.0236560 6.0709680 Al 2.0236560 0.0000000 6.0709680 Al 4.0509582 2.0236560 6.0709680 Al 2.0236560 4.0513593 6.0709680 C 2.0236560 2.0236560 7.0709680 C 2.0236560 2.0236560 8.3381604 C 2.0236560 2.0236560 9.6053528 C 2.0236560 2.0236560 10.8725452 C 2.0236560 2.0236560 12.1397376 C 2.0236560 2.0236560 13.40693 C 2.0236560 2.0236560 14.6741224 Al 0.0000000 2.0236560 15.6741224 Al 2.0236560 0.0000000 15.6741224 Al 4.0513593 2.0236560 15.6741224 Al 2.0236560 4.0513593 15.6741224 Al 0.0000000 0.0000000 17.6937311 Al 4.0513593 0.0000000 17.6937311 Al 0.0000000 4.0513593 17.6937311 Al 4.0513593 4.0513593 17.6937311 Al 2.0236560 2.0236560 17.6937311 Al 0.0000000 2.0236560 19.7133398 Al 2.0236560 0.0000000 19.7133398 Al 4.0513593 2.0236560 19.7133398 Al 2.0236560 4.0513593 19.7133398 K_POINTS (automatic) 1 1 2 1 1 1 ------------------------------------------------------ al-c7-al.cond &inputcond outdir = '/home/sagar/sagar_espresso_tmp/' tran_file ='al-c7-al-tran' ikind = 1 iofspin = 1 prefixl = 'al-nc' prefixs = 'al_7c_al-nc' energy0 = 4.0 denergy = -0.1 ewind = 2 ecut2d = 20.0 epsproj = 1.d-8 nz1 = 7 delgep = 5.d-10 / 1 0.0 0.0 1.0 11 ---------------------------------------------------------- Thank you Sagar Ambavale Research Student The M.S. Uni. Of Baroda India Add more friends to your messenger and enjoy! Go to http://in.messenger..yahoo.com/invite/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080910/e7a96f56/attachment-0001.htm From smogunov at sissa.it Wed Sep 10 15:58:28 2008 From: smogunov at sissa.it (Alexander) Date: Wed, 10 Sep 2008 15:58:28 +0200 Subject: [Pw_forum] Parallel PWCOND problem In-Reply-To: <442651.34866.qm@web94616.mail.in2.yahoo.com> References: <442651.34866.qm@web94616.mail.in2.yahoo.com> Message-ID: <200809101558.28169.smogunov@sissa.it> Dear Sagar Could you also send the PWCOND output files for different number of CPU when you have different results? Alexander On Wednesday 10 September 2008 16:04, ambavale sagar wrote: > Hello all, > I am getting different results for PWCOND calculations of same system if > different number of processors are used. My system is of x86_64 > architacture with dual cpu quad core xeon machine (total 8 cores). Total > memory is 8 GB. Here LAM-MPI-7.1.4 is compiled with intel fortran-10.1.12 > and intel MKL-10.0.1.013 is used. More Surprisingly, we can get same > results with 2 and 8 cores while for results (value of Transmission > coefficient and eigen channels) with 4 processors are very different. In > some cases results are same and sometimes they differ. Where the things > may going wrong? Input files for al.scf, al-c7-al.scf and al-c7-al.cond are > here: > ------------------------------- > al.scf > > &CONTROL > calculation = 'scf' > restart_mode= 'from_scratch' > prefix = 'al_nc' > pseudo_dir = '/home/sagar/espresso-4.0/pseudo/' > outdir = '/home/sagar/sagar_espresso_tmp/' > / > &SYSTEM > ibrav = 6, > celldm(1) = 22.9679, > celldm(3) = 0.333, > nat = 9, > ntyp = 1, > ecutwfc = 30.0, > ecutrho = 180.0, > occupations = "smearing", > smearing = "methfessel-paxton", > degauss = 0.01, > / > &ELECTRONS > conv_thr = 1.d-12, > mixing_beta = 0.5, > / > ATOMIC_SPECIES > Al 26.98 Al.vbc.UPF > ATOMIC_POSITIONS > Al 0.0 0.0 0.0 > Al 0.333 0.0 0.0 > Al 0.0 0.333 0..0 > Al 0.333 0.333 0.0 > Al .1665 .1665 0.0 > Al 0.0 0.1665 0.1665 > Al 0.1665 0.0 0.1665 > Al 0.333 0.1665 0.1665 > Al 0.1665 0.333 0.1665 > K_POINTS (automatic) > 1 1 2 1 1 1 > > > --------------------------------------- > al-c7-al.scf > > &CONTROL > calculation = 'scf' > restart_mode= 'from_scratch' > prefix = 'al_7c_al-nc' > pseudo_dir = '/home/sagar/espresso-4.0/pseudo/' > outdir = '/home/sagar/sagar_espresso_tmp/' > / > &SYSTEM > ibrav = 6, > celldm(1) = 22.9679, > celldm(3) = 1.7884518095, > nat = 38, > ntyp = 2, > ecutwfc = 30.0, > ecutrho = 180.0, > occupations = 'smearing' > smearing = 'methfessel-paxton' > degauss = 0.01, > / > &ELECTRONS > conv_thr = 1.d-8, > mixing_beta = 0.5, > / > ATOMIC_SPECIES > Al 26.98 Al.vbc.UPF > C 12.01 C.pz-vbc.UPF > ATOMIC_POSITIONS (angstrom) > Al 0.0000000 0.0000000 0.0000000 > Al 4.0509582 0.0000000 0.0000000 > Al 0.0000000 4.0473120 0.0000000 > Al 4.0509582 4.0509582 0.0000000 > Al 2.0236560 2.0236560 0.0000000 > Al 0.0000000 2.0236560 2.0236560 > Al 2.0236560 0.0000000 2.0236560 > Al 4.0509582 2.0236560 2.0236560 > Al 2.0236560 4.0513593 2.0236560 > Al 0.0000000 0.0000000 4.0513593 > Al 4.0509582 0.0000000 4.0513593 > Al 0.0000000 4.0473120 4.0513593 > Al 4.0509582 4.0509582 4.0513593 > Al 2.0236560 2.0236560 4.0513593 > Al 0.0000000 2.0236560 6.0709680 > Al 2.0236560 0.0000000 6.0709680 > Al 4.0509582 2.0236560 6.0709680 > Al 2.0236560 4.0513593 6.0709680 > C 2.0236560 2.0236560 7.0709680 > C 2.0236560 2.0236560 8.3381604 > C 2.0236560 2.0236560 9.6053528 > C 2.0236560 2.0236560 10.8725452 > C 2.0236560 2.0236560 12.1397376 > C 2.0236560 2.0236560 13.40693 > C 2.0236560 2.0236560 14.6741224 > Al 0.0000000 2.0236560 15.6741224 > Al 2.0236560 0.0000000 15.6741224 > Al 4.0513593 2.0236560 15.6741224 > Al 2.0236560 4.0513593 15.6741224 > Al 0.0000000 0.0000000 17.6937311 > Al 4.0513593 0.0000000 17.6937311 > Al 0.0000000 4.0513593 17.6937311 > Al 4.0513593 4.0513593 17.6937311 > Al 2.0236560 2.0236560 17.6937311 > Al 0.0000000 2.0236560 19.7133398 > Al 2.0236560 0.0000000 19.7133398 > Al 4.0513593 2.0236560 19.7133398 > Al 2.0236560 4.0513593 19.7133398 > K_POINTS (automatic) > 1 1 2 1 1 1 > ------------------------------------------------------ > al-c7-al.cond > > &inputcond > outdir = '/home/sagar/sagar_espresso_tmp/' > tran_file ='al-c7-al-tran' > ikind = 1 > iofspin = 1 > prefixl = 'al-nc' > prefixs = 'al_7c_al-nc' > energy0 = 4.0 > denergy = -0.1 > ewind = 2 > ecut2d = 20.0 > epsproj = 1.d-8 > nz1 = 7 > delgep = 5.d-10 > / > 1 > 0.0 0.0 1.0 > 11 > ---------------------------------------------------------- > Thank you > > > Sagar Ambavale > Research Student > The M.S. Uni. Of Baroda > India > > > Add more friends to your messenger and enjoy! Go to > http://in.messenger..yahoo.com/invite/ From szs_naghavi at yahoo.com Wed Sep 10 19:24:18 2008 From: szs_naghavi at yahoo.com (zahra sadat naghavi) Date: Wed, 10 Sep 2008 10:24:18 -0700 (PDT) Subject: [Pw_forum] (no subject) Message-ID: <731392.91964.qm@web63007.mail.re1.yahoo.com> Dear all ! Can any body help me about ? from phq_readin : error #???????? 9 ???? k-points are odd when I run ph.x for apoint other than gamma? this is my ph.in: phonons of nanotube &inputph ? tr2_ph=1.0d-16, ? prefix='CNT6', ? lnscf=.true., ? amass(1)=12.0107 ? outdir='/scratch19/test3/tmp6', ? fildyn='nano.dyn3', ?/ ?0.0?? 0.0?? 0.931079277303714E+00 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080910/63914ed8/attachment.htm From vegalew at hotmail.com Thu Sep 11 06:45:15 2008 From: vegalew at hotmail.com (vega) Date: Thu, 11 Sep 2008 12:45:15 +0800 Subject: [Pw_forum] thank you In-Reply-To: References: Message-ID: >i tried the '-in' option instead of '<' in the command, its working nicely, I'm glad to here that. >thank you very much for your help you're welcome vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: kola srinivas Sent: Wednesday, September 10, 2008 8:45 PM To: pw_forum at pwscf.org Subject: [Pw_forum] thank you dear vega lew, i tried the '-in' option instead of '<' in the command, its working nicely, thank you very much for your help, thanks to pw_forum. yours sincerely, K.Srinivas ---------- Forwarded message ---------- From: Date: Wed, Sep 10, 2008 at 12:56 AM Subject: Pw_forum Digest, Vol 15, Issue 19 To: pw_forum at pwscf.org Send Pw_forum mailing list submissions to pw_forum at pwscf.org To subscribe or unsubscribe via the World Wide Web, visit http://www.democritos.it/mailman/listinfo/pw_forum or, via email, send a message with subject or body 'help' to pw_forum-request at pwscf.org You can reach the person managing the list at pw_forum-owner at pwscf.org When replying, please edit your Subject line so it is more specific than "Re: Contents of Pw_forum digest..." Today's Topics: 1. Re: need help (vega) ---------------------------------------------------------------------- Message: 1 Date: Wed, 10 Sep 2008 14:09:44 +0800 From: "vega" Subject: Re: [Pw_forum] need help To: "PWSCF Forum" Message-ID: Content-Type: text/plain; charset="iso-8859-1" >According to the process pid it is showing that in all nodes the job has distributed, but doing nothing. what do you mean by 'doing nothing'? I think you can try to use '-in' instead of '<' in command line. for example, if you type pw.x < inputfile > outputfile, nothing printed on the screen. Then you could try to type pw.x -in inputfile > outputfile. hope helps. by the way I also encounter another situation. If the subnet mask is not consistent among the clusters, problems may be happen. so you may check the subnet mask. ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: kola srinivas Sent: Wednesday, September 10, 2008 2:00 PM To: pw_forum at pwscf.org Subject: [Pw_forum] need help Respected Sir, I am doing some adsorption studies with anatase(101) surface with ultrasoft pseudopotentials. my surface model is 2x1x2 with 15 angstroms vaccum space. I am able to run in parallel with water adsorbed on the TiO2 surface. the moment when i replaced water molecule with 2-Cyano-3-phenyl-acrylic acid it is not doing any calculations in parallel. According to the process pid it is showing that in all nodes the job has distributed, but doing nothing. I tried with several options with npool and number of processors. please help me in this regard. yours sincerely, K.Srinivas -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080911/602cbfa7/attachment.htm From giannozz at democritos.it Thu Sep 11 09:36:24 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 11 Sep 2008 09:36:24 +0200 Subject: [Pw_forum] (no subject) In-Reply-To: <731392.91964.qm@web63007.mail.re1.yahoo.com> References: <731392.91964.qm@web63007.mail.re1.yahoo.com> Message-ID: <48C8CA78.2030106@democritos.it> zahra sadat naghavi wrote: > Can any body help me about > from phq_readin : error # 9 > k-points are odd > when I run ph.x for apoint other than gamma? > this is my ph.in: please provide the input for the scf calculation as well Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From lataprem29 at gmail.com Thu Sep 11 13:41:16 2008 From: lataprem29 at gmail.com (premlata pandit) Date: Thu, 11 Sep 2008 17:11:16 +0530 Subject: [Pw_forum] phonon calculation Message-ID: <6f289440809110441x7262fa72t70a266a9a866e548@mail.gmail.com> respected sir , i calculating phonon for Na2S, in parallel process, the input files are na2s.scf.in &CONTROL calculation = 'scf' , restart_mode = 'from_scratch' , outdir = './OUT', pseudo_dir = '/root/bipul/espresso-4.0.1/pseudo/Na2S', prefix = 'Na2S', etot_conv_thr = 1.0d-5 , forc_conv_thr = 1.0d-4 , tstress = .true. , tprnfor = .true. , dt = 80 / &SYSTEM ibrav = 2, celldm(1) = 12.41600, nat = 3, ntyp = 2 , ecutwfc = 28, ecutrho = 600, occupations = 'smearing' , smearing = 'methfessel-paxton', degauss = 0.01 / &ELECTRONS mixing_beta = 0.5 , conv_thr = 1.0d-8 diagonalization = 'david' , / &IONS / ATOMIC_SPECIES S 32.066 S.pw91-van_ak.UPF Na 22.988 Na.pw91-sp-van_ak.UPF ATOMIC_POSITIONS (crystal) S 0.0 0.0 0.0 Na 0.25 0.25 0.25 Na 0.75 0.75 0.75 K_POINTS automatic 8 8 8 1 1 1 and na2S.ph.in phonons of Na2S &inputph tr2_ph=1.0d-12, alpha_mix(1)=0.7, prefix='Na2S', ! fildvscf='na2sdv', amass(1)=32.066, amass(2)=22.988, outdir='./OUT', fildyn='na2s.dyn', ! elph=.true., ! trans=.true., ldisp=.true. nq1=4, nq2=4, nq3=4 / first ,i run mpd & command then we run na2s.scf.in file with mpirun -np 2 /root/bipul/parallel/esspresso-cvs/bin/pw.x command this file will be run but when i run na2s.ph.in file with this command , it show error from read_namelists : error # 1 reading namelist control sir , i check all things but, i can not find where is mistake any one suggest me where is mistake thank you sir -- Premlata Pandit Ph.d. student, Barkatullah university, Bhopal 462026, MP, India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080911/abd2c8a6/attachment.htm From giannozz at democritos.it Thu Sep 11 14:31:06 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 11 Sep 2008 14:31:06 +0200 Subject: [Pw_forum] phonon calculation In-Reply-To: <6f289440809110441x7262fa72t70a266a9a866e548@mail.gmail.com> References: <6f289440809110441x7262fa72t70a266a9a866e548@mail.gmail.com> Message-ID: <48C90F8A.1020100@democritos.it> premlata pandit wrote: > first ,i run mpd & command then we run na2s.scf.in > file with mpirun -np 2 /root/bipul/parallel/esspresso-cvs/bin/pw.x command > this file will be run > but when i run na2s.ph.in file with this command , > it show error > from read_namelists : error # 1 > reading namelist control > > sir , i check all things but ...you didn't. The code that performs phonon calculation is called "ph.x" Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From huang87 at purdue.edu Thu Sep 11 20:05:38 2008 From: huang87 at purdue.edu (Zhen Huang) Date: Thu, 11 Sep 2008 14:05:38 -0400 Subject: [Pw_forum] *.fc file in PW_SCF help. Message-ID: <1221156338.48c95df20c43e@webmail.purdue.edu> *.fc file in PW_SCF help. Dear Friends- Can anyone help me explain the real space force constant file *.fc. What does each index mean for example (I am doing a 4 by4 by4 calculation): What is the second row 1 1 1 1 means what is saying 1.45522e+00 is this force constant in ev/A^2? Thank you every much. 4 4 4 1 1 1 1 1 1 1 1.45522064500E+00 2 1 1 -6.48059987500E-02 3 1 1 1.15693187500E-03 4 1 1 -6.48059987500E-02 1 2 1 -6.48059987500E-02 2 2 1 -9.61647875000E-03 3 2 1 -2.75985062500E-03 4 2 1 5.30879975000E-02 1 3 1 1.15693187500E-03 2 3 1 -2.75985062500E-03 3 3 1 1.03823687500E-02 4 3 1 -2.75985062500E-03 1 4 1 -6.48059987500E-02 2 4 1 5.30879975000E-02 3 4 1 -2.75985062500E-03 4 4 1 -9.61647875000E-03 1 1 2 -2.61743750000E-04 2 1 2 -3.88856250000E-05 3 1 2 1.02087500000E-04 4 1 2 -5.68562500000E-05 --------------- Zhen Huang Ph.D. Student Nanoscale Transport Research Group School of Mechanical Engineering Purdue University Tel: 765 426 9302 From wangqj1 at 126.com Fri Sep 12 03:33:40 2008 From: wangqj1 at 126.com (wangqj1) Date: Fri, 12 Sep 2008 09:33:40 +0800 (CST) Subject: [Pw_forum] the relax is to slow Message-ID: <20230225.357191221183220440.JavaMail.coremail@bj126app29.126.com> Dear pwscf users When I use 'relax' to optimize the suppercell of 48 atoms ,it runs too slow ,in 9 hours the iterative is only 5,I think it will need 1 month to relax the suppercell like this,but I remember several months ago ,I relax the suppercell of 32 atoms ,it only needs 8 atoms .I don't know what's the reason cause this ,how can I to promote this . Any advice will be appreciated ! sincerely qjwang XiangTan University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080912/6102a677/attachment.htm From vegalew at hotmail.com Fri Sep 12 06:14:47 2008 From: vegalew at hotmail.com (vega) Date: Fri, 12 Sep 2008 12:14:47 +0800 Subject: [Pw_forum] the relax is to slow In-Reply-To: <20230225.357191221183220440.JavaMail.coremail@bj126app29.126.com> References: <20230225.357191221183220440.JavaMail.coremail@bj126app29.126.com> Message-ID: pay more attention on you CPU user% ratio and see whether your cpu is waiting for IO or virtual memory. if IO wait% quite high, this may be because of the network. as far as I know, the network will quite busy if you use very few k points. you may need infiniband for gamma point calculation for a large system. If your memory is not enough, this will also happen. More atoms in the system will demand much more memory. Virtual memory will be employed if the physical memory is not enough. And the exchanging of data between the virtual memory and physical memory need a lot of time. So you may suffer from this. I know nothing about your hardware, and input file. so I can't tell you the further about your problem. hope helps vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: wangqj1 Sent: Friday, September 12, 2008 9:33 AM To: pw_forum Subject: [Pw_forum] the relax is to slow Dear pwscf users When I use 'relax' to optimize the suppercell of 48 atoms ,it runs too slow ,in 9 hours the iterative is only 5,I think it will need 1 month to relax the suppercell like this,but I remember several months ago ,I relax the suppercell of 32 atoms ,it only needs 8 atoms .I don't know what's the reason cause this ,how can I to promote this . Any advice will be appreciated ! sincerely qjwang XiangTan University -------------------------------------------------------------------------------- _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080912/5c488ea6/attachment.htm From willykohn at gmail.com Fri Sep 12 08:16:45 2008 From: willykohn at gmail.com (willy kohn) Date: Fri, 12 Sep 2008 02:16:45 -0400 Subject: [Pw_forum] How is the eigenvectors stored in the parallel version? Message-ID: <8a1b6d820809112316r2fe4197bp2c0088a44fe01b51@mail.gmail.com> Dear PWSCF users: I'm trying to write all the eigenvectors into a file for further analysis. I found they are stored in evc , but got confused for the partition format in the parallel version (v4.0.1). Each cpu seems get only the eigenvectors on a subset of the "global" G space, and all the cpus have all the kpoints. I'm really puzzled by the way the evc stored in each cpu. My question is: Do I need to collect evc's from all the cpu's in order to get a eigenvector for a perticular k and perticular band? If yes, how? Best, Wei Georgia Tech -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080912/bbc28e5e/attachment-0001.htm From masoudnahali at gmail.com Fri Sep 12 08:33:11 2008 From: masoudnahali at gmail.com (masoud nahali) Date: Thu, 11 Sep 2008 23:33:11 -0700 Subject: [Pw_forum] sum-band.f90 Message-ID: Dear users after configuring quantum espresso, while I wrote "make all" in the shell, all things was proper and compiling has been continued until I gave this error: "sum-band.f90 (code 3)". and installing cropped because of this error. would you please give me some comment? -- Sincerely Yours Masoud Iran -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080911/2ae9d3af/attachment.htm From giannozz at democritos.it Fri Sep 12 11:02:42 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 12 Sep 2008 11:02:42 +0200 Subject: [Pw_forum] How is the eigenvectors stored in the parallel version? In-Reply-To: <8a1b6d820809112316r2fe4197bp2c0088a44fe01b51@mail.gmail.com> References: <8a1b6d820809112316r2fe4197bp2c0088a44fe01b51@mail.gmail.com> Message-ID: <48CA3032.2060502@democritos.it> willy kohn wrote: > I'm trying to write all the eigenvectors into a file for further > analysis. the code already does this. Use the available machinery to write and read eigenvectors. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Fri Sep 12 11:03:19 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 12 Sep 2008 11:03:19 +0200 Subject: [Pw_forum] sum-band.f90 In-Reply-To: References: Message-ID: <48CA3057.6040006@democritos.it> masoud nahali wrote: > "sum-band.f90 (code 3)" there must be an error message preceding this -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Fri Sep 12 11:15:25 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 12 Sep 2008 11:15:25 +0200 Subject: [Pw_forum] *.fc file in PW_SCF help. In-Reply-To: <1221156338.48c95df20c43e@webmail.purdue.edu> References: <1221156338.48c95df20c43e@webmail.purdue.edu> Message-ID: <48CA332D.3010102@democritos.it> Zhen Huang wrote: > Can anyone help me explain the real space force constant file *.fc. What does > each index mean for example (I am doing a 4 by4 by4 calculation): What is the > second row 1 1 1 1 means what is saying 1.45522e+00 never heard about the "search" function of the pw_forum archives? look for "force constant file" and you will find this: http://www.democritos.it/pipermail/pw_forum/2005-April/002408.html > is this force constant in ev/A^2? atomic (Rydberg) units Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From vegalew at hotmail.com Fri Sep 12 11:26:01 2008 From: vegalew at hotmail.com (vega) Date: Fri, 12 Sep 2008 17:26:01 +0800 Subject: [Pw_forum] Fw: relax is too slow Message-ID: >I think it is the too much k-points that caused the problem .when I set the k-point to Gamma ,it iterated very quickly.but I don't know whether it is right or wrong that I set it to Gamma. I don't know either. I think you could compare the Gamma results and 2X2X2 or 4X4X4 if necessary. If the energy different and coordinate is nearly the same. you may could use a small K-mesh to save computional cost. As far as I tested, the energy would be within 1% for large supercells, but the coordinates of several atoms may differ 200%. So I have the same question like you. Could anyone help us, please~ what is the benchmark for judging whether the k-mesh is adequate? the most important thing is the energy or the coordinates of atoms? If the energy difference is with 1% and 80% of atoms or more in the system is nearly at the same position, but coordinate some of atoms very different for the two case. Could we trust the results calculated by relative small number of k points? vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: wangqj1 Sent: Friday, September 12, 2008 3:34 PM To: vegalew at hotmail.com Subject: relax is too slow Thank you very much for your advice ! My input file as follows: &CONTROL title = ZnOrelax , calculation = 'relax' , restart_mode = 'from_scratch' , pseudo_dir = '/root/tool/espresso-3.2.3/pseudo/' , prefix = ZnO , forc_conv_thr = 1.D-4 , tstress = .true. , tprnfor = .true. , dt = 100 , / &SYSTEM ibrav = 14, celldm(1) = 18.6488, celldm(2) = 0.666667, celldm(3) = 1.0691, celldm(4) = 0, celldm(5) = 0, celldm(6) = -0.5, nat = 48, ntyp = 2, ecutwfc = 40 , ecutrho = 200 , nosym = .true. , nbnd = 220, occupations = 'smearing' , degauss = 0.002 , smearing = 'methfessel-paxton' , nspin = 1 , lda_plus_u = .false. , / &ELECTRONS startingpot = 'atomic' , startingwfc = 'atomic' , mixing_mode = 'plain' , mixing_beta = 0.7 , diagonalization = 'cg' , / &IONS ion_dynamics = 'damp' , phase_space = 'full' , pot_extrapolation = 'second_order' , wfc_extrapolation = 'second_order' , / ATOMIC_SPECIES Zn 65.00000 Zn.pbe-van.UPF O 16.00000 O.pbe-rrkjus.UPF ATOMIC_POSITIONS angstrom Zn 0.000000000 1.899193710 0.000000000 1 1 1 O 0.000000000 1.899193710 2.018253260 1 1 1 Zn 1.644750000 0.949596855 2.637550000 1 1 1 O 1.644750000 0.949596855 4.655803260 1 1 1 Zn 3.289500000 1.899193710 0.000000000 1 1 1 O 3.289500000 1.899193710 2.018253260 1 1 1 Zn 4.934250000 0.949596855 2.637550000 0 0 0 O 4.934250000 0.949596855 4.655803260 0 0 0 Zn 6.579000000 1.899193710 0.000000000 0 0 0 O 6.579000000 1.899193710 2.018253260 0 0 0 Zn 8.223750000 0.949596855 2.637550000 0 0 0 O 8.223750000 0.949596855 4.655803260 0 0 0 Zn -1.644750000 4.747984276 0.000000000 1 1 1 O -1.644750000 4.747984276 2.018253260 1 1 1 Zn 0.000000000 3.798387421 2.637550000 1 1 1 O 0.000000000 3.798387421 4.655803260 1 1 1 Zn 1.644750000 4.747984276 0.000000000 1 1 1 O 1.644750000 4.747984276 2.018253260 1 1 1 Zn 3.289500000 3.798387421 2.637550000 0 0 0 O 3.289500000 3.798387421 4.655803260 0 0 0 Zn 4.934250000 4.747984276 0.000000000 0 0 0 O 4.934250000 4.747984276 2.018253260 0 0 0 Zn 6.579000000 3.798387421 2.637550000 0 0 0 O 6.579000000 3.798387421 4.655803260 0 0 0 Zn 0.000000000 1.899193710 5.275100000 1 1 1 O 0.000000000 1.899193710 7.293353260 1 1 1 Zn 1.644750000 0.949596855 7.912650000 1 1 1 O 1.644750000 0.949596855 9.930903260 1 1 1 Zn 3.289500000 1.899193710 5.275100000 1 1 1 O 3.289500000 1.899193710 7.293353260 1 1 1 Zn 4.934250000 0.949596855 7.912650000 0 0 0 O 4.934250000 0.949596855 9.930903260 0 0 0 Zn 6.579000000 1.899193710 5.275100000 0 0 0 O 6.579000000 1.899193710 7.293353260 0 0 0 Zn 8.223750000 0.949596855 7.912650000 0 0 0 O 8.223750000 0.949596855 9.930903260 0 0 0 Zn -1.644750000 4.747984276 5.275100000 1 1 1 O -1.644750000 4.747984276 7.293353260 1 1 1 Zn 0.000000000 3.798387421 7.912650000 1 1 1 O 0.000000000 3.798387421 9.930903260 1 1 1 Zn 1.644750000 4.747984276 5.275100000 1 1 1 O 1.644750000 4.747984276 7.293353260 1 1 1 Zn 3.289500000 3.798387421 7.912650000 0 0 0 O 3.289500000 3.798387421 9.930903260 0 0 0 Zn 4.934250000 4.747984276 5.275100000 0 0 0 O 4.934250000 4.747984276 7.293353260 0 0 0 Zn 6.579000000 3.798387421 7.912650000 0 0 0 O 6.579000000 3.798387421 9.930903260 0 0 0 K_POINTS automatic 4 4 4 0 0 0 According your advice I found the CPU user% ratio is about 98% and the swap is about 16%.I think it is the too much k-points that caused the problem .when I set the k-point to Gamma ,it iterated very quickly.but I don't know whether it is right or wrong that I set it to Gamma. Sincerely -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080912/e29e69ea/attachment-0001.htm From giannozz at democritos.it Fri Sep 12 11:27:44 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 12 Sep 2008 11:27:44 +0200 Subject: [Pw_forum] Fw: relax is too slow In-Reply-To: References: Message-ID: <48CA3610.1070704@democritos.it> vega wrote: > As far as I tested, the energy would be within 1% for large supercells, > but the coordinates of several atoms may differ 200%. doesn't make sense -- Paolo Giannozzi, Democritos and University of Udine, Italy From vegalew at hotmail.com Fri Sep 12 11:36:34 2008 From: vegalew at hotmail.com (vega) Date: Fri, 12 Sep 2008 17:36:34 +0800 Subject: [Pw_forum] Fw: relax is too slow In-Reply-To: <48CA3610.1070704@democritos.it> References: <48CA3610.1070704@democritos.it> Message-ID: Dear sir > vega wrote: > >> As far as I tested, the energy would be within 1% for large supercells, >> but the coordinates of several atoms may differ 200%. > > doesn't make sense could you tell me more about how to choose the k point mesh? vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------------------------------------------- From: "Paolo Giannozzi" Sent: Friday, September 12, 2008 5:27 PM To: "PWSCF Forum" Subject: Re: [Pw_forum] Fw: relax is too slow > -- > 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 masoudnahali at gmail.com Fri Sep 12 12:27:13 2008 From: masoudnahali at gmail.com (masoud nahali) Date: Fri, 12 Sep 2008 03:27:13 -0700 Subject: [Pw_forum] error in compiling Quantum Espresso Message-ID: Dear Users error in compiling Quantum Espresso: ./configure was successful but after using make all I have received this: fortcom: server: ** internal compiler error :segmentation violation signal raised ** please report this error along with the circumstances in which it occured in a software problem report. note : file and line given may not be explicit cause this error. compilation aborted for sum_band.f90 make[1]: *** [sum_band.o] error 3 make[1]: living directory /home/quantum espresso/. . /PW make: *** [PW} error 2 my compiler is "ifort 9.0" and I have installed "MKL 9.0" so. c++ has been installed too. would you please help me? many thanks. -- Sincerely Yours Masoud Iran -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080912/26bdfd34/attachment.htm From masoudnahali at gmail.com Fri Sep 12 14:13:54 2008 From: masoudnahali at gmail.com (masoud nahali) Date: Fri, 12 Sep 2008 05:13:54 -0700 Subject: [Pw_forum] error in compiling Quantum Espresso Message-ID: Dear Users error in compiling Quantum Espresso: ./configure was successful but after using make all I have received this: fortcom: server: ** internal compiler error :segmentation violation signal raised ** please report this error along with the circumstances in which it occured in a software problem report. note : file and line given may not be explicit cause this error. compilation aborted for sum_band.f90 make[1]: *** [sum_band.o] error 3 make[1]: living directory /home/quantum espresso/. . /PW make: *** [PW} error 2 my compiler is "ifort 9.0" and I have installed "MKL 9.0" so. c++ has been installed too. would you please help me? many thanks. -- Sincerely Yours Masoud Nahali Sharif University of Tehran Iran -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080912/f0469050/attachment.htm From huang87 at purdue.edu Fri Sep 12 15:47:53 2008 From: huang87 at purdue.edu (Zhen Huang) Date: Fri, 12 Sep 2008 09:47:53 -0400 Subject: [Pw_forum] Can anyone help me explain what do the terms in *.fc file means? Message-ID: <1221227273.48ca7309639ff@webmail.purdue.edu> Dear PW_SCFers, If you use q2r.x command to calculate IFC's in real space and the output file is gra.fc. What does the output mean in the gra.fc file. for example: T (what does T mean) 13.7429076 0.0000000 0.0000000 0.0000000 13.7429076 0.0000000 0.0000000 0.0000000 13.7429076 (what does this 3 by 3 matrix represent) 1 2.5582694 0.0000000 0.0000000 0.0000000 2.5582694 0.0000000 0.0000000 0.0000000 2.5582694 2 -2.5582694 0.0000000 0.0000000 0.0000000 -2.5582694 0.0000000 0.0000000 0.0000000 -2.5582694 4 4 4 1 1 1 1 (what do the indexes represent? 1 for what) 1 1 1 2.24617477868E-01 (what is this value and in what unit?) 2 1 1 -6.30036204778E-03 3 1 1 3.67142905560E-05 4 1 1 -6.30036204778E-03 1 2 1 1.16235628668E-02 2 2 1 -5.93301773304E-04 3 2 1 -2.47983120458E-04 4 2 1 -6.30036204778E-03 1 3 1 4.15237248762E-03 2 3 1 -5.93301773304E-04 3 3 1 3.67142905560E-05 4 3 1 -5.93301773304E-04 1 4 1 1.16235628668E-02 2 4 1 -6.30036204778E-03 3 4 1 -2.47983120458E-04 4 4 1 -5.93301773304E-04 1 1 2 -6.30036204778E-03 --------------- Zhen Huang Ph.D. Student Nanoscale Transport Research Group School of Mechanical Engineering Purdue University Tel: 765 426 9302 From zucco at dipteris.unige.it Fri Sep 12 16:00:06 2008 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Fri, 12 Sep 2008 16:00:06 +0200 Subject: [Pw_forum] Runtime error in vc-relax Message-ID: Hello to all users, I'm trying to do a variable cell relaxation of the iron carbonate sent to the forum few days ago, with v 4.0.2 and 4.1CVS. Just after the computation of the stress matrix of the second step this error appear on the console, and the job crash suddenly: At line 151 of file vcsmd.f90 Fortran runtime error: Range error during floating point read It seems that it is not able to read a file generated by itself... Does anyone can say what is happening? Thanks to all m. ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From giannozz at democritos.it Fri Sep 12 16:03:22 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 12 Sep 2008 16:03:22 +0200 Subject: [Pw_forum] Can anyone help me explain what do the terms in *.fc file means? In-Reply-To: <1221227273.48ca7309639ff@webmail.purdue.edu> References: <1221227273.48ca7309639ff@webmail.purdue.edu> Message-ID: <48CA76AA.60608@democritos.it> Zhen Huang wrote: > What does the output mean in the gra.fc file. it means nothing: none of your business, as long as there are no error messages. Once force constants are calculated, you use them via the matdyn.x code. > 1 1 1 1 (what do the indexes represent? 1 for what) > 1 1 1 2.24617477868E-01 (what is this value and in what unit?) what did you do instead of readinfg my previous message? -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Fri Sep 12 16:33:12 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 12 Sep 2008 16:33:12 +0200 Subject: [Pw_forum] Runtime error in vc-relax In-Reply-To: References: Message-ID: <48CA7DA8.3030506@democritos.it> Marino Vetuschi Zuccolini wrote: > At line 151 of file vcsmd.f90 > Fortran runtime error: Range error during floating point read this is line 151 of file vcsmd.f90: READ( 4, * ) rati, ratd, rat2d, rat2di, tauold Unit 4 should be "siderite.opt.md" (I am assuming that you are running the sederite_espresso job you posted a few days ago). Look in that file - it is formatted and quite small - for anthing looking suspicious: NaN, missing spaces between numbers, etc. > It seems that it is not able to read a file generated by itself... once upon a time, on a Cray machine, read(5,*) didn't accept the format N*value that write(6,*) used when N consecuting variables had the same value. Really! Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Fri Sep 12 17:16:47 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 12 Sep 2008 17:16:47 +0200 Subject: [Pw_forum] Runtime error in vc-relax In-Reply-To: References: Message-ID: <48CA87DF.60504@democritos.it> ...and by the way: - postprocessing with the "tqr" option currently doesn't work (can be easily fixed, though) - davidson is much faster than conjugate-gradient: don't use the latter unless you have a good reason - starting_magnetization doesn't work the way you are using it: its index refers to the species, not to the atoms (I think it is ignored anyway with tot_magnetization) - self-consistency should be faster with a small broadening Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From zucco at dipteris.unige.it Fri Sep 12 17:38:31 2008 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Fri, 12 Sep 2008 17:38:31 +0200 Subject: [Pw_forum] Runtime error in vc-relax In-Reply-To: <48CA87DF.60504@democritos.it> References: <48CA87DF.60504@democritos.it> Message-ID: <3DE87689-B3EC-486C-BAEA-B59D451D536A@dipteris.unige.it> Yes, thank you very much Paolo. Everything you remember me below has been, little by little introduced after a better comprehension of the keywords (starting_magnetization were a non-sense) and tqr was a very strange solution to all other inconsistencies... The file I'm using now is a little bit different (I hope without too much errors) and after a control of what will be written in Unit 4 I will submit to the forum to try to solve the problem if persists. On 12/set/08, at 17:16, Paolo Giannozzi wrote: > ...and by the way: > - postprocessing with the "tqr" option currently doesn't work > (can be easily fixed, though) > - davidson is much faster than conjugate-gradient: don't use the > latter unless you have a good reason It seems that cg with atomic+random wfcn is much more reliable and the convergence path very stable...but the floating error avoid the good ending. > > - starting_magnetization doesn't work the way you are using it: > its index refers to the species, not to the atoms (I think it > is ignored anyway with tot_magnetization) > - self-consistency should be faster with a small broadening > > Grazie marino > 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 ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From baroni at sissa.it Sat Sep 13 12:04:19 2008 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 13 Sep 2008 12:04:19 +0200 Subject: [Pw_forum] Fw: relax is too slow In-Reply-To: References: <48CA3610.1070704@democritos.it> Message-ID: <9493BC1C-FC16-47FC-8AB2-4DC6D3497E9F@sissa.it> the same way as you would choose any other parameter controlling the accuracy of a numerical calculation. try a value; try a few others corresponding to increasing accuracy, and stop when you see that the increase in the accuracy is not worth the extra workload. never forget to use common sense (whenever it is available). SB On Sep 12, 2008, at 11:36 AM, vega wrote: > Dear sir >> vega wrote: >> >>> As far as I tested, the energy would be within 1% for large >>> supercells, >>> but the coordinates of several atoms may differ 200%. >> >> doesn't make sense > > could you tell me more about how to choose the k point mesh? > > vega > = > = > = > = > = > = > = > = > = > = > = > ====================================================================== > Vega Lew (weijia liu) > PH.D Candidate in Chemical Engineering > State Key Laboratory of Materials-oriented Chemical Engineering > College of Chemistry and Chemical Engineering > Nanjing University of Technology, 210009, Nanjing, Jiangsu, China > > -------------------------------------------------- > From: "Paolo Giannozzi" > Sent: Friday, September 12, 2008 5:27 PM > To: "PWSCF Forum" > Subject: Re: [Pw_forum] Fw: relax is too slow > > >> -- >> 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 >> > _______________________________________________ > 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 http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080913/cfb996b3/attachment.htm From vegalew at hotmail.com Sat Sep 13 14:20:26 2008 From: vegalew at hotmail.com (vega) Date: Sat, 13 Sep 2008 20:20:26 +0800 Subject: [Pw_forum] Fw: relax is too slow In-Reply-To: <48CA3610.1070704@democritos.it> <9493BC1C-FC16-47FC-8AB2-4DC6D3497E9F@sissa.it> References: <48CA3610.1070704@democritos.it> <9493BC1C-FC16-47FC-8AB2-4DC6D3497E9F@sissa.it> Message-ID: Dear sir, Thank you for your responding. Could you tell me what is the most important thing to judge whether the accuracy of a numerical calculation is enough. The energy? force? atomic coordinates? or any other things? vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: Stefano Baroni Sent: Saturday, September 13, 2008 6:04 PM To: PWSCF Forum Subject: Re: [Pw_forum] Fw: relax is too slow the same way as you would choose any other parameter controlling the accuracy of a numerical calculation. try a value; try a few others corresponding to increasing accuracy, and stop when you see that the increase in the accuracy is not worth the extra workload. never forget to use common sense (whenever it is available). SB On Sep 12, 2008, at 11:36 AM, vega wrote: Dear sir vega wrote: As far as I tested, the energy would be within 1% for large supercells, but the coordinates of several atoms may differ 200%. doesn't make sense could you tell me more about how to choose the k point mesh? vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------------------------------------------- From: "Paolo Giannozzi" Sent: Friday, September 12, 2008 5:27 PM To: "PWSCF Forum" Subject: Re: [Pw_forum] Fw: relax is too slow -- 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 _______________________________________________ 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 http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080913/6da81d92/attachment.htm From baroni at sissa.it Sat Sep 13 14:40:27 2008 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 13 Sep 2008 14:40:27 +0200 Subject: [Pw_forum] Fw: