From dmitry at korotin.name Thu Jan 1 15:00:20 2009 From: dmitry at korotin.name (Dmitry Korotin) Date: Thu, 1 Jan 2009 19:00:20 +0500 Subject: [Pw_forum] Coordinate system In-Reply-To: <200812300958.41002.giannozz@democritos.it> References: <166cd7c60812292212l7cd07e09o11f5a295b247de10@mail.gmail.com> <200812300958.41002.giannozz@democritos.it> Message-ID: <166cd7c60901010600m6b486818tc2c76a46c95cc49a@mail.gmail.com> >> Does QE use left-hand triple of coordinate vectors or right-hand one? > > it uses the "traditional" one, i.e. right-hand side if I remember correctly. Thank you for the answer. > Not that it makes any difference, except in some very specific cases. Sure. But I have that "specific case" - I am trying to play with a local coordinate system - so I was need to know. Best regards, Dmitry Korotin Ph. D. Student, Institute of Metal Physics S. Kovalevskaya, 18 620041 Ekaterinburg GSP-170 Russia From bipulrr at gmail.com Thu Jan 1 16:22:03 2009 From: bipulrr at gmail.com (Bipul Rakshit) Date: Thu, 1 Jan 2009 20:52:03 +0530 Subject: [Pw_forum] small energy difference in B1 and B2 structure Message-ID: <3a749910901010722w19cbe5bevafe6ae6bdf41f25d@mail.gmail.com> Dear PWscf user, First of all Happy New year to all of you. I am doing electronic calculation of ScSb in B1 and B2 phase. From the calculation, I am getting the following results. ScSb(B1) a0=11.019 Bohr, E0= -290.2578 Ry ScSb(B2) a0=6.834 Bohr, E0= -290.20279 Ry As you can see the energy difference is very small for large volume compression. Why this happen? The input file for B1 and B2 is attached. Kindly check it and suggest me what mistake I am doing. with regards. -- 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/20090101/ce0ff0c8/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: scsb.relax.in Type: application/octet-stream Size: 2662 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20090101/ce0ff0c8/attachment.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: scsb-B2.relax.in Type: application/octet-stream Size: 2723 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20090101/ce0ff0c8/attachment-0001.obj From baroni at sissa.it Thu Jan 1 17:49:10 2009 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 1 Jan 2009 17:49:10 +0100 Subject: [Pw_forum] small energy difference in B1 and B2 structure In-Reply-To: <3a749910901010722w19cbe5bevafe6ae6bdf41f25d@mail.gmail.com> References: <3a749910901010722w19cbe5bevafe6ae6bdf41f25d@mail.gmail.com> Message-ID: <81B9BD10-F8BD-4C2A-8781-EB6A5B0C516D@sissa.it> On Jan 1, 2009, at 4:22 PM, Bipul Rakshit wrote: > Dear PWscf user, Dear Bibul Rakshit: > First of all Happy New year to all of you. all the same to you! > I am doing electronic calculation of ScSb in B1 and B2 phase. From > the calculation, I am getting the following results. > > ScSb(B1) a0=11.019 Bohr, E0= -290.2578 Ry > > ScSb(B2) a0=6.834 Bohr, E0= -290.20279 Ry > > As you can see the energy difference is very small for large volume > compression. I do not quite understand the question: why do you say that there is a "large volume compression"? There is none: 11.019 au (volume=11.019^3/4=334.5) is rather close to the experimental lattice parameter that I have been able to find in the literature (11.06), while the volume that you report for the B2 stucture (6.834^3=319.2) is certainly smaller, but certainly not "very compressed". Isn't it by any chance because of the _very_ different lattice parameter? If so, have you doble checked in you preferred solid state textbook the formules that give the unit-cell volumes for the B1 (FCC) and B2 (SC) structures? > Why this happen? The input file for B1 and B2 is attached. I was not able to detect anything manifestly wrong in your input files, assuming that you have carefully checked the convergence of your results wrt the number of k points, as needed particularly in view of the rather small value for the Gaussian broadening that you are using. take care 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/20090101/3159afd4/attachment.htm From eyvaz_isaev at yahoo.com Thu Jan 1 19:00:50 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 1 Jan 2009 10:00:50 -0800 (PST) Subject: [Pw_forum] small energy difference in B1 and B2 structure In-Reply-To: <3a749910901010722w19cbe5bevafe6ae6bdf41f25d@mail.gmail.com> Message-ID: <877228.31721.qm@web65713.mail.ac4.yahoo.com> Hi, Just to add to Stefano's reply: In fact, the difference is quite large. Let us to calculate: (290.2578-290.20279)*13.6055*11600 = 8682K!!! 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 Thu, 1/1/09, Bipul Rakshit wrote: > From: Bipul Rakshit > Subject: [Pw_forum] small energy difference in B1 and B2 structure > To: "PWSCF Forum" > Date: Thursday, January 1, 2009, 6:22 PM > Dear PWscf user, > First of all Happy New year to all of you. > > I am doing electronic calculation of ScSb in B1 and B2 > phase. From the > calculation, I am getting the following results. > > ScSb(B1) a0=11.019 Bohr, E0= -290.2578 Ry > > ScSb(B2) a0=6.834 Bohr, E0= -290.20279 Ry > > As you can see the energy difference is very small for > large volume > compression. Why this happen? The input file for B1 and B2 > is attached. > Kindly check it and suggest me what mistake I am doing. > with regards. > > -- > 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 giannozz at democritos.it Thu Jan 1 22:54:31 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 1 Jan 2009 22:54:31 +0100 Subject: [Pw_forum] epsilon In-Reply-To: <83020.21942.qm@web33104.mail.mud.yahoo.com> References: <83020.21942.qm@web33104.mail.mud.yahoo.com> Message-ID: <200901012254.31365.giannozz@democritos.it> On Wednesday 31 December 2008 21:51, ali kazempoor wrote: > When I run epsilon.x for silicon bulk after doing scf this error happen: > ban number of bands "ban band number" (the exact form is important: it allows to locate where the error happens with a command like "grep error-message *f90"). It means that you have less bands than the number of electrons divided by 2 P. -- Paolo Giannozzi, Democritos and Udine University From j.vansijl at gmail.com Fri Jan 2 09:48:25 2009 From: j.vansijl at gmail.com (Jelle van Sijl) Date: Fri, 02 Jan 2009 09:48:25 +0100 Subject: [Pw_forum] PAW parallel error Message-ID: <1230886105.24427.21.camel@krafla> Dear All, first of, happy 2009! I have compiled QE-4.0.4 (parallel) with intel fortran (v.11) and the appropriate blas and lapack libraries (tried it with and without MKL). Jobs that use pseudopotentials from the online database run smoothly. I am now testing a PAW pseudo potential made with ld1. The calculation proceeds perfectly on QE-4.0.4 compiled in serial on my desktop, but dies in parallel right at the start: Program PWSCF v.4.0.4 starts ... Today is 2Jan2009 at 8:31:38 Parallel version (MPI) Number of processors in use: 4 R & G space division: proc/pool = 4 For Norm-Conserving or Ultrasoft (Vanderbilt) Pseudopotentials or PAW Current dimensions of program pwscf are: Max number of different atomic species (ntypx) = 10 Max number of k-points (npk) = 40000 Max angular momentum in pseudopotentials (lmaxx) = 3 rank 0 in job 1 caused collective abort of all ranks And the following error message appears: ######################################################################################################################## # FROM IOTK LIBRARY, VERSION 1.1.0development # UNRECOVERABLE ERROR (ierr=1) # ERROR IN: iotk_getline (iotk_scan.spp:812) # CVS Revision: 1.7 # iostat=-1 # ERROR IN: iotk_scan_tag (iotk_scan.spp:581) # CVS