relax is too slow In-Reply-To: References: <48CA3610.1070704@democritos.it> <9493BC1C-FC16-47FC-8AB2-4DC6D3497E9F@sissa.it> Message-ID: <9A8A1DD6-1EF3-4991-A59E-83C19C6286D4@sissa.it> On Sep 13, 2008, at 2:20 PM, vega wrote: > Dear sir, > > Thank you for your responding. > > Could you tell me what is the most important thing to judge whether > the accuracy of a numerical calculation is enough. > The energy? force? atomic coordinates? or any other things? It depends on what you are interested in, of course. What you are interested in (geometry, energy, other properties ...) is what you should look at ... for instance, it is known that obtaining well converged values for the dielectric constant may require k-point meshes way finer than required for a plain energy calculation. Here as well as in many other circumstances, the correct answer is not one that the beginner may like to hear: "it depends" ... life is beatiful and hard. if it were not as hard, it would not be as beautiful cheers - SB > > vega > > = > = > = > = > = > = > = > = > = > = > = > ====================================================================== > Vega Lew (weijia liu) > PH.D Candidate in Chemical Engineering > State Key Laboratory of Materials-oriented Chemical Engineering > College of Chemistry and Chemical Engineering > Nanjing University of Technology, 210009, Nanjing, Jiangsu, China > > From: Stefano Baroni > Sent: Saturday, September 13, 2008 6:04 PM > To: PWSCF Forum > Subject: Re: [Pw_forum] Fw: relax is too slow > > the same way as you would choose any other parameter controlling the > accuracy of a numerical calculation. try a value; try a few others > corresponding to increasing accuracy, and stop when you see that the > increase in the accuracy is not worth the extra workload. never > forget to use common sense (whenever it is available). > > SB > > On Sep 12, 2008, at 11:36 AM, vega wrote: > >> Dear sir >>> vega wrote: >>> >>>> As far as I tested, the energy would be within 1% for large >>>> supercells, >>>> but the coordinates of several atoms may differ 200%. >>> >>> doesn't make sense >> >> could you tell me more about how to choose the k point mesh? >> >> vega >> = >> = >> = >> = >> = >> = >> = >> = >> = >> = >> = >> = >> ===================================================================== >> Vega Lew (weijia liu) >> PH.D Candidate in Chemical Engineering >> State Key Laboratory of Materials-oriented Chemical Engineering >> College of Chemistry and Chemical Engineering >> Nanjing University of Technology, 210009, Nanjing, Jiangsu, China >> >> -------------------------------------------------- >> From: "Paolo Giannozzi" >> Sent: Friday, September 12, 2008 5:27 PM >> To: "PWSCF Forum" >> Subject: Re: [Pw_forum] Fw: relax is too slow >> >> >>> -- >>> 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 >>> >> _______________________________________________ >> 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 > http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / > stefanobaroni (skype) > > La morale est une logique de l'action comme la logique est une > morale de la pens?e - Jean Piaget > > Please, if possible, don't send me MS Word or PowerPoint attachments > Why? See: http://www.gnu.org/philosophy/no-word-attachments.html > > > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080913/4d7330a7/attachment-0001.htm From vegalew at hotmail.com Sat Sep 13 15:08:41 2008 From: vegalew at hotmail.com (vega) Date: Sat, 13 Sep 2008 21:08:41 +0800 Subject: [Pw_forum] Fw: relax is too slow In-Reply-To: <48CA3610.1070704@democritos.it><9493BC1C-FC16-47FC-8AB2-4DC6D3497E9F@sissa.it> <9A8A1DD6-1EF3-4991-A59E-83C19C6286D4@sissa.it> References: <48CA3610.1070704@democritos.it><9493BC1C-FC16-47FC-8AB2-4DC6D3497E9F@sissa.it> <9A8A1DD6-1EF3-4991-A59E-83C19C6286D4@sissa.it> Message-ID: Dear sir, Thank you. Now, I'm better understanding on this issue. vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: Stefano Baroni Sent: Saturday, September 13, 2008 8:40 PM To: PWSCF Forum Subject: Re: [Pw_forum] Fw: relax is too slow On Sep 13, 2008, at 2:20 PM, vega wrote: Dear sir, Thank you for your responding. Could you tell me what is the most important thing to judge whether the accuracy of a numerical calculation is enough. The energy? force? atomic coordinates? or any other things? It depends on what you are interested in, of course. What you are interested in (geometry, energy, other properties ...) is what you should look at ... for instance, it is known that obtaining well converged values for the dielectric constant may require k-point meshes way finer than required for a plain energy calculation. Here as well as in many other circumstances, the correct answer is not one that the beginner may like to hear: "it depends" ... life is beatiful and hard. if it were not as hard, it would not be as beautiful cheers - SB vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: Stefano Baroni Sent: Saturday, September 13, 2008 6:04 PM To: PWSCF Forum Subject: Re: [Pw_forum] Fw: relax is too slow the same way as you would choose any other parameter controlling the accuracy of a numerical calculation. try a value; try a few others corresponding to increasing accuracy, and stop when you see that the increase in the accuracy is not worth the extra workload. never forget to use common sense (whenever it is available). SB On Sep 12, 2008, at 11:36 AM, vega wrote: Dear sir vega wrote: As far as I tested, the energy would be within 1% for large supercells, but the coordinates of several atoms may differ 200%. doesn't make sense could you tell me more about how to choose the k point mesh? vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------------------------------------------- From: "Paolo Giannozzi" Sent: Friday, September 12, 2008 5:27 PM To: "PWSCF Forum" Subject: Re: [Pw_forum] Fw: relax is too slow -- 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 _______________________________________________ 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 http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html ------------------------------------------------------------------------------ _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------------------------------------------------------------------------- _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080913/c498dea3/attachment.htm From akohlmey at cmm.chem.upenn.edu Sat Sep 13 16:04:43 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 13 Sep 2008 10:04:43 -0400 (EDT) Subject: [Pw_forum] error in compiling Quantum Espresso In-Reply-To: References: Message-ID: On Fri, 12 Sep 2008, masoud nahali wrote: MN> Dear Users MN> MN> error in compiling Quantum Espresso: MN> MN> ./configure was successful MN> but after using make all I have received this: MN> MN> fortcom: server: ** internal compiler error :segmentation violation signal MN> raised ** please report this error along with the circumstances in which it MN> occured in a software problem report. note : file and line given may not be MN> explicit cause this error. an "internal compiler error", is (by definition) an error of the _compiler_, not the compiled software. so you should contact the compiler vendor and or the respective support forum. MN> compilation aborted for sum_band.f90 MN> make[1]: *** [sum_band.o] error 3 MN> make[1]: living directory /home/quantum espresso/. . /PW MN> make: *** [PW} error 2 MN> MN> MN> my compiler is "ifort 9.0" and I have installed "MKL 9.0" so. c++ has been MN> installed too. MN> would you please help me? you have to help yourself to a correctly working version of the compiler. cheers, axel. MN> MN> many thanks. MN> MN> MN> MN> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From tone.kokalj at ijs.si Sat Sep 13 21:15:56 2008 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Sat, 13 Sep 2008 21:15:56 +0200 Subject: [Pw_forum] How to plot the 2D-charge density for a triclinic or monoclinic box In-Reply-To: References: Message-ID: <1221333356.32134.6.camel@walk.ijs.si> On Mon, 2008-09-08 at 19:39 +0800, vega wrote: > Dear all, > > I need to plot 2D charge density for a triclinic or monoclinic box. > while the section plane along the c-axis is not a rectangle but a > parallelogram. so the vector of the section plane e1 and e2 is not > orthogonal. how could I specify the e1(1)~e1(3) and e2(1)~e(3) in this > case. Quite simple, pick one vector, say e1, and make the second one, say e2, orthogonal to it, see http://en.wikipedia.org/wiki/Gram%E2%80%93Schmidt_process Then you may scale the so obtained vector to suit your needs. Regards, Tone -- Tone Kokalj J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia (tel: +386-1-477-3523 // fax:+386-1-477-3822) From tone.kokalj at ijs.si Sat Sep 13 21:28:49 2008 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Sat, 13 Sep 2008 21:28:49 +0200 Subject: [Pw_forum] About the atomic radii used for visuliztion of a structre. In-Reply-To: <1039503785.20080908172815@yahoo.com.cn> References: <1039503785.20080908172815@yahoo.com.cn> Message-ID: <1221334129.32134.19.camel@walk.ijs.si> On Mon, 2008-09-08 at 17:28 +0800, Hongsheng Zhao wrote: > My issue is as follows: > > What's relationship between the atomic radii used for visuliztion of a > structre and the atomic real radii of the atoms/ions? This issue arise > from the fact that I find many crystal visuliztion tools, i.e., diamond, > rasmol, and etc., don't treat the atomic radii according to the real radii > of the atoms/ions belong to the structure, never so much as proportionate > to the counterpart real radii. What should we choose for the real atomic radius: atomic, covalent, ionic, wan der Waals, ... Usually the vizualizers do plot atoms according to some preset radii, but there are many types of radii. Say that they are plotted according to covalent radii, and that you are plotting some ionic crystal. The radii will not be what you expect ... Of course you know it is ionic, but the dummy visualizer do not, and merely plots the radii from preselected table. Some programs do have the possibility to choose among different types of radii, and even to set the radii manually. Regards, Tone -- Tone Kokalj J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia (tel: +386-1-477-3523 // fax:+386-1-477-3822) From szs_naghavi at yahoo.com Sat Sep 13 21:54:07 2008 From: szs_naghavi at yahoo.com (zahra sadat naghavi) Date: Sat, 13 Sep 2008 12:54:07 -0700 (PDT) Subject: [Pw_forum] HELP ME! Message-ID: <121226.12935.qm@web63005.mail.re1.yahoo.com> I am mixed by ?? Program PHONON??? v.3.2.3? starts ... ???? Today is 13Sep2008 at 20:16: 5 ???? Parallel version (MPI) ???? Number of processors in use:????? 16 ???? R & G space division:? proc/pool =?? 16 ???? Ultrasoft (Vanderbilt) Pseudopotentials ?%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ???? from phq_readin : error #???????? 1 ???? reading inputph namelist ?%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ???? stopping ... can anybody help?this is my ph.in: &inputph ? tr2_ph=1.0d-16, ? prefix='CNT6', ? ldis=.true., ? nq1=1 nq2=1 nq3=8, ? amass(1)=12.0107 ? outdir='/scratch19/test3/scratch', ? fildyn='cnt.dyn', ?/ ? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080913/3cc6625a/attachment.htm From baroni at sissa.it Sat Sep 13 23:00:56 2008 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 13 Sep 2008 23:00:56 +0200 Subject: [Pw_forum] HELP ME! In-Reply-To: <121226.12935.qm@web63005.mail.re1.yahoo.com> References: <121226.12935.qm@web63005.mail.re1.yahoo.com> Message-ID: <929D9380-7530-4B41-A0FE-CE1C706FD78A@sissa.it> On Sep 13, 2008, at 9:54 PM, zahra sadat naghavi wrote: > I am mixed by > Program PHONON v.3.2.3 starts ... > Today is 13Sep2008 at 20:16: 5 > Parallel version (MPI) > Number of processors in use: 16 > R & G space division: proc/pool = 16 > Ultrasoft (Vanderbilt) Pseudopotentials > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > %%%%%%%%%% > from phq_readin : error # 1 > reading inputph namelist > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > %%%%%%%%%% > stopping ... > can anybody help?this is my ph.in: > &inputph > tr2_ph=1.0d-16, > prefix='CNT6', > ldis=.true., > nq1=1 nq2=1 nq3=8, not quite sure, but it seems to me that there are a few commas missing, here ... SB > > amass(1)=12.0107 > outdir='/scratch19/test3/scratch', > fildyn='cnt.dyn', > / > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080913/310c8205/attachment.htm From nhviet at sissa.it Sat Sep 13 23:06:45 2008 From: nhviet at sissa.it (Viet Huy Nguyen) Date: Sat, 13 Sep 2008 23:06:45 +0200 Subject: [Pw_forum] HELP ME! In-Reply-To: <121226.12935.qm@web63005.mail.re1.yahoo.com> References: <121226.12935.qm@web63005.mail.re1.yahoo.com> Message-ID: On Sat, Sep 13, 2008 at 9:54 PM, zahra sadat naghavi > * > * > ldis=.true., > I think it should be ldisp instead of ldis. Huy Viet -- Mr. Viet Huy Nguyen PhD student, Condensed Matter Theory Sector SISSA/ISAS Via Beirut 2-4, 34014, Trieste, Italy Tel: +39 040 3787 492 Fax: +39 040 3787 528 Email: nhviet at sissa.it hviet.nguyen at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080913/876e7a0f/attachment.htm From rajanpnano at gmail.com Sun Sep 14 02:01:00 2008 From: rajanpnano at gmail.com (Raj Pandey) Date: Sun, 14 Sep 2008 05:31:00 +0530 Subject: [Pw_forum] Error while calculationg complex band structure of silicon Message-ID: <48b8698d0809131701s6ce0103pa8d2b78727f9ca07@mail.gmail.com> Dear pwscf developers and users, I am trying to calculate complex band structure of silicon using espresso-4.0.2. The output returns with the error %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from init_orbitals : error # 1 Too big atomic spheres %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... Below is the input for scf calculation as in example 05: &control calculation='scf' restart_mode='from_scratch', prefix='si' pseudo_dir = '/home/rajan/espresso-4.0.2_P/pseudo/', outdir='/home/rajan/tmp/' / &system ibrav= 2, celldm(1)= 10.2, nat= 2, ntyp= 1, ecutwfc =18.0 / &electrons conv_thr = 1.0d-8 mixing_beta = 0.7 / ATOMIC_SPECIES Si 28.086 Si.vbc.UPF ATOMIC_POSITIONS Si 0.00 0.00 0.00 Si 0.25 0.25 0.25 K_POINTS 10 0.1250000 0.1250000 0.1250000 1.00 0.1250000 0.1250000 0.3750000 3.00 0.1250000 0.1250000 0.6250000 3.00 0.1250000 0.1250000 0.8750000 3.00 0.1250000 0.3750000 0.3750000 3.00 0.1250000 0.3750000 0.6250000 6.00 0.1250000 0.3750000 0.8750000 6.00 0.1250000 0.6250000 0.6250000 3.00 0.3750000 0.3750000 0.3750000 1.00 0.3750000 0.3750000 0.6250000 3.00 and the input to calculate the complex band structure is &inputcond outdir = '/home/rajan/tmp/' prefixl = 'si' band_file = 'bands.si' ikind = 0 energy0 = 10.d0 denergy = -0.4d0 ewind = 3.d0 epsproj = 1.d-7 delgep = 1.d-10 cutplot = 3.d0 / 1 0.0 0.0 1.0 60 The output returns with the following error: Program POST-PROC v.4.0.2 starts ... Today is 14Sep2008 at 5:20:33 Parallel version (MPI) Number of processors in use: 1 Planes per process (thick) : nr3 = 20 npp = 20 ncplane = 400 Proc/ planes cols G planes cols G columns G Pool (dense grid) (smooth grid) (wavefct grid) 1 20 253 2733 20 253 2733 91 609 ===== INPUT FILE containing the left lead ===== GEOMETRY: lattice parameter (a_0) = 10.2000 a.u. the volume = 265.3020 (a.u.)^3 the cross section = 26.0100 (a.u.)^2 l of the unit cell = 0.0000 (a_0) number of atoms/cell = 2 number of atomic types = 1 crystal axes: (cart. coord. in units of a_0) a(1) = ( -0.5000 0.0000 0.5000 ) a(2) = ( 0.0000 0.5000 0.5000 ) a(3) = ( -0.5000 0.5000 0.0000 ) Cartesian axes site n. atom positions (a_0 units) 1 Si tau( 1)=( 0.0000 0.0000 0.0000 ) 2 Si tau( 2)=( 0.2500 0.2500 0.2500 ) nr1s = 20 nr2s = 20 nr3s = 20 nrx1s = 20 nrx2s = 20 nrx3s = 20 nr1 = 20 nr2 = 20 nr3 = 20 nrx1 = 20 nrx2 = 20 nrx3 = 20 _______________________________ Radii of nonlocal spheres: type ibeta ang. mom. radius (a_0 units) Si 1 0 0.2832 Si 2 1 0.2978 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from init_orbitals : error # 1 Too big atomic spheres %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... I appreciate any help/suggestion. Thanks, Rajan -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080914/a73f59f7/attachment.htm From manoj at phys.ufl.edu Sun Sep 14 04:40:28 2008 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Sat, 13 Sep 2008 22:40:28 -0400 (EDT) Subject: [Pw_forum] Error while calculationg complex band structure of silicon In-Reply-To: <48b8698d0809131701s6ce0103pa8d2b78727f9ca07@mail.gmail.com> Message-ID: Dear Raj, I guess this problem is arising from not choosing the correct unit cell, I had similar problem too, but it got fixed. you should probably double check your unit cell and the atomic positions. Cheers, Manoj Srivastava On Sun, 14 Sep 2008, Raj Pandey wrote: > Dear pwscf developers and users, > > I am trying to calculate complex band structure of silicon using > espresso-4.0.2. The output returns with the > error > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from init_orbitals : error # 1 > Too big atomic spheres > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > Below is the input for scf calculation as in example 05: > > &control > calculation='scf' > restart_mode='from_scratch', > prefix='si' > pseudo_dir = '/home/rajan/espresso-4.0.2_P/pseudo/', > outdir='/home/rajan/tmp/' > / > &system > ibrav= 2, celldm(1)= 10.2, nat= 2, ntyp= 1, > ecutwfc =18.0 > / > &electrons > conv_thr = 1.0d-8 > mixing_beta = 0.7 > / > ATOMIC_SPECIES > Si 28.086 Si.vbc.UPF > ATOMIC_POSITIONS > Si 0.00 0.00 0.00 > Si 0.25 0.25 0.25 > K_POINTS > 10 > 0.1250000 0.1250000 0.1250000 1.00 > 0.1250000 0.1250000 0.3750000 3.00 > 0.1250000 0.1250000 0.6250000 3.00 > 0.1250000 0.1250000 0.8750000 3.00 > 0.1250000 0.3750000 0.3750000 3.00 > 0.1250000 0.3750000 0.6250000 6.00 > 0.1250000 0.3750000 0.8750000 6.00 > 0.1250000 0.6250000 0.6250000 3.00 > 0.3750000 0.3750000 0.3750000 1.00 > 0.3750000 0.3750000 0.6250000 3.00 > > and the input to calculate the complex band structure is > > &inputcond > outdir = '/home/rajan/tmp/' > prefixl = 'si' > band_file = 'bands.si' > ikind = 0 > energy0 = 10.d0 > denergy = -0.4d0 > ewind = 3.d0 > epsproj = 1.d-7 > delgep = 1.d-10 > cutplot = 3.d0 > / > 1 > 0.0 0.0 1.0 > 60 > > The output returns with the following error: > > Program POST-PROC v.4.0.2 starts ... > Today is 14Sep2008 at 5:20:33 > > Parallel version (MPI) > > Number of processors in use: 1 > > Planes per process (thick) : nr3 = 20 npp = 20 ncplane = 400 > > Proc/ planes cols G planes cols G columns G > Pool (dense grid) (smooth grid) (wavefct grid) > 1 20 253 2733 20 253 2733 91 609 > > > ===== INPUT FILE containing the left lead ===== > > GEOMETRY: > > lattice parameter (a_0) = 10.2000 a.u. > the volume = 265.3020 (a.u.)^3 > the cross section = 26.0100 (a.u.)^2 > l of the unit cell = 0.0000 (a_0) > number of atoms/cell = 2 > number of atomic types = 1 > > crystal axes: (cart. coord. in units of a_0) > a(1) = ( -0.5000 0.0000 0.5000 ) > a(2) = ( 0.0000 0.5000 0.5000 ) > a(3) = ( -0.5000 0.5000 0.0000 ) > > Cartesian axes > > site n. atom positions (a_0 units) > 1 Si tau( 1)=( 0.0000 0.0000 0.0000 ) > 2 Si tau( 2)=( 0.2500 0.2500 0.2500 ) > > nr1s = 20 > nr2s = 20 > nr3s = 20 > nrx1s = 20 > nrx2s = 20 > nrx3s = 20 > nr1 = 20 > nr2 = 20 > nr3 = 20 > nrx1 = 20 > nrx2 = 20 > nrx3 = 20 > > _______________________________ > Radii of nonlocal spheres: > > type ibeta ang. mom. radius (a_0 units) > Si 1 0 0.2832 > Si 2 1 0.2978 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from init_orbitals : error # 1 > Too big atomic spheres > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > > I appreciate any help/suggestion. > > Thanks, > > Rajan > From zhaohscas at yahoo.com.cn Sun Sep 14 05:50:26 2008 From: zhaohscas at yahoo.com.cn (Hongsheng Zhao) Date: Sun, 14 Sep 2008 11:50:26 +0800 Subject: [Pw_forum] How to plot the 2D-charge density for a triclinic or monoclinic box In-Reply-To: <1221333356.32134.6.camel@walk.ijs.si> References: <1221333356.32134.6.camel@walk.ijs.si> Message-ID: <3510209594.20080914115026@yahoo.com.cn> On Sunday, September 14, 2008 at 3:15, tone.kokalj at ijs.si wrote: > On Mon, 2008-09-08 at 19:39 +0800, vega wrote: >> Dear all, >> >> I need to plot 2D charge density for a triclinic or monoclinic box. >> while the section plane along the c-axis is not a rectangle but a >> parallelogram. so the vector of the section plane e1 and e2 is not >> orthogonal. how could I specify the e1(1)~e1(3) and e2(1)~e(3) in this >> case. > Quite simple, pick one vector, say e1, and make the second one, say e2, > orthogonal to it, see > http://en.wikipedia.org/wiki/Gram%E2%80%93Schmidt_process Thanks for this hints. > Then you may scale the so obtained vector to suit your needs. I cann't figure can the meanning and the process of *scale the so obtained vector to suit your needs*. In detail, in order to plot 2D charge density for a triclinic or monoclinic box on a arbitrary plane, the xcrysden will use the two span vectors to construct the parallelogram. In the above case, we can *scale* from the e1 and e2 to obtain the the corresponding two span vectors easily, but for the 2D charge density using by xcrysden, how can I *scale* from those obtained by e1 and e2 to suit the xcrysden's needs within the plane decided by the two span vectors? > Regards, Tone Sincerely yours, -- Hongsheng Zhao Xinjiang Technical Institute of Physics and Chemistry Chinese Academy of Sciences GnuPG DSA: 0xD108493 2008-9-14 From rajanpnano at gmail.com Sun Sep 14 06:00:31 2008 From: rajanpnano at gmail.com (Raj Pandey) Date: Sun, 14 Sep 2008 09:30:31 +0530 Subject: [Pw_forum] On error while calculating complex band structure of silicon Message-ID: <48b8698d0809132100j2862faefwda7bc4360075ce6e@mail.gmail.com> Dear Manoj, Thanks for your reply. But I am sorry, I did not quite get you. I am able to compute real band structure of Si from the input of example05. Similar to bands.x (which gives real band structure), pwcond.x is a post-processor (gives complex band structure). Therefore the question is that if I can get the real band structure then why can't the complex band structure for the same unit cell? Perhaps, something is wrong with my input file for complex band which I can't figure out! Thanks, Rajan -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080914/d1a7d034/attachment.htm From tone.kokalj at ijs.si Sun Sep 14 12:02:44 2008 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Sun, 14 Sep 2008 12:02:44 +0200 Subject: [Pw_forum] How to plot the 2D-charge density for a triclinic or monoclinic box In-Reply-To: <3510209594.20080914115026@yahoo.com.cn> References: <1221333356.32134.6.camel@walk.ijs.si> <3510209594.20080914115026@yahoo.com.cn> Message-ID: <1221386564.10695.9.camel@walk.ijs.si> On Sun, 2008-09-14 at 11:50 +0800, Hongsheng Zhao wrote: > In the above case, we can *scale* from the e1 and e2 to obtain the the corresponding two > span vectors easily, but for the 2D charge density using by xcrysden, > how can I *scale* from those obtained by e1 and e2 to suit the > xcrysden's needs within the plane decided by the two span vectors? OK, I see your point. These are two distinct issues. (i) you can calculate 2D plane using the pp.x program. This requires the orthogonal spanning vectors + origin. (my yesterday emails referred to this). When you manually calculate the two orthogonal vectors, then you need to choose the origin + the size of the two vectors (the "scaling" referred to that). (ii) you may calculate the whole 3D unit cell by pp.x, and then xcrysden can plot the three basal planes out of it (ab,ac,bc), and here ortogonality doesn't metter. With xcrysden, all you can do further is to "clone" these planes with periodic images so as to obtain plot on a larger area (use "Expand" tab for this). Regards, Tone -- Tone Kokalj J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia (tel: +386-1-477-3523 // fax:+386-1-477-3822) From zhaohscas at yahoo.com.cn Sun Sep 14 13:03:01 2008 From: zhaohscas at yahoo.com.cn (Hongsheng Zhao) Date: Sun, 14 Sep 2008 19:03:01 +0800 Subject: [Pw_forum] How to plot the 2D-charge density for a triclinic or monoclinic box In-Reply-To: <1221386564.10695.9.camel@walk.ijs.si> References: <1221333356.32134.6.camel@walk.ijs.si> <3510209594.20080914115026@yahoo.com.cn> <1221386564.10695.9.camel@walk.ijs.si> Message-ID: <1943302099.20080914190301@yahoo.com.cn> On Sunday, September 14, 2008 at 18:02, tone.kokalj at ijs.si wrote: > On Sun, 2008-09-14 at 11:50 +0800, Hongsheng Zhao wrote: >> In the above case, we can *scale* from the e1 and e2 to obtain the the corresponding two >> span vectors easily, but for the 2D charge density using by xcrysden, >> how can I *scale* from those obtained by e1 and e2 to suit the >> xcrysden's needs within the plane decided by the two span vectors? > OK, I see your point. These are two distinct issues. > (i) you can calculate 2D plane using the pp.x program. This requires the > orthogonal spanning vectors + origin. (my yesterday emails referred to > this). When you manually calculate the two orthogonal vectors, then you > need to choose the origin + the size of the two vectors (the "scaling" > referred to that). > (ii) you may calculate the whole 3D unit cell by pp.x, and then xcrysden > can plot the three basal planes out of it (ab,ac,bc), and here > ortogonality doesn't metter. With xcrysden, all you can do further is to > "clone" these planes with periodic images so as to obtain plot on a > larger area (use "Expand" tab for this). I see your point too :-) IMO, in order to obtain the 2D charge density for arbitrary plane with the above method by using xcrysden, use a the whole 3D unit cell is unnecessary. According to the above description given by you, the "Expand" tab can also give the adequate large plot for the specific plane constructed by pp.x. am I right? Sincerely yours, -- Hongsheng Zhao Xinjiang Technical Institute of Physics and Chemistry Chinese Academy of Sciences GnuPG DSA: 0xD108493 2008-9-14 From tone.kokalj at ijs.si Sun Sep 14 13:16:18 2008 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Sun, 14 Sep 2008 13:16:18 +0200 Subject: [Pw_forum] How to plot the 2D-charge density for a triclinic or monoclinic box In-Reply-To: <1943302099.20080914190301@yahoo.com.cn> References: <1221333356.32134.6.camel@walk.ijs.si> <3510209594.20080914115026@yahoo.com.cn> <1221386564.10695.9.camel@walk.ijs.si> <1943302099.20080914190301@yahoo.com.cn> Message-ID: <1221390978.10695.18.camel@walk.ijs.si> On Sun, 2008-09-14 at 19:03 +0800, Hongsheng Zhao wrote: > > IMO, in order to obtain the 2D charge density for arbitrary plane with > the above method by using xcrysden, use a the whole 3D unit cell is > unnecessary. According to the above description given by you, the > "Expand" tab can also give the adequate large plot for the specific > plane constructed by pp.x. am I right? Yes. However the "expand" can be used for the whole unit cell 3D pp.x plot: then you can plot the three basal planes and expand them. However if you calculate 2D by pp.x, then "expand" is not available (because an arbitrary chosen rectangle does not fulfill translational symmetry of crystal). So you need to chose proper origin + size of rectange. But you can specify an arbitrary oriented plane! Regards, Tone -- Tone Kokalj J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia (tel: +386-1-477-3523 // fax:+386-1-477-3822) From zhaohscas at yahoo.com.cn Sun Sep 14 14:30:09 2008 From: zhaohscas at yahoo.com.cn (Hongsheng Zhao) Date: Sun, 14 Sep 2008 20:30:09 +0800 Subject: [Pw_forum] How to plot the 2D-charge density for a triclinic or monoclinic box In-Reply-To: <1221386564.10695.9.camel@walk.ijs.si> References: <1221333356.32134.6.camel@walk.ijs.si> <3510209594.20080914115026@yahoo.com.cn> <1221386564.10695.9.camel@walk.ijs.si> Message-ID: <391493881.20080914203009@yahoo.com.cn> On Sunday, September 14, 2008 at 18:02, tone.kokalj at ijs.si wrote: > On Sun, 2008-09-14 at 11:50 +0800, Hongsheng Zhao wrote: >> In the above case, we can *scale* from the e1 and e2 to obtain the the corresponding two >> span vectors easily, but for the 2D charge density using by xcrysden, >> how can I *scale* from those obtained by e1 and e2 to suit the >> xcrysden's needs within the plane decided by the two span vectors? > OK, I see your point. These are two distinct issues. > (i) you can calculate 2D plane using the pp.x program. This requires the > orthogonal spanning vectors + origin. (my yesterday emails referred to > this). When you manually calculate the two orthogonal vectors, then you > need to choose the origin + the size of the two vectors (the "scaling" > referred to that). > (ii) you may calculate the whole 3D unit cell by pp.x, and then xcrysden > can plot the three basal planes out of it (ab,ac,bc), and here > ortogonality doesn't metter. With xcrysden, all you can do further is to > "clone" these planes with periodic images so as to obtain plot on a > larger area (use "Expand" tab for this). Due to the pp.x program require that all of the three span vectors must be orthogonal to each other, so this will induce the 3D unit cell by pp.x with the three span vectors determined by Gram?CSchmidt process not always can be expands to the corresponding crystallographic cell. I mean, the "Expand" tab of xcrysden will not always give the corresponding three basal crystallographic cell planes by using the the abobe three span vectors given by Gram?CSchmidt process. Am I right? Sincerely yours, -- Hongsheng Zhao Xinjiang Technical Institute of Physics and Chemistry Chinese Academy of Sciences GnuPG DSA: 0xD108493 2008-9-14 From w2agz at pacbell.net Mon Sep 15 05:00:33 2008 From: w2agz at pacbell.net (Paul M. Grant) Date: Sun, 14 Sep 2008 20:00:33 -0700 Subject: [Pw_forum] Units of Charge Density - Again Message-ID: <01ff01c916df$38fc0b90$aaf422b0$@net> There have been a number of posts to pw_forum regarding the units employed for charge density throughout PWscf package and XcrySDen. This is an important issue, and I still find the situation and answers provided in the forum archives ambiguous. Before slogging through relevant source codes, and testing by writing time-consuming homemade examples, I'm hoping someone in the community has preceded me. In a perfect SI world, these units would be coulombs per cubic meter. In a practical, or rather intuitive, world, they would be in units of total fractional number of electrons per unit cell volume (in cubic angstroms or nanometers or bohrs (ugh)). So.what is it? And does it depend on the division of states between core and valence taken when constructing a given pseudopotential? Finally, I've been unable (by visual inspection) to determine from the documentation for XcrySDen (a marvelous tool.marvelous), exactly how the "rho grid input" is scaled, especially regarding the computation of the 3D min-max grid values, the "isovalue".and.particularly the mysterious "delta n(r)" displayed in the temperature inset box. Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford (2005-2008) EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080914/c820a4ea/attachment.htm From w2agz at pacbell.net Mon Sep 15 05:30:02 2008 From: w2agz at pacbell.net (Paul M. Grant) Date: Sun, 14 Sep 2008 20:30:02 -0700 Subject: [Pw_forum] k-point grids...and all that Message-ID: <020401c916e3$571a2bd0$054e8370$@net> This post is mainly a question regarding available resources in "o-sesame CVS or espresso-4.0" enabling calculation of "k-point," or "k-space" volumetric grids within the first Brillouin Zone of a given real space structure for use in the computation of various "post-processing" applications in the PWscf package (including pwtools, e.g., Fermi surface and optical properties). The ability to compute weighted 1st BZ grids according to the Monkhorst-Pack scheme is included somewhere within espresso-4.0, but exactly where is unclear (k-points.f90? No). (Sure, I can fake it with an appropriate *.in input to pw.x) The nearest I find is Eyvaz Isaev's kvect_FS.f which (seems to) compute a volumetric grid (unweigted) for subsequent Fermi surface determination (Is the output symmetrized? I.e., does the code at least account for inversion symmetry?) Now, how do I compute the "grid" to input to a tool such as epsilon.x, e.g., like grid.data for the example of aluminum? And what about the "60-point" grid for Ni in Example 08? With or without MP weighting. Finally, is there a visualization tool to display the k-point distribution within the 1st BZ? XcrysDen? Paul M. Grant, PhD Principal, W2AGZ Technologies Visiting Scholar, Applied Physics, Stanford (2005-2008) EPRI Science Fellow (Retired) IBM Research Staff Member Emeritus w2agz at pacbell.net http://www.w2agz.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080914/81a629fb/attachment-0001.htm From baroni at sissa.it Mon Sep 15 08:06:51 2008 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 15 Sep 2008 08:06:51 +0200 Subject: [Pw_forum] Units of Charge Density - Again In-Reply-To: <01ff01c916df$38fc0b90$aaf422b0$@net> References: <01ff01c916df$38fc0b90$aaf422b0$@net> Message-ID: <36BF0F75-67C9-459F-A19F-568E560887C6@sissa.it> On Sep 15, 2008, at 5:00 AM, Paul M. Grant wrote: > There have been a number of posts to pw_forum regarding the units > employed for charge density throughout PWscf package and XcrySDen. > This is an important issue, and I still find the situation and > answers provided in the forum archives ambiguous. Before slogging > through relevant source codes, and testing by writing time-consuming > homemade examples, I?m hoping someone in the community has preceded > me. > > In a perfect SI world, these units would be coulombs per cubic > meter. In a practical, or rather intuitive, world, they would be in > units of total fractional number of electrons per unit cell volume > (in cubic angstroms or nanometers or bohrs (ugh)). So?what is it? > Paul: when I was familiar with the code (well, it is some time ago ...) it used to be (bohrs)^(-3). i.e. "densities" used to be NUMBER densities (length^(-3)), in the units used by the code. the only ambiguity about units is between the "pwscf" and "CP" branches of the distribution. The first uses Rydberg au (e^2/2=2m=hbar=1), the second Hartree au (e=m=hbar=1). in both sets of units, the unit length is the bohr, hence no ambiguity here. > And does it depend on the division of states between core and > valence taken when constructing a given pseudopotential? > sure it does. the integral ofer the unit (super-) cell of the density is the number ov valence electrons. > Finally, I?ve been unable (by visual inspection) to determine from > the documentation for XcrySDen (a marvelous tool?marvelous), exactly > how the ?rho grid input? is scaled, especially regarding the > computation of the 3D min-max grid values, the ?isovalue??and? > particularly the mysterious ?delta n(r)? displayed in the > temperature inset box. > no hint here cheers - SB --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080915/244e0df4/attachment.htm From ustc0200 at gmail.com Mon Sep 15 08:44:37 2008 From: ustc0200 at gmail.com (Zhu Xi) Date: Mon, 15 Sep 2008 14:44:37 +0800 Subject: [Pw_forum] ntyp too large, increase NSX Message-ID: Hi users I caculate 82 toms, but the code( V3.2.3) returns errors, which says " ntyp too large, increase NSX", how to solve this error ? Thanks. &control calculation='relax', PSEUDO_DIR='./', prefix='GR', forc_conv_thr=1.0D-4, outdir='./', / &SYSTEM ibrav = 0, celldm(1) = 34.015070944, nat = 2, ntyp = 82, ecutwfc = 50.0 , nosym = .true. , / &electrons conv_thr = 1.D-6, mixing_beta = 0.5D0, / CELL_PARAMETERS 1.0 0.000000000 0.000000000 0.000000000 0.68333333 0.000000000 0.000000000 0.000000000 0.6666666 ATOMIC_SPECIES C 12.00000 C.pz-vbc.UPF H 1.00 H.pz-vbc.UPF ATOMIC_POSITIONS (crystal) H 0.8789045011003104 0.0999999999999998 0.4999999999999997 H 0.8789045011003104 0.3000000000000000 0.4999999999999997 H 0.1210954988996892 0.9000000000000001 0.4999999999999998 H 0.1210954988996892 0.6999999999999997 0.4999999999999998 H 0.1210954988996892 0.0999999999999999 0.4999999999999999 H 0.1210954988996892 0.3000000000000002 0.4999999999999999 H 0.8789045011003104 0.9000000000000000 0.4999999999999998 H 0.8789045011003104 0.6999999999999996 0.4999999999999998 H 0.8789045011003112 0.4999999999999999 0.4999999999999998 H 0.1210954988996885 0.4999999999999999 0.4999999999999998 H 0.3621981616111121 0.4999999999999999 0.2083333333333333 H 0.6378018383888874 0.4999999999999999 0.2083333333333332 H 0.4311263939444431 0.3252703332520321 0.2083333333333333 H 0.5688736060555563 0.6747296667479678 0.2083333333333332 H 0.5688736060555563 0.3252703332520321 0.2083333333333332 H 0.4311263939444431 0.6747296667479678 0.2083333333333332 C 0.2238810822533565 -0.0000000000000001 0.4999999999999997 C 0.4605477489200225 -0.0000000000000001 0.4999999999999998 C 0.6972144155866894 -0.0000000000000001 0.4999999999999996 C 0.7761189177466430 -0.0000000000000001 0.4999999999999996 C 0.5394522510799771 -0.0000000000000001 0.4999999999999998 ............ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080915/58b93b02/attachment.htm From lanhaiping at gmail.com Mon Sep 15 09:04:23 2008 From: lanhaiping at gmail.com (lan haiping) Date: Mon, 15 Sep 2008 15:04:23 +0800 Subject: [Pw_forum] ntyp too large, increase NSX In-Reply-To: References: Message-ID: The error information is so obviously !! You mistake the ntyp setting with natom parameter ! why do you believe a system with 2 atoms can have 82 types ? On Mon, Sep 15, 2008 at 2:44 PM, Zhu Xi wrote: > Hi users > > I caculate 82 toms, but the code( V3.2.3) returns errors, which says " > ntyp too large, increase NSX", > > how to solve this error ? Thanks. > > > > > &control > calculation='relax', > PSEUDO_DIR='./', > prefix='GR', > forc_conv_thr=1.0D-4, > outdir='./', > / > &SYSTEM > ibrav = 0, > celldm(1) = 34.015070944, > nat = 2, > ntyp = 82, > ecutwfc = 50.0 , > nosym = .true. , > / > &electrons > conv_thr = 1.D-6, > mixing_beta = 0.5D0, > / > CELL_PARAMETERS > 1.0 0.000000000 0.000000000 > 0.000000000 0.68333333 0.000000000 > 0.000000000 0.000000000 0.6666666 > ATOMIC_SPECIES > C 12.00000 C.pz-vbc.UPF > H 1.00 H.pz-vbc.UPF > ATOMIC_POSITIONS (crystal) > H 0.8789045011003104 0.0999999999999998 0.4999999999999997 > H 0.8789045011003104 0.3000000000000000 0.4999999999999997 > H 0.1210954988996892 0.9000000000000001 0.4999999999999998 > H 0.1210954988996892 0.6999999999999997 0.4999999999999998 > H 0.1210954988996892 0.0999999999999999 0.4999999999999999 > H 0.1210954988996892 0.3000000000000002 0.4999999999999999 > H 0.8789045011003104 0.9000000000000000 0.4999999999999998 > H 0.8789045011003104 0.6999999999999996 0.4999999999999998 > H 0.8789045011003112 0.4999999999999999 0.4999999999999998 > H 0.1210954988996885 0.4999999999999999 0.4999999999999998 > H 0.3621981616111121 0.4999999999999999 0.2083333333333333 > H 0.6378018383888874 0.4999999999999999 0.2083333333333332 > H 0.4311263939444431 0.3252703332520321 0.2083333333333333 > H 0.5688736060555563 0.6747296667479678 0.2083333333333332 > H 0.5688736060555563 0.3252703332520321 0.2083333333333332 > H 0.4311263939444431 0.6747296667479678 0.2083333333333332 > C 0.2238810822533565 -0.0000000000000001 0.4999999999999997 > C 0.4605477489200225 -0.0000000000000001 0.4999999999999998 > C 0.6972144155866894 -0.0000000000000001 0.4999999999999996 > C 0.7761189177466430 -0.0000000000000001 0.4999999999999996 > C 0.5394522510799771 -0.0000000000000001 0.4999999999999998 > ............ > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080915/3156078b/attachment.htm From hande at newton.physics.metu.edu.tr Mon Sep 15 09:05:05 2008 From: hande at newton.physics.metu.edu.tr (Hande Ustunel) Date: Mon, 15 Sep 2008 10:05:05 +0300 (EEST) Subject: [Pw_forum] ntyp too large, increase NSX In-Reply-To: Message-ID: You have confused the meanings of ntyp and nat. Just switch them. Best, Hande On Mon, 15 Sep 2008, Zhu Xi wrote: > Hi users > > I caculate 82 toms, but the code( V3.2.3) returns errors, which says " ntyp > too large, increase NSX", > > how to solve this error ? Thanks. > > > > > &control > calculation='relax', > PSEUDO_DIR='./', > prefix='GR', > forc_conv_thr=1.0D-4, > outdir='./', > / > &SYSTEM > ibrav = 0, > celldm(1) = 34.015070944, > nat = 2, > ntyp = 82, > ecutwfc = 50.0 , > nosym = .true. , > / > &electrons > conv_thr = 1.D-6, > mixing_beta = 0.5D0, > / > CELL_PARAMETERS > 1.0 0.000000000 0.000000000 > 0.000000000 0.68333333 0.000000000 > 0.000000000 0.000000000 0.6666666 > ATOMIC_SPECIES > C 12.00000 C.pz-vbc.UPF > H 1.00 H.pz-vbc.UPF > ATOMIC_POSITIONS (crystal) > H 0.8789045011003104 0.0999999999999998 0.4999999999999997 > H 0.8789045011003104 0.3000000000000000 0.4999999999999997 > H 0.1210954988996892 0.9000000000000001 0.4999999999999998 > H 0.1210954988996892 0.6999999999999997 0.4999999999999998 > H 0.1210954988996892 0.0999999999999999 0.4999999999999999 > H 0.1210954988996892 0.3000000000000002 0.4999999999999999 > H 0.8789045011003104 0.9000000000000000 0.4999999999999998 > H 0.8789045011003104 0.6999999999999996 0.4999999999999998 > H 0.8789045011003112 0.4999999999999999 0.4999999999999998 > H 0.1210954988996885 0.4999999999999999 0.4999999999999998 > H 0.3621981616111121 0.4999999999999999 0.2083333333333333 > H 0.6378018383888874 0.4999999999999999 0.2083333333333332 > H 0.4311263939444431 0.3252703332520321 0.2083333333333333 > H 0.5688736060555563 0.6747296667479678 0.2083333333333332 > H 0.5688736060555563 0.3252703332520321 0.2083333333333332 > H 0.4311263939444431 0.6747296667479678 0.2083333333333332 > C 0.2238810822533565 -0.0000000000000001 0.4999999999999997 > C 0.4605477489200225 -0.0000000000000001 0.4999999999999998 > C 0.6972144155866894 -0.0000000000000001 0.4999999999999996 > C 0.7761189177466430 -0.0000000000000001 0.4999999999999996 > C 0.5394522510799771 -0.0000000000000001 0.4999999999999998 > ............ > -- Hande Toffoli 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 holger.hesske at chem.ethz.ch Mon Sep 15 09:06:09 2008 From: holger.hesske at chem.ethz.ch (Holger Hesske) Date: Mon, 15 Sep 2008 09:06:09 +0200 Subject: [Pw_forum] ntyp too large, increase NSX In-Reply-To: References: Message-ID: <200809150906.10152.holger.hesske@chem.ethz.ch> Am Montag, 15. September 2008 08.44:37 schrieb Zhu Xi: > Hi users > Hi Zhu, > I caculate 82 toms, but the code( V3.2.3) returns errors, which says " > ntyp too large, increase NSX", > > how to solve this error ? Thanks. > It is what the code tells you: to many TYPES of atoms! If you refer to the user guide nat = number of atoms ntyp = number od types of atoms... Seems you mixed up these things. Hope this helps! And btw: would you be so kind to tell us your name and affiliation in the textbody?! Cheers, Holger > > > > &control > calculation='relax', > PSEUDO_DIR='./', > prefix='GR', > forc_conv_thr=1.0D-4, > outdir='./', > / > &SYSTEM > ibrav = 0, > celldm(1) = 34.015070944, > nat = 2, > ntyp = 82, > ecutwfc = 50.0 , > nosym = .true. , > / > &electrons > conv_thr = 1.D-6, > mixing_beta = 0.5D0, > / > CELL_PARAMETERS > 1.0 0.000000000 0.000000000 > 0.000000000 0.68333333 0.000000000 > 0.000000000 0.000000000 0.6666666 > ATOMIC_SPECIES > C 12.00000 C.pz-vbc.UPF > H 1.00 H.pz-vbc.UPF > ATOMIC_POSITIONS (crystal) > H 0.8789045011003104 0.0999999999999998 0.4999999999999997 > H 0.8789045011003104 0.3000000000000000 0.4999999999999997 > H 0.1210954988996892 0.9000000000000001 0.4999999999999998 > H 0.1210954988996892 0.6999999999999997 0.4999999999999998 > H 0.1210954988996892 0.0999999999999999 0.4999999999999999 > H 0.1210954988996892 0.3000000000000002 0.4999999999999999 > H 0.8789045011003104 0.9000000000000000 0.4999999999999998 > H 0.8789045011003104 0.6999999999999996 0.4999999999999998 > H 0.8789045011003112 0.4999999999999999 0.4999999999999998 > H 0.1210954988996885 0.4999999999999999 0.4999999999999998 > H 0.3621981616111121 0.4999999999999999 0.2083333333333333 > H 0.6378018383888874 0.4999999999999999 0.2083333333333332 > H 0.4311263939444431 0.3252703332520321 0.2083333333333333 > H 0.5688736060555563 0.6747296667479678 0.2083333333333332 > H 0.5688736060555563 0.3252703332520321 0.2083333333333332 > H 0.4311263939444431 0.6747296667479678 0.2083333333333332 > C 0.2238810822533565 -0.0000000000000001 0.4999999999999997 > C 0.4605477489200225 -0.0000000000000001 0.4999999999999998 > C 0.6972144155866894 -0.0000000000000001 0.4999999999999996 > C 0.7761189177466430 -0.0000000000000001 0.4999999999999996 > C 0.5394522510799771 -0.0000000000000001 0.4999999999999998 > ............ -- Dr. Holger Hesske Inst.f.Chemie-/Bioingenieurwissenschaft. Wolfgang-Pauli-Str. 10 ETH H?nggerberg, HCI E 133 CH-8093 Z?rich Tel: +41 44 633 66 19 Fax: +41 44 632 11 63 E-Mail: holger.hesske at chem.ethz.ch From giannozz at democritos.it Mon Sep 15 10:33:21 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 15 Sep 2008 10:33:21 +0200 Subject: [Pw_forum] k-point grids...and all that In-Reply-To: <020401c916e3$571a2bd0$054e8370$@net> References: <020401c916e3$571a2bd0$054e8370$@net> Message-ID: On Sep 15, 2008, at 5:30 , Paul M. Grant wrote: > The ability to compute weighted 1st BZ grids according to the > Monkhorst-Pack scheme is included somewhere within espresso-4.0, > but exactly where is unclear (k-points.f90? No). > of course no, since it is non-existent: $ ls -l */k-points.f90 ls: */k-points.f90: No such file or directory There is however an auxiliary code, pwtools/kpoints.f, that produces Monkhorst-Pack grids with weights (executable: kpoints.x). Inside q-e, there are several routines for the generation of automatic k-point grids. See PW/setup.f90, under ! ! ... Automatic generation of k-points (if required) ! Routines kpoint_grid_efield and kp_strings are specific for macroscopic field and Berry phase calculations; kpoint_grid and setupkpoint (undocumented) take care of the typical scf case. The difference between the two is just aesthetic. > Finally, is there a visualization tool to display the k-point > distribution within the 1st BZ? XcrysDen? > for sure: see http://www.xcrysden.org/img/xcrysden-in-action.png, first picture Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From rmargine at gmail.com Mon Sep 15 15:51:05 2008 From: rmargine at gmail.com (Elena Roxana Margine) Date: Mon, 15 Sep 2008 15:51:05 +0200 Subject: [Pw_forum] error cannot remap grid on k-point list Message-ID: Dear all, I have been trying to perform an electron-phonon calculation for a hexagonal structure but the calculation crashes at second point on a 2x2x2 q-mesh in the elphon subroutine with the error "cannot remap grid on k-point list" do n=1,nkh do nk=1,nkBZ if (eqBZ(nk) == n) go to 20 end do ! this failure of the algorithm may indicate that the displaced grid ! (with k1,k2,k3.ne.0) does not have the full symmetry of the lattice call errore('lint','cannot remap grid on k-point list',n) 20 continue end do The problem is that the code finds more irreducible k-points than there should be. In this example k(27) is the same as k(25) and k(29) is the same as k(9). So, the correct number of k points should be 26 not 30. number of k points= 30 gaussian broad. (Ry)= 0.0050 ngauss = 1 cart. coord. in units 2pi/a_0 k( 1) = ( 0.0000000 0.0000000 0.0000000), wk = 0.0312500 k( 2) = ( 0.0000000 0.0000000 -0.3048780), wk = 0.0000000 k( 3) = ( 0.0000000 0.0000000 0.1524390), wk = 0.0625000 k( 4) = ( 0.0000000 0.0000000 -0.1524390), wk = 0.0000000 k( 5) = ( 0.0000000 0.0000000 -0.3048780), wk = 0.0312500 k( 6) = ( 0.0000000 0.0000000 -0.6097561), wk = 0.0000000 k( 7) = ( 0.0000000 0.2886751 0.0000000), wk = 0.1875000 k( 8) = ( 0.0000000 0.2886751 -0.3048780), wk = 0.0000000 k( 9) = ( 0.0000000 0.2886751 0.1524390), wk = 0.0937500 k( 10) = ( 0.0000000 0.2886751 -0.1524390), wk = 0.0000000 k( 11) = ( 0.0000000 0.2886751 -0.3048780), wk = 0.1875000 k( 12) = ( 0.0000000 0.2886751 -0.6097561), wk = 0.0000000 k( 13) = ( 0.0000000 -0.5773503 0.0000000), wk = 0.0937500 k( 14) = ( 0.0000000 -0.5773503 -0.3048780), wk = 0.0000000 k( 15) = ( 0.0000000 -0.5773503 0.1524390), wk = 0.1875000 k( 16) = ( 0.0000000 -0.5773503 -0.1524390), wk = 0.0000000 k( 17) = ( 0.0000000 -0.5773503 -0.3048780), wk = 0.0937500 k( 18) = ( 0.0000000 -0.5773503 -0.6097561), wk = 0.0000000 k( 19) = ( 0.2500000 0.4330127 0.0000000), wk = 0.1875000 k( 20) = ( 0.2500000 0.4330127 -0.3048780), wk = 0.0000000 k( 21) = ( 0.2500000 0.4330127 0.1524390), wk = 0.3750000 k( 22) = ( 0.2500000 0.4330127 -0.1524390), wk = 0.0000000 k( 23) = ( 0.2500000 0.4330127 -0.3048780), wk = 0.1875000 k( 24) = ( 0.2500000 0.4330127 -0.6097561), wk = 0.0000000 k( 25) = ( 0.0000000 -0.2886751 0.1524390), wk = 0.0937500 k( 26) = ( 0.0000000 -0.2886751 -0.1524390), wk = 0.0000000 k( 27) = ( 0.0000000 -0.2886751 0.1524390), wk = 0.0937500 k( 28) = ( 0.0000000 -0.2886751 -0.1524390), wk = 0.0000000 k( 29) = ( 0.0000000 0.2886751 0.1524390), wk = 0.0937500 k( 30) = ( 0.0000000 0.2886751 -0.1524390), wk = 0.0000000 These are the input files for this simple test that I can be easily reproduces. sic.scf.fit.in --------------------------- &control calculation = 'scf' tstress = .true. tprnfor = .true. pseudo_dir = './', prefix='sic2', outdir='./' etot_conv_thr = 1.0d-5, / &system ibrav= 4, celldm(1) = 5.81, celldm(3) = 1.64, nat= 4, ntyp=3, ecutwfc =25.0,ecutrho=200.0, occupations='smearing', smearing='methfessel-paxton', degauss=0.005, la2F=.true. / &electrons diagonalization='david' mixing_mode = 'plain' mixing_beta = 0.7 conv_thr = 1.0d-8 / ATOMIC_SPECIES C 12.011 C.pbe-rrkjus.UPF Si 28.086 Si.pbe-n-van.UPF B 10.811 B.pbe-n-van.UPF ATOMIC_POSITIONS crystal C 0.666666667 0.333333333 0.500000000 Si 0.000000000 0.000000000 0.375000000 Si 0.666666667 0.333333333 0.875000000 B 0.000000000 0.000000000 0.000000000 K_POINTS automatic 8 8 8 0 0 0 sic.scf.in ---------------- &control calculation = 'scf' tstress = .true. tprnfor = .true. pseudo_dir = './', prefix='sic2', outdir='./' etot_conv_thr = 1.0d-5, / &system ibrav= 4, celldm(1) = 5.81, celldm(3) = 1.64, nat= 4, ntyp=3, ecutwfc =25.0,ecutrho=200.0, occupations='smearing', smearing='methfessel-paxton', degauss=0.005, / &electrons diagonalization='david' mixing_mode = 'plain' mixing_beta = 0.7 conv_thr = 1.0d-8 / ATOMIC_SPECIES C 12.011 C.pbe-rrkjus.UPF Si 28.086 Si.pbe-n-van.UPF B 10.811 B.pbe-n-van.UPF ATOMIC_POSITIONS crystal C 0.666666667 0.333333333 0.500000000 Si 0.000000000 0.000000000 0.375000000 Si 0.666666667 0.333333333 0.875000000 B 0.000000000 0.000000000 0.000000000 K_POINTS automatic 4 4 4 0 0 0 sic.elph.in ---------------- Electron-phonon coefficients &inputph tr2_ph=1.0d-10, prefix='sic2', fildvscf='dv', amass(1)=12.011, amass(2)=28.086, amass(3)=10.811, outdir='./', fildyn='sic2.dyn', elph=.true., trans=.true., ldisp=.true., nq1=2, nq2=2,nq3=2 / Thank you, Roxana Margine ==================================================== Elena Roxana Margine, Ph.D. Laboratoire de Physique de la Mati?re Condens?e et Nanostructures Universit? Claude Bernard Lyon 1, 43 bd du 11 novembre 1918 69622 Villeurbanne - France E-mail: rmargine at lpmcn.univ-lyon1.fr Phone: +33 (0) 4 724 482 37 ==================================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080915/24edd576/attachment-0001.htm From edulsa at quimica.ufpr.br Mon Sep 15 19:53:31 2008 From: edulsa at quimica.ufpr.br (Eduardo Lemos de Sa) Date: Mon, 15 Sep 2008 14:53:31 -0300 (BRT) Subject: [Pw_forum] trying to compile espresso4.0.2 Message-ID: <54494.200.17.215.86.1221501211.squirrel@www.quimica.ufpr.br> Sirs I ask your pardon by disturber you with a so silly issue. I have tried to compile espresso 4.0.2 in a FreeBSD (amd64) box. I guess that the hints made for linux are usable to FreeBSD. In general lines: 1) ./configure ended with no errors. Make log file (edu.config.log) and make.sys generated are attached in this email. I have installed further fftw, atlas (located in /usr/local/lib/ and /usr/local/lib/atlas), but the makefile does not use it (instead, it prefer fftw). 2) Continuing, I did make all. The log file generated (edu.make.log) is attached too. The problem seems to be in a clib library build. Please, can you help me? Thanks a lot in advance Eduardo -- Eduardo Lemos de Sa edulsa at quimica.ufpr.br Professor Assoc. I - Departamento de Quimica Universidade Federal do Parana C. P. 19081 81531-990 Curitiba/PR - Brazil Fone (41) 3361 3300 -------------- next part -------------- A non-text attachment was scrubbed... Name: edu.config.log Type: text/x-log Size: 5625 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20080915/c7b9f70d/attachment.bin From edulsa at quimica.ufpr.br Mon Sep 15 19:55:29 2008 From: edulsa at quimica.ufpr.br (Eduardo Lemos de Sa) Date: Mon, 15 Sep 2008 14:55:29 -0300 (BRT) Subject: [Pw_forum] [Fwd: trying to compile espresso4.0.2] Message-ID: <61314.200.17.215.86.1221501329.squirrel@www.quimica.ufpr.br> Sirs Sorry: I forgot two files. Here it goes Yours Eduardo Sirs I ask your pardon by disturber you with a so silly issue. I have tried to compile espresso 4.0.2 in a FreeBSD (amd64) box. I guess that the hints made for linux are usable to FreeBSD. In general lines: 1) ./configure ended with no errors. Make log file (edu.config.log) and make.sys generated are attached in this email. I have installed further fftw, atlas (located in /usr/local/lib/ and /usr/local/lib/atlas), but the makefile does not use it (instead, it prefer fftw). 2) Continuing, I did make all. The log file generated (edu.make.log) is attached too. The problem seems to be in a clib library build. Please, can you help me? Thanks a lot in advance Eduardo -- Eduardo Lemos de Sa edulsa at quimica.ufpr.br Professor Assoc. I - Departamento de Quimica Universidade Federal do Parana C. P. 19081 81531-990 Curitiba/PR - Brazil Fone (41) 3361 3300 -- Eduardo Lemos de Sa edulsa at quimica.ufpr.br Professor Assoc. I - Departamento de Quimica Universidade Federal do Parana C. P. 19081 81531-990 Curitiba/PR - Brazil Fone (41) 3361 3300 -------------- next part -------------- A non-text attachment was scrubbed... Name: edu.config.log Type: text/x-log Size: 5625 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20080915/23eff5db/attachment.bin -------------- next part -------------- A non-text attachment was scrubbed... Name: make.sys Type: application/octet-stream Size: 4522 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20080915/23eff5db/attachment.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: edu.make.log Type: text/x-log Size: 719 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20080915/23eff5db/attachment-0001.bin From shahn at iastate.edu Mon Sep 15 20:10:03 2008 From: shahn at iastate.edu (Steven Hahn) Date: Mon, 15 Sep 2008 13:10:03 -0500 Subject: [Pw_forum] phonon structure factor Message-ID: <0BD8874C-90A6-4097-BF45-6CC5DED97AD9@iastate.edu> Dear all, I am trying to calculate the phonon structure factor from the matdyn.modes file. The textbook I am using shows two different definitions of the polarization vector that differ by exp(iqd), where d is the position vector to the dth atom within the unit cell. How is the polarization vector defined in PWSCF? Steven Hahn Graduate Student Department of Physics and Astronomy Iowa State University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080915/a367370f/attachment.htm From giannozz at democritos.it Mon Sep 15 23:29:20 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 15 Sep 2008 23:29:20 +0200 Subject: [Pw_forum] error cannot remap grid on k-point list In-Reply-To: References: Message-ID: <5DE56847-D212-45F6-9B81-20BC2A4F7C2D@democritos.it> On Sep 15, 2008, at 15:51 , Elena Roxana Margine wrote: > I have been trying to perform an electron-phonon calculation for a > hexagonal > structure but the calculation crashes at second point on a 2x2x2 q- > mesh in > the elphon subroutine with the error "cannot remap grid on k-point > list" > [...] > The problem is that the code finds more irreducible k-points than > there should > be. In this example k(27) is the same as k(25) and k(29) is the > same as k(9). > So, the correct number of k points should be 26 not 30. it is a known[*] problem of all recent versions of the phonon code, fixed in the CVS version (which is undergoing some heavy restructuring of the symmetry part). Fixing it in the stable version requires some nontrivial changes. As a workaround, try to run the intermediate non-scf calculation with the 'phonon' keyword instead of the "ldisp" option: you should get the correct number of k-points. Thank you for bringing this to our attention and for providing a simple and useful bug report Paolo [*] that is, known to me and to a couple of other guys --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From adrainzhou at yahoo.com.cn Tue Sep 16 07:06:34 2008 From: adrainzhou at yahoo.com.cn (Adrain Zhou) Date: Tue, 16 Sep 2008 13:06:34 +0800 (CST) Subject: [Pw_forum] Hybrid functional calcualtion Message-ID: <446560.12696.qm@web15805.mail.cnb.yahoo.com> Dear all, I found the hybrid functional calculation became very slow when I increase nqx1, nqx2, nqx3 from 2 to 4. Is there any trick for such hybrid functional calculation? Many thanks in advance! Regards, Adrian ___________________________________________________________ ???????????? http://cn.mail.yahoo.com/ From lawrence_lee_lee at yahoo.com.hk Tue Sep 16 07:39:47 2008 From: lawrence_lee_lee at yahoo.com.hk (Lawrence Lee) Date: Tue, 16 Sep 2008 13:39:47 +0800 Subject: [Pw_forum] The meaning of "a" in the phonon calculation Message-ID: <1221543587.18888.18.camel@scg11> Dear all, In the pw.x code, if I choose "calculation = phonon" calculation. According to the INPUT_PW instruction, the xqq(1) (and 2 and 3) flags in the namelist "&PHONON" are in the unit of 2pi/a. I would like to ask whether that "a" is the conventional cell parameter, or the magnitude of the primitive lattice vector? This is important for me to know whether my specification in the namelist "&PHONON" is correct. For example, if I want to specify a calculation at the K point or X point. I need to know the meaning of "a" Thanks for help very much! -- S.H. Lee Research Assistant Physics Division The CUHK Graduate School _______________________________________ ?x?????? - Yahoo! Messenger ???????]?????W?????????????????????????o???????????W?r???????????????????f???????????? http://messenger.yahoo.com.hk From giannozz at democritos.it Tue Sep 16 08:43:43 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 16 Sep 2008 08:43:43 +0200 Subject: [Pw_forum] trying to compile espresso4.0.2 In-Reply-To: <54494.200.17.215.86.1221501211.squirrel@www.quimica.ufpr.br> References: <54494.200.17.215.86.1221501211.squirrel@www.quimica.ufpr.br> Message-ID: <9FDCCB6D-1691-419F-B274-077EF7726D60@democritos.it> On Sep 15, 2008, at 19:53 , Eduardo Lemos de Sa wrote: > 2) Continuing, I did make all. The log file generated > (edu.make.log) is > attached too. The problem seems to be in a clib library build. not really: the problem seems to be the presence of spurious *.o or *.mod files > Fatal Error: While reading module 'kinds' found module version 0, expected 5. "make clean" first (removes all objects) Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From degironc at sissa.it Tue Sep 16 08:47:56 2008 From: degironc at sissa.it (Stefano de Gironcoli) Date: Tue, 16 Sep 2008 08:47:56 +0200 Subject: [Pw_forum] Hybrid functional calcualtion In-Reply-To: <446560.12696.qm@web15805.mail.cnb.yahoo.com> References: <446560.12696.qm@web15805.mail.cnb.yahoo.com> Message-ID: <48CF569C.7040909@sissa.it> the nqx1,nqx2,nqx3 are there just for that... sometime (especially in insulators where Wannier functions are localized) the structure of the 1-particle density-matrix allows the Fock operator to be computed on a coarser q-point grid saving a lot of time... However as usual this is something that needs to be checked carefully case by case on the basis of the desired accuracy, which typically depends on the desired quantity: geometry, energy difference or absolute energy etc... stefano Adrain Zhou wrote: > Dear all, > > I found the hybrid functional calculation became very > slow when I increase nqx1, nqx2, nqx3 from 2 to 4. > > Is there any trick for such hybrid functional > calculation? Many thanks in advance! > > Regards, > Adrian > > > > ___________________________________________________________ > ???????????? > http://cn.mail.yahoo.com/ > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From giannozz at democritos.it Tue Sep 16 08:48:05 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 16 Sep 2008 08:48:05 +0200 Subject: [Pw_forum] The meaning of "a" in the phonon calculation In-Reply-To: <1221543587.18888.18.camel@scg11> References: <1221543587.18888.18.camel@scg11> Message-ID: <1614F586-49E8-48F0-9B0E-68988BB7781E@democritos.it> On Sep 16, 2008, at 7:39 , Lawrence Lee wrote: > The meaning of "a" in the phonon calculation the same as in a scf calculation: a is the lattice parameter Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From ski2 at mail.uh.edu Tue Sep 16 18:16:34 2008 From: ski2 at mail.uh.edu (ski2 at mail.uh.edu) Date: Tue, 16 Sep 2008 11:16:34 -0500 Subject: [Pw_forum] SCF calculations aborted Message-ID: Dear All, I am trying to calculate a spin polarized state using USPP with GGA+U. But the calculation was aborted with an message as the following: ???? iteration #? 1???? ecut=??? 37.00 Ry???? beta=0.20 ???? CG style diagonalization ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 2 eigenvalues not converged ???? c_bands:? 1 eigenvalues not converged ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 1 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 1 eigenvalues not converged ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 2 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 1 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 1 eigenvalues not converged ???? c_bands:? 3 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ???? c_bands:? 4 eigenvalues not converged ???? c_bands:? 5 eigenvalues not converged ?%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ???? from davcio : error #??????? 10 ???? error while reading from file ?%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ???? stopping ... Please let me know how to avoid this problem. Just in case, I also pasted a part of my input file (see below). &control ??? calculation='scf' ??? restart_mode='restart', ??? prefix='vso' ??? pseudo_dir = '$PSEUDO_DIR/', ??? outdir='$TMP_DIR/' ?/ ?&system ??? ibrav= 6, A= 11.2338, B= 11.2338, C= 7.8711, cosAB= 0, cosAC= 0, cosBC= 0, nat= 72, ntyp= 3, ??? ecutwfc =37.0, ecutrho= 296, nbnd= 342, occupations='smearing', smearing='gauss', degauss=0.01, ??? nspin= 2, starting_magnetization(1)= 0.5, report=1, ??? lda_plus_u=.true.? Hubbard_U(1)=1 ?/ ?&electrons ??? electron_maxstep= 300 ??? mixing_mode = 'plain' ??? mixing_beta = 0.2 ??? conv_thr =? 1.0d-6 ??? diagonalization= 'cg' ??? mixing_fixed_ns= 0 ?/ Best regards, Sang-Hwan Kim Department of Chemistry University of Houston 136 Fleming Building Houston, TX 77204-5003 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080916/8ad06862/attachment.htm From giannozz at democritos.it Tue Sep 16 18:36:11 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 16 Sep 2008 18:36:11 +0200 Subject: [Pw_forum] SCF calculations aborted In-Reply-To: References: Message-ID: <2B1922CA-FA44-4331-8FFB-A1E78961E835@democritos.it> On Sep 16, 2008, at 18:16 , ski2 at mail.uh.edu wrote: > CG style diagonalization much slower than Davidson, FYI > from davcio : error # 10 > error while reading from file > > Please let me know how to avoid this problem. http://www.quantum-espresso.org/wiki/index.php/Troubleshooting_% 28PWscf%29 item 1.12 Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From eyvaz_isaev at yahoo.com Wed Sep 17 00:32:23 2008 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 16 Sep 2008 15:32:23 -0700 (PDT) Subject: [Pw_forum] *.fc file in PW_SCF help. In-Reply-To: <1221156338.48c95df20c43e@webmail.purdue.edu> Message-ID: <358211.87394.qm@web65704.mail.ac4.yahoo.com> Dear Zhen, --- On Thu, 9/11/08, Zhen Huang wrote: > > Can anyone help me explain the real space force constant > file *.fc. What does each index mean for example (I am > doing a 4 by4 by4 calculation): What is the second row > 1 1 1 1 means what is saying 1.45522e+00 is this > force constant in ev/A^2? > > 4 4 4 > 1 1 1 1 > 1 1 1 1.45522064500E+00 > 2 1 1 -6.48059987500E-02 > 3 1 1 1.15693187500E-03 > 4 1 1 -6.48059987500E-02 ................................. > 1 1 2 -2.61743750000E-04 > 2 1 2 -3.88856250000E-05 > 3 1 2 1.02087500000E-04 > 4 1 2 -5.68562500000E-05 FC is not a PW_SCF file, it is generated by q2r.x program, so, it is rather PW_PH file. Well, let us to recall the notation for the force constants matrix, F_{\alpha,\beta}^{i,j}(R), where \alpha, \beta are atomic numbers, i,j are polarization vectors (x,y,z), and R is the distance between two atoms in units of lattice vectors. Below there are examples how these lines should be treated. 4 4 4 - you used 4x4x4 q-points to generate FC 1 3 1 2 - x, z, 1st atom, 2nd atom 4 3 2 0.1234 - R=i*a+j*b+k*c, where i,j,k are integers from 1 to 4 (you used 4x4x4 mesh), a,b,c are lattice vectors. 0.1234 is F_{1,2}^{x,z}(R(4,3,2)). As QE uses atomic units, F should be in units of Ry/au^2, but not in eV/A^2. Hope this helps. Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com From huangchunlin1984 at 163.com Wed Sep 17 03:55:27 2008 From: huangchunlin1984 at 163.com (huangchunlin1984) Date: Wed, 17 Sep 2008 09:55:27 +0800 (CST) Subject: [Pw_forum] some problem in lowdin charges Message-ID: <30396433.91311221616527794.JavaMail.coremail@bj163app26.163.com> Doctor: I have done Lowdin population analysis on the Fe clusters. From the Lowdin charges, I find that the total charge of the cluster is a little lower than the accurate value. Take the Fe3 cluster for example,using 3d4s for Fe. Lowdin charges: Atom # 1: total charge = 7.8407 ... Atom # 2: total charge = 7.8407 ... Atom # 3: total charge = 8.0938 ... The sum charge of the three atoms is 23.7752 which is not 24. So where is the other charges ? Or is this accurate enough ? the input is as follows: &inputpp outdir='./' prefix='Fe3' Emin=-8.0, Emax=-1, DeltaE=0.1 ngauss=0, degauss=0.001 / thanks for your help. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/2de30cfb/attachment-0001.htm From sagarambavale at yahoo.co.in Wed Sep 17 06:18:07 2008 From: sagarambavale at yahoo.co.in (ambavale sagar) Date: Wed, 17 Sep 2008 09:48:07 +0530 (IST) Subject: [Pw_forum] problems computing cholesky decomposition Message-ID: <653592.2380.qm@web94615.mail.in2.yahoo.com> Dear all, I am running 'VC-relax calculation of simple si-c infinite wire using espresso-4.0 on Xeon quad core dual CPU (X86_64) workstation with intel mkl -10, lam-mpi -7.1.4. I am getting:? (1) ??' from pzpotrf : error # 1 ? ?????problems computing cholesky decomposition' ?while using cell-dynamics = 'bfgs'. (2) from rdiaghg : error ?# 27 ?????info =/= 0 when ran the same input with -ndiag 1 (3) run successfully with cell-dynamics = 'damp-pr' and -ndiag 1 in command line. (4) run successfully with cell-dynamics = 'damp-pr' and? without -ndiag 1 in command line. My input file is : &CONTROL ?? calculation = 'vc-relax' ?? restart_mode= 'from_scratch' ?? prefix = 'c-si-inf-vcrlx' ?? pseudo_dir? = '/home/phy/espresso-4.0/pseudo/' ?? outdir = '/home/phy/sagar_espresso_tmp/' ?/ ?&SYSTEM ?? ibrav?????? = 6, ?? celldm(1)?? = 25, ?? celldm(3)?? = 0.241884944, ?? nat???????? = 2, ?? ntyp??????? = 2, ?? ecutwfc???? = 35.0, ?? ecutrho???? = 150.0, ?? occupations = 'smearing' ?? smearing??? = 'methfessel-paxton' ?? degauss???? = 0.01, ?/ ?&ELECTRONS ?? conv_thr??? = 1.d-12, ?? mixing_beta = 0.7, ?/ ?&IONS ?? ion_dynamics = 'damp' ?? wfc_extrapolation = 'second_order' ?? pot_extrapolation = 'second_order' ?/ ?&CELL ?? cell_dynamics = 'damp-pr' ?/ ?ATOMIC_SPECIES ?C?? 12.01? C.pz-vbc.UPF ?Si? 28.086 Si.vbc.UPF ATOMIC_POSITIONS (angstrom) C??? 0.000000000??? 0.000000000??? 0.000000000 Si??? 0.000000000??? 0.000000000??? 1.6 K_POINTS (gamma) ********************************************************************************** Thanx. Sagar Ambavale Research Student, The M.S. University of Baroda, India Unlimited freedom, unlimited storage. Get it now, on http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/1f0dd92f/attachment.htm From meisam_a84 at yahoo.com Wed Sep 17 07:01:27 2008 From: meisam_a84 at yahoo.com (meisam aghtar) Date: Tue, 16 Sep 2008 22:01:27 -0700 (PDT) Subject: [Pw_forum] relaxation & breaking symmetry Message-ID: <409053.74950.qm@web30606.mail.mud.yahoo.com> Dear all In relaxing calculation, if I break the symmetry could the system go to a higher symmetry? ? Sincerely ? M.aghtar Msc student, physics department of University of Kashan. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080916/3946dc72/attachment.htm From baroni at sissa.it Wed Sep 17 07:46:51 2008 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 17 Sep 2008 07:46:51 +0200 Subject: [Pw_forum] relaxation & breaking symmetry In-Reply-To: <409053.74950.qm@web30606.mail.mud.yahoo.com> References: <409053.74950.qm@web30606.mail.mud.yahoo.com> Message-ID: <5596FCB7-EF8F-4408-8E5A-FFD7C56B1B59@sissa.it> Dear M.aghtar: > > In relaxing calculation, if I break the symmetry could the system go > to a higher symmetry? sure it can. a high-simmetry configuration is a particular case of a low-simmetry one (the other way around is obviously untrue) SB --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/c61fdba1/attachment.htm From baroni at sissa.it Wed Sep 17 07:51:01 2008 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 17 Sep 2008 07:51:01 +0200 Subject: [Pw_forum] some problem in lowdin charges In-Reply-To: <30396433.91311221616527794.JavaMail.coremail@bj163app26.163.com> References: <30396433.91311221616527794.JavaMail.coremail@bj163app26.163.com> Message-ID: Dear Mr. unnamed: Lowdin charges (as well as other conventional atomic charges) do not satisfy any sum rule. You can easily convince yourself that ths is the case because the atomic orbitals that are used to calculate them are arbitrary to some extent. If yu like, you can think that the missing charge is "delocalized" or "bonding" charge, but this would be another way of naming the conventional (to some extent) character of Lowdin charge. Please, do not forget to use the courtesy of signing your correspondence to this forum amd of providing your academic affiliation. Have a good day SB On Sep 17, 2008, at 3:55 AM, huangchunlin1984 wrote: > Doctor: > I have done Lowdin population analysis on the Fe clusters. > From the Lowdin charges, I find that the total charge of the cluster > is a little lower than the accurate value. Take the Fe3 cluster for > example,using 3d4s for Fe. > Lowdin charges: > Atom # 1: total charge = 7.8407 ... > Atom # 2: total charge = 7.8407 ... > Atom # 3: total charge = 8.0938 ... > The sum charge of the three atoms is 23.7752 which is not 24. So > where is the other charges ? Or is this accurate enough ? > the input is as follows: > &inputpp > outdir='./' > prefix='Fe3' > Emin=-8.0, Emax=-1, DeltaE=0.1 > ngauss=0, degauss=0.001 > / > thanks for your help. > > > > [??] ?????????????? > ?"???"?????_______________________________________________ > 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 http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/5a134168/attachment-0001.htm From wangqj1 at 126.com Wed Sep 17 07:55:24 2008 From: wangqj1 at 126.com (wangqj1) Date: Wed, 17 Sep 2008 13:55:24 +0800 (CST) Subject: [Pw_forum] about atom magnetic moment Message-ID: <11683171.144051221630924148.JavaMail.coremail@bj126app27.126.com> Dear pwscf users: I know that from the projwfc.x can get the atom magnetic moment ,but I don't know the specific process to get the atom magnetic moment from the result of pdos. could someone give me some advice ? Your kind heart will be appreciated . sincerely XiangTang university -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/a2c4b7ab/attachment.htm From vegalew at hotmail.com Wed Sep 17 10:33:56 2008 From: vegalew at hotmail.com (vega) Date: Wed, 17 Sep 2008 16:33:56 +0800 Subject: [Pw_forum] Is there a small tools to sum the pdos of selected atoms? Message-ID: Dear all, There is a lot of atoms in my system. I want to sum the pdos of selected atoms. But doing this by hand is quite boring. I know there is a lot of small tools in QE for various purposes. But I can't find them. Is there a list or description about them? Where is the list? thank you for reading vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/7da339ec/attachment.htm From lanhaiping at gmail.com Wed Sep 17 13:21:55 2008 From: lanhaiping at gmail.com (lan haiping) Date: Wed, 17 Sep 2008 19:21:55 +0800 Subject: [Pw_forum] Is there a small tools to sum the pdos of selected atoms? In-Reply-To: References: Message-ID: Just take a look at $QE/PP, you can find some routines. On Wed, Sep 17, 2008 at 4:33 PM, vega wrote: > Dear all, > > There is a lot of atoms in my system. I want to sum the pdos of selected > atoms. But doing this by hand is quite boring. > I know there is a lot of small tools in QE for various purposes. But I > can't find them. Is there a list or description about them? > Where is the list? > > thank you for reading > > vega > > > > ================================================================================= > Vega Lew (weijia liu) > PH.D Candidate in Chemical Engineering > State Key Laboratory of Materials-oriented Chemical Engineering > College of Chemistry and Chemical Engineering > Nanjing University of Technology, 210009, Nanjing, Jiangsu, China > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/47ab9e80/attachment.htm From vegalew at hotmail.com Wed Sep 17 13:24:57 2008 From: vegalew at hotmail.com (vega) Date: Wed, 17 Sep 2008 19:24:57 +0800 Subject: [Pw_forum] Is there a small tools to sum the pdos of selectedatoms? In-Reply-To: References: Message-ID: Thanks, I'll check it. ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China From: lan haiping Sent: Wednesday, September 17, 2008 7:21 PM To: PWSCF Forum Subject: Re: [Pw_forum] Is there a small tools to sum the pdos of selectedatoms? Just take a look at $QE/PP, you can find some routines. On Wed, Sep 17, 2008 at 4:33 PM, vega wrote: Dear all, There is a lot of atoms in my system. I want to sum the pdos of selected atoms. But doing this by hand is quite boring. I know there is a lot of small tools in QE for various purposes. But I can't find them. Is there a list or description about them? Where is the list? thank you for reading vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------------------------------------------------------------------------- _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/8da57ffa/attachment.htm From giannozz at democritos.it Wed Sep 17 13:58:38 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 17 Sep 2008 13:58:38 +0200 Subject: [Pw_forum] Is there a small tools to sum the pdos of selected atoms? In-Reply-To: References: Message-ID: PP/sumpdos.f90 --- Paolo Giannozzi, Democritos and University of Udine, Italy From vegalew at hotmail.com Wed Sep 17 14:01:47 2008 From: vegalew at hotmail.com (vega) Date: Wed, 17 Sep 2008 20:01:47 +0800 Subject: [Pw_forum] Is there a small tools to sum the pdos of selectedatoms? In-Reply-To: References: Message-ID: Dear sir, > PP/sumpdos.f90 thank you. I'll try it vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------------------------------------------- From: "Paolo Giannozzi" Sent: Wednesday, September 17, 2008 7:58 PM To: "PWSCF Forum" Subject: Re: [Pw_forum] Is there a small tools to sum the pdos of selectedatoms? > PP/sumpdos.f90 > --- > 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 huangchunlin1984 at 163.com Wed Sep 17 15:09:47 2008 From: huangchunlin1984 at 163.com (huangchunlin1984) Date: Wed, 17 Sep 2008 21:09:47 +0800 (CST) Subject: [Pw_forum] I am sorry. Message-ID: <11876247.512351221656987593.JavaMail.coremail@bj163app121.163.com> Dear doctor: I apologize to the uncomfortable words in my correspondence. I am very sorry. Today I send a correspondence to ask help and I am very happy to see your email coming. I feel sad when I find that I wrote an email impolitely. I am not mean to order anyone. I really respect all of you and the pw_forum. Please forgive me, and I will study English more accurately to avoid the same things. It's my fault. I am sorry. student: huang chun lin -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/9d5b8f52/attachment.htm From Steven.Kirk at hv.se Wed Sep 17 15:19:55 2008 From: Steven.Kirk at hv.se (Steven Kirk) Date: Wed, 17 Sep 2008 15:19:55 +0200 Subject: [Pw_forum] Al example for ensemble DFT in QE wiki for CP program Message-ID: <48D103FB.9010708@hv.se> Hello! There is a description in the QE wiki of methods for doing CP dynamics using ensemble DFT. The example is for an Aluminum surface. The example given is for a simulation cell of 21.694 x 21.694 x 49.009 bohr, containing 96 atoms, at 500K. Unfortunately the example instruction file provided contains only the information necessary for the last (e.g. the ensemble-DFT part). I would be extremely grateful to be pointed to a set of input files that contains all steps of this particular calculation, as important information (e.g. ATOMIC_POSITIONS for the initial configuration, equillibration time before the production run, etc. ) are missing. I intend to do something similar for an Ni surface, but above the melting point, to evaluate the surface tension of a liquid surface as a function of temperature. All advice (papers, input files, etc) would be very greatly appreciated. Many thanks in advance, Steve Kirk -- Dr. Steven R. Kirk Dept. of Technology, Mathematics & Computer Science (P)+46 520 223215 University West (F)+46 520 223299 Trollhattan 461 86 SWEDEN http://beacon.webhop.org From paulatto at sissa.it Wed Sep 17 17:01:44 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Wed, 17 Sep 2008 17:01:44 +0200 (CEST) Subject: [Pw_forum] problems computing cholesky decomposition In-Reply-To: <653592.2380.qm@web94615.mail.in2.yahoo.com> References: <653592.2380.qm@web94615.mail.in2.yahoo.com> Message-ID: <50986.147.122.5.32.1221663704.squirrel@webmail.sissa.it> On Wed, September 17, 2008 06:18, ambavale sagar wrote: > while using cell-dynamics = 'bfgs'. Hi Ambavale, This question have been asked a few times, several answers have been given, I think you can find some help by searching the mailing list archives. I have met those errors a lot of times, usually when the cell shrinks too much during a variable cell relaxation, and a few times when two atoms got very close toghether during a bfgs relaxation. Furthermore it usually always happens on the first iteration with bad pseudopotential (which is not your case, if you used one from the pwscf web site) Usually changing the initial conditions a bit solves the problem. e.g. taking a more reasonable first guess on the structure, or slowing down the initial steps of the relax algorithm. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From lmoreira at fc.unesp.br Wed Sep 17 20:29:29 2008 From: lmoreira at fc.unesp.br (Leandro Moreira de Campos Pinto) Date: Wed, 17 Sep 2008 15:29:29 -0300 Subject: [Pw_forum] Calculations for Au Message-ID: <20080917152929.14421psym7a0b349@webmail.fc.unesp.br> Dear all, I'm trying to perform a scf calculation for Au and I obtained a value of Fermi Energy about ~16 eV, and in the literature I found this value like ~5.5 eV. What should be causing this discrepancy? Another problem occur when I use pseudopotentials generated with PWGGA xc. The error is: "Some wfcs were lost". I will appreciate the suggestions. Best regards, leandro -- ***************************************** * Leandro Moreira de Campos Pinto * * Lab. Eletrocatalise, Fac. de Ciencias * * UNESP, P.O. Box 473 * * 17033-360, Bauru, SP, Brazil * * http://www.fc.unesp.br/grupo_dafc * ***************************************** ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From baroni at sissa.it Wed Sep 17 21:59:21 2008 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 17 Sep 2008 21:59:21 +0200 Subject: [Pw_forum] Calculations for Au In-Reply-To: <20080917152929.14421psym7a0b349@webmail.fc.unesp.br> References: <20080917152929.14421psym7a0b349@webmail.fc.unesp.br> Message-ID: Leandro: the value of the Fermi energy (as well as of any energy, for that matter) depends of the reference level. What you are referring to is probably the "Fermi energy referred to the vacuum level" (i.e. the work function). In order to obtain that, you need to know what the vacuum level is, which cannot be said from a bulk calculation only. This has been discussed many times on this forum, I think, as well as in many solid-state/electronic-structure/surface-science textbooks. If after reading some relevant literature you still do not understand why the value of the work function cannot be obtained form a bulk calculation, please revert to us and I will try to give you some more explanations ... cheers - SB On Sep 17, 2008, at 8:29 PM, Leandro Moreira de Campos Pinto wrote: > Dear all, > > I'm trying to perform a scf calculation for Au and I obtained a value > of Fermi Energy about ~16 eV, and in the literature I found this value > like ~5.5 eV. What should be causing this discrepancy? > > Another problem occur when I use pseudopotentials generated with PWGGA > xc. The error is: "Some wfcs were lost". > > I will appreciate the suggestions. > > Best regards, > > leandro > > > -- > ***************************************** > * Leandro Moreira de Campos Pinto * > * Lab. Eletrocatalise, Fac. de Ciencias * > * UNESP, P.O. Box 473 * > * 17033-360, Bauru, SP, Brazil * > * http://www.fc.unesp.br/grupo_dafc * > ***************************************** > > ---------------------------------------------------------------- > This message was sent using IMP, the Internet Messaging Program. > > > _______________________________________________ > 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 http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080917/bee05649/attachment.htm From sathya_sheela1985001 at rediffmail.com Thu Sep 18 06:23:12 2008 From: sathya_sheela1985001 at rediffmail.com (sathya subramanyan) Date: 18 Sep 2008 04:23:12 -0000 Subject: [Pw_forum] pseudopotential for Bi, PBE approximation Message-ID: <20080918042312.50550.qmail@f4mail-235-129.rediffmail.com> ? Dear Proff Eyvaz On Thu Mar 23 16:41:43 CET 2006 Sergy had asked for pseudopotential for Bi, PBE approximation to the forum and you replied that you have it. I checked the Pwscf for US Bi PBE pseudopotential but could not find or generate it myself . Thus could you please send me Bi US pseudopotential with PBE approximation. SATHYA SHEELA.S Grad. Student Department of Physics National Institute of Technology Tiruchirapalli - 620015 India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080918/04d81af1/attachment.htm From giannozz at democritos.it Thu Sep 18 08:53:09 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 18 Sep 2008 08:53:09 +0200 Subject: [Pw_forum] Calculations for Au In-Reply-To: <20080917152929.14421psym7a0b349@webmail.fc.unesp.br> References: <20080917152929.14421psym7a0b349@webmail.fc.unesp.br> Message-ID: <48D1FAD5.8040306@democritos.it> Leandro Moreira de Campos Pinto wrote: > Another problem occur when I use pseudopotentials generated with PWGGA > xc. The error is: "Some wfcs were lost". please provide a test job (with pseudopotentials). This is marked as 'internal error', meaning that it should never happen Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From paulatto at sissa.it Thu Sep 18 09:34:32 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Thu, 18 Sep 2008 09:34:32 +0200 Subject: [Pw_forum] Units of Charge Density - Again In-Reply-To: <01ff01c916df$38fc0b90$aaf422b0$@net> References: <01ff01c916df$38fc0b90$aaf422b0$@net> Message-ID: <48D20488.2000303@sissa.it> Paul M. Grant ha scritto: > > In a perfect SI world, these units would be coulombs per cubic meter. > In a practical, or rather intuitive, world, they would be in units of > total fractional number of electrons per unit cell volume (in cubic > angstroms or nanometers or bohrs (ugh)). So...what is it? And does it > depend on the division of states between core and valence taken when > constructing a given pseudopotential? > The internal unit is q/bohr^3, where q is the electron charge. It is a bit awkward, but the correct unit of measure depend on the output file format. but when it is exported to xcd files the cell axis are rescaled to Angstrom, but the charge density is *not* rescaled, hence the unit of charge density in the xcd files is q/A^3 * (A^3/bohr^3). In order to get a consistent representation inside the xcd file you should multiply rho times (bohr/A)^3 = 0.14818474. In other kind of output files there should be no such problem (but I've not checked all of them). Furthermore, xcd format is not suitable for any subsequent postprocessing, as it is written with only 6 significant digits. > Finally, I've been unable (by visual inspection) to determine from the > documentation for XcrySDen (a marvelous tool...marvelous), exactly how > the "rho grid input" is scaled, especially regarding the computation > of the 3D min-max grid values, the "isovalue"...and...particularly the > mysterious "delta n(r)" displayed in the temperature inset box. > I don't know if it's scaled at all, I think it is at most a side effect of interpolation. If you use the minimum level of interpolation do you get the extremes you expected? Anyway, I think it is better to wait for Tone to reply on this topic regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From paulatto at sissa.it Thu Sep 18 09:42:30 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Thu, 18 Sep 2008 09:42:30 +0200 Subject: [Pw_forum] [Fwd: trying to compile espresso4.0.2] In-Reply-To: <61314.200.17.215.86.1221501329.squirrel@www.quimica.ufpr.br> References: <61314.200.17.215.86.1221501329.squirrel@www.quimica.ufpr.br> Message-ID: <48D20666.7050102@sissa.it> Dear Eduardo, it looks like the configure script has correctly detected the fftw libraries you have installed in the system, but they were compiled with a different compiler than the one you are using to compile QE, and they are incompatible. Probably system's fftw were compiled with gfortran. Instruct QE to use internal fftw libraries should solve the problem (there is no difference in performance). Please notice that linking QE with atlas libraries may be very tricky, and be sure to use the last atlas version, as the older ones contained a couple of bugs which caused random crashes and NaN's in the code. If you have any more problems don't be afraid to ask on the list, some time ago I've tried to make the compilation of QE + atlas easier; I've never released the patches for apparent lack of interest (and possibility of unwanted side-effects), but they may come handy to you. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From paulatto at sissa.it Thu Sep 18 09:46:00 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Thu, 18 Sep 2008 09:46:00 +0200 Subject: [Pw_forum] [Fwd: trying to compile espresso4.0.2] In-Reply-To: <48D20666.7050102@sissa.it> References: <61314.200.17.215.86.1221501329.squirrel@www.quimica.ufpr.br> <48D20666.7050102@sissa.it> Message-ID: <48D20738.6010501@sissa.it> Rereading your log I've noticed that this part of my reply is completely wrong: > it looks like the configure script has correctly detected the fftw > libraries you have installed in the system, but they were compiled with > a different compiler than the one you are using to compile QE, and they > are incompatible. a `make clean` (like Paolo suggested) should solve the problem. -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From sathya_sheela1985001 at rediffmail.com Thu Sep 18 11:09:17 2008 From: sathya_sheela1985001 at rediffmail.com (sathya subramanyan) Date: 18 Sep 2008 09:09:17 -0000 Subject: [Pw_forum] monoclinic structure with unique c axis Message-ID: <20080918090917.39825.qmail@f4mail-235-245.rediffmail.com> ? Dear All, I'm trying to optimize a monoclinic structure(C2/c). I have the experimental atomic postions(wycoff)with b axis as unique.I think in espresso the monoclinic structure positions have to be given taking c axis as unique. My question is can the y axis value(b axis unique) of 4e wycoff site be taken as the z axis value of 4e site (with unique axis c).I have put the wycoff postions both with c and b a=xis unique for C2/c structure. Please do help me out.Thank You in advance Wyckoff Positions of Group 15 (C2/c) [unique axis c] Multiplicity Coordinates 8 f 1 (x,y,z) (-x+1/2,-y,z) (-x,-y,-z) (x+1/2,y,-z) 4 e 2 (1/4,0,z) (3/4,0,-z) 4 d -1 (1/2,1/4,1/4) (0,3/4,1/4) 4 c -1 (0,1/4,1/4) (1/2,3/4,1/4) 4 b -1 (0,0,1/2) (1/2,0,1/2) 4 a -1 (0,0,0) (1/2,0,0) Wyckoff Positions of Group 15 (C2/c) [unique axis b] Multiplicity Coordinates 8 f 1 (x,y,z) (-x,y,-z+1/2) (-x,-y,-z) (x,-y,z+1/2) 4 e 2 (0,y,1/4) (0,-y,3/4) 4 d -1 (1/4,1/4,1/2) (3/4,1/4,0) 4 c -1 (1/4,1/4,0) (3/4,1/4,1/2) 4 b -1 (0,1/2,0) (0,1/2,1/2) 4 a -1 (0,0,0) (0,0,1/2) Regards SATHYA SHEELA.S Grad. Student Department of Physics National Institute of Technology Tiruchirapalli - 620015 India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080918/f74d8fc5/attachment.htm From J.Kendrick at Bradford.ac.uk Thu Sep 18 12:49:06 2008 From: J.Kendrick at Bradford.ac.uk (John Kendrick) Date: Thu, 18 Sep 2008 11:49:06 +0100 Subject: [Pw_forum] 3rd Order Properties and GGA's Message-ID: <48D23222.7050909@bradford.ac.uk> Dear All, I have recently been looking at calculating Raman scattering intensities at q=0 for organic crystals. I was aware that ultra soft potentials are not supported for the calculation of 3rd order properties. I recently tried out the CVS version of Quantum Espresso and discovered that there is a check in phq_setup.f90 that aborts the calculation if the density functional is a GGA. No such check exists in 4.0.1. I had a look at the manual and wiki, but I can't find this restriction described anywhere. Could someone let me know what the current restrictions are on the calculation of Raman scattering intensities at q=0? Thanks for your help. John Kendrick From giannozz at democritos.it Thu Sep 18 16:16:54 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 18 Sep 2008 16:16:54 +0200 Subject: [Pw_forum] 3rd Order Properties and GGA's In-Reply-To: <48D23222.7050909@bradford.ac.uk> References: <48D23222.7050909@bradford.ac.uk> Message-ID: <48D262D6.6070300@democritos.it> John Kendrick wrote: > > I have recently been looking at calculating Raman scattering intensities > at q=0 for organic crystals. I was aware that ultra soft potentials are > not supported for the calculation of 3rd order properties. I recently > tried out the CVS version of Quantum Espresso and discovered that there > is a check in phq_setup.f90 that aborts the calculation if the density > functional is a GGA. No such check exists in 4.0.1. the problem is that the calculation of third derivatives wrt the charge density of the exchange-correlation energy was never implemented. Only the LDA part is calculated, and only for the Perdew-Zunger parameterization. I do not expect this to introduce a big error, though, so you might try (at your own risk) to comment out that check Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From kazempoor2000 at yahoo.com Thu Sep 18 18:37:45 2008 From: kazempoor2000 at yahoo.com (ali kazempoor) Date: Thu, 18 Sep 2008 09:37:45 -0700 (PDT) Subject: [Pw_forum] bent system Message-ID: <752289.18951.qm@web33108.mail.mud.yahoo.com> Dear all ?Is there (in dft codes like espresso) any method like supercell method to simulate bent system like nanobelt(bent one dimensional system)? or it must be? used?from ?non-periodic condition (open boundary condition)? thanks Ali kazempour,Isfahan university of technology,Isfahan, Iran? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080918/a04629ce/attachment.htm From lanhaiping at gmail.com Thu Sep 18 19:47:17 2008 From: lanhaiping at gmail.com (lan haiping) Date: Fri, 19 Sep 2008 01:47:17 +0800 Subject: [Pw_forum] bent system In-Reply-To: <752289.18951.qm@web33108.mail.mud.yahoo.com> References: <752289.18951.qm@web33108.mail.mud.yahoo.com> Message-ID: Hi, for periodic boudary code, the only possible way is to build a large enough cell . It means that you can simulate something 'periodic ' bent systems. By the way, BigDFT in abinit community can handle with free boundary systems. On Fri, Sep 19, 2008 at 12:37 AM, ali kazempoor wrote: > Dear all > Is there (in dft codes like espresso) any method like supercell method to > simulate bent system like nanobelt(bent one dimensional system)? > or it must be used from non-periodic condition (open boundary condition)? > thanks > Ali kazempour,Isfahan university of technology,Isfahan, Iran > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080919/33203d97/attachment.htm From polynmr at physics.unc.edu Thu Sep 18 21:48:39 2008 From: polynmr at physics.unc.edu (Charles Chen) Date: Thu, 18 Sep 2008 15:48:39 -0400 Subject: [Pw_forum] bent system In-Reply-To: References: <752289.18951.qm@web33108.mail.mud.yahoo.com> Message-ID: <48D2B097.5000302@physics.unc.edu> lan haiping wrote: > > Hi, for periodic boudary code, the only possible way is to > build a large enough cell . It means that you can simulate > something 'periodic ' bent systems. > > By the way, BigDFT in abinit community can handle > with free boundary systems. I think the problem is not just free boundary, it is how many atoms the code can deal with. Charles > > On Fri, Sep 19, 2008 at 12:37 AM, ali kazempoor > > wrote: > > Dear all > Is there (in dft codes like espresso) any method like supercell > method to simulate bent system like nanobelt(bent one dimensional > system)? > or it must be used from non-periodic condition (open boundary > condition)? > thanks > Ali kazempour,Isfahan university of technology,Isfahan, Iran > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > -- > Hai-Ping Lan > Department of Electronics , > Peking University , Bejing, 100871 > lanhaiping at gmail.com , hplan at pku.edu.cn > > ------------------------------------------------------------------------ > > Spam > Not spam > Forget previous vote > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- NMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMR NMR Dr. Qiang "Charles" Chen NMR NMR Department of Physics and Astronomy NMR NMR University of North Carolina NMR NMR Phillips Hall, CB#3255 NMR NMR Chapel Hill, NC 27599-3255 NMR NMR Tel: 919-962-1571 (O) NMR NMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMR From anna.ferrari at unito.it Fri Sep 19 12:29:21 2008 From: anna.ferrari at unito.it (Anna Maria Ferrari) Date: Fri, 19 Sep 2008 11:29:21 +0100 Subject: [Pw_forum] bent system References: <752289.18951.qm@web33108.mail.mud.yahoo.com> <48D2B097.5000302@physics.unc.edu> Message-ID: <002b01c91a42$965227a0$2376c082@desdemona> i'm not sure what you need but there are codes, no plane waves, (CRYSTAL, GAUSSIAN) that can deal with one-dimensional periodic system with supercell (new need of course a translationally invariant unit to be selected) anna ferrari ----- Original Message ----- From: "Charles Chen" To: "PWSCF Forum" Sent: Thursday, September 18, 2008 8:48 PM Subject: Re: [Pw_forum] bent system > lan haiping wrote: >> >> Hi, for periodic boudary code, the only possible way is to >> build a large enough cell . It means that you can simulate >> something 'periodic ' bent systems. >> >> By the way, BigDFT in abinit community can handle >> with free boundary systems. > I think the problem is not just free boundary, it is how many atoms the > code can deal with. > > Charles >> >> On Fri, Sep 19, 2008 at 12:37 AM, ali kazempoor >> > wrote: >> >> Dear all >> Is there (in dft codes like espresso) any method like supercell >> method to simulate bent system like nanobelt(bent one dimensional >> system)? >> or it must be used from non-periodic condition (open boundary >> condition)? >> thanks >> Ali kazempour,Isfahan university of technology,Isfahan, Iran >> >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> >> >> -- >> Hai-Ping Lan >> Department of Electronics , >> Peking University , Bejing, 100871 >> lanhaiping at gmail.com , hplan at pku.edu.cn >> >> ------------------------------------------------------------------------ >> >> Spam >> Not spam >> Forget previous vote >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> > > > -- > NMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMR > NMR Dr. Qiang "Charles" Chen NMR > NMR Department of Physics and Astronomy NMR > NMR University of North Carolina NMR > NMR Phillips Hall, CB#3255 NMR > NMR Chapel Hill, NC 27599-3255 NMR > NMR Tel: 919-962-1571 (O) NMR > NMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMRNMR > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > From vegalew at hotmail.com Fri Sep 19 18:54:08 2008 From: vegalew at hotmail.com (vega) Date: Sat, 20 Sep 2008 00:54:08 +0800 Subject: [Pw_forum] why my pw.x run with low efficiency? Message-ID: Dear all, I just finished a relax calculation for 120 atoms. After calculation was done, the outputfile reported as follows, Program PWSCF v.4.0.1 starts ... Today is 16Sep2008 at 19:14:42 Parallel version (MPI) Number of processors in use: 78 K-points division: npool = 3 R & G space division: proc/pool = 26 For Norm-Conserving or Ultrasoft (Vanderbilt) Pseudopotentials or PAW ................................ per-process dynamical memory: 129.6 Mb ................................ PWSCF : 0d 14h46m CPU time, 2d 18h 4m wall time init_run : 91.49s CPU electrons : 47137.56s CPU ( 27 calls,1745.836 s avg) update_pot : 187.80s CPU ( 26 calls, 7.223 s avg) forces : 4492.20s CPU ( 27 calls, 166.378 s avg) Called by init_run: wfcinit : 23.68s CPU potinit : 3.15s CPU Called by electrons: c_bands : 23198.29s CPU ( 258 calls, 89.916 s avg) sum_band : 11159.67s CPU ( 258 calls, 43.255 s avg) v_of_rho : 167.39s CPU ( 280 calls, 0.598 s avg) newd : 13679.79s CPU ( 280 calls, 48.856 s avg) mix_rho : 30.14s CPU ( 258 calls, 0.117 s avg) Called by c_bands: init_us_2 : 48.95s CPU ( 517 calls, 0.095 s avg) cegterg : 23038.94s CPU ( 258 calls, 89.298 s avg) Called by *egterg: h_psi : 8629.82s CPU ( 1459 calls, 5.915 s avg) s_psi : 2230.78s CPU ( 1459 calls, 1.529 s avg) g_psi : 34.68s CPU ( 1200 calls, 0.029 s avg) cdiaghg : 5929.74s CPU ( 1427 calls, 4.155 s avg) Called by h_psi: add_vuspsi : 2209.17s CPU ( 1459 calls, 1.514 s avg) General routines calbec : 2904.12s CPU ( 1744 calls, 1.665 s avg) cft3s : 4337.89s CPU ( 950068 calls, 0.005 s avg) interpolate : 34.87s CPU ( 538 calls, 0.065 s avg) Parallel routines fft_scatter : 538.83s CPU ( 950068 calls, 0.001 s avg) From paulatto at sissa.it Sat Sep 20 02:00:12 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Sat, 20 Sep 2008 02:00:12 +0200 (CEST) Subject: [Pw_forum] why my pw.x run with low efficiency? In-Reply-To: References: Message-ID: <27030.82.50.174.28.1221868812.squirrel@webmail.sissa.it> On Ven, Settembre 19, 2008 18:54, vega wrote: > PWSCF : 0d 14h46m CPU time, 2d 18h 4m wall > This is indeed low! > Could you tell me what make my CPUs run in a such low efficiency style? I cannot be sure, but I will try to give a reasonable guess > Do you think 10G infiniband is good enough for 39 nodes? I'm not an hardware junkie, but infiniband has a fame to be quite performant, it should be ok. > Could tell me which folders must be on a NFS location so that all the > nodes can load and write? Only the pseudopotentials and the input files, all the rest can be written on local scratch. If you are using the wf_collect option, you may have to use a network file system for the wavefunctions (I think there is an input variable to specify the wavefunctions directory). I'm not 100% sure about that, and it may be improved in future versions... > Do you think the pw.x underestimate greatly for the memory? It is possible indeed, the estimate is hardcoded by hand, unluckily many of the contributors (me included) haven't updated it to reflect their changes to the code. > any hints on my problem will be deeply appreciated. Youare already using the maximum number of k-points (one per pool), there is little you can do. Do not set the outdir to a network filesystem will help, especially if the filesystem passes on the same network as mpi. Maybe openmpi is more efficient than the mpich, but I don't really know it (I'm just assuming than new is better than old). cheers -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From zdw2000 at gmail.com Sat Sep 20 02:38:35 2008 From: zdw2000 at gmail.com (Wei Zhou) Date: Sat, 20 Sep 2008 08:38:35 +0800 Subject: [Pw_forum] Na pseudo potential with d states? Message-ID: <3e36a8fc0809191738m74fb566bj3b328ccdcf57bbe@mail.gmail.com> hell every one I want to do some calcualations about sodium under ultrahigh pressure , and the d shell became important ,but I can not find the corresponding pseudo potential? what I need is sodium pseudo potential with 2s2 2p6 3p d0,would someone can give me a such pseudo potential? LDA or GGA . thank you in advance,any help will be appreciated. -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080920/d76bc4f1/attachment.htm From vegalew at hotmail.com Sat Sep 20 05:30:02 2008 From: vegalew at hotmail.com (vega) Date: Sat, 20 Sep 2008 11:30:02 +0800 Subject: [Pw_forum] why my pw.x run with low efficiency? In-Reply-To: <27030.82.50.174.28.1221868812.squirrel@webmail.sissa.it> References: <27030.82.50.174.28.1221868812.squirrel@webmail.sissa.it> Message-ID: Dear sir, Thank you for your suggestion. >> Do you think 10G infiniband is good enough for 39 nodes? > > I'm not an hardware junkie, but infiniband has a fame to be quite > performant, it should be ok. thank god... > on local scratch. If you are using the wf_collect option, you may have to > use a network file system for the wavefunctions (I think there is an input > variable to specify the wavefunctions directory). to reduce the IO, I didn't use it. > is little you can do. Do not set the outdir to a network filesystem will > help, especially if the filesystem passes on the same network as mpi. > Maybe openmpi is more efficient than the mpich, but I don't really know it > (I'm just assuming than new is better than old). OK, I'll try to set a outdir to a local location. I used to run QE with openmpi but openmpi seems to be something wrong with invoking the virtual memory for my cluster. I think so. Only mpich2 could run the calculation with large systems up to 120 atoms or more. Do you think I should copy the compilers, MKL, fftw, mpi, and QE files to local disk for every node? Do you think this will reduce the IO greatly? thank you for reading vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------------------------------------------- From: "Lorenzo Paulatto" Sent: Saturday, September 20, 2008 8:00 AM To: "PWSCF Forum" Subject: Re: [Pw_forum] why my pw.x run with low efficiency? > > On Ven, Settembre 19, 2008 18:54, vega wrote: >> PWSCF : 0d 14h46m CPU time, 2d 18h 4m wall > > > This is indeed low! > >> Could you tell me what make my CPUs run in a such low efficiency style? > > I cannot be sure, but I will try to give a reasonable guess > >> Do you think 10G infiniband is good enough for 39 nodes? > > I'm not an hardware junkie, but infiniband has a fame to be quite > performant, it should be ok. > >> Could tell me which folders must be on a NFS location so that all the >> nodes can load and write? > > Only the pseudopotentials and the input files, all the rest can be written > on local scratch. If you are using the wf_collect option, you may have to > use a network file system for the wavefunctions (I think there is an input > variable to specify the wavefunctions directory). I'm not 100% sure about > that, and it may be improved in future versions... > >> Do you think the pw.x underestimate greatly for the memory? > > It is possible indeed, the estimate is hardcoded by hand, unluckily many > of the contributors (me included) haven't updated it to reflect their > changes to the code. > >> any hints on my problem will be deeply appreciated. > > Youare already using the maximum number of k-points (one per pool), there > > cheers > > -- > Lorenzo Paulatto > SISSA & DEMOCRITOS (Trieste) > phone: +39 040 3787 511 > skype: paulatz > www: http://people.sissa.it/~paulatto/ > > > > ---------------------------------------------------------------- > SISSA Webmail https://webmail.sissa.it/ > Powered by SquirrelMail http://www.squirrelmail.org/ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From paulatto at sissa.it Sat Sep 20 08:41:57 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Sat, 20 Sep 2008 08:41:57 +0200 (CEST) Subject: [Pw_forum] why my pw.x run with low efficiency? In-Reply-To: References: <27030.82.50.174.28.1221868812.squirrel@webmail.sissa.it> Message-ID: <28704.82.50.174.28.1221892917.squirrel@webmail.sissa.it> On Sab, Settembre 20, 2008 05:30, vega wrote: > Do you think I should copy the compilers, MKL, fftw, mpi, and QE files to > local disk for every node? Do you think this will reduce the IO greatly? > No, I don't think it would change at all (maybe a few seconds at the beginning of the job) -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From lawrence_lee_lee at yahoo.com.hk Sat Sep 20 16:43:41 2008 From: lawrence_lee_lee at yahoo.com.hk (Lawrence Lee) Date: Sat, 20 Sep 2008 22:43:41 +0800 Subject: [Pw_forum] Any convenient tool to find the phonon DOS? Message-ID: <1221921821.6273.11.camel@scg11> Dear all, I have calculated the phonon dispersion for certain material, and would like to compute the phonon DOS. Are there any convenient codes/tools to perform this? -- S.H. Lee Research Assistant Physics Division The CUHK Graduate School _______________________________________ ?x?????? - Yahoo! Messenger ???????]?????W?????????????????????????o???????????W?r???????????????????f???????????? http://messenger.yahoo.com.hk From eyvaz_isaev at yahoo.com Sat Sep 20 21:37:04 2008 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sat, 20 Sep 2008 12:37:04 -0700 (PDT) Subject: [Pw_forum] Any convenient tool to find the phonon DOS? In-Reply-To: <1221921821.6273.11.camel@scg11> Message-ID: <486315.72679.qm@web65714.mail.ac4.yahoo.com> Dear Lawrence, Please have a look at /examples/example07/run_example for calculation of the total phonon DOS. If you need atom projected DOS, then I can send you the QHA code desined to calculate some thermodynamic properties. Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com --- On Sat, 9/20/08, Lawrence Lee wrote: > From: Lawrence Lee > Subject: [Pw_forum] Any convenient tool to find the phonon DOS? > To: "PW_Forum" > Date: Saturday, September 20, 2008, 6:43 PM > Dear all, > > I have calculated the phonon dispersion for certain > material, and would > like to compute the phonon DOS. Are there any convenient > codes/tools to > perform this? > > -- > S.H. Lee > Research Assistant > Physics Division > The CUHK Graduate School > > _______________________________________ > ???? - Yahoo! Messenger > > ??????????????????????????????????????????? > http://messenger.yahoo.com.hk > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From wangqj1 at 126.com Sun Sep 21 05:49:41 2008 From: wangqj1 at 126.com (wangqj1) Date: Sun, 21 Sep 2008 11:49:41 +0800 (CST) Subject: [Pw_forum] Are there the instruction of every modules ? Message-ID: <23663678.47301221968981420.JavaMail.coremail@bj126app103.126.com> Dear pwscf developers Do you have the intruction of evrey modules ? with the instructions ,PWSCF users can study quickly .Yes,there are some examples in the example3.2.3,but it's not enough if deeply study . Can you give some more information ? Your advice will be appreciated. sincerely -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080921/107c4007/attachment.htm From niuli1978 at yahoo.com.cn Sun Sep 21 10:59:04 2008 From: niuli1978 at yahoo.com.cn (li niu) Date: Sun, 21 Sep 2008 16:59:04 +0800 (CST) Subject: [Pw_forum] about partial Raman Message-ID: <541770.9522.qm@web15005.mail.cnb.yahoo.com> Hello pwscf users, I am doing Raman calculation on boron doped diamond by second-order response. I am using the dynmat.x code to obtain Raman cross sections, but how to perform a Raman projected on specific atoms? (partial Raman) thanks Li Niu Harbin Institue of Technology China --------------------------------- ???????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080921/086fd7f2/attachment.htm From vegalew at hotmail.com Sun Sep 21 11:44:15 2008 From: vegalew at hotmail.com (vega) Date: Sun, 21 Sep 2008 17:44:15 +0800 Subject: [Pw_forum] why my pw.x run with low efficiency? In-Reply-To: <27030.82.50.174.28.1221868812.squirrel@webmail.sissa.it> <28704.82.50.174.28.1221892917.squirrel@webmail.sissa.it> References: <27030.82.50.174.28.1221868812.squirrel@webmail.sissa.it> <28704.82.50.174.28.1221892917.squirrel@webmail.sissa.it> Message-ID: Dear sir, thank you for your advice. I set the ourdir at local disk following your suggestions. and I also set the disk_io='none' and no wf_collect options. But the pw.x still run at low efficiency. I think it's most probable because of lack of memory. Most of time was wasted by exchanging the data between the virtual memory and physical memory. So could you estimate that how many memory will be good enough for 40 [TiO2] units( 120 atoms)? vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------------------------------------------- From: "Lorenzo Paulatto" Sent: Saturday, September 20, 2008 2:41 PM To: "PWSCF Forum" Subject: Re: [Pw_forum] why my pw.x run with low efficiency? > > On Sab, Settembre 20, 2008 05:30, vega wrote: >> Do you think I should copy the compilers, MKL, fftw, mpi, and QE files >> to >> local disk for every node? Do you think this will reduce the IO greatly? >> > > No, I don't think it would change at all (maybe a few seconds at the > beginning of the job) > > -- > Lorenzo Paulatto > SISSA & DEMOCRITOS (Trieste) > phone: +39 040 3787 511 > skype: paulatz > www: http://people.sissa.it/~paulatto/ > > > > ---------------------------------------------------------------- > SISSA Webmail https://webmail.sissa.it/ > Powered by SquirrelMail http://www.squirrelmail.org/ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From paulatto at sissa.it Sun Sep 21 13:59:54 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Sun, 21 Sep 2008 13:59:54 +0200 (CEST) Subject: [Pw_forum] why my pw.x run with low efficiency? In-Reply-To: References: <27030.82.50.174.28.1221868812.squirrel@webmail.sissa.it> <28704.82.50.174.28.1221892917.squirrel@webmail.sissa.it> Message-ID: <15583.82.50.174.28.1221998394.squirrel@webmail.sissa.it> On Dom, Settembre 21, 2008 11:44, vega wrote: > So could you estimate > that how many memory > will be good enough for 40 [TiO2] units( 120 atoms)? No way, it depends strongly on cutoff and cell size but weakly on number and types of atoms. Furthermore, I can't do it even with the correct data avilable: if I could I would put it in the code! I'm not really convinced that you have a swap problem, truly your performance are degraded, but not as much as I would expect from swap. You should ask the sys administrator to disable swap usage, or you can use the following hack to prevent the code from using swap space. Before running the pw code you can set the max memory limit per process equal to the amount of free RAM: FREE_RAM=$(free|grep buffers/cache|awk '{print $4}') ulimit -v $FREE_RAM In this way if the program will try to use more RAM than available it will be killed, and you'll surely notice. Read the ``ulimit'' help and ``free '' manpage for details on what you are doing. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From manoj at phys.ufl.edu Sun Sep 21 19:43:43 2008 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Sun, 21 Sep 2008 13:43:43 -0400 (EDT) Subject: [Pw_forum] array bound check Message-ID: Dear PWSCF users, I have a question about array bound check. How can I compile espresso, with array bound check. Is this some kind of switch, I need to turn it on. If so, where can I find it? Any help will be appreciated. Regards, Manoj Srivastava Graduate Student Deapartment of Physics, University of Florida, Gainesville, USA. From lanhaiping at gmail.com Sun Sep 21 21:36:06 2008 From: lanhaiping at gmail.com (lan haiping) Date: Mon, 22 Sep 2008 03:36:06 +0800 Subject: [Pw_forum] About compilation of cvs version Message-ID: Dear Developers: I just checked out source code via cvs, and found a compilation error due to routine ptoolkit.f90 with information below : " IF( ( ii - 1 )/nproc + 1 > SIZE(a,1) ) CALL errore( msg, ' ( ii - 1 )/npr The upper bound in the last dimension must appear in the reference to the assumed size array 'a' at (1). " My compilor is gfortran v4.1.2. I then examined this routine ,and found subroutine check_sndbuf_index only called by cyc2blk_*redist in parallel computation. But no mpi macros are defined for this subroutine. I just wonder we should modify this routine by adding mpi macro as following : #if defined (_MPI) IF( j > SIZE(sndbuf,2) ) CALL errore( msg, ' j > SIZE(sndbuf,2) ', ip+1 ) IF( il > SIZE(sndbuf,1) ) CALL errore( msg, ' il > SIZE(sndbuf,1) ', ip+1 ) IF( ( ii - 1 )/nproc + 1 < 1 ) CALL errore( msg, ' ( ii - 1 )/nproc + 1 < 1 ', ip+1 ) IF( ( ii - 1 )/nproc + 1 > SIZE(a,1) ) CALL errore( msg, ' ( ii - 1 )/nproc + 1 > SIZE(a,1) ', ip+1 ) IF( jj < 1 ) CALL errore( msg, ' jj < 1 ', ip+1 ) IF( jj > n ) CALL errore( msg, ' jj > n ', ip+1 ) #endif RETURN I then recompiled it, and found no error occured again during compiling. I would like to know whether it is reasonable by adding this mpi macro for ptoolkit subroutine. Best, Hai-Ping -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080922/6b148915/attachment.htm From giannozz at democritos.it Sun Sep 21 22:32:21 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sun, 21 Sep 2008 22:32:21 +0200 Subject: [Pw_forum] array bound check In-Reply-To: References: Message-ID: <037B80E5-F912-43E1-AB28-EC4895A5C70E@democritos.it> On Sep 21, 2008, at 19:43 , Manoj Srivastava wrote: > How can I compile espresso, with array bound check. it depends on the compiler you are using Paolo -- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at democritos.it Sun Sep 21 22:51:54 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sun, 21 Sep 2008 22:51:54 +0200 Subject: [Pw_forum] why my pw.x run with low efficiency? In-Reply-To: References: Message-ID: <7FE9D3D1-A41D-44E5-A06B-34138FAC9F2F@democritos.it> On Sep 19, 2008, at 18:54 , vega wrote: > PWSCF : 0d 14h46m CPU time, 2d 18h 4m > wall time I am quite sure Axel Kohlmeeyer has already answered 2n+1 times to the sameor similar question. Please look in particular for OMP_NUM_THREADS in the archives of the mailing list Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From akohlmey at cmm.chem.upenn.edu Mon Sep 22 00:08:30 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sun, 21 Sep 2008 18:08:30 -0400 (EDT) Subject: [Pw_forum] why my pw.x run with low efficiency? In-Reply-To: <7FE9D3D1-A41D-44E5-A06B-34138FAC9F2F@democritos.it> References: <7FE9D3D1-A41D-44E5-A06B-34138FAC9F2F@democritos.it> Message-ID: On Sun, 21 Sep 2008, Paolo Giannozzi wrote: PG> PG> On Sep 19, 2008, at 18:54 , vega wrote: PG> PG> > PWSCF : 0d 14h46m CPU time, 2d 18h 4m PG> > wall time PG> PG> I am quite sure Axel Kohlmeeyer has already answered 2n+1 times PG> to the sameor similar question. Please look in particular for PG> OMP_NUM_THREADS in the archives of the mailing list sorry paolo, but that is only the case when the CPU time is much _higher_ than the wall time. here it looks as if the jobs is either swapping like crazy or the communication is stalling. on that note, it would be nice to also see the wall time spent in the individual routines, as the cpu time is usually a somewhat inadequate descriptor, except for serial calculations. i'd have commented on the discussion earlier on, as several assessments are unsubstantiated or don't make much sense at all. but the only way to find out for sure what is happening, would be to run that very same job on a machine where i know for certain that the hardware and software is set up correctly and QE is compiled in the best possible way. this however is very unpractical from addis ababa where reasonable internet access is only available intermittently and at very high costs. thus only a bunch of questions regarding observations and explanations. vega mentioned that openmpi didn't work because of "lack of memory". i suspect that this is due to incorrect setup of the infiniband fabric and user limits. ulimit -a should produce on the compute nodes something like this. particularly the "max locked memory" entry is very important, and not setting it high enough will result in severely degraded performance (cf. ofed documentation). core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited max nice (-e) 0 file size (blocks, -f) unlimited pending signals (-i) 71680 max locked memory (kbytes, -l) 4096000 max memory size (kbytes, -m) unlimited open files (-n) 1024 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) 819200 max rt priority (-r) 0 stack size (kbytes, -s) 1024000 cpu time (seconds, -t) unlimited max user processes (-u) 71680 virtual memory (kbytes, -v) unlimited file locks (-x) unlimited another option to enhance OpenMPI performance for QE over infiniband, while reducing its memory requirements is to use the flag: --mca btl_openib_use_srq 1 with mpirun or set it for all jobs in the etc/openmpi-mca-params.conf within the openmpi installation directory. it was then mentioned, that the job was calculated using MPICH instead, which brings up the question, does this MPICH compile actually support infiniband. as far as i know, not. the usual OFED distributions contain either OpenMPI or MVAPICH. the next step to evaluate is whether the job actually does scale that far, and if memory requirements are really killing the performance. the reported virtual memory size is a bad indicator, as it reports the reserved address space, but does not say anything about how much of it is actually used. with openmpi over infiniband, this tends to always be ridiculously high, but that has no impact on the performance (symptom: large usage of SWAP for the process without swap actually being used). the real information about lack of available physical memory would come from comparing the data set size with the resident set size (you have to change the default configuration of top to see it). also one can see how much swapping is going on by simply logging into the node and monitoring whether kswapd processes actually consume a lot of time. in the (somewhat unlikely) case that one is running out of memory, it may help to not use all processors per node. i don't remember what the hardware exactly was, but, e.g., on nodes with 2x intel quad core processors, the (absolute!) performance and scaling is much better when running with half the available cpu cores. i have posted some benchmark graphs on cp2k here: http://cp2k.googlegroups.com/web/cp2k-water-bench-cmm.png but the outcome is similar for other DFT codes (CPMD, Q-E). using less MPI tasks should reduce the total memory requirements a bit, since not all data in QE can be or is distributed amongst the nodes. finally, it may just be that the input at hand doesn't scale to that many cpus, so in case of seeing performance problems it always pays off to check whether running with less processors or nodes is actually faster. bigger is not always better!! in summary, giving a recommendation on what is causing the degraded performance, is impossible without knowing more details and particularly a more systematic performance analysis. i've given already an example of how to do this a while ago, comparing scaling behavior with using -npools and without and comparing gigabit to myrinet and demonstrating the scaling limits of QE in both cases. i suggest to look up that e-mail from the mailing list archives and do a similar assessment with special consideration of the additional concerns listed in this e-mail. cheers, axel. PG> PG> Paolo PG> --- PG> Paolo Giannozzi, Dept of Physics, University of Udine PG> via delle Scienze 208, 33100 Udine, Italy PG> Phone +39-0432-558216, fax +39-0432-558222 PG> PG> PG> PG> _______________________________________________ PG> Pw_forum mailing list PG> Pw_forum at pwscf.org PG> http://www.democritos.it/mailman/listinfo/pw_forum PG> -- ======================================================================= 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 vegalew at hotmail.com Mon Sep 22 05:31:52 2008 From: vegalew at hotmail.com (vega) Date: Mon, 22 Sep 2008 11:31:52 +0800 Subject: [Pw_forum] why my pw.x run with low efficiency? In-Reply-To: <7FE9D3D1-A41D-44E5-A06B-34138FAC9F2F@democritos.it> References: <7FE9D3D1-A41D-44E5-A06B-34138FAC9F2F@democritos.it> Message-ID: Dear sir, Thank you so much for your responding. I do appreciate your help. > than the wall time. here it looks as if the jobs is either > swapping like crazy or the communication is stalling. on that I believe it. So do you think 10G infiniband is good enough for my job? By the way, there is also another two parallel job on the line. one is lammps, a classical MD code. the other is VASP. the lammps and vasp are using 32 cpus in total. my job is using 39 nodes and two cpu for each node. >vega mentioned that openmpi didn't work because > of "lack of memory". i suspect that this is due to incorrect > setup of the infiniband fabric and user limits. ulimit -a > should produce on the compute nodes something like this. > particularly the "max locked memory" entry is very important, > and not setting it high enough will result in severely > degraded performance (cf. ofed documentation). I'm really admiring your experiences and sense about the parallel job. the max lock memory option on my machines is 4. Now I asked my system administrator to reset it to 1024000. My physical memory is 2G. Do you think set it to 2048000 will be better? I'll try openmpi again, later, using flags --mca btl_openib_use_srq 1 with mpirun. By the way I'm using MPICH2 now, not MPICH. > to see it). also one can see how much swapping is going on by > simply logging into the node and monitoring whether kswapd > processes actually consume a lot of time. I checked it as your suggested. kswapd did not consume a lot of time. I think the problem was probably caused by the network. > in the (somewhat unlikely) case that one is running out of > memory, it may help to not use all processors per node. i don't your words remind me of a previous job. I run the job for one thread per core, the job failed. When I reduce the threads, the job can be done smoothly. > finally, it may just be that the input at hand doesn't scale > to that many cpus, so in case of seeing performance problems > it always pays off to check whether running with less processors > or nodes is actually faster. bigger is not always better!! > limits of QE in both cases. i suggest to look up that e-mail > from the mailing list archives and do a similar assessment > with special consideration of the additional concerns listed > in this e-mail. Thank you so much for sharing your knowledge and experiences to me. best wishes, vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------------------------------------------- From: "Axel Kohlmeyer" Sent: Monday, September 22, 2008 6:08 AM To: "Paolo Giannozzi" Cc: "PWSCF Forum" Subject: Re: [Pw_forum] why my pw.x run with low efficiency? > On Sun, 21 Sep 2008, Paolo Giannozzi wrote: > > PG> > PG> On Sep 19, 2008, at 18:54 , vega wrote: > PG> > PG> > PWSCF : 0d 14h46m CPU time, 2d 18h 4m > PG> > wall time > PG> > PG> I am quite sure Axel Kohlmeeyer has already answered 2n+1 times > PG> to the sameor similar question. Please look in particular for > PG> OMP_NUM_THREADS in the archives of the mailing list > > sorry paolo, > but that is only the case when the CPU time is much _higher_ > than the wall time. here it looks as if the jobs is either > swapping like crazy or the communication is stalling. on that > note, it would be nice to also see the wall time spent in the > individual routines, as the cpu time is usually a somewhat > inadequate descriptor, except for serial calculations. > > i'd have commented on the discussion earlier on, as several > assessments are unsubstantiated or don't make much sense at all. > > but the only way to find out for sure what is happening, > would be to run that very same job on a machine where i > know for certain that the hardware and software is set up > correctly and QE is compiled in the best possible way. > this however is very unpractical from addis ababa where > reasonable internet access is only available intermittently > and at very high costs. > > thus only a bunch of questions regarding observations and > explanations. vega mentioned that openmpi didn't work because > of "lack of memory". i suspect that this is due to incorrect > setup of the infiniband fabric and user limits. ulimit -a > should produce on the compute nodes something like this. > particularly the "max locked memory" entry is very important, > and not setting it high enough will result in severely > degraded performance (cf. ofed documentation). > > core file size (blocks, -c) 0 > data seg size (kbytes, -d) unlimited > max nice (-e) 0 > file size (blocks, -f) unlimited > pending signals (-i) 71680 > max locked memory (kbytes, -l) 4096000 > max memory size (kbytes, -m) unlimited > open files (-n) 1024 > pipe size (512 bytes, -p) 8 > POSIX message queues (bytes, -q) 819200 > max rt priority (-r) 0 > stack size (kbytes, -s) 1024000 > cpu time (seconds, -t) unlimited > max user processes (-u) 71680 > virtual memory (kbytes, -v) unlimited > file locks (-x) unlimited > > another option to enhance OpenMPI performance for QE over > infiniband, while reducing its memory requirements is to > use the flag: --mca btl_openib_use_srq 1 with mpirun or > set it for all jobs in the etc/openmpi-mca-params.conf > within the openmpi installation directory. > > it was then mentioned, that the job was calculated using MPICH > instead, which brings up the question, does this MPICH compile > actually support infiniband. as far as i know, not. the usual > OFED distributions contain either OpenMPI or MVAPICH. > > the next step to evaluate is whether the job actually does > scale that far, and if memory requirements are really killing > the performance. the reported virtual memory size is a bad > indicator, as it reports the reserved address space, but does > not say anything about how much of it is actually used. with > openmpi over infiniband, this tends to always be ridiculously > high, but that has no impact on the performance (symptom: large > usage of SWAP for the process without swap actually being used). > the real information about lack of available physical memory > would come from comparing the data set size with the resident > set size (you have to change the default configuration of top > to see it). also one can see how much swapping is going on by > simply logging into the node and monitoring whether kswapd > processes actually consume a lot of time. > > in the (somewhat unlikely) case that one is running out of > memory, it may help to not use all processors per node. i don't > remember what the hardware exactly was, but, e.g., on nodes with > 2x intel quad core processors, the (absolute!) performance and > scaling is much better when running with half the available cpu > cores. i have posted some benchmark graphs on cp2k here: > > http://cp2k.googlegroups.com/web/cp2k-water-bench-cmm.png > > but the outcome is similar for other DFT codes (CPMD, Q-E). > using less MPI tasks should reduce the total memory requirements > a bit, since not all data in QE can be or is distributed amongst > the nodes. > > finally, it may just be that the input at hand doesn't scale > to that many cpus, so in case of seeing performance problems > it always pays off to check whether running with less processors > or nodes is actually faster. bigger is not always better!! > > in summary, giving a recommendation on what is causing the > degraded performance, is impossible without knowing more details > and particularly a more systematic performance analysis. > i've given already an example of how to do this a while ago, > comparing scaling behavior with using -npools and without and > comparing gigabit to myrinet and demonstrating the scaling > limits of QE in both cases. i suggest to look up that e-mail > from the mailing list archives and do a similar assessment > with special consideration of the additional concerns listed > in this e-mail. > > cheers, > axel. > > PG> > PG> Paolo > PG> --- > PG> Paolo Giannozzi, Dept of Physics, University of Udine > PG> via delle Scienze 208, 33100 Udine, Italy > PG> Phone +39-0432-558216, fax +39-0432-558222 > PG> > PG> > PG> > PG> _______________________________________________ > PG> Pw_forum mailing list > PG> Pw_forum at pwscf.org > PG> http://www.democritos.it/mailman/listinfo/pw_forum > PG> > > -- > ======================================================================= > 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 > From asafis at yahoo.com.br Mon Sep 22 06:26:36 2008 From: asafis at yahoo.com.br (=?iso-8859-1?Q?=C1lvaro_Alves?=) Date: Sun, 21 Sep 2008 21:26:36 -0700 (PDT) Subject: [Pw_forum] Pseudopotential Co Message-ID: <613179.44395.qm@web52307.mail.re2.yahoo.com> Hello everyone. I'm trying to create a fully relativistic pseudopotential for Co. But the following error appears ?%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ???? from ld1_setup : error #???????? 7 ???? mismatched all-electron/pseudo occupations ?%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ? Anyone can check my input and tell me where is the error???? ?&input ??? title='Co', ??? zed=27., ??? rel=2, ??? nld=3, ??? config='[Ar] 3d7.0 4s2.0 4p0.0', ??? iswitch=3, ??? dft='PBE' ?/ ?&inputp ?? lloc=0, ?? pseudotype=3, ?? file_pseudopw='Co.rel-pbe3-rrkjus.UPF', ?? nlcc=.true., ?? rcore=0.9, ?? author='ASA', ?/ 9? 4P? 2? 1? 0.00? 0.00? 2.40? 2.40? 0.5 4P? 2? 1? 0.00? 0.00? 2.40? 2.40? 1.5 4P? 2? 1? 0.00? 0.00? 2.40? 2.40? 0.5 4P? 2? 1? 0.00? 0.00? 2.40? 2.40? 1.5 3D? 3? 2? 4.00? 0.00? 1.50? 2.10? 1.5 3D? 3? 2? 0.00 -0.30? 1.50? 2.10? 1.5 3D? 3? 2? 4.00? 0.00? 1.50? 2.10? 2.5 3D? 3? 2? 0.00 -0.30? 1.50? 2.10? 2.5? 4S? 1? 0? 2.00? 0.00? 2.20? 2.20? 0.5 This is the first time you try to make a pseudopotencial. Thus, any help is welcome. ?????????????????????????? Sincerely A. S. Santos Universidade Federal Fluminense - Rio de Janeiro - Brasil Novos endere?os, o Yahoo! que voc? conhece. Crie um email novo com a sua cara @ymail.com ou @rocketmail.com. http://br.new.mail.yahoo.com/addresses -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080921/216d8524/attachment.htm From akohlmey at cmm.chem.upenn.edu Mon Sep 22 06:42:03 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 22 Sep 2008 00:42:03 -0400 (EDT) Subject: [Pw_forum] why my pw.x run with low efficiency? In-Reply-To: References: <7FE9D3D1-A41D-44E5-A06B-34138FAC9F2F@democritos.it> Message-ID: On Mon, 22 Sep 2008, vega wrote: VL> Dear sir, VL> VL> Thank you so much for your responding. I do appreciate your help. VL> VL> > than the wall time. here it looks as if the jobs is either VL> > swapping like crazy or the communication is stalling. on that VL> VL> I believe it. So do you think 10G infiniband is good enough for my job? no. i think your machine is not used properly. this can be either due to the setup of the machine or due to the way jobs are run. based on previous experience, it is a combination of both. VL> By the way, there is also another two parallel job on the line. one is VL> lammps, a classical MD code. the other is VASP. the lammps and vasp are VL> using 32 cpus in total. my job is using 39 nodes and two cpu for each node. for as long as those jobs are running on _seperate_ nodes, there should be no problem. again, it is very hard to give _any_ useful advice without having the proper information at hand, so without knowing all details about your machine (number of nodes, hardware) and how nodes are assigned for use (batch system, how you submit jobs etc.). VL> >vega mentioned that openmpi didn't work because VL> > of "lack of memory". i suspect that this is due to incorrect VL> > setup of the infiniband fabric and user limits. ulimit -a VL> > should produce on the compute nodes something like this. VL> > particularly the "max locked memory" entry is very important, VL> > and not setting it high enough will result in severely VL> > degraded performance (cf. ofed documentation). VL> VL> I'm really admiring your experiences and sense about the parallel job. this is not a very special skill, but actually _reading_ the output on the screen and _looking it up_ in the documentation. VL> the max lock memory option on my machines is 4. Now I asked my VL> system administrator to reset it to 1024000. My physical memory VL> is 2G. Do you think set it to 2048000 will be better? 1GB should be ok. in case you are having quad-core cpus, you are pretty much memory starved for large DFT jobs, and you should run with -npernode 4, i.e. using only half the cores. VL> I'll try openmpi again, later, using flags --mca btl_openib_use_srq 1 VL> with mpirun. VL> VL> By the way I'm using MPICH2 now, not MPICH. same difference. it don't remember it supporting infiniband. so it is quite possible, that you have not been using the infiniband at all. this can be _easily_ verified, but running a smaller test job with both MPICH and OpenMPI across a few nodes and comparing the timings (not using -npools, to make the difference more evident). in total, please don't jump to conclusions and guess what is going wrong when running parallel jobs, but identify the performance bottlenecks systematically and back each claim you make up with hard data, so that people don't have to speculate about what could have gone wrong, but can confirm your findings and give proper advice. you simply have to apply the same strategies and problem solving skills that you need for proper scientific research. just doing some random calculations and guessing conclusions from them doesn't give you scientific insight either. 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 kyin.cn at gmail.com Mon Sep 22 08:05:58 2008 From: kyin.cn at gmail.com (Kun Yin) Date: Mon, 22 Sep 2008 14:05:58 +0800 Subject: [Pw_forum] Unit of dielectric constant tensor Message-ID: <383ADE06A3B84A538D6F80A579B98EEC@workhorse> Dear PWSCF users, I am doing the phonon calculation of MgO. Part of the output is > -------------- Output of PWSCF ph.x ------------------------------------- > Dielectric constant in cartesian axis > > ( 3.23010 0.00000 0.00000 ) > ( 0.00000 3.23010 0.00000 ) > ( 0.00000 0.00000 3.23010 ) > > Effective charges E-U in cartesian axis > > atom 1 > ( 2.01867 0.00000 0.00000 ) > ( 0.00000 2.01867 0.00000 ) > ( 0.00000 0.00000 2.01867 ) > atom 2 > ( -2.03905 0.00000 0.00000 ) > ( 0.00000 -2.03905 0.00000 ) > ( 0.00000 0.00000 -2.03905 ) > --------------------------------------------------------------------------- My question is what are the units of the above physical quantities, the dielectric constant and effective charge, in PWSCF? Any help will be appreciated. Regards, -- Kun YIN, Department of Geoengineering and Geoinformatics, School of Earth Sciences and Engineering, Nanjing University, Hankou Road 22, Geological Lab Building, Office 502, Nanjing, 210093, P.R. China Email : kyin.cn at gmail.com From giannozz at democritos.it Mon Sep 22 09:28:58 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 22 Sep 2008 09:28:58 +0200 Subject: [Pw_forum] Unit of dielectric constant tensor In-Reply-To: <383ADE06A3B84A538D6F80A579B98EEC@workhorse> References: <383ADE06A3B84A538D6F80A579B98EEC@workhorse> Message-ID: <48D7493A.5060208@democritos.it> Kun Yin wrote: > My question is what are the units of the above physical quantities, > the dielectric constant and effective charge, in PWSCF? interesting Zen koan: "what are the units of adimensional quantities"? -- Paolo Giannozzi, Democritos and University of Udine, Italy From wangqj1 at 126.com Mon Sep 22 09:48:37 2008 From: wangqj1 at 126.com (wangqj1) Date: Mon, 22 Sep 2008 15:48:37 +0800 (CST) Subject: [Pw_forum] about the output of projwfc.x Message-ID: <7809188.384251222069717365.JavaMail.coremail@bj126app104.126.com> Dear All After I performing the consequent projwfc.x calculation,I am puzzled about the output as follows: Atom O 1 total charge = 7.4091, s, p, d, f = 0.8321 6.5770 000000 Atom Zn 1 total charge = 5.8953, s, p, d, f = 2.8952 0.0000 3.0001 we know that the electron configuration of O is 1s2 2s2 2P4.and Zn is 4S2 3d10 ,from the result of output we kown that the charge of S orbit of O is 0.8321,but I don't know whether it is the value of 1s2 ,2s2 or the sum of 1s2 and 2s2.About the s orbit of Zn ,the charge value is 2.8952,whether it is the value of 4S2 or the sum of 1s2,2s2,3s2.and 4s2 ? sincerely wang XiangTan University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080922/1b1080f5/attachment.htm From tone.kokalj at ijs.si Mon Sep 22 09:59:53 2008 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Mon, 22 Sep 2008 09:59:53 +0200 Subject: [Pw_forum] Units of Charge Density - Again In-Reply-To: <48D20488.2000303@sissa.it> References: <01ff01c916df$38fc0b90$aaf422b0$@net> <48D20488.2000303@sissa.it> Message-ID: <1222070393.498.15.camel@walk.ijs.si> On Thu, 2008-09-18 at 09:34 +0200, Lorenzo Paulatto wrote: > > Finally, I've been unable (by visual inspection) to determine from the > > documentation for XcrySDen (a marvelous tool...marvelous), exactly how > > the "rho grid input" is scaled, especially regarding the computation > > of the 3D min-max grid values, the "isovalue"...and...particularly the > > mysterious "delta n(r)" displayed in the temperature inset box. This \Delta n(r) label is mysterious indeed, and is there solely for convenience: if one plots the charge density difference, then one would like to have "\Delta" sign, but it is not possible to input it with keyboard (unless one knows its ascii code), so I put it there. If one plots total charge density or some other property like potential, then this label can be simply deleted and replaced by the corresponding one. > > > I don't know if it's scaled at all, I think it is at most a side effect > of interpolation. If you use the minimum level of interpolation do you > get the extremes you expected? Anyway, I think it is better to wait for > Tone to reply on this topic Here I am (I was off-line previous week). Xcrysden does not assume any unit of charge, and simply leaves the values as stated in the file. In PW the unit is e/bohr^3, so this unit remains intact in xcrysden (this means that in this case the unit is e/bohr^3, e being the electron charge). The fact that the spanning vectors + origin for the grid are specified in Ansgtroms does not affect the unit of charge at all. This data is merely used to locate the grid points in the space. They could be equally well represented in meter or bohr units: the position of points would still be the same, and at each point the value of charge as supplied by the file is used. Regards, Tone -- Tone Kokalj J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia (tel: +386-1-477-3523 // fax:+386-1-477-3822) From tone.kokalj at ijs.si Mon Sep 22 10:16:38 2008 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Mon, 22 Sep 2008 10:16:38 +0200 Subject: [Pw_forum] How to plot the 2D-charge density for a triclinic or monoclinic box In-Reply-To: <391493881.20080914203009@yahoo.com.cn> References: <1221333356.32134.6.camel@walk.ijs.si> <3510209594.20080914115026@yahoo.com.cn> <1221386564.10695.9.camel@walk.ijs.si> <391493881.20080914203009@yahoo.com.cn> Message-ID: <1222071398.498.30.camel@walk.ijs.si> On Sun, 2008-09-14 at 20:30 +0800, Hongsheng Zhao wrote: > Due to the pp.x program require that all of the three span vectors > must be orthogonal to each other, ... Not if you specify output_format = 5. In this case the whole unit cell will be used and corresponding 3D XSF file constructed (even for triclinic lattice). You may try to use PWgui, it provides some guidelines: select "XCRYSDEN's XSF format (whole unit cell)". Regards, Tone -- Tone Kokalj J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia (tel: +386-1-477-3523 // fax:+386-1-477-3822) From paulatto at sissa.it Mon Sep 22 09:53:01 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Mon, 22 Sep 2008 09:53:01 +0200 Subject: [Pw_forum] Pseudopotential Co In-Reply-To: <613179.44395.qm@web52307.mail.re2.yahoo.com> References: <613179.44395.qm@web52307.mail.re2.yahoo.com> Message-ID: <48D74EDD.8070307@sissa.it> ?lvaro Alves ha scritto: Anyone can check my input and tell me where is the error? when doing fully relativistic pseudos, you have to specify the occupation of up and down orbitals even in the all electron configuration. This can be done in several ways, but if you prefer to use the compact notation, like this: config='[Ar] 3d7.0 4s2.0 4p0.0', then you should specify the occupations in the following way: config='[Ar] 3d4.0 3d3.0 4s1.0 4s1.0 4p0.0 4p0.0', but it will still be wrong, because you have made a mistake in the pseudo occupations, putting one excess electron in the 3d-up channel. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From giannozz at democritos.it Mon Sep 22 10:46:00 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 22 Sep 2008 10:46:00 +0200 Subject: [Pw_forum] Pseudopotential Co In-Reply-To: <613179.44395.qm@web52307.mail.re2.yahoo.com> References: <613179.44395.qm@web52307.mail.re2.yahoo.com> Message-ID: <48D75B48.9010508@democritos.it> ?lvaro Alves wrote: > from ld1_setup : error # 7 > mismatched all-electron/pseudo occupations > config='[Ar] 3d7.0 4s2.0 4p0.0', 3d7 > 3D 3 2 4.00 0.00 1.50 2.10 1.5 > 3D 3 2 0.00 -0.30 1.50 2.10 1.5 > 3D 3 2 4.00 0.00 1.50 2.10 2.5 > 3D 3 2 0.00 -0.30 1.50 2.10 2.5 3d4+4 = ? -- Paolo Giannozzi, Democritos and University of Udine, Italy From paulatto at sissa.it Mon Sep 22 09:57:33 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Mon, 22 Sep 2008 09:57:33 +0200 Subject: [Pw_forum] array bound check In-Reply-To: <037B80E5-F912-43E1-AB28-EC4895A5C70E@democritos.it> References: <037B80E5-F912-43E1-AB28-EC4895A5C70E@democritos.it> Message-ID: <48D74FED.1000308@sissa.it> Paolo Giannozzi ha scritto: > it depends on the compiler you are using > > So you can find it in the compiler manual. bye -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From baroni at sissa.it Mon Sep 22 10:49:44 2008 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 22 Sep 2008 10:49:44 +0200 Subject: [Pw_forum] array bound check In-Reply-To: <48D74FED.1000308@sissa.it> References: <037B80E5-F912-43E1-AB28-EC4895A5C70E@democritos.it> <48D74FED.1000308@sissa.it> Message-ID: be aware, though, that the code compiled with bound check on is way slower ... SB On Sep 22, 2008, at 9:57 AM, Lorenzo Paulatto wrote: > Paolo Giannozzi ha scritto: >> it depends on the compiler you are using >> >> > So you can find it in the compiler manual. > > bye > > -- > Lorenzo Paulatto > SISSA & DEMOCRITOS (Trieste) > phone: +39 040 3787 511 > skype: paulatz > www: http://people.sissa.it/~paulatto/ > > > _______________________________________________ > 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) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080922/818576bd/attachment-0001.htm From paulatto at sissa.it Mon Sep 22 10:02:37 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Mon, 22 Sep 2008 10:02:37 +0200 Subject: [Pw_forum] about the output of projwfc.x In-Reply-To: <7809188.384251222069717365.JavaMail.coremail@bj126app104.126.com> References: <7809188.384251222069717365.JavaMail.coremail@bj126app104.126.com> Message-ID: <48D7511D.2070103@sissa.it> wangqj1 ha scritto: > > I don't know whether it is the value of 1s2 ,2s2 or the sum of > > 1s2 and 2s2. > The 1s orbital of oxygen is deep in the core, it is not even included in the pseudopotential. > > is the value of 4S2 or the sum of 1s2,2s2,3s2.and 4s2 ? > I don't know, the Zn pseudopotential may include the semicore, if this was the case than 3s and 4s would be considered. Otherwise only the 4s is included. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From giannozz at democritos.it Mon Sep 22 10:55:29 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 22 Sep 2008 10:55:29 +0200 Subject: [Pw_forum] array bound check In-Reply-To: References: <037B80E5-F912-43E1-AB28-EC4895A5C70E@democritos.it> <48D74FED.1000308@sissa.it> Message-ID: <48D75D81.5010908@democritos.it> Stefano Baroni wrote: > be aware, though, that the code compiled with bound check on is way > slower ... ..and, of course, before the code is released, at least a few test cases are run with bound checks on Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From zhaohscas at yahoo.com.cn Mon Sep 22 11:04:32 2008 From: zhaohscas at yahoo.com.cn (Hongsheng Zhao) Date: Mon, 22 Sep 2008 17:04:32 +0800 Subject: [Pw_forum] How to plot the 2D-charge density for a triclinic or monoclinic box In-Reply-To: <1222071398.498.30.camel@walk.ijs.si> References: <1221333356.32134.6.camel@walk.ijs.si> <3510209594.20080914115026@yahoo.com.cn> <1221386564.10695.9.camel@walk.ijs.si> <391493881.20080914203009@yahoo.com.cn> <1222071398.498.30.camel@walk.ijs.si> Message-ID: <544823852.20080922170432@yahoo.com.cn> On Monday, September 22, 2008 at 16:16, tone.kokalj at ijs.si wrote: > On Sun, 2008-09-14 at 20:30 +0800, Hongsheng Zhao wrote: >> Due to the pp.x program require that all of the three span vectors >> must be orthogonal to each other, ... > Not if you specify output_format = 5. In this case the whole unit cell > will be used and corresponding 3D XSF file constructed (even for > triclinic lattice). You may try to use PWgui, it provides some > guidelines: select "XCRYSDEN's XSF format (whole unit cell)". Thanks a lot, I've sorted it out basically now. Sincerely yours, -- Hongsheng Zhao Xinjiang Technical Institute of Physics and Chemistry Chinese Academy of Sciences GnuPG DSA: 0xD108493 2008-9-22 From giannozz at democritos.it Mon Sep 22 11:14:39 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 22 Sep 2008 11:14:39 +0200 Subject: [Pw_forum] Are there the instruction of every modules ? In-Reply-To: <23663678.47301221968981420.JavaMail.coremail@bj126app103.126.com> References: <23663678.47301221968981420.JavaMail.coremail@bj126app103.126.com> Message-ID: <48D761FF.5070004@democritos.it> wangqj1 wrote: > Do you have the intruction of evrey modules ? all available documentation is either in the code or in the wiki Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Mon Sep 22 11:28:50 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 22 Sep 2008 11:28:50 +0200 Subject: [Pw_forum] About compilation of cvs version In-Reply-To: References: Message-ID: <48D76552.2040503@democritos.it> lan haiping wrote: > I would like to know whether it is reasonable by adding this mpi > macro for ptoolkit subroutine. if you plan to use the code in serial mode, it is: the "p" in "ptoolkit" stands for "parallel". Routines in ptoolkit.f90 are used only in parallel execution. Other compilers do not complain, but the syntax used in that part of code is questionable, so it will be fixed sooner or later Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Mon Sep 22 11:52:21 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 22 Sep 2008 11:52:21 +0200 Subject: [Pw_forum] about the output of projwfc.x In-Reply-To: <7809188.384251222069717365.JavaMail.coremail@bj126app104.126.com> References: <7809188.384251222069717365.JavaMail.coremail@bj126app104.126.com> Message-ID: <48D76AD5.8090408@democritos.it> wangqj1 wrote: > Atom O 1 total charge = 7.4091, s, p, d, f = 0.8321 6.5770 000000 > Atom Zn 1 total charge = 5.8953, s, p, d, f = 2.8952 0.0000 3.0001 funny. Could you please provide a test job? P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From lanhaiping at gmail.com Mon Sep 22 12:02:22 2008 From: lanhaiping at gmail.com (lan haiping) Date: Mon, 22 Sep 2008 18:02:22 +0800 Subject: [Pw_forum] About compilation of cvs version In-Reply-To: <48D76552.2040503@democritos.it> References: <48D76552.2040503@democritos.it> Message-ID: Hi, Paolo . I knew ptoolkit is for parallel executation, but i donot have much experience in fortran preprocessing. Thanks for your confirmation ! Best, Hai-Ping On Mon, Sep 22, 2008 at 5:28 PM, Paolo Giannozzi wrote: > lan haiping wrote: > > > I would like to know whether it is reasonable by adding this mpi > > macro for ptoolkit subroutine. > > if you plan to use the code in serial mode, it is: the "p" in > "ptoolkit" stands for "parallel". Routines in ptoolkit.f90 > are used only in parallel execution. > > Other compilers do not complain, but the syntax used in that part > of code is questionable, so it will be fixed sooner or later > > 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 > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080922/c1b709c6/attachment.htm From giannozz at democritos.it Mon Sep 22 12:07:23 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 22 Sep 2008 12:07:23 +0200 Subject: [Pw_forum] About compilation of cvs version In-Reply-To: References: <48D76552.2040503@democritos.it> Message-ID: <48D76E5B.102@democritos.it> lan haiping wrote: > I knew ptoolkit is for parallel executation, but i donot > have much experience in fortran preprocessing. it is not "fortran" preprocessing, actually: it is C preprocessing, applied to fortran code Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From manoj at phys.ufl.edu Mon Sep 22 16:14:02 2008 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Mon, 22 Sep 2008 10:14:02 -0400 (EDT) Subject: [Pw_forum] array bound check In-Reply-To: <48D75D81.5010908@democritos.it> Message-ID: Dear all, Thank you very much for the reply and suggestions. I appreciate it :) Regards, Manoj On Mon, 22 Sep 2008, Paolo Giannozzi wrote: > Stefano Baroni wrote: > > > be aware, though, that the code compiled with bound check on is way > > slower ... > > ..and, of course, before the code is released, at least a few test > cases are run with bound checks on > > 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 asafis at yahoo.com.br Mon Sep 22 20:09:28 2008 From: asafis at yahoo.com.br (=?iso-8859-1?Q?=C1lvaro_Alves?=) Date: Mon, 22 Sep 2008 11:09:28 -0700 (PDT) Subject: [Pw_forum] Pseudopotential Co In-Reply-To: <48D74EDD.8070307@sissa.it> Message-ID: <941714.81305.qm@web52311.mail.re2.yahoo.com> Hi Lorenzo.? What are other ways to specify the occupation of up and down orbitals ? And where can I find information about these other ways? regardsA. S. Santos Universidade Federal Fluminense - Rio de Janeiro - Brasil --- Em seg, 22/9/08, Lorenzo Paulatto escreveu: De: Lorenzo Paulatto Assunto: Re: [Pw_forum] Pseudopotential Co Para: asafis at yahoo.com.br, "PWSCF Forum" Data: Segunda-feira, 22 de Setembro de 2008, 4:53 ?lvaro Alves ha scritto: Anyone can check my input and tell me where is the error? when doing fully relativistic pseudos, you have to specify the occupation of up and down orbitals even in the all electron configuration. This can be done in several ways, but if you prefer to use the compact notation, like this: config='[Ar] 3d7.0 4s2.0 4p0.0', then you should specify the occupations in the following way: config='[Ar] 3d4.0 3d3.0 4s1.0 4s1.0 4p0.0 4p0.0', but it will still be wrong, because you have made a mistake in the pseudo occupations, putting one excess electron in the 3d-up channel. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum Novos endere?os, o Yahoo! que voc? conhece. Crie um email novo com a sua cara @ymail.com ou @rocketmail.com. http://br.new.mail.yahoo.com/addresses -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080922/f45c6dba/attachment.htm From paulatto at sissa.it Mon Sep 22 21:28:32 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Mon, 22 Sep 2008 21:28:32 +0200 (CEST) Subject: [Pw_forum] Pseudopotential Co In-Reply-To: <941714.81305.qm@web52311.mail.re2.yahoo.com> References: <941714.81305.qm@web52311.mail.re2.yahoo.com> Message-ID: <24904.82.50.174.28.1222111712.squirrel@webmail.sissa.it> On Lun, Settembre 22, 2008 20:09, ?lvaro Alves wrote: > And where can I find information about these other ways? In the ld1 documentation, in atomic_doc directory. cheers -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From wenchenai at tom.com Tue Sep 23 06:01:24 2008 From: wenchenai at tom.com (Wenxiaodong) Date: Tue, 23 Sep 2008 00:01:24 -0400 Subject: [Pw_forum] some trouble in Phonon dispersion calculation Message-ID: <20080923043138.246B5112D10@democritos.sissa.it> Dear everybody Recently, I did a phonon dispersion calculation using PWscf. However, I can not obtain smooth phonon dispersion. That is to say, the phonon dispersion looks like zigzag shape. For making sure, I tried to use the example 07 (in PWscf), while there are same phenomena for phonon dispersion. By the way, I used the plotband.x to handle data. If you have some experience about this, could you tell me the reasons? I will very very appreciate you! Best Xiaodong From lanhaiping at gmail.com Tue Sep 23 07:26:45 2008 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 23 Sep 2008 13:26:45 +0800 Subject: [Pw_forum] Question about Fortran 77 Message-ID: Dear Developers: I am trying to compile cvs version, and have a question about lib building with $F77 compilor. As far as i know, F77 doesnot allow dynamical array. With regards to $QE/Muligrid , there are several routines with dynamical arrays declaration. When I try to build muligrid.a lib with pgf77, the compilor complains about syntax error. I think these syntax errors should be due to the dynamical declarations for several arrays used in routines. I confirm it with pgf90 compilor without such errors. So , I just wonder whether these syntax errors are just related with pgf77 compilor ? Should we change multigrid.a lib's building rule in the makefile ? Best , Hai-Ping -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080923/57fa0034/attachment.htm From lfhuang at theory.issp.ac.cn Tue Sep 23 11:22:16 2008 From: lfhuang at theory.issp.ac.cn (=?utf-8?B?bGZodWFuZw==?=) Date: Tue, 23 Sep 2008 17:22:16 +0800 Subject: [Pw_forum] =?utf-8?q?some_trouble_in_Phonon_dispersion_calculatio?= =?utf-8?q?n?= Message-ID: <20080923092216.19381.qmail@ms.hfcas.ac.cn> Dear Xiaodong: Could you have paid attention to the coordinations that were chosen by you and the code? Maybe they are different(x-y coordination or reciprocal lattice vector coordination). If you haven't, you can transform between them with the transformation matrix that can be found in the pw.x output file. Best Wishes! Yours Sincerely L.F.Huang ====================================================================== L.F.Huang(???) lfhuang at theory.issp.ac.cn ====================================================================== Add: Research Laboratory for Computational Materials Sciences, Instutue of Solid State Physics,the Chinese Academy of Sciences, P.O.Box 1129, Hefei 230031, P.R.China Tel: 86-551-5591464-328(office) Fax: 86-551-5591434 Web: http://theory.issp.ac.cn (website of our theory group) http://www.issp.ac.cn (website of our institute) ====================================================================== > From: "Wenxiaodong" > Subject: [Pw_forum] some trouble in Phonon dispersion calculation > To: "pw_forum" > Message-ID: > Content-Type: text/plain; charset="gb2312" > > Dear everybody > Recently, I did a phonon dispersion calculation using PWscf. However, I can not obtain smooth phonon dispersion. That is to say, the phonon dispersion looks like zigzag shape. For making sure, I tried to use the example 07 (in PWscf), while there are same phenomena for phonon dispersion. By the way, I used the plotband.x to handle data. > If you have some experience about this, could you tell me the reasons? I will very very appreciate you! > Best > Xiaodong ------ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080923/fd27ff48/attachment.htm From eyvaz_isaev at yahoo.com Tue Sep 23 13:49:37 2008 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 23 Sep 2008 04:49:37 -0700 (PDT) Subject: [Pw_forum] some trouble in Phonon dispersion calculation In-Reply-To: <20080923043138.246B5112D10@democritos.sissa.it> Message-ID: <7392.1113.qm@web65702.mail.ac4.yahoo.com> Dear Xiaodong, Can you please submit the script you used and figure you obtained? If you submit frequency file it will be much easier to fix the problem. Please also provide your affiliation. Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com --- On Tue, 9/23/08, Wenxiaodong wrote: > From: Wenxiaodong > Subject: [Pw_forum] some trouble in Phonon dispersion calculation > To: "pw_forum" > Date: Tuesday, September 23, 2008, 8:01 AM > Dear everybody > Recently, I did a phonon dispersion calculation using > PWscf. However, I can not obtain smooth phonon dispersion. > That is to say, the phonon dispersion looks like zigzag > shape. For making sure, I tried to use the example 07 (in > PWscf), while there are same phenomena for phonon > dispersion. By the way, I used the plotband.x to handle > data. > If you have some experience about this, could you tell me > the reasons? I will very very appreciate you! > Best > Xiaodong > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From giannozz at democritos.it Tue Sep 23 13:56:19 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 23 Sep 2008 13:56:19 +0200 Subject: [Pw_forum] Question about Fortran 77 In-Reply-To: References: Message-ID: <48D8D963.7010500@democritos.it> lan haiping wrote: > With regards to $QE/Muligrid , there are several routines with dynamical > arrays declaration. there is actually only one file containing such declarations. It is however true that the naming convention is inconsistent with the rest of q-e and this may be a source of trouble with some compilers (pgi being, unsurprisingly, one of these). Please update your cvs and retry. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From lanhaiping at gmail.com Tue Sep 23 19:06:02 2008 From: lanhaiping at gmail.com (lan haiping) Date: Wed, 24 Sep 2008 01:06:02 +0800 Subject: [Pw_forum] Question about Fortran 77 In-Reply-To: <48D8D963.7010500@democritos.it> References: <48D8D963.7010500@democritos.it> Message-ID: Thank you , Paolo ! Best , Hai-Ping On Tue, Sep 23, 2008 at 7:56 PM, Paolo Giannozzi wrote: > lan haiping wrote: > > > With regards to $QE/Muligrid , there are several routines with dynamical > > arrays declaration. > > there is actually only one file containing such declarations. It is > however true that the naming convention is inconsistent with the rest > of q-e and this may be a source of trouble with some compilers (pgi > being, unsurprisingly, one of these). Please update your cvs and retry. > > 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 > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080924/966057cc/attachment.htm From manoj at phys.ufl.edu Tue Sep 23 19:47:47 2008 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Tue, 23 Sep 2008 13:47:47 -0400 (EDT) Subject: [Pw_forum] array bound check In-Reply-To: Message-ID: Dear PWSCF users, Following your suggestions, I am being able to locate what needs to be done for array bound check, but I still have not figured out where do I need to make changes. I am using ifort compiler. For array bound check i need to do - check bounds but where do i need to make this change? I am new to linux and most of these things are beyond my current knowledge. Any help would be appreciated. Regards, Manoj Srivastava Graduate Student Department of Physics University of Florida, Gainesville. USA. On Mon, 22 Sep 2008, Manoj Srivastava wrote: > Dear all, > Thank you very much for the reply and suggestions. I appreciate it :) > > Regards, > Manoj > > On Mon, 22 Sep 2008, Paolo Giannozzi wrote: > > > Stefano Baroni wrote: > > > > > be aware, though, that the code compiled with bound check on is way > > > slower ... > > > > ..and, of course, before the code is released, at least a few test > > cases are run with bound checks on > > > > 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 paulatto at sissa.it Tue Sep 23 20:01:58 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Tue, 23 Sep 2008 20:01:58 +0200 (CEST) Subject: [Pw_forum] array bound check In-Reply-To: References: Message-ID: <14882.82.50.174.28.1222192918.squirrel@webmail.sissa.it> On Mar, Settembre 23, 2008 19:47, Manoj Srivastava wrote: > but I still have not figured out where do I need to make changes. After running the configure script, edit make.sys and look for the line starting setting the F90FLAGS variable, add "-check bounds" to it. I'll advice to add -traceback as well, otherwise you'll not be able to know the line of the code where the check is triggered. Remeber to make clean, if needed, and recompile after modifying make.sys. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From manoj at phys.ufl.edu Wed Sep 24 06:41:07 2008 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Wed, 24 Sep 2008 00:41:07 -0400 (EDT) Subject: [Pw_forum] array bound check In-Reply-To: <14882.82.50.174.28.1222192918.squirrel@webmail.sissa.it> Message-ID: Dear Dr. Paulatto and PWSCF users, Thank you very much for your kind of response. with your help I have been able to complile the code with array bound check. But it has introduced an additional problem. I can run some of the examples but I cannot run others e.g. example 12. I could run example 12 before where i did not have bound check and traceback options on. Following is what I did - I ran configure file and then i edited make.sys file, which was 90FLAGS = $(FFLAGS) -nomodule -fpp $(FDFLAGS) $(IFLAGS) $(MODFLAGS) to 90FLAGS = $(FFLAGS) -nomodule -fpp -check bounds -traceback $(FDFLAGS) $(IFLAGS) $(MODFLAGS) then i recompiled the code again. Am I doing what you suggested or is there something more to it? I am using ifort compiler. Regards, Manoj Srivastava On Tue, 23 Sep 2008, Lorenzo Paulatto wrote: > > On Mar, Settembre 23, 2008 19:47, Manoj Srivastava wrote: > > but I still have not figured out where do I need to make changes. > > After running the configure script, edit make.sys and look for the line > starting setting the F90FLAGS variable, add "-check bounds" to it. I'll > advice to add -traceback as well, otherwise you'll not be able to know the > line of the code where the check is triggered. > Remeber to make clean, if needed, and recompile after modifying make.sys. > > regards > > -- > Lorenzo Paulatto > SISSA & DEMOCRITOS (Trieste) > phone: +39 040 3787 511 > skype: paulatz > www: http://people.sissa.it/~paulatto/ > > > > ---------------------------------------------------------------- > SISSA Webmail https://webmail.sissa.it/ > Powered by SquirrelMail http://www.squirrelmail.org/ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From giannozz at democritos.it Wed Sep 24 10:19:24 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 24 Sep 2008 10:19:24 +0200 Subject: [Pw_forum] array bound check In-Reply-To: References: Message-ID: On Sep 24, 2008, at 6:41 , Manoj Srivastava wrote: > I can run some of the examples but I cannot run others e.g. example > 12. what do you mean by "| cannot run" ? Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From smogunov at sissa.it Wed Sep 24 12:19:02 2008 From: smogunov at sissa.it (Alexander) Date: Wed, 24 Sep 2008 12:19:02 +0200 Subject: [Pw_forum] array bound check In-Reply-To: References: Message-ID: <200809241219.02506.smogunov@sissa.it> Dear Manoj As far as the Ex. 12 is concerned, there was indeed a problem with few arrays in the subroutine PWCOND/init_orbitals.f90. The problem seems to be solved now in the cvs version and the example runs properly (at least on my PC). Thank you. best regards, Alexander On Wednesday 24 September 2008 06:41, Manoj Srivastava wrote: > Dear Dr. Paulatto and PWSCF users, > Thank you very much for your kind of response. with your help I have been > able to complile the code with array bound check. But it has introduced an > additional problem. I can run some of the examples but I cannot run others > e.g. example 12. I could run example 12 before where i did not > have bound check and traceback options on. > Following is what I did - > I ran configure file and then i edited make.sys file, which was > 90FLAGS = $(FFLAGS) -nomodule -fpp $(FDFLAGS) $(IFLAGS) $(MODFLAGS) > to > 90FLAGS = $(FFLAGS) -nomodule -fpp -check bounds -traceback $(FDFLAGS) > $(IFLAGS) $(MODFLAGS) > > then i recompiled the code again. Am I doing what you suggested or is > there something more to it? I am using ifort compiler. > > Regards, > Manoj Srivastava > > On Tue, 23 Sep 2008, Lorenzo Paulatto wrote: > > On Mar, Settembre 23, 2008 19:47, Manoj Srivastava wrote: > > > but I still have not figured out where do I need to make changes. > > > > After running the configure script, edit make.sys and look for the line > > starting setting the F90FLAGS variable, add "-check bounds" to it. I'll > > advice to add -traceback as well, otherwise you'll not be able to know > > the line of the code where the check is triggered. > > Remeber to make clean, if needed, and recompile after modifying make.sys. > > > > regards > > > > -- > > Lorenzo Paulatto > > SISSA & DEMOCRITOS (Trieste) > > phone: +39 040 3787 511 > > skype: paulatz > > www: http://people.sissa.it/~paulatto/ > > > > > > > > ---------------------------------------------------------------- > > SISSA Webmail https://webmail.sissa.it/ > > Powered by SquirrelMail http://www.squirrelmail.org/ > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From manoj at phys.ufl.edu Wed Sep 24 15:24:35 2008 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Wed, 24 Sep 2008 09:24:35 -0400 (EDT) Subject: [Pw_forum] array bound check In-Reply-To: Message-ID: Dear Paolo, Alexander and PWSCF users, Thank you very much for your response and I am sorry I was not clear before, when i said 'i cannot run example12'.Following is what i get when i run example12, with espresso compiled with array bound check- pwcond.x 0000000000471304 poten_ 69 poten.f90 pwcond.x 000000000043709E init_cond_ 102 init_cond.f90 pwcond.x 000000000042583F do_cond_ 250 do_cond.f90 pwcond.x 0000000000420E6F MAIN__ 20 condmain.f90 pwcond.x 0000000000403762 Unknown Unknown Unknown libc.so.6 0000002A9580F40B Unknown Unknown Unknown pwcond.x 00000000004036AA Unknown Unknown Unknown while with out array bound check, i can run this example without any error message and right results. My guess is the way i am doing array bound check is introducing some kind of bug. My question here is whether I am doing the array bound check right? This is what i did for array bound check, as suggested by Dr.Paulatto I ran configure file and then i edited make.sys file, which was 90FLAGS = $(FFLAGS) -nomodule -fpp $(FDFLAGS) $(IFLAGS) $(MODFLAGS) to 90FLAGS = $(FFLAGS) -nomodule -fpp -check bounds -traceback $(FDFLAGS) $(IFLAGS) $(MODFLAGS) then i recompiled the code again. intel ifort compiler is being used. Am I doing it right? Regards, Manoj Srivastava On Wed, 24 Sep 2008, Paolo Giannozzi wrote: > > On Sep 24, 2008, at 6:41 , Manoj Srivastava wrote: > > > I can run some of the examples but I cannot run others e.g. example > > 12. > > what do you mean by "| cannot run" ? > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From chenweiguang82 at gmail.com Wed Sep 24 16:00:38 2008 From: chenweiguang82 at gmail.com (Weiguang Chen) Date: Wed, 24 Sep 2008 22:00:38 +0800 Subject: [Pw_forum] compile error Message-ID: <9e6d976a0809240700g253ef299l8644f0ff504842a1@mail.gmail.com> Hi, The following error happened while i compiled the parallel pwscf with mpich. How can i do? -------------------------------------------------------------------------------------------------- iotk_dat.spp: In function 'iotk_write_dat_character1_0': iotk_dat.spp:142: internal compiler error: in gfc_conv_function_call, at fortran/trans-expr.c:1105 Please submit a full bug report, with preprocessed source if appropriate. See for instructions. make[2]: *** [iotk_dat+CHARACTER1_0.o] ?? 1 make[2]: Leaving directory `/root/espresso-4.0.1/iotk/src' make[1]: *** [lib+util] ?? 2 make[1]: Leaving directory `/root/espresso-4.0.1/iotk' make: *** [libiotk] ?? 2 -------------------------------------------------------------------------------------------------- Thanks -- Best Wishes ChenWeiguang -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080924/9f68204b/attachment.htm From zucco at dipteris.unige.it Wed Sep 24 16:46:30 2008 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Wed, 24 Sep 2008 16:46:30 +0200 Subject: [Pw_forum] vc-relax and LDA+U: Segmentation fault Message-ID: Hello to all, from a very recent email posted on the forum it seem to me that relaxation (maybe vc-relax) is possible using LDA+U, and that experimental implementation of phonons with LDA+U is going on (for this I hope to be a beta-tester if my structure converge..). My question is: does anyone of you have experience a Segmentation fault using vc-relax and LDA+U just after the wfc convergence has been achieved, starting 'from_scratch'? I'm using 4.0CVS in serial mode, but it also in 4.0.2 version it happens. Any help is welcome. Marino ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From hande at newton.physics.metu.edu.tr Wed Sep 24 16:49:09 2008 From: hande at newton.physics.metu.edu.tr (Hande Ustunel) Date: Wed, 24 Sep 2008 17:49:09 +0300 (EEST) Subject: [Pw_forum] test pseudo bug? Message-ID: Dear PWSCF developers, I have been trying to generate an O pseudopotential using the 4.0 version. In the test section I noticed that the AE and PS values come out to be identical no matter how outrageous the test configuration is. The AE energies are also somewhat different. I include below my input file for the test and the relevant part of the output for both the 3.2 and the 4.0 versions. (The problem persists also in the 4.0.2 version) After some code mining, I MIGHT HAVE identified a possible source where in the ascheqps.f90, the fourth input (and output) variable, which is later called from the write_resultsps.f90 subroutine as elts(n) doesn't seem to be modified after being set to the ae energy in the beginning. I am, of course, really not sure if this is the real cause so I was wondering if you could help me see my mistake (if I'm making one) or if you could confirm that this is a bug. Thank you very much in advance. Best wishes, Hande ------------------------------------------------------------ Test input : &input atom='O', rlderiv=4.0, eminld=-2.5, emaxld= 2.5, deld = 0.02, nld = 3, config='[He] 2s2 2p2 3s2.0 3p2.0', iswitch=2, prefix='o' dft='pbe', / &test file_pseudo='O.pbe.uspp-n-rrkj.UPF' nconf=3, configts(1)=' 2s2.00 2p1.75 ', configts(2)=' 2s1.80 2p2.50 ', configts(3)=' 2s1.80 2p1.50 ', / ------------------------------------------------------------ Energy diffences in the 3.2 version Conf 1 : n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) 1 0 2S 1( 2.00) -2.72810 -4.65642 1.92832 2 1 2P 1( 1.75) -1.62655 -3.55432 1.92777 Conf 2 : n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) 1 0 2S 1( 1.80) -1.78231 -3.85038 2.06808 2 1 2P 1( 2.50) -0.67236 -2.74216 2.06981 Conf 3 : n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) 1 0 2S 1( 1.80) -3.02350 -5.34542 2.32192 2 1 2P 1( 1.50) -1.93748 -4.25302 2.31555 ------------------------------------------------------------ Energy diffences in the 4.0 version Conf 1 : n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) 1 0 2S 1( 2.00) -2.25668 -2.25668 0.00000 2 1 2P 1( 1.75) -1.15425 -1.15425 0.00000 Conf 2 : n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) 1 0 2S 1( 1.80) -1.61762 -1.61762 0.00000 2 1 2P 1( 2.50) -0.50726 -0.50726 0.00000 Conf 3 : n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) 1 0 2S 1( 1.80) -3.20787 -3.20787 0.00000 2 1 2P 1( 1.50) -2.12306 -2.12306 0.00000 -- Hande Toffoli 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 matteo at umn.edu Wed Sep 24 16:59:26 2008 From: matteo at umn.edu (Matteo Cococcioni) Date: Wed, 24 Sep 2008 09:59:26 -0500 Subject: [Pw_forum] vc-relax and LDA+U: Segmentation fault In-Reply-To: References: Message-ID: <48DA55CE.2000308@umn.edu> Dear Marino, I have never run a vc-relax simulation with LDA+U but I know of people who have done that successfully. if U_projection_type = 'atomic' the code should compute forces and stresses with LDA+U no problem. Do you obtain forces and stresses? Also I'm not sure I have understood where the code stops. Does the segmentation fault appear after the first scf cycle (and after the first calculation of stresses and forces)? Also what kind of dynamics have you picked? regards, Matteo Marino Vetuschi Zuccolini wrote: > Hello to all, > from a very recent email posted on the forum it seem to me that > relaxation (maybe vc-relax) is possible using LDA+U, and that > experimental implementation of phonons with LDA+U is going on (for > this I hope to be a beta-tester if my structure converge..). > My question is: does anyone of you have experience a Segmentation > fault using vc-relax and LDA+U just after the wfc convergence has been > achieved, starting 'from_scratch'? > I'm using 4.0CVS in serial mode, but it also in 4.0.2 version it > happens. > Any help is welcome. > > Marino > > > ******************************************************* > Marino Vetuschi Zuccolini > zucco at dipteris.unige.it > Researcher / Geochemist > Laboratory of Geochemistry > > DIPartimento per lo studio della TErra e delle sue RISorse - > Universit? di Genova > Tel. ++39 010 3538136 Fax. ++39 010 352169 > Corso Europa 26, 16132 - Genova - Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- A non-text attachment was scrubbed... Name: matteo.vcf Type: text/x-vcard Size: 294 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20080924/04ac5959/attachment.vcf From giannozz at democritos.it Wed Sep 24 16:59:35 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 24 Sep 2008 16:59:35 +0200 Subject: [Pw_forum] compile error In-Reply-To: <9e6d976a0809240700g253ef299l8644f0ff504842a1@mail.gmail.com> References: <9e6d976a0809240700g253ef299l8644f0ff504842a1@mail.gmail.com> Message-ID: <0C07BD53-6FD3-4472-AB9A-D44E856F83F1@democritos.it> On Sep 24, 2008, at 16:00 , Weiguang Chen wrote: > internal compiler error http://www.quantum-espresso.org/wiki/index.php/ Frequently_Asked_Questions item 1.1.6 --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From dalcorso at sissa.it Wed Sep 24 17:02:34 2008 From: dalcorso at sissa.it (Dal Corso Andrea) Date: Wed, 24 Sep 2008 17:02:34 +0200 Subject: [Pw_forum] test pseudo bug? In-Reply-To: References: Message-ID: <1222268554.3304.7.camel@dhpc-5-53.sissa.it> On Wed, 2008-09-24 at 17:49 +0300, Hande Ustunel wrote: > Dear PWSCF developers, > > I have been trying to generate an O pseudopotential using the 4.0 > version. In the test section I noticed that the AE and PS values come out > to be identical no matter how outrageous the test configuration is. The AE > energies are also somewhat different. > > I include below my input file for the test and the relevant part of the > output for both the 3.2 and the 4.0 versions. (The problem persists also in > the 4.0.2 version) > > After some code mining, I MIGHT HAVE identified a possible source where in > the ascheqps.f90, the fourth input (and output) variable, which is later > called from the write_resultsps.f90 subroutine as elts(n) doesn't seem to > be modified after being set to the ae energy in the beginning. I am, of > course, really not sure if this is the real cause so I was wondering if you > could help me see my mistake (if I'm making one) or if you could confirm > that this is a bug. Please check carefully the output. There should be an error message saying Errors in the PS-KS equations. Your pseudo has a ghost and the new routine is not able to find it so it does not change the eigenvalues but it writes the error message. In the test case the program does not stop because the next test configuration might converge. Andrea > Thank you very much in advance. > Best wishes, > Hande > > ------------------------------------------------------------ > > Test input : > > &input > atom='O', > rlderiv=4.0, > eminld=-2.5, > emaxld= 2.5, > deld = 0.02, > nld = 3, > config='[He] 2s2 2p2 3s2.0 3p2.0', > iswitch=2, > prefix='o' > dft='pbe', > / > &test > file_pseudo='O.pbe.uspp-n-rrkj.UPF' > nconf=3, > configts(1)=' 2s2.00 2p1.75 ', > configts(2)=' 2s1.80 2p2.50 ', > configts(3)=' 2s1.80 2p1.50 ', > / > > ------------------------------------------------------------ > > Energy diffences in the 3.2 version > > Conf 1 : > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > 1 0 2S 1( 2.00) -2.72810 -4.65642 1.92832 > 2 1 2P 1( 1.75) -1.62655 -3.55432 1.92777 > Conf 2 : > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > 1 0 2S 1( 1.80) -1.78231 -3.85038 2.06808 > 2 1 2P 1( 2.50) -0.67236 -2.74216 2.06981 > Conf 3 : > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > 1 0 2S 1( 1.80) -3.02350 -5.34542 2.32192 > 2 1 2P 1( 1.50) -1.93748 -4.25302 2.31555 > > > ------------------------------------------------------------ > > Energy diffences in the 4.0 version > Conf 1 : > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > 1 0 2S 1( 2.00) -2.25668 -2.25668 0.00000 > 2 1 2P 1( 1.75) -1.15425 -1.15425 0.00000 > Conf 2 : > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > 1 0 2S 1( 1.80) -1.61762 -1.61762 0.00000 > 2 1 2P 1( 2.50) -0.50726 -0.50726 0.00000 > Conf 3 : > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > 1 0 2S 1( 1.80) -3.20787 -3.20787 0.00000 > 2 1 2P 1( 1.50) -2.12306 -2.12306 0.00000 > -- 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 paulatto at sissa.it Wed Sep 24 16:56:35 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Wed, 24 Sep 2008 16:56:35 +0200 Subject: [Pw_forum] vc-relax and LDA+U: Segmentation fault In-Reply-To: References: Message-ID: <48DA5523.2020001@sissa.it> Marino Vetuschi Zuccolini ha scritto: > My question is: does anyone of you have experience a Segmentation > fault using vc-relax and LDA+U just after the wfc convergence has been > achieved, starting 'from_scratch'? > Probably some array (may be related to a some specific/newer option of lda+U) get deallocated at the end of the first scf step but then it is not reallocated. If you send us a test job we can have a look at it, otherwise you should compile the code with debug options and find the exact place where it is crashing. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From giannozz at democritos.it Wed Sep 24 17:32:51 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 24 Sep 2008 17:32:51 +0200 Subject: [Pw_forum] array bound check In-Reply-To: References: Message-ID: <94132985-7B03-4E4E-A52D-3087F311FBAD@democritos.it> On Sep 24, 2008, at 15:24 , Manoj Srivastava wrote: > pwcond.x 0000000000471304 poten_ 69 > poten.f90 > pwcond.x 000000000043709E init_cond_ 102 > init_cond.f90 > pwcond.x 000000000042583F do_cond_ 250 > do_cond.f90 > pwcond.x 0000000000420E6F MAIN__ 20 > condmain.f90 it also says which array index has exceeded which bound > My question here is whether I am doing the array bound check right? and my question is: why are you doing this? array bound check is ONLY used to locate bugs. It should be never used in production work. If a code compiled with array bound check signals a problem, that's exactly what it is supposed to do. In most cases the errors it signal are not real errors, or they are harmless, but sometimes thay are true bugs. By the way: I didn't get any error (with the CVS version) after Alexander's fix was applied Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From zucco at dipteris.unige.it Wed Sep 24 17:39:51 2008 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Wed, 24 Sep 2008 17:39:51 +0200 Subject: [Pw_forum] vc-relax and LDA+U: Segmentation fault In-Reply-To: <48DA55CE.2000308@umn.edu> References: <48DA55CE.2000308@umn.edu> Message-ID: <6C145550-1A49-470F-B609-F3D1BE0BDE94@dipteris.unige.it> Dear Matteo, from documentation appears that U_projection_type = 'atomic' is the default and then I never declared it in the SYSTEM. When tprnfor and tstress were declared .true. the segmentation fault happened. If .false. happens again. Then it stop here at ... 0.2500 0.2500 0.0028 0.0027 -0.0044 -0.0004 -0.0004 -0.0002 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 the Fermi energy is 9.3699 ev ! total energy = -2166.89264188 Ry Harris-Foulkes estimate = -2166.89264402 Ry estimated scf accuracy < 0.00000264 Ry total magnetization = 0.00 Bohr mag/cell absolute magnetization = 22.98 Bohr mag/cell I used damp-w as cell-dynamic and default for ion-dynamic. I will try bfgs next. Thanks, m. On 24/set/08, at 16:59, Matteo Cococcioni wrote: > > Dear Marino, > > I have never run a vc-relax simulation with LDA+U but I know of > people who have done that successfully. > if U_projection_type = 'atomic' the code should compute forces and > stresses with LDA+U no problem. Do you obtain forces and stresses? > Also I'm not sure I have understood where the code stops. Does the > segmentation fault appear after the first scf cycle (and after the > first calculation > of stresses and forces)? > Also what kind of dynamics have you picked? > > regards, > > Matteo > > > Marino Vetuschi Zuccolini wrote: >> Hello to all, >> from a very recent email posted on the forum it seem to me that >> relaxation (maybe vc-relax) is possible using LDA+U, and that >> experimental implementation of phonons with LDA+U is going on (for >> this I hope to be a beta-tester if my structure converge..). >> My question is: does anyone of you have experience a Segmentation >> fault using vc-relax and LDA+U just after the wfc convergence has >> been achieved, starting 'from_scratch'? >> I'm using 4.0CVS in serial mode, but it also in 4.0.2 version it >> happens. >> Any help is welcome. >> >> Marino >> >> >> ******************************************************* >> Marino Vetuschi Zuccolini >> zucco at dipteris.unige.it >> Researcher / Geochemist >> Laboratory of Geochemistry >> >> DIPartimento per lo studio della TErra e delle sue RISorse - >> Universit? di Genova >> Tel. ++39 010 3538136 Fax. ++39 010 352169 >> Corso Europa 26, 16132 - Genova - Italy >> >> >> _______________________________________________ >> 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 ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From wenchenai at tom.com Wed Sep 24 20:35:39 2008 From: wenchenai at tom.com (Xiaodong) Date: Wed, 24 Sep 2008 14:35:39 -0400 Subject: [Pw_forum] some trouble in Phonon dispersion calculation Message-ID: <48DA8FA4.3D2F81.02238> Dear Prof. Eyvaz Isaev and Liangfeng I have handled this trouble about phonon dispersion. The reason is from K-piont. In the phonon dispersion calculation, high symmetry k-piont (q-vector) should be chosen and determined. However, for phonon density calculation, in general, there are only common k-piont (nk1,nk2,nk3). This is my fault. Thank you very much for your answer and response. Great! Best Xiaodong Cornell University http://www.chem.cornell.edu/rh34/ ======= 2008-09-24 08:35:00 ======= >---------------------------------------------------------------------- > >Message: 1 >Date: Tue, 23 Sep 2008 17:22:16 +0800 >From: " lfhuang " >Subject: Re: [Pw_forum] some trouble in Phonon dispersion calculation >To: " pw_forum " >Message-ID: <20080923092216.19381.qmail at ms.hfcas.ac.cn> >Content-Type: text/plain; charset="utf-8" > >Dear Xiaodong: > Could you have paid attention to the coordinations that were chosen by you and the code? Maybe they are different(x-y coordination or reciprocal lattice vector coordination). If you haven't, you can transform between >them with the transformation matrix that can be found in the pw.x output file. >Best Wishes! > Yours Sincerely > L.F.Huang >====================================================================== >L.F.Huang(???) lfhuang at theory.issp.ac.cn >====================================================================== >Add: Research Laboratory for Computational Materials Sciences, >Instutue of Solid State Physics,the Chinese Academy of Sciences, >P.O.Box 1129, Hefei 230031, P.R.China >Tel: 86-551-5591464-328(office) >Fax: 86-551-5591434 >Web: http://theory.issp.ac.cn (website of our theory group) >http://www.issp.ac.cn (website of our institute) >====================================================================== > >> From: "Wenxiaodong" >> Subject: [Pw_forum] some trouble in Phonon dispersion calculation >> To: "pw_forum" >> Message-ID: >> Content-Type: text/plain; charset="gb2312" >> >> Dear everybody >> Recently, I did a phonon dispersion calculation using PWscf. However, I can not obtain smooth phonon dispersion. That is to say, the phonon dispersion looks like zigzag shape. For making sure, I tried to use the example 07 (in PWscf), while there are same phenomena for phonon dispersion. By the way, I used the plotband.x to handle data. >> If you have some experience about this, could you tell me the reasons? I will very very appreciate you! >> Best >> Xiaodong > > >------ > > > >-------------- next part -------------- >An HTML attachment was scrubbed... >URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080923/fd27ff48/attachment.html > >------------------------------ > >Message: 2 >Date: Tue, 23 Sep 2008 04:49:37 -0700 (PDT) >From: Eyvaz Isaev >Subject: Re: [Pw_forum] some trouble in Phonon dispersion calculation >To: PWSCF Forum >Message-ID: <7392.1113.qm at web65702.mail.ac4.yahoo.com> >Content-Type: text/plain; charset=us-ascii > >Dear Xiaodong, > >Can you please submit the script you used and figure you obtained? If you submit frequency file it will be much easier to fix the problem. > >Please also provide your affiliation. > >Bests, >Eyvaz. > >------------------------------------------------------------------- >Prof. Eyvaz Isaev, >Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, >Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden >Condensed Matter Theory Group, Uppsala University, Sweden >Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com > > >--- On Tue, 9/23/08, Wenxiaodong wrote: > >> From: Wenxiaodong >> Subject: [Pw_forum] some trouble in Phonon dispersion calculation >> To: "pw_forum" >> Date: Tuesday, September 23, 2008, 8:01 AM >> Dear everybody >> Recently, I did a phonon dispersion calculation using >> PWscf. However, I can not obtain smooth phonon dispersion. >> That is to say, the phonon dispersion looks like zigzag >> shape. For making sure, I tried to use the example 07 (in >> PWscf), while there are same phenomena for phonon >> dispersion. By the way, I used the plotband.x to handle >> data. >> If you have some experience about this, could you tell me >> the reasons? I will very very appreciate you! >> Best >> Xiaodong >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum > > > > > >------------------------------ From matteo at umn.edu Thu Sep 25 06:30:43 2008 From: matteo at umn.edu (Matteo Cococcioni) Date: Wed, 24 Sep 2008 23:30:43 -0500 Subject: [Pw_forum] vc-relax and LDA+U: Segmentation fault In-Reply-To: <6C145550-1A49-470F-B609-F3D1BE0BDE94@dipteris.unige.it> References: <48DA55CE.2000308@umn.edu> <6C145550-1A49-470F-B609-F3D1BE0BDE94@dipteris.unige.it> Message-ID: <48DB13F3.4090409@umn.edu> Dear Marino, I have asked people in R. Wentzcovitch's group and they said that damp-w vc-relax had no problem with LDA+U using 3.2.3. I think that something like what Lorenzo was explaining happened in between. So I guess the problem can be tracked and solved just looking at the list of last madifications. don't take this as an encouragement to use previous versions of the code..... regards, Matteo Marino Vetuschi Zuccolini wrote: > Dear Matteo, > > from documentation appears that U_projection_type = 'atomic' is the > default and then I never declared it in the SYSTEM. When tprnfor and > tstress were declared .true. the segmentation fault happened. > If .false. happens again. > > Then it stop here at > ... > 0.2500 0.2500 0.0028 0.0027 -0.0044 -0.0004 -0.0004 > -0.0002 > 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 > 0.0000 > 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 > 0.0000 > 0.0000 > > the Fermi energy is 9.3699 ev > > ! total energy = -2166.89264188 Ry > Harris-Foulkes estimate = -2166.89264402 Ry > estimated scf accuracy < 0.00000264 Ry > > total magnetization = 0.00 Bohr mag/cell > absolute magnetization = 22.98 Bohr mag/cell > > I used damp-w as cell-dynamic and default for ion-dynamic. I will try > bfgs next. > > Thanks, > > m. > > On 24/set/08, at 16:59, Matteo Cococcioni wrote: > > >> Dear Marino, >> >> I have never run a vc-relax simulation with LDA+U but I know of >> people who have done that successfully. >> if U_projection_type = 'atomic' the code should compute forces and >> stresses with LDA+U no problem. Do you obtain forces and stresses? >> Also I'm not sure I have understood where the code stops. Does the >> segmentation fault appear after the first scf cycle (and after the >> first calculation >> of stresses and forces)? >> Also what kind of dynamics have you picked? >> >> regards, >> >> Matteo >> >> >> Marino Vetuschi Zuccolini wrote: >> >>> Hello to all, >>> from a very recent email posted on the forum it seem to me that >>> relaxation (maybe vc-relax) is possible using LDA+U, and that >>> experimental implementation of phonons with LDA+U is going on (for >>> this I hope to be a beta-tester if my structure converge..). >>> My question is: does anyone of you have experience a Segmentation >>> fault using vc-relax and LDA+U just after the wfc convergence has >>> been achieved, starting 'from_scratch'? >>> I'm using 4.0CVS in serial mode, but it also in 4.0.2 version it >>> happens. >>> Any help is welcome. >>> >>> Marino >>> >>> >>> ******************************************************* >>> Marino Vetuschi Zuccolini >>> zucco at dipteris.unige.it >>> Researcher / Geochemist >>> Laboratory of Geochemistry >>> >>> DIPartimento per lo studio della TErra e delle sue RISorse - >>> Universit? di Genova >>> Tel. ++39 010 3538136 Fax. ++39 010 352169 >>> Corso Europa 26, 16132 - Genova - Italy >>> >>> >>> _______________________________________________ >>> 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 >> > > > > ******************************************************* > Marino Vetuschi Zuccolini > zucco at dipteris.unige.it > Researcher / Geochemist > Laboratory of Geochemistry > > DIPartimento per lo studio della TErra e delle sue RISorse - > Universit? di Genova > Tel. ++39 010 3538136 Fax. ++39 010 352169 > Corso Europa 26, 16132 - Genova - Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From manoj at phys.ufl.edu Thu Sep 25 06:41:51 2008 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Thu, 25 Sep 2008 00:41:51 -0400 (EDT) Subject: [Pw_forum] array bound check In-Reply-To: <94132985-7B03-4E4E-A52D-3087F311FBAD@democritos.it> Message-ID: Dear Dr. Giannozzi, Thank you very much for your reply. The reason why I wanted to do the array bound check is that I have been trying to make some changes in the PWCOND part of the code to get the correct complex band for cu(111), but implementation of those changes was giving me segmentation fault. Now with the array bound check i perform, the code is even giving me a problem where there was no problem before :( . I guess I will try the CVS version and see what i get. By the way, is there a way to do array bound check in a particular subroutine I am interested rather than doing the whole code? Regards, Manoj Srivastava On Wed, 24 Sep 2008, Paolo Giannozzi wrote: > > On Sep 24, 2008, at 15:24 , Manoj Srivastava wrote: > > > pwcond.x 0000000000471304 poten_ 69 > > poten.f90 > > pwcond.x 000000000043709E init_cond_ 102 > > init_cond.f90 > > pwcond.x 000000000042583F do_cond_ 250 > > do_cond.f90 > > pwcond.x 0000000000420E6F MAIN__ 20 > > condmain.f90 > > it also says which array index has exceeded which bound > > > My question here is whether I am doing the array bound check right? > > and my question is: why are you doing this? array bound check is > ONLY used to locate bugs. It should be never used in production > work. If a code compiled with array bound check signals a problem, > that's exactly what it is supposed to do. In most cases the errors > it signal are not real errors, or they are harmless, but sometimes > thay are true bugs. By the way: I didn't get any error (with the CVS > version) after Alexander's fix was applied > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From baroni at sissa.it Thu Sep 25 08:06:17 2008 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 25 Sep 2008 08:06:17 +0200 Subject: [Pw_forum] array bound check In-Reply-To: References: Message-ID: <0CA9A2D0-312E-4FB5-AB9A-00B23E3E54C0@sissa.it> On Sep 25, 2008, at 6:41 AM, Manoj Srivastava wrote: > > try the CVS version and see what i get. By the way, is there a way > to do > array bound check in a particular subroutine I am interested rather > than > doing the whole code? there shoud bem I guess, but this is going to be a very compiler- specific issue that you can only address studying the the compiler's users' guide. most compilers have "directives" (i.e. comment lines with a special syntax) that are only read and interpreted when specic options are passed to the compiler itself. it depends very much on the specific compiler you are using ... SB --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080925/685de4dc/attachment-0001.htm From hande at newton.physics.metu.edu.tr Thu Sep 25 09:16:32 2008 From: hande at newton.physics.metu.edu.tr (Hande Ustunel) Date: Thu, 25 Sep 2008 10:16:32 +0300 (EEST) Subject: [Pw_forum] test pseudo bug? In-Reply-To: <1222268554.3304.7.camel@dhpc-5-53.sissa.it> Message-ID: Dear Andrea, I do in fact have this error -- not in the output of the original pseudo generation but in the output of the test. Thank you very much for your prompt reply. Best wishes, Hande On Wed, 24 Sep 2008, Dal Corso Andrea wrote: > On Wed, 2008-09-24 at 17:49 +0300, Hande Ustunel wrote: > > Dear PWSCF developers, > > > > I have been trying to generate an O pseudopotential using the 4.0 > > version. In the test section I noticed that the AE and PS values come out > > to be identical no matter how outrageous the test configuration is. The AE > > energies are also somewhat different. > > > > I include below my input file for the test and the relevant part of the > > output for both the 3.2 and the 4.0 versions. (The problem persists also in > > the 4.0.2 version) > > > > After some code mining, I MIGHT HAVE identified a possible source where in > > the ascheqps.f90, the fourth input (and output) variable, which is later > > called from the write_resultsps.f90 subroutine as elts(n) doesn't seem to > > be modified after being set to the ae energy in the beginning. I am, of > > course, really not sure if this is the real cause so I was wondering if you > > could help me see my mistake (if I'm making one) or if you could confirm > > that this is a bug. > > > Please check carefully the output. There should be an error message > saying Errors in the PS-KS equations. Your pseudo has a ghost and the > new routine is not able to find it so it does not change the eigenvalues > but it writes the error message. In the test case the program does not > stop because the next test configuration might converge. > > Andrea > > > > > Thank you very much in advance. > > Best wishes, > > Hande > > > > ------------------------------------------------------------ > > > > Test input : > > > > &input > > atom='O', > > rlderiv=4.0, > > eminld=-2.5, > > emaxld= 2.5, > > deld = 0.02, > > nld = 3, > > config='[He] 2s2 2p2 3s2.0 3p2.0', > > iswitch=2, > > prefix='o' > > dft='pbe', > > / > > &test > > file_pseudo='O.pbe.uspp-n-rrkj.UPF' > > nconf=3, > > configts(1)=' 2s2.00 2p1.75 ', > > configts(2)=' 2s1.80 2p2.50 ', > > configts(3)=' 2s1.80 2p1.50 ', > > / > > > > ------------------------------------------------------------ > > > > Energy diffences in the 3.2 version > > > > Conf 1 : > > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > > 1 0 2S 1( 2.00) -2.72810 -4.65642 1.92832 > > 2 1 2P 1( 1.75) -1.62655 -3.55432 1.92777 > > Conf 2 : > > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > > 1 0 2S 1( 1.80) -1.78231 -3.85038 2.06808 > > 2 1 2P 1( 2.50) -0.67236 -2.74216 2.06981 > > Conf 3 : > > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > > 1 0 2S 1( 1.80) -3.02350 -5.34542 2.32192 > > 2 1 2P 1( 1.50) -1.93748 -4.25302 2.31555 > > > > > > ------------------------------------------------------------ > > > > Energy diffences in the 4.0 version > > Conf 1 : > > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > > 1 0 2S 1( 2.00) -2.25668 -2.25668 0.00000 > > 2 1 2P 1( 1.75) -1.15425 -1.15425 0.00000 > > Conf 2 : > > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > > 1 0 2S 1( 1.80) -1.61762 -1.61762 0.00000 > > 2 1 2P 1( 2.50) -0.50726 -0.50726 0.00000 > > Conf 3 : > > n l nl e AE (Ry) e PS (Ry) De AE-PS (Ry) > > 1 0 2S 1( 1.80) -3.20787 -3.20787 0.00000 > > 2 1 2P 1( 1.50) -2.12306 -2.12306 0.00000 > > > -- Hande Toffoli 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 yuminqian at gmail.com Thu Sep 25 11:10:31 2008 From: yuminqian at gmail.com (yumin qian) Date: Thu, 25 Sep 2008 17:10:31 +0800 Subject: [Pw_forum] how to use PWSCF forum Message-ID: Dear I subscribed to the pw_forum mailing list, But I don't know how to use this forum to post my question and problem thank you -- Sincerely Y. M. Qian Lab.of Condensed Matter Theory and Materials Computation Institute of Physics Chinese Academy of Sciences Tel: + 8610 8264 9147 E-Mail:yuminqian at gmail.com P.O.Box 603 Beijing 100190 China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080925/195c6501/attachment.htm From giannozz at democritos.it Thu Sep 25 11:26:08 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 25 Sep 2008 11:26:08 +0200 Subject: [Pw_forum] array bound check In-Reply-To: References: Message-ID: <4F7560E1-34E1-4F5B-BD70-82CC9E674ECB@democritos.it> On Sep 25, 2008, at 6:41 , Manoj Srivastava wrote: > I guess I will try the CVS version and see what i get. I guess you should rather see what happens at the line of the code where the error happens. If you did things properly (likely), and if the compiler work properly (less likely), you should have gotten a message saying something very self-explanatory like Subscript #1 of the array R has value 172 which is greater than the upper bound of 171 in addition to a traceback with the line number where the offending command is called. Sometimes such messages are bogus, i.e. they do not signal a true out of bound array but usage of obsolete syntax that prevents the compiler to know about the true size of arrays > By the way, is there a way to do array bound check in a particular > subroutine I am interested rather than doing the whole code? there is the safe and slow way: compile normally the entire code, change the flags in make.sys to switch bound check on, remove the object (*.o) corresponding to the routine you want to check, "make" again Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From lanhaiping at gmail.com Thu Sep 25 11:26:30 2008 From: lanhaiping at gmail.com (lan haiping) Date: Thu, 25 Sep 2008 17:26:30 +0800 Subject: [Pw_forum] how to use PWSCF forum In-Reply-To: References: Message-ID: You already posted your question successfully !! Just Send your question to pw_forum at pwscf.org, but you should dig out archives since there are many repeated questions you may come to . On Thu, Sep 25, 2008 at 5:10 PM, yumin qian wrote: > Dear > I subscribed to the pw_forum mailing list, But I don't know how to > use this forum to post my question and problem > thank you > -- > Sincerely Y. M. Qian > Lab.of Condensed Matter Theory and Materials Computation > Institute of Physics > Chinese Academy of Sciences > Tel: + 8610 8264 9147 > E-Mail:yuminqian at gmail.com > P.O.Box 603 Beijing 100190 > China > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080925/44f9cd83/attachment.htm From paulatto at sissa.it Thu Sep 25 13:39:41 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Thu, 25 Sep 2008 13:39:41 +0200 Subject: [Pw_forum] array bound check In-Reply-To: References: Message-ID: <48DB787D.8070201@sissa.it> Manoj Srivastava ha scritto: > I guess I will > try the CVS version and see what i get. I should be enough for you to apply the fix committed by Alexander Smogunov a few days ago: http://www.democritos.it:8888/O-sesame/filediff?f=O-sesame/PWCOND/init_orbitals.f90&v1=1.12&v2=1.13 -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ From bsomnath at mrc.iisc.ernet.in Thu Sep 25 20:27:21 2008 From: bsomnath at mrc.iisc.ernet.in (Somnath) Date: Thu, 25 Sep 2008 23:57:21 +0530 Subject: [Pw_forum] Post processing of spin polarization calculation Message-ID: <1222367241.31544.18.camel@Abacus> Hi, I am learning to use quantum espresso. I want to do a spin polarized calculation of graphene antiferromagnetic nanoribbon calculation. I have used the input file from your forum - http://www.democritos.it/pipermail/pw_forum/2008-May/008891.html. After the scf run I am using the following input file for post processing (using pp.x) to get the spin polarization (rho(up)-rho(down)). &INPUTPP prefix = 'graphene' , outdir = '$outdir, filplot = 'sd.dat' , plot_num = 6, / &PLOT nfile = 1 , fileout = 'sd.xsf' , iflag = 3 , output_format = 5 , / I am getting the following output in the terminal Program POST-PROC v.4.0.1 starts ... Today is 25Sep2008 at 23:53:54 Parallel version (MPI) Number of processors in use: 1 Planes per process (thick) : nr3 = 81 npp = 81 ncplane =10368 Proc/ planes cols G planes cols G columns G Pool (dense grid) (smooth grid) (wavefct grid) 1 81 7949 427613 81 7949 427613 2071 56883 negative rho (up, down): 0.222E+00 0.222E+00 Calling punch_plot, plot_num = 6 Writing data to file sd.dat Reading data from file sd.dat Writing data to be plotted to file sd.xsf Plot Type: 3D Output format: XCrySDen But I am unable to visualize the spin polarization using XCrySDen. I am trying "xcrysden --xsf sd.xsf". But it is only showing the atoms. Can someone please help me regarding this ? Thanks, -- ---------------- Somnath Bhowmick Room No. 104, Materials Research Centre Indian Institute of Science Bangalore - 560012 e-mail: bsomnath at mrc.iisc.ernet.in somnath.bhowmick at gmail.com -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From manoj at phys.ufl.edu Thu Sep 25 20:52:14 2008 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Thu, 25 Sep 2008 14:52:14 -0400 (EDT) Subject: [Pw_forum] array bound check In-Reply-To: <4F7560E1-34E1-4F5B-BD70-82CC9E674ECB@democritos.it> Message-ID: Dear Dr. Giannozzi,Dr. Baroni and Dr. Paulatto, Thank you very much for your help. I finally have done the array bound check in the desired subroutine. But the array that goes out of bound, at least that is the message I get, I dont think it could go out of bound. Could this be problem of the compiler? I have noticed that sometimes when I run a job, it gives me some random error, but when i run the same job again, it runs without any complain! I am using intel ifort compiler on x86_64 architecture machine. Regards, Manoj Srivastava On Thu, 25 Sep 2008, Paolo Giannozzi wrote: > > On Sep 25, 2008, at 6:41 , Manoj Srivastava wrote: > > > I guess I will try the CVS version and see what i get. > > I guess you should rather see what happens at the line of the > code where the error happens. If you did things properly (likely), > and if the compiler work properly (less likely), you should have > gotten a message saying something very self-explanatory like > > Subscript #1 of the array R has value 172 which is greater than the > upper bound of 171 > > in addition to a traceback with the line number where the offending > command is called. > > Sometimes such messages are bogus, i.e. they do not signal a true > out of bound array but usage of obsolete syntax that prevents the > compiler to know about the true size of arrays > > > By the way, is there a way to do array bound check in a particular > > subroutine I am interested rather than doing the whole code? > > there is the safe and slow way: compile normally the entire code, > change the flags in make.sys to switch bound check on, remove > the object (*.o) corresponding to the routine you want to check, > "make" again > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From tone.kokalj at ijs.si Thu Sep 25 21:10:34 2008 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Thu, 25 Sep 2008 21:10:34 +0200 Subject: [Pw_forum] Post processing of spin polarization calculation In-Reply-To: <1222367241.31544.18.camel@Abacus> References: <1222367241.31544.18.camel@Abacus> Message-ID: <1222369834.27111.2.camel@walk.ijs.si> On Thu, 2008-09-25 at 23:57 +0530, Somnath wrote: > But I am unable to visualize the spin polarization using XCrySDen. > I am trying "xcrysden --xsf sd.xsf". But it is only showing the atoms. > Can someone please help me regarding this ? See: http://www.xcrysden.org/doc/density.html http://www.xcrysden.org/doc/isocontrol.html http://www.xcrysden.org/doc/plane1control.html Regards, Tone -- Tone Kokalj J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia (tel: +386-1-477-3523 // fax:+386-1-477-3822) From giannozz at democritos.it Fri Sep 26 10:00:50 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 26 Sep 2008 10:00:50 +0200 Subject: [Pw_forum] array bound check In-Reply-To: References: Message-ID: <48DC96B2.7000004@democritos.it> Manoj Srivastava wrote: > [...] the array that goes out of bound, at > least that is the message I get, I dont think it could go out of bound. > Could this be problem of the compiler? I have noticed that sometimes when > I run a job, it gives me some random error, but when i run the same job > again, it runs without any complain! I am using intel ifort compiler on > x86_64 architecture machine. me too, with no problem at all, but it depend on the specific version of the intel compiler and maybe also of the operating system (I never had problem with v.10.1.008, while Axel says that it miscompiles a lot of codes) Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From yuminqian at gmail.com Fri Sep 26 11:01:14 2008 From: yuminqian at gmail.com (yumin qian) Date: Fri, 26 Sep 2008 17:01:14 +0800 Subject: [Pw_forum] phonon symmetry and mode Message-ID: One thing that confuse me is the irreducible BZ k-points in phonon calculation. We use a q-points mesh in order to get the interatomic force constant. becuase of crystal symmetry we need not to calcaculate all the q points but a minimum of irreducible q-points. And in non-selfconsistent calculation of electron structure with fixed charge for a give q point . why we should use the symmetry of samll group of S(q) ? The group of S(q) is a subgroup of crystall lattice symmetry group S(L) ,what it the difference between these two symmetry group, how to determine S(q) from S(L)? so is this subgroup of S(q) determines the phonon mode at point q ? -- Sincerely Y. M. Qian Lab.of Condensed Matter Theory and Materials Computation Institute of Physics Chinese Academy of Sciences Tel: + 8610 8264 9147 E-Mail:yuminqian at gmail.com P.O.Box 603 Beijing 100190 China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080926/f59b10af/attachment.htm From chenweiguang82 at gmail.com Fri Sep 26 11:14:23 2008 From: chenweiguang82 at gmail.com (Weiguang Chen) Date: Fri, 26 Sep 2008 17:14:23 +0800 Subject: [Pw_forum] A puzzle about the aborption energy curve calculation Message-ID: <9e6d976a0809260214v2f559e81v95e021d9b285a698@mail.gmail.com> Hi friends, Many authors show the absorption energy curve function as the change of adsorption heigth in their papers for looking for the potential barrier, and they usually hold the system unrelaxed. But i have a puzzle about it, i think the system may be unstable which is placed different heigth, and some even change largely especially near the surface, which mean the system structure is not the initial anymore. How do i solve my puzzle? Thanks -- Best Wishes ChenWeiguang -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080926/7f8e0e67/attachment.htm From manoj at phys.ufl.edu Fri Sep 26 15:05:55 2008 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Fri, 26 Sep 2008 09:05:55 -0400 (EDT) Subject: [Pw_forum] array bound check In-Reply-To: <48DC96B2.7000004@democritos.it> Message-ID: Thank you Dr.Giannozzi. I will give it a shot with different compiler. -manoj On Fri, 26 Sep 2008, Paolo Giannozzi wrote: > Manoj Srivastava wrote: > > > [...] the array that goes out of bound, at > > least that is the message I get, I dont think it could go out of bound. > > Could this be problem of the compiler? I have noticed that sometimes when > > I run a job, it gives me some random error, but when i run the same job > > again, it runs without any complain! I am using intel ifort compiler on > > x86_64 architecture machine. > > me too, with no problem at all, but it depend on the specific version > of the intel compiler and maybe also of the operating system (I never > had problem with v.10.1.008, while Axel says that it miscompiles a lot > of codes) > > 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 wenchenai at tom.com Fri Sep 26 21:23:58 2008 From: wenchenai at tom.com (Xiaodong) Date: Fri, 26 Sep 2008 15:23:58 -0400 Subject: [Pw_forum] a error from davcio! Message-ID: <48DD36E7.6DA916.28572> Dear everybody Sorry for bothering you! Recently, I am running a phonon calculation. There is an error "from davcio". I do not know why. It is very strange. The job have run for one day very well. But, suddenly, the job dead! If you feel free, could you tell me what happened here? Thank you very much. The error have been listed: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% task # 4 from davcio : error # 22 error while writing to file %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Xiaodong Hoffmann group, Chemistry and Chemical Biology Cornell University, Ithaca, NY, USA. http://www.chem.cornell.edu/rh34/ From giannozz at democritos.it Fri Sep 26 21:57:35 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 26 Sep 2008 21:57:35 +0200 Subject: [Pw_forum] a error from davcio! In-Reply-To: <48DD36E7.6DA916.28572> References: <48DD36E7.6DA916.28572> Message-ID: <20D37532-AF99-4755-B47E-C47CBC011DCB@democritos.it> On Sep 26, 2008, at 21:23 , Xiaodong wrote: > Recently, I am running a phonon calculation. There is an error > "from davcio". > I do not know why. nor do we, but it is a question that has been asked at least N! times with a rather large N value. See item 1.12 at http://www.quantum-espresso.org/wiki/index.php/Troubleshooting_% 28PWscf%29 > Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at democritos.it Fri Sep 26 22:24:38 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 26 Sep 2008 22:24:38 +0200 Subject: [Pw_forum] A puzzle about the aborption energy curve calculation In-Reply-To: <9e6d976a0809260214v2f559e81v95e021d9b285a698@mail.gmail.com> References: <9e6d976a0809260214v2f559e81v95e021d9b285a698@mail.gmail.com> Message-ID: On Sep 26, 2008, at 11:14 , Weiguang Chen wrote: > Many authors show the absorption energy curve function as the change > of adsorption heigth in their papers for looking for the potential > barrier, > and they usually hold the system unrelaxed. But i have a puzzle > about it, > i think the system may be unstable which is placed different > heigth, and > some even change largely especially near the surface, which mean the > system structure is not the initial anymore. if I understand what you mean, you are referreing to the case of a molecule over a surface and on the energy vs distance curve E(d). The correct way to calculate E(d) is to impose a constraint to the (apporpriately defined) distance between the molecule and the surface and to relax the remaining degrees of freedom. In some cases you can get a decent approximated curve by keeping the molecule in its geometry and vary d. Of course it is an approximation. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at democritos.it Fri Sep 26 22:54:44 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 26 Sep 2008 22:54:44 +0200 Subject: [Pw_forum] phonon symmetry and mode In-Reply-To: References: Message-ID: On Sep 26, 2008, at 11:01 , yumin qian wrote: > The group of S(q) is a subgroup of crystal lattice symmetry group S > (L), S(q) is the subgroup of the point group S of the crystal, that comprises all symmetry operations that leave q unchanged ("small group of q"). S may be smaller than the symmetry group of the lattice alone. > how to determine S(q) from S(L)? by applying the definition. It is automatically done by the code. > so is this subgroup of S(q) determines the phonon mode at point q ? it determines the symmetry of phonon modes at point q, according to group theory. The symmetry in the phonon codes works this way: - for each q wavevector, the S(q) group is found - a new IBZ (Irreducible Brillouin Zone) is determined, using the S(q) symmetry instead of the symmetry S of the crystal point group - the irreproducible representations (irreps) of S(q) are found. Irreps are typically 1 to 3-dimensional (although there are a few cases of 4- and even of 6-dimensional irreps) - the 3N-dimensional representation generated by the 3N coordinates of the N atoms in the unit cell are decomposed into irreps. Linear- response calculations are performed for displacements along these irreps . - all sums over the IBZ are performed using the k-points in the IBZ of S(q) (typically larger than the IBZ used for the scf calculation, so more k-points are needed) - symmetrization of the potential and of the dynamical matrix is performed, using the S(q) group and the fact that each irrep transforms into itself - Finally, dynamical matrices at a different q in the star of q (i.e. all vectors generated by applying operations of S to q) are generated, by applying operations of S to the dynamical matrix Why is it so complicated? because it is a compromise between CPU and memory requirements. Disclaimer: I didn't write that part, so what written above reflects my present understanding and may not bear any resemblance with reality Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From yuminqian at gmail.com Sat Sep 27 04:40:50 2008 From: yuminqian at gmail.com (yumin qian) Date: Sat, 27 Sep 2008 10:40:50 +0800 Subject: [Pw_forum] phonon symmetry and mode In-Reply-To: References: Message-ID: thanks for your reply 2008/9/27 Paolo Giannozzi > On Sep 26, 2008, at 11:01 , yumin qian wrote: > > > The group of S(q) is a subgroup of crystal lattice symmetry group S > > (L), > > S(q) is the subgroup of the point group S of the crystal, that > comprises all > symmetry operations that leave q unchanged ("small group of q"). S may > be smaller than the symmetry group of the lattice alone. > > > how to determine S(q) from S(L)? > > by applying the definition. It is automatically done by the code. > > > so is this subgroup of S(q) determines the phonon mode at point q ? > > it determines the symmetry of phonon modes at point q, according to > group theory. > > The symmetry in the phonon codes works this way: > - for each q wavevector, the S(q) group is found > - a new IBZ (Irreducible Brillouin Zone) is determined, using the S(q) > symmetry instead of the symmetry S of the crystal point group > - the irreproducible representations (irreps) of S(q) are found. > Irreps are typically 1 to 3-dimensional (although there are a few > cases > of 4- and even of 6-dimensional irreps) > - the 3N-dimensional representation generated by the 3N coordinates of > the N atoms in the unit cell are decomposed into irreps. Linear- > response > calculations are performed for displacements along these irreps . > - all sums over the IBZ are performed using the k-points in the IBZ of > S(q) (typically larger than the IBZ used for the scf calculation, > so more > k-points are needed) > - symmetrization of the potential and of the dynamical matrix is > performed, > using the S(q) group and the fact that each irrep transforms into > itself > - Finally, dynamical matrices at a different q in the star of q (i.e. > all vectors > generated by applying operations of S to q) are generated, by > applying > operations of S to the dynamical matrix > > Why is it so complicated? because it is a compromise between CPU and > memory requirements. > > Disclaimer: I didn't write that part, so what written above reflects my > present understanding and may not bear any resemblance with reality > > Paolo > --- > Paolo Giannozzi, Dept of Physics, University of Udine > via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Sincerely Y. M. Qian Lab.of Condensed Matter Theory and Materials Computation Institute of Physics Chinese Academy of Sciences Tel: + 8610 8264 9147 E-Mail:yuminqian at gmail.com P.O.Box 603 Beijing 100190 China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080927/7df47dc0/attachment-0001.htm From mansourehp at gmail.com Sat Sep 27 14:07:37 2008 From: mansourehp at gmail.com (Mansoureh Pashangpour) Date: Sat, 27 Sep 2008 15:37:37 +0330 Subject: [Pw_forum] error Message-ID: Dear all I recieved this error when I was running CP molecular dynamics [cli_1]: aborting job: application called MPI_Abort(MPI_COMM_WORLD, 8072) - process 1 and rank 1 in job 1 n18.nanolab_53697 caused collective abort of all ranks exit status of rank 1: return code 136 what does it mean? thanks Mansoureh -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080927/0cf84538/attachment.htm From akohlmey at cmm.chem.upenn.edu Sat Sep 27 17:25:46 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 27 Sep 2008 11:25:46 -0400 (EDT) Subject: [Pw_forum] error In-Reply-To: References: Message-ID: On Sat, 27 Sep 2008, Mansoureh Pashangpour wrote: MP> Dear all MP> I recieved this error when I was running CP molecular dynamics MP> [cli_1]: aborting job: MP> application called MPI_Abort(MPI_COMM_WORLD, 8072) - process 1 MP> MP> and MP> MP> rank 1 in job 1 n18.nanolab_53697 caused collective abort of all ranks MP> exit status of rank 1: return code 136 MP> what does it mean? it means that your job has crashed and that the MPI library has terminated it because of that. axel. MP> thanks MP> Mansoureh MP> -- ======================================================================= 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 shrek_826 at yahoo.com.cn Sun Sep 28 05:01:45 2008 From: shrek_826 at yahoo.com.cn (Fan Yang) Date: Sun, 28 Sep 2008 11:01:45 +0800 (CST) Subject: [Pw_forum] how to add barrier potentials Message-ID: <575036.80905.qm@web15304.mail.cnb.yahoo.com> ?Dear PWscf-users, I'd like?to restrict the water molecules within a given distance from the?Metal surface, namely, preventing water molecules from leaking out of the interface slab. According to some literatures, confining potentials (usually parabolic barrier potentials) should be applied, does anyone know how to realize it in PWscf? Thanks & Best Regards, Fan Yang ******************************************* Fan Yang PH.D Candidate?in Electrochemistry College of Chemistry and Molecular Science Wuhan University, 430072, Hubei Province, China E-mail:shrek_826 at yahoo.com.cn ******************************************* -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080928/d497bed7/attachment.htm From baroni at sissa.it Sun Sep 28 09:15:09 2008 From: baroni at sissa.it (Stefano Baroni) Date: Sun, 28 Sep 2008 09:15:09 +0200 Subject: [Pw_forum] how to add barrier potentials In-Reply-To: <575036.80905.qm@web15304.mail.cnb.yahoo.com> References: <575036.80905.qm@web15304.mail.cnb.yahoo.com> Message-ID: Dear Fan Yang: there used to be an option in the past to add an arbitrary external potential. Not sure what the situation is now, but somebody can chip in on this issue. Stay tuned. SB On Sep 28, 2008, at 5:01 AM, Fan Yang wrote: > > Dear PWscf-users, > > I'd like to restrict the water molecules within a given distance > from the Metal surface, namely, preventing water molecules from > leaking out of the interface slab. According to some literatures, > confining potentials (usually parabolic barrier potentials) should > be applied, does anyone know how to realize it in PWscf? > > Thanks & Best Regards, > Fan Yang > ******************************************* > Fan Yang > PH.D Candidate in Electrochemistry > College of Chemistry and Molecular Science > Wuhan University, 430072, Hubei Province, China > E-mail:shrek_826 at yahoo.com.cn > ******************************************* > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080928/0e3f7b54/attachment.htm From wangqj1 at 126.com Sun Sep 28 15:07:28 2008 From: wangqj1 at 126.com (wangqj1) Date: Sun, 28 Sep 2008 21:07:28 +0800 (CST) Subject: [Pw_forum] how to compute the energy level ? Message-ID: <27712502.223911222607248109.JavaMail.coremail@bj126app75.126.com> Dear pwscf users Does somebody know how to compute the energy level ? Any advice will be appreciated . sincerely wangqj -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080928/36ece31a/attachment.htm From giannozz at democritos.it Sun Sep 28 15:20:47 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sun, 28 Sep 2008 15:20:47 +0200 Subject: [Pw_forum] how to compute the energy level ? In-Reply-To: <27712502.223911222607248109.JavaMail.coremail@bj126app75.126.com> References: <27712502.223911222607248109.JavaMail.coremail@bj126app75.126.com> Message-ID: On Sep 28, 2008, at 15:07 , wangqj1 wrote: > Does somebody know how to compute the energy level ? of what? --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From rezaebraahimi at yahoo.com Sun Sep 28 19:45:47 2008 From: rezaebraahimi at yahoo.com (Reza Ebraahimi) Date: Sun, 28 Sep 2008 10:45:47 -0700 (PDT) Subject: [Pw_forum] about superposition of atomic rho and atomic wfc's Message-ID: <138754.30673.qm@web59811.mail.ac4.yahoo.com> Hi I am new with pwscf, so sorry if my question seems basic to you! i have noticed that the code calculates the superpositon of charge density and wfc's from atomic configurations by two subroutines: "atomic_rho.f90" and "atomic_wfc.f90". Could anyone refer me to a text or something to know exactly what happens in these two( i mean mathematically)? And i thought that we need the superposition of charge density from atomic configuration to build up the KS Hamiltonian( with the knowledge of n(r) )?and obtain the solid wfc's by diagonalizing it at the first iteration. Why do we need the superposition of wfc's from atomic configuration? Thank You -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080928/04c05cd7/attachment.htm From degironc at sissa.it Sun Sep 28 20:33:04 2008 From: degironc at sissa.it (Stefano de Gironcoli) Date: Sun, 28 Sep 2008 20:33:04 +0200 Subject: [Pw_forum] about superposition of atomic rho and atomic wfc's In-Reply-To: <138754.30673.qm@web59811.mail.ac4.yahoo.com> References: <138754.30673.qm@web59811.mail.ac4.yahoo.com> Message-ID: <20080928203304.cjg8pdskgkkcw44s@webmail.sissa.it> Dear Reza Ebraahimi, Quoting Reza Ebraahimi : > Hi > I am new with pwscf, so sorry if my question seems basic to you! > i have noticed that the code calculates the superpositon of charge > density and wfc's from atomic configurations by two subroutines: > "atomic_rho.f90" and "atomic_wfc.f90". Could anyone refer me to a > text or something to know exactly what happens in these two( i mean > mathematically)? atomic_rho builds the charge density from superposition of atomic charges in reciprocal space as the product of the spherical fourier transform of the atomic charge and the structure factor of each atomic species and then tranform them in real space. atomic_wfc does the same for the wfcs with some variations due to the fact that wfcs correspond to a given angular momentum. To understand the coded formulas it may be useful to have at hand the expansion of a plane wave in terms of spherical harmonics and bessel functions that one can find in basic book on quantum mechanics. > And i thought that we need the superposition of charge density from > atomic configuration to build up the KS Hamiltonian( with the > knowledge of n(r) ) and obtain the solid wfc's by diagonalizing it > at the first iteration. Why do we need the superposition of wfc's > from atomic configuration? Iterative diagonalizations require a starting guess. The code allows you to choose among atomic or random... > Thank You Don't forget to provide your affiliation best regards, Stefano de Gironcoli - SISSA and DEMOCRITOS ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by Horde http://www.horde.org/ From vegalew at hotmail.com Mon Sep 29 06:24:16 2008 From: vegalew at hotmail.com (vega) Date: Mon, 29 Sep 2008 12:24:16 +0800 Subject: [Pw_forum] questions about the scf convergence and accuracy Message-ID: Dear all, Now I'm suffering the convergence problems about my system. My system is the surface of TiO2 and noble metal atoms. First I relaxed surface of TiO2 with a slab geometry. Most of parameter of the calculations was set to default. The convergence of SCF and BFGS a bit slower but they both converged finally. When I add the noble metals on the surface, the convergence of SCF can't converge any more. Then I checked the output of the not converged systems. I found the self-consistency error fluctuated, some thing like the manual describing, 'A typical sign of such behavior is that the self-consistency error goes down, down, down, than all of a sudden up again, and so on'. So I think maybe I should increase the empty bands to the system because of the adding of noble metal atoms. Then I add about 20% more empty bands to the system. I want to know whether the accuracy will be affected by doing this? Do you think adding more empty bands to the system will significantly affect the results of relax calculation for TiO2 and noble metals? And I also found another suggestions to using mixing mode=?local-TF? for slab systems, and Reduce mixing beta. Do you think there will be a big difference between the relax calculation using default setting and adjusted settings to improve the SCF convergence? thank you for reading. any hints will be appreciated. vega ================================================================================= Vega Lew (weijia liu) PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080929/592e5aef/attachment-0001.htm From sathya_sheela1985001 at rediffmail.com Mon Sep 29 08:00:33 2008 From: sathya_sheela1985001 at rediffmail.com (sathya subramanyan) Date: 29 Sep 2008 06:00:33 -0000 Subject: [Pw_forum] symmetry and character table for C2/C Message-ID: <20080929060033.3104.qmail@f4mail-235-220.rediffmail.com> ? Dear All, I am doing the optimization of a C2/c structure but I'm not getting correct the symmetry operations or the character table.I'm using espresso-4.0.2.There was a similar question in the forum about a R3c group and I even tried the suggestion given on Tue Sep 9 17:32:48 CEST 2008 by Andrea .On Tue Sep 9 17:32:48 CEST 2008 Andrea wrote.... Thank you for reporting this. Actually there was a bug with the improper rotations. I have commited the fix on the cvs version. To correct the bug you can substitute the two files PW/divide_class.f90 PW/divide_class_so.f90 with those of the cvs version. I have attached the input for reference This is the Output Im getting.Please do help .Thank you in advance LDA+U calculation, Hubbard_lmax = 2 atomic species L Hubbard U Hubbard alpha Mn 2 0.166842 0.064679 No symmetry! s frac. trans. isym = 1 identity cryst. s( 1) = ( 1 0 0 ) ( 0 1 0 ) ( 0 0 1 ) cart. s( 1) = ( 1.0000000 -0.0000000 0.0000000 ) ( 0.0000000 1.0000000 0.0000000 ) ( 0.0000000 0.0000000 1.0000000 ) point group C_1 (1) there are 1 classes the character table: E A 1.00 the symmetry operations in each class: E 1 Regards SATHYA SHEELA.S Grad. Student Department of Physics National Institute of Technology Tiruchirapalli - 620015 India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080929/a991f969/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: bimno3.rx.in Type: application/octet-stream Size: 2349 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20080929/a991f969/attachment.obj From giannozz at democritos.it Mon Sep 29 08:33:32 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 29 Sep 2008 08:33:32 +0200 Subject: [Pw_forum] symmetry and character table for C2/C In-Reply-To: <20080929060033.3104.qmail@f4mail-235-220.rediffmail.com> References: <20080929060033.3104.qmail@f4mail-235-220.rediffmail.com> Message-ID: <48E076BC.5080705@democritos.it> sathya subramanyan wrote: > I am doing the optimization of a C2/c structure but I'm not > getting correct the symmetry operations http://www.quantum-espresso.org/wiki/index.php/Troubleshooting_(PWscf)#pw.x_doesn.E2.80.99t_.EF.AC.81nd_all_the_symmetries_you_expected -- Paolo Giannozzi, Democritos and University of Udine, Italy From baroni at sissa.it Mon Sep 29 08:33:21 2008 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 29 Sep 2008 08:33:21 +0200 Subject: [Pw_forum] about superposition of atomic rho and atomic wfc's In-Reply-To: <138754.30673.qm@web59811.mail.ac4.yahoo.com> References: <138754.30673.qm@web59811.mail.ac4.yahoo.com> Message-ID: On Sep 28, 2008, at 7:45 PM, Reza Ebraahimi wrote: > Hi > I am new with pwscf, so sorry if my question seems basic to you! > i have noticed that the code calculates the superpositon of charge > density and wfc's from atomic configurations by two subroutines: > "atomic_rho.f90" and "atomic_wfc.f90". Could anyone refer me to a > text or something to know exactly what happens in these two( i mean > mathematically)? > And i thought that we need the superposition of charge density from > atomic configuration to build up the KS Hamiltonian( with the > knowledge of n(r) ) and obtain the solid wfc's by diagonalizing it > at the first iteration. Why do we need the superposition of wfc's > from atomic configuration? as a first guess for iterative diagonalization of the KS Hamiltonian > Thank You you are most welcome - you would be even more so if you were to use the courtesy commonly adopted in this forum to provide your name with affiliation SB --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080929/993f2fd9/attachment.htm From marcel at physik.tu-berlin.de Mon Sep 29 10:28:30 2008 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Mon, 29 Sep 2008 10:28:30 +0200 (CEST) Subject: [Pw_forum] Two binaries for serial and parallel use needed In-Reply-To: References: <138754.30673.qm@web59811.mail.ac4.yahoo.com> Message-ID: Dear linux experts I have newly set up our computer cluster with new mpich2 (1.07) and ifort. In the last cluster the binary pw.x (compiled with mpich2 1.04) could be executed directly or with mpiexec for serial or parallel use, respectively. Now, this is not possible anymore, I always have to start an smpd and run pw.x with mpiexec, even for serial usage. It is a bit uncomfortable, so if anyone knows a switch for the compiler, or could comment I would greatly appreciate it. Cheers Marcel From sathya_sheela1985001 at rediffmail.com Mon Sep 29 13:49:32 2008 From: sathya_sheela1985001 at rediffmail.com (sathya subramanyan) Date: 29 Sep 2008 11:49:32 -0000 Subject: [Pw_forum] symmetry and character table for C2/C Message-ID: <20080929114932.39284.qmail@f4mail-235-128.rediffmail.com> ? Dear Sir, Sorry I forgot to mention that I had read your answer to a question asked in the forum on Tue May 23 18:25:53 CEST 2006 about symmetry operations and have already tried with different values of accep in PW/eqvect.f90 but I got the same output..No Symmetry and the wrong symmetry operations.Does it mean that the compound contains a fractional translation that is incompatible with the FFT grid .Please do help out. Regards SATHYA SHEELA.S Grad. Student Department of Physics National Institute of Technology Tiruchirapalli - 620015 India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080929/48048520/attachment.htm From giannozz at democritos.it Mon Sep 29 14:30:28 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 29 Sep 2008 14:30:28 +0200 Subject: [Pw_forum] symmetry and character table for C2/C In-Reply-To: <20080929114932.39284.qmail@f4mail-235-128.rediffmail.com> References: <20080929114932.39284.qmail@f4mail-235-128.rediffmail.com> Message-ID: <48E0CA64.4040806@democritos.it> sathya subramanyan wrote: > Does it mean that the compound contains a fractional translation > that is incompatible with the FFT grid if this is what is happening, the code prints a message. If not, the symmetry of the crystal is not what you think it is. Xcrysden allows a quick and easy way to visualize the crystal as generated by input data. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From akohlmey at cmm.chem.upenn.edu Mon Sep 29 16:21:35 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 29 Sep 2008 10:21:35 -0400 (EDT) Subject: [Pw_forum] Two binaries for serial and parallel use needed In-Reply-To: References: <138754.30673.qm@web59811.mail.ac4.yahoo.com> Message-ID: On Mon, 29 Sep 2008, Marcel Mohr wrote: MM> Dear linux experts MM> MM> I have newly set up our computer cluster with new mpich2 (1.07) MM> and ifort. MM> In the last cluster the binary pw.x (compiled with mpich2 1.04) could be MM> executed directly or with mpiexec for serial or parallel use, MM> respectively. MM> Now, this is not possible anymore, I always have to start an smpd MM> and run pw.x with mpiexec, even for serial usage. MM> It is a bit uncomfortable, so if anyone knows a switch for the compiler, MM> or could comment I would greatly appreciate it. this has to be a "feature" of MPICH, or the way you compiled it. i'm using OpenMPI and don't have that problem. if you are after convenience, you should seriously consider switching to OpenMPI. particularly on PBS/Torque managed clusters, it is _extremely_ convenient (job launch through torque, flexible choice of node assignment, high efficiency, independence of the executable from the underlying communication layer (i can compile executable on my desktop with no network and run them on myrinet, or infiniband machines directly)). cheers, axel. MM> MM> Cheers MM> Marcel MM> _______________________________________________ MM> Pw_forum mailing list MM> Pw_forum at pwscf.org MM> http://www.democritos.it/mailman/listinfo/pw_forum MM> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From wenchenai at tom.com Mon Sep 29 16:30:55 2008 From: wenchenai at tom.com (Xiaodong) Date: Mon, 29 Sep 2008 10:30:55 -0400 Subject: [Pw_forum] a error from davcio! Message-ID: <48E0E6B5.8CCB5A.28261> Hi Paolo Thank you very much for the information. I thought that this problems are from the directory "tmp". There are wrong reading data from the tmp one setup! try it again! Best Xiaodong ##################################################### >Message: 6 >Date: Fri, 26 Sep 2008 21:57:35 +0200 >From: Paolo Giannozzi >Subject: Re: [Pw_forum] a error from davcio! >To: PWSCF Forum >Message-ID: <20D37532-AF99-4755-B47E-C47CBC011DCB at democritos.it> >Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed > > >On Sep 26, 2008, at 21:23 , Xiaodong wrote: > >> Recently, I am running a phonon calculation. There is an error >> "from davcio". >> I do not know why. > >nor do we, but it is a question that has been asked at least N! times >with >a rather large N value. See item 1.12 at >http://www.quantum-espresso.org/wiki/index.php/Troubleshooting_% >28PWscf%29 >> > > >Paolo >--- >Paolo Giannozzi, Dept of Physics, University of Udine >via delle Scienze 208, 33100 Udine, Italy >Phone +39-0432-558216, fax +39-0432-558222 ####################################################### From nakhmanson at anl.gov Tue Sep 30 01:48:00 2008 From: nakhmanson at anl.gov (Serge Nakhmanson) Date: Mon, 29 Sep 2008 18:48:00 -0500 Subject: [Pw_forum] Compiling with intel fortran, MKL and MKL FFT wrappers Message-ID: <48E16930.1000506@anl.gov> Dear All, Here's a question to those of you who succeeded doing the Subj. Here's how MKL distribution looks like on my cluster: serafina {1323} > pwd /usr/local/intel/mkl/10.0.4.023/interfaces serafina {1324} > ls blas95 fftw2xc fftw2x_cdft fftw2xf fftw3xc fftw3xf lapack95 serafina {1325} > We (me and our sysadmin) compiled all versions of the MKL FFT wrappers, but we assumed that the parallel one (fftw2x_cdft) is the one to be used when compiling Espresso. However, linking with that one produces a bunch of undefined references in {espresso}/clib/fft_stick.c. When, instead, we link with fftw2xc (which is serial), everything compiles and works just fine. My question is: is the latter (serial) library really the one intended to be used with Espresso, or is it the parallel one that should work but fails due to some function renaming in the latest MKL version? Or is there something else that could be wrong? Just in case, here are a few informative lines from our make.sys: DFLAGS = -D__INTEL -D__FFTW -D__MPI -D__PARA IFLAGS = -I../include -I/usr/local/intel/mkl/10.0.4.023/include/fftw BLAS_LIBS = -L/usr/local/intel/mkl/10.0.4.023/lib/em64t -lmkl_intel_lp64 -lmkl_sequential -lmkl_core LAPACK_LIBS = -L/usr/local/intel/mkl/10.0.4.023/lib/em64t -lmkl_intel_lp64 -lmkl_sequential -lmkl_core #FFT_LIBS = -L/usr/local/intel/mkl/10.0.4.023/lib/64 -lfftw2x_cdft_DOUBLE FFT_LIBS = -L/usr/local/intel/mkl/10.0.4.023/lib/64 -lfftw2xc_intel THX, Serge -- ********************************************************* Serge M. Nakhmanson phone: (630) 252-5205 Assistant Scientist fax: (630) 252-4798 MSD-212, Rm. C-224 Argonne National Laboratory 9700 S. Cass Ave. Argonne, IL 60439 ********************************************************* From paulatto at sissa.it Tue Sep 30 09:27:05 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Tue, 30 Sep 2008 09:27:05 +0200 (CEST) Subject: [Pw_forum] Compiling with intel fortran, MKL and MKL FFT wrappers In-Reply-To: <48E16930.1000506@anl.gov> References: <48E16930.1000506@anl.gov> Message-ID: <14069.82.50.174.28.1222759625.squirrel@webmail.sissa.it> On Mar, Settembre 30, 2008 01:48, Serge Nakhmanson wrote: > My question is: is the latter (serial) library really the > one intended to be used with Espresso, or is it the parallel > one that should work but fails due to some function renaming > in the latest MKL version Dear Serge, QE itself is parallel. Using a parallel version of any library would reduce performance as the different processes, on which QE is parallelized, would spawn multiple threads, on which the libraries are parallelized, using more processors than available. If you look for OMP_NUM_THREAD in the list archive you will find extensive discussions on this matter. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/ From vegalew at hotmail.com Tue Sep 30 10:38:05 2008 From: vegalew at hotmail.com (vega) Date: Tue, 30 Sep 2008 16:38:05 +0800 Subject: [Pw_forum] Compiling with intel fortran, MKL and MKL FFT wrappers In-Reply-To: <48E16930.1000506@anl.gov> References: <48E16930.1000506@anl.gov> Message-ID: Dear Serge, I'm very anxious to know how to link the intel FFTW in MKL to the QE. From conor.hogan at roma2.infn.it Tue Sep 30 11:57:08 2008 From: conor.hogan at roma2.infn.it (Conor Hogan) Date: Tue, 30 Sep 2008 11:57:08 +0200 (CEST) Subject: [Pw_forum] Problems with nscf diago with many bands Message-ID: <43053.141.108.248.231.1222768628.squirrel@141.108.248.231> Dear forum, I'm doing a slab calculation, and for reasons I wont explain, I need to obtain the nscf KS states for many many unoccupied bands. There are 96 occupied bands: using the davidson I can diagonalize ok for 500 bands and 1000 bands, but if I push it up to 1500 I get problems with (3.2.3): from cdiaghg : error # 2684 info =/= 0 or (4.0): from pzpotrf : error # 94 problems computing cholesky decomposition I fear the conjugate gradient will be very slow, so before I try that, any suggestions for either approach with so many bands? I'm using the system compiled pw.x on BCX. Cheers, Conor &system ecutwfc =12.0, nbnd=1500, / &electrons diago_thr_init=5.0e-6, diago_full_acc=.true. electron_maxstep = 70, diagonalization='david' mixing_mode = 'plain' mixing_beta = 0.7 conv_thr = 1.0d-8, / ---- Dr. Conor Hogan --- ----- --- --- Dipartimento di Fisica e CNR-INFM | : | : Universita' di Roma "Tor Vergata" |-- : --- |--- Tel: +39 06 7259 4121 *NEW* | : | : Fax: +39 06 2023507 --- : --- : http://www.fisica.uniroma2.it/~cmtheo-group/ European Theoretical Spectroscopy Facility http://www.etsf.eu Q. What do you call a fish with no eyes? A. A fsh. (Tommy Cooper) The early bird gets the worm, but the second mouse gets the cheese. - S. Wright If you go through a lot of hammers each month, I don't think it necessarily means you're a hard worker. It may just mean that you have a lot to learn about proper hammer maintenance - J. Handey From akohlmey at cmm.chem.upenn.edu Tue Sep 30 15:33:56 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Tue, 30 Sep 2008 09:33:56 -0400 (EDT) Subject: [Pw_forum] Compiling with intel fortran, MKL and MKL FFT wrappers In-Reply-To: <48E16930.1000506@anl.gov> References: <48E16930.1000506@anl.gov> Message-ID: On Mon, 29 Sep 2008, Serge Nakhmanson wrote: SN> Dear All, SN> SN> Here's a question to those of you who succeeded doing the Subj. SN> SN> Here's how MKL distribution looks like on my cluster: SN> SN> serafina {1323} > pwd SN> /usr/local/intel/mkl/10.0.4.023/interfaces SN> serafina {1324} > ls SN> blas95 fftw2xc fftw2x_cdft fftw2xf fftw3xc fftw3xf lapack95 SN> serafina {1325} > SN> SN> We (me and our sysadmin) compiled all versions of the MKL FFT SN> wrappers, but we assumed that the parallel one (fftw2x_cdft) SN> is the one to be used when compiling Espresso. no it is not. Q-E does its own parallelization of the 3d-FFT. SN> However, linking with that one produces a bunch of undefined SN> references in {espresso}/clib/fft_stick.c. SN> SN> When, instead, we link with fftw2xc (which is serial), everything SN> compiles and works just fine. last time i tested them _neither_ of the intel MKL fftw wrappers worked reliably under all circumstances and - with very few exceptions - they had no advantage over using the regular FFTW. your version of MKL is newer than the one that i had the opportunity to check, so the fftw wrapper problems may have been resolved, but it is quite likely, that plain fftw will still be faster in most cases. the major advantage of the MKL fftw would be in using multi-threaded MKL in combination with MPI, but so far a significant speedup vs. the MPI parallelization has only been found for a _serial_ compile of Q-E in combination with a system with a large number of states. cheers, axel. SN> My question is: is the latter (serial) library really the SN> one intended to be used with Espresso, or is it the parallel SN> one that should work but fails due to some function renaming SN> in the latest MKL version? Or is there something else that SN> could be wrong? SN> SN> Just in case, here are a few informative lines from our make.sys: SN> SN> DFLAGS = -D__INTEL -D__FFTW -D__MPI -D__PARA SN> IFLAGS = -I../include -I/usr/local/intel/mkl/10.0.4.023/include/fftw SN> BLAS_LIBS = -L/usr/local/intel/mkl/10.0.4.023/lib/em64t -lmkl_intel_lp64 SN> -lmkl_sequential -lmkl_core SN> LAPACK_LIBS = -L/usr/local/intel/mkl/10.0.4.023/lib/em64t -lmkl_intel_lp64 SN> -lmkl_sequential -lmkl_core SN> #FFT_LIBS = -L/usr/local/intel/mkl/10.0.4.023/lib/64 -lfftw2x_cdft_DOUBLE SN> FFT_LIBS = -L/usr/local/intel/mkl/10.0.4.023/lib/64 -lfftw2xc_intel SN> SN> SN> THX, SN> SN> Serge SN> SN> SN> -- ======================================================================= 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 Tue Sep 30 15:34:43 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Tue, 30 Sep 2008 09:34:43 -0400 (EDT) Subject: [Pw_forum] Compiling with intel fortran, MKL and MKL FFT wrappers In-Reply-To: <14069.82.50.174.28.1222759625.squirrel@webmail.sissa.it> References: <48E16930.1000506@anl.gov> <14069.82.50.174.28.1222759625.squirrel@webmail.sissa.it> Message-ID: On Tue, 30 Sep 2008, Lorenzo Paulatto wrote: LP> LP> On Mar, Settembre 30, 2008 01:48, Serge Nakhmanson wrote: LP> > My question is: is the latter (serial) library really the LP> > one intended to be used with Espresso, or is it the parallel LP> > one that should work but fails due to some function renaming LP> > in the latest MKL version LP> LP> Dear Serge, LP> QE itself is parallel. Using a parallel version of any library would LP> reduce performance as the different processes, on which QE is LP> parallelized, would spawn multiple threads, on which the libraries are LP> parallelized, using more processors than available. LP> LP> If you look for OMP_NUM_THREAD in the list archive you will find extensive LP> discussions on this matter. sorry, but this is a completely unrelated matter. cheers, axel. LP> LP> regards LP> LP> -- ======================================================================= 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 Tue Sep 30 15:36:31 2008 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Tue, 30 Sep 2008 09:36:31 -0400 (EDT) Subject: [Pw_forum] Compiling with intel fortran, MKL and MKL FFT wrappers In-Reply-To: References: <48E16930.1000506@anl.gov> Message-ID: On Tue, 30 Sep 2008, vega wrote: VL> Dear Serge, VL> VL> I'm very anxious to know how to link the intel FFTW in MKL to the QE. just read the MKL documentation. it is explained in detail. RTFM. but then again, it is a rather pointless exercise. FFTW works fine and very reliable and MKL's fft (through fftw) is rarely faster and crashes occasionally. axel. VL> >From your make.sys, I know you might link the intel FFTW to QE successfully. VL> VL> could you tell me which dir should be added to the LD_LIBRARY_PATH, or which VL> flags should be added after the QE configure command? VL> VL> thank you VL> VL> vega VL> VL> ================================================================================= VL> Vega Lew (weijia liu) VL> PH.D Candidate in Chemical Engineering VL> State Key Laboratory of Materials-oriented Chemical Engineering VL> College of Chemistry and Chemical Engineering VL> Nanjing University of Technology, 210009, Nanjing, Jiangsu, China VL> VL> -------------------------------------------------- VL> From: "Serge Nakhmanson" VL> Sent: Tuesday, September 30, 2008 7:48 AM VL> To: "PWSCF Forum" VL> Subject: [Pw_forum] Compiling with intel fortran, MKL and MKL FFT wrappers VL> VL> > Dear All, VL> > VL> > Here's a question to those of you who succeeded doing the Subj. VL> > VL> > Here's how MKL distribution looks like on my cluster: VL> > VL> > serafina {1323} > pwd VL> > /usr/local/intel/mkl/10.0.4.023/interfaces VL> > serafina {1324} > ls VL> > blas95 fftw2xc fftw2x_cdft fftw2xf fftw3xc fftw3xf lapack95 VL> > serafina {1325} > VL> > VL> > We (me and our sysadmin) compiled all versions of the MKL FFT VL> > wrappers, but we assumed that the parallel one (fftw2x_cdft) VL> > is the one to be used when compiling Espresso. VL> > VL> > However, linking with that one produces a bunch of undefined VL> > references in {espresso}/clib/fft_stick.c. VL> > VL> > When, instead, we link with fftw2xc (which is serial), everything VL> > compiles and works just fine. VL> > VL> > My question is: is the latter (serial) library really the VL> > one intended to be used with Espresso, or is it the parallel VL> > one that should work but fails due to some function renaming VL> > in the latest MKL version? Or is there something else that VL> > could be wrong? VL> > VL> > Just in case, here are a few informative lines from our make.sys: VL> > VL> > DFLAGS = -D__INTEL -D__FFTW -D__MPI -D__PARA VL> > IFLAGS VL> > = -I../include -I/usr/local/intel/mkl/10.0.4.023/include/fftw VL> > BLAS_LIBS VL> > = -L/usr/local/intel/mkl/10.0.4.023/lib/em64t -lmkl_intel_lp64 VL> > -lmkl_sequential -lmkl_core VL> > LAPACK_LIBS VL> > = -L/usr/local/intel/mkl/10.0.4.023/lib/em64t -lmkl_intel_lp64 VL> > -lmkl_sequential -lmkl_core VL> > #FFT_LIBS VL> > = -L/usr/local/intel/mkl/10.0.4.023/lib/64 -lfftw2x_cdft_DOUBLE VL> > FFT_LIBS = -L/usr/local/intel/mkl/10.0.4.023/lib/64 -lfftw2xc_intel VL> > VL> > VL> > THX, VL> > VL> > Serge VL> > VL> > VL> > -- VL> > ********************************************************* VL> > Serge M. Nakhmanson phone: (630) 252-5205 VL> > Assistant Scientist fax: (630) 252-4798 VL> > MSD-212, Rm. C-224 VL> > Argonne National Laboratory VL> > 9700 S. Cass Ave. VL> > Argonne, IL 60439 VL> > ********************************************************* VL> > _______________________________________________ VL> > Pw_forum mailing list VL> > Pw_forum at pwscf.org VL> > http://www.democritos.it/mailman/listinfo/pw_forum VL> > VL> _______________________________________________ VL> Pw_forum mailing list VL> Pw_forum at pwscf.org VL> http://www.democritos.it/mailman/listinfo/pw_forum VL> -- ======================================================================= 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 conor.hogan at roma2.infn.it Tue Sep 30 16:13:00 2008 From: conor.hogan at roma2.infn.it (Conor Hogan) Date: Tue, 30 Sep 2008 16:13:00 +0200 (CEST) Subject: [Pw_forum] Problems with nscf diago with many bands In-Reply-To: <43053.141.108.248.231.1222768628.squirrel@141.108.248.231> References: <43053.141.108.248.231.1222768628.squirrel@141.108.248.231> Message-ID: <51706.141.108.248.231.1222783980.squirrel@141.108.248.231> Hello again, forget that last message, problem solved: < CG style diagonalization < ethr = 5.00E-06, avg # of iterations = 34.7 < PWSCF : 3m47.44s CPU time, 3m59.01s wall time > Davidson diagonalization with overlap > ethr = 5.00E-06, avg # of iterations = 27.2 > PWSCF : 7m41.49s CPU time, 7m47.46s wall time Wasn't expecting to see CG faster! PW continues to amaze me with its brutal efficiency :) cheers Conor On Tue, September 30, 2008 11:57 am, Conor Hogan said: > Dear forum, > I'm doing a slab calculation, and for reasons I wont explain, I need to > obtain the nscf KS states for many many unoccupied bands. There are 96 > occupied bands: using the davidson I can diagonalize ok for 500 bands and > 1000 bands, but if I push it up to 1500 I get problems with (3.2.3): > from cdiaghg : error # 2684 > info =/= 0 > or (4.0): > from pzpotrf : error # 94 > problems computing cholesky decomposition > > I fear the conjugate gradient will be very slow, so before I try that, any > suggestions for either approach with so many bands? > > I'm using the system compiled pw.x on BCX. > Cheers, > Conor > > &system > ecutwfc =12.0, > nbnd=1500, > / > &electrons > diago_thr_init=5.0e-6, > diago_full_acc=.true. > electron_maxstep = 70, > diagonalization='david' > mixing_mode = 'plain' > mixing_beta = 0.7 > conv_thr = 1.0d-8, > / > > > > ---- > > Dr. Conor Hogan --- ----- --- --- > Dipartimento di Fisica e CNR-INFM | : | : > Universita' di Roma "Tor Vergata" |-- : --- |--- > Tel: +39 06 7259 4121 *NEW* | : | : > Fax: +39 06 2023507 --- : --- : > http://www.fisica.uniroma2.it/~cmtheo-group/ European Theoretical > Spectroscopy Facility > http://www.etsf.eu > > Q. What do you call a fish with no eyes? > A. A fsh. (Tommy Cooper) > > The early bird gets the worm, but the second mouse gets the cheese. - S. > Wright > > If you go through a lot of hammers each month, I don't think it > necessarily > means you're a hard worker. It may just mean that you have a lot to learn > about proper hammer maintenance - J. Handey > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ---- Dr. Conor Hogan --- ----- --- --- Dipartimento di Fisica e CNR-INFM | : | : Universita' di Roma "Tor Vergata" |-- : --- |--- Tel: +39 06 7259 4121 *NEW* | : | : Fax: +39 06 2023507 --- : --- : http://www.fisica.uniroma2.it/~cmtheo-group/ European Theoretical Spectroscopy Facility http://www.etsf.eu Q. What do you call a fish with no eyes? A. A fsh. (Tommy Cooper) The early bird gets the worm, but the second mouse gets the cheese. - S. Wright If you go through a lot of hammers each month, I don't think it necessarily means you're a hard worker. It may just mean that you have a lot to learn about proper hammer maintenance - J. Handey From chenhanghuipwscf at gmail.com Tue Sep 30 17:43:44 2008 From: chenhanghuipwscf at gmail.com (alan chen) Date: Tue, 30 Sep 2008 11:43:44 -0400 Subject: [Pw_forum] relaxation in the presence of external electric field Message-ID: <22ae3ca40809300843q792a3089u3291e50ae87de22d@mail.gmail.com> Dear PWSCF users, I turn on the 'tefield' option to relax an isolated slab system in the presence of external electric field. The input file is attached below. I use 'bfgs' for the ionic dynamics. However I always got the following error: 'bfgs history already reset at previous step'. Why 'bfgs' method does not work for this 'sawlike' electric field? Shall I change some parameters? Thank you very much. Hanghui Chen Department of Physics, Yale University &CONTROL calculation='relax' wf_collect=.true. pseudo_dir = './psp' outdir='/home1/hc336/scratch' wfcdir='/home1/hc336/scratch' prefix='STO-efield-r-E0.01' tprnfor = .true. tstress = .true. disk_io='low' verbosity='high' dt=80.D0 tefield=.true. / &SYSTEM ibrav= 6 celldm(1) = 7.02 celldm(3) = 5.0 nat= 8 ntyp= 3 ecutwfc = 30.0 ecutrho = 180.0 occupations='smearing' smearing=gauss degauss=0.005 edir=3 emaxpos=0.8 eopreg=0.2 eamp=0.01 / &ELECTRONS diagonalization='david' mixing_beta = 0.7D0 diago_david_ndim = 4 startingwfc='atomic' startingpot='atomic' conv_thr = 1.0d-8 / &IONS ion_dynamics = 'bfgs' phase_space = 'full' pot_extrapolation = 'second_order' wfc_extrapolation = 'second_order' / &CELL cell_dynamics = 'damp-w' / ATOMIC_SPECIES Sr 87.62 038-Sr-ca-sp-vgrp.uspp.format.UPF Ti 47.90 022-Ti-ca-sp-vgrp.uspp.format.UPF O 16.00 008-O-ca--vgrp.uspp.format.UPF ATOMIC_POSITIONS alat Ti 0.50 0.50 0.50 O 0.00 0.50 0.50 O 0.50 0.00 0.50 Sr 0.00 0.00 1.00 O 0.50 0.50 1.00 Ti 0.50 0.50 1.50 O 0.00 0.50 1.50 O 0.50 0.00 1.50 K_POINTS {automatic} 10 10 1 0 0 0 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080930/cf47a22a/attachment.htm From giannozz at democritos.it Tue Sep 30 17:47:24 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 30 Sep 2008 17:47:24 +0200 Subject: [Pw_forum] relaxation in the presence of external electric field In-Reply-To: <22ae3ca40809300843q792a3089u3291e50ae87de22d@mail.gmail.com> References: <22ae3ca40809300843q792a3089u3291e50ae87de22d@mail.gmail.com> Message-ID: <48E24A0C.1030409@democritos.it> alan chen wrote: > I turn on the 'tefield' option to relax an isolated slab system > in the presence of external electric field. The input file is attached > below. I use 'bfgs' for the ionic dynamics. However I always got the > following error: 'bfgs history already reset at previous step'. code version? in the most recent Doc/BUGS file, you find this: Fixed in version 4.0.2: * Structural optimization with external sawtooth potential was not working correctly (electric field disappeared after first run). All versions after october 2005 affected. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From nakhmanson at anl.gov Tue Sep 30 20:18:35 2008 From: nakhmanson at anl.gov (Serge Nakhmanson) Date: Tue, 30 Sep 2008 13:18:35 -0500 Subject: [Pw_forum] Compiling with intel fortran, MKL and MKL FFT wrappers Message-ID: <48E26D7B.2050809@anl.gov> Lorenzo, Axel, Thank you for your comments. I expected the serial lib to be "it" but decided to check in with the forum just to be sure. I have not tested the new "MKL FTT" binary extensively yet but I will. As I understand, I should not expect a tremendous performance improvement over plain vanilla fftw2 lib, however, if I see something interesting I will drop a line here. What is given below is a comment to answer Vega's questions. It could be skipped unless one's interested in the nitty-gritty details and/or wants to find and expose flaws in my operating procedures. > could you tell me which dir should be added to the LD_LIBRARY_PATH, or which > flags should be added after the QE configure command? Vega, Here's what you can do, if, against Axel's advice, you just gotta have an MKL FFT binary in your {espresso}/bin. I usually do not use configure. Instead I edit make.sys directly, which seems to be much more convenient. If you take a look at the end of my original message you will see how I set all the important flags (including include and lib paths) to compile everything without problems. The only thing you have to do is put in your own paths to MKL instead of my paths. Regarding the FFT_LIBS flag, by default, my sysadmin placed all the compiled MKL FFT libs into {mkl}/{number}/lib/64 but you can place them anywhere as long as you provide a correct path to them in make.sys. Regarding the LD_LIBRARY_PATH, I usually set it inside my PBS job script, e.g., # # Set up some compiler-dependent things: # if [ $COMPILER = intel ]; then # Set up some things to run an Intel-compiler binary: MPIRUN=/usr/local/bin/mpiexec MPIFLAGS='-verbose -kill' export LD_LIBRARY_PATH=/usr/local/intel/mkl/10.0.4.023/lib/em64t/:$LD_LIBRARY_PATH elif [ $COMPILER = pgi64 ]; then {do something else here} else echo "Compiler $COMPILER unsupported for parallel execution" exit 1 fi You can then check if all the appropriate libs are found correctly by doing "ldd your_binary.x" (in the same PBS script, of course) before you run it: # Run the job in //: DATE=`date` echo "" echo "Time at the start of the job is: $DATE" echo "" cd $SCRATCH/$RUNDIR echo "" echo "ldd pw.x:" ldd $SCRATCH/$RUNDIR/pw.$ARCH.x echo "" echo "" echo "Running the job:" echo "$MPIRUN $MPIFLAGS $SCRATCH/$RUNDIR/pw.$ARCH.x $NPOOL < $SCRATCH/$RUNDIR/$NAME.$PREFIX.pw.inp > $SCRATCH/$RUNDIR/$NAME.$PREFIX.pw.out" echo "" $MPIRUN $MPIFLAGS $SCRATCH/$RUNDIR/pw.$ARCH.x $NPOOL < $SCRATCH/$RUNDIR/$NAME.$PREFIX.pw.inp > $SCRATCH/$RUNDIR/$NAME.$PREFIX.pw.out DATE=`date` echo "" echo "Time at the end of the job is: $DATE" echo "Doing some housekeeping duties..." echo "" Here's a sample output from this part of the script that I get on my cluster, which shows that at least the MKL libraries from BLAS_LIBS and LAPACK_LIBS are found correctly: > Time at the start of the job is: Mon Sep 29 17:37:18 CDT 2008 > > > ldd pw.x: > libmkl_intel_lp64.so => /usr/local/intel/mkl/10.0.4.023/lib/em64t/libmkl_intel_lp64.so (0x0000002a95557000) > libmkl_sequential.so => /usr/local/intel/mkl/10.0.4.023/lib/em64t/libmkl_sequential.so (0x0000002a9586d000) > libmkl_core.so => /usr/local/intel/mkl/10.0.4.023/lib/em64t/libmkl_core.so (0x0000002a95a1a000) > libpthread.so.0 => /lib64/tls/libpthread.so.0 (0x000000375f300000) > librt.so.1 => /lib64/tls/librt.so.1 (0x000000375e900000) > libm.so.6 => /lib64/tls/libm.so.6 (0x000000375e300000) > libc.so.6 => /lib64/tls/libc.so.6 (0x000000375de00000) > libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x000000375eb00000) > libdl.so.2 => /lib64/libdl.so.2 (0x000000375e100000) > /lib64/ld-linux-x86-64.so.2 (0x000000375dc00000) > > > Running the job: > /usr/local/bin/mpiexec -verbose -kill /scratch/nakhmanson/PWtest3/pw.test.x -npool 2 < /scratch/nakhmanson/PWtest3/C1c.relaxed.pw.inp > /scratch/nakhmanson/PWt > est3/C1c.relaxed.pw.out > > mpiexec: resolve_exe: using absolute path "/scratch/nakhmanson/PWtest3/pw.test.x". > mpiexec: process_start_event: evt 2 task 0 on node09.cluster. > mpiexec: read_p4_master_port: waiting for port from master. > mpiexec: read_p4_master_port: got port 33007. > mpiexec: process_start_event: evt 4 task 1 on node09.cluster. > mpiexec: All 2 tasks (spawn 0) started. > mpiexec: wait_tasks: waiting for node09.cluster node09.cluster. > mpiexec: process_obit_event: evt 3 task 0 on node09.cluster stat 0. > mpiexec: process_obit_event: evt 5 task 1 on node09.cluster stat 0. > > Time at the end of the job is: Mon Sep 29 17:41:03 CDT 2008 > Doing some housekeeping duties... Hope this helps, Serge -- ********************************************************* Serge M. Nakhmanson phone: (630) 252-5205 Assistant Scientist fax: (630) 252-4798 MSD-212, Rm. C-224 Argonne National Laboratory 9700 S. Cass Ave. Argonne, IL 60439 ********************************************************* From brad.malone at gmail.com Tue Sep 30 21:49:49 2008 From: brad.malone at gmail.com (Brad Malone) Date: Tue, 30 Sep 2008 12:49:49 -0700 Subject: [Pw_forum] Out of range subscript for igk_l2g_kdip Message-ID: Hi, I noticed some unexpected behavior when running pw.x on more than 1 pool. Specifically this happens in PW/pw_restart.f90 if a calculation is done with wf_collect set to true and if the number of pools is not equal to 1. I believe the offending lines are IF ( lwfc ) THEN ! CALL write_gk(iunout,ik,filename) CALL write_this_wfc(iunout,ik) ! END IF The error comes about because in the subroutine write_this_wfc the subroutine write_wfc is called, which has as one of its arguments igk_l2g_kdip(:, ik-iks+1). iks is the first index of the first k point in a particular pool. However, ik in this routine runs over ALL the kpoints, thus indexing negative values (e.g., ik goes from 1-28 while iks goes from 17-18). If debugging flags are not turned on the program doesn't crash and I haven't tested to see what difference in might make in the output. But turning debugging flags on catches this error and exits the program. Anyhow, I thought I'd bring this to your attention in case you weren't aware of it. Best, Brad Malone UC Berkeley -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080930/1726a749/attachment.htm From giannozz at democritos.it Tue Sep 30 22:15:29 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 30 Sep 2008 22:15:29 +0200 Subject: [Pw_forum] Problems with nscf diago with many bands In-Reply-To: <51706.141.108.248.231.1222783980.squirrel@141.108.248.231> References: <43053.141.108.248.231.1222768628.squirrel@141.108.248.231> <51706.141.108.248.231.1222783980.squirrel@141.108.248.231> Message-ID: On Sep 30, 2008, at 16:13 , Conor Hogan wrote: > Hello again, forget that last message, problem solved: > > < CG style diagonalization > < ethr = 5.00E-06, avg # of iterations = 34.7 > < PWSCF : 3m47.44s CPU time, 3m59.01s wall time > >> Davidson diagonalization with overlap >> ethr = 5.00E-06, avg # of iterations = 27.2 >> PWSCF : 7m41.49s CPU time, 7m47.46s wall time > > Wasn't expecting to see CG faster! PW continues to amaze me with its > brutal efficiency :) it is not CG that is faster, it is Davidson that is slower :) Iterative diagonalization is designed to be fast when the number M of desired solutions (i.e. KS state) is much smaller than the dimension of vectors (i.e. the number of PW). If this is not true, iterative diagonalization loses much of its appeal. Likely, the CPU usage in your iterative diagonalization is dominated by conventional subspace diagonalization, requiring O(M^3) operations (see CPU time spent in *diagh). Davidson performs more of them and with larger matrix size than CG. This does not explain the failures you get: my experience is that Davidson is usually better but more sensitive to numerical problems. Anyway: is CG is good for such cases, yet another reason to keep it alive. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From giannozz at democritos.it Tue Sep 30 22:24:39 2008 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 30 Sep 2008 22:24:39 +0200 Subject: [Pw_forum] Out of range subscript for igk_l2g_kdip In-Reply-To: References: Message-ID: On Sep 30, 2008, at 21:49 , Brad Malone wrote: > The error comes about because in the subroutine write_this_wfc the > subroutine write_wfc is called, which has as one of its arguments > igk_l2g_kdip(:, ik-iks+1). iks is the first index of the first k > point in a particular pool. However, ik in this routine runs over > ALL the kpoints, thus indexing negative values (e.g., ik goes from > 1-28 while iks goes from 17-18). If debugging flags are not turned > on the program doesn't crash and I haven't tested to see what > difference in might make in the output. But turning debugging flags > on catches this error and exits the program. it is a known problem, or more exactly, it is known not to be a problem in practise. The subscript goes out of bound only on those processors that do not use the information contained in the out-of-bound array. It is ugly but harmless. Thank you anyway: we'll think how to fix it. I know a simple solution that would simplify a lot the horrible mess of collecting the wavefunctions with pools of processors, but it works well only in presence of parallel file systems Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From paulatto at sissa.it Tue Sep 30 23:12:26 2008 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Tue, 30 Sep 2008 23:12:26 +0200 (CEST) Subject: [Pw_forum] Out of range subscript for igk_l2g_kdip In-Reply-To: References: Message-ID: <25345.82.50.174.28.1222809146.squirrel@webmail.sissa.it> On Mar, Settembre 30, 2008 22:24, Paolo Giannozzi wrote: > it is a known problem, or more exactly, it is known not to be a > problem in practise. To comment a bit further: if you are using run-time boundary checks they *may* find those lines offending and prevent you from spotting the real problem. I have a modified version of the file that prevent it. I haven't released it as it is a quick hack, even more messy than the current one. If you need it I'll send it to you. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ ---------------------------------------------------------------- SISSA Webmail https://webmail.sissa.it/ Powered by SquirrelMail http://www.squirrelmail.org/