Revision: 1.7 # ERROR IN: iotk_scan (iotk_scan.spp:742) # CVS Revision: 1.7 # ERROR IN: iotk_scan_begin (iotk_scan.spp:94) # CVS Revision: 1.7 ######################################################################################################################## application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0[cli_0]: aborting job: application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0 One of the issues may be that environmental variables are not passed to all slave jobes on all nodes (from QE-wiki, comment by Axel I think). However, even using a single node and 1 processor results in this error. Another possible solution involving modifying a string in xml_io_base.f90 (in ./Modules) did not work because that particular change has already been included in this release. I am at a bit at a loss how to resolve it, hope someone here can help me. cheers, Jelle van Sijl -- drs. Jelle van Sijl PhD student, Department of Petrology Faculty of Earth and Life Sciences (FALW) VU University Amsterdam, De Boelelaan 1085 1081 HV Amsterdam, The Netherlands Phone +31 (0)20 5987403 http://www.falw.vu.nl/petrology From paulatto at sissa.it Fri Jan 2 10:03:54 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Fri, 02 Jan 2009 10:03:54 +0100 Subject: [Pw_forum] PAW parallel error In-Reply-To: <1230886105.24427.21.camel@krafla> References: <1230886105.24427.21.camel@krafla> Message-ID: On Fri, 02 Jan 2009 09:48:25 +0100, Jelle van Sijl wrote: > Dear Jelle, > first of, happy 2009! Happy new year to you as well, if the problem arises even after cleaning the temporary files belonging to the job (i.e. $outdir/$prefix.*), than it may be you are using a old PAW pseudopotential generated with a pre-release version of ld1.x Please try to clean the job files and let us now if the problem is still there, in this case you can try and regenerate the pseudopotentials with the most recent version of ld1 and/or send them to me so I can check if they are ok. cheers -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From j.vansijl at gmail.com Fri Jan 2 11:29:14 2009 From: j.vansijl at gmail.com (Jelle van Sijl) Date: Fri, 02 Jan 2009 11:29:14 +0100 Subject: [Pw_forum] PAW parallel error In-Reply-To: References: <1230886105.24427.21.camel@krafla> Message-ID: <1230892154.24427.26.camel@krafla> verdict: I did not pay attention. One cannot use a PAW pseudo made with ld1.x from a previous QE release in QE-4.0.4 parallel. Thanks! -Jelle On Fri, 2009-01-02 at 10:03 +0100, Lorenzo Paulatto wrote: > On Fri, 02 Jan 2009 09:48:25 +0100, Jelle van Sijl wrote: > > Dear Jelle, > > first of, happy 2009! > > Happy new year to you as well, > if the problem arises even after cleaning the temporary files belonging to the job (i.e. $outdir/$prefix.*), than it may be you are using a old PAW pseudopotential generated with a pre-release version of ld1.x > > Please try to clean the job files and let us now if the problem is still there, in this case you can try and regenerate the pseudopotentials with the most recent version of ld1 and/or send them to me so I can check if they are ok. > > cheers > -- drs. Jelle van Sijl PhD student, Department of Petrology Faculty of Earth and Life Sciences (FALW) VU University Amsterdam, De Boelelaan 1085 1081 HV Amsterdam, The Netherlands Phone +31 (0)20 5987403 http://www.falw.vu.nl/petrology From bipulrr at gmail.com Fri Jan 2 14:52:10 2009 From: bipulrr at gmail.com (Bipul Rakshit) Date: Fri, 2 Jan 2009 19:22:10 +0530 Subject: [Pw_forum] question about vc-relax Message-ID: <3a749910901020552v3a3cc06bua95b46a7bb5a07a1@mail.gmail.com> Dear Pwscf users, I have done the vc-relax calculation of ScSb. At the end of the file, I am getting relaxed cell volume, CELL_PARAMETER AND relaxed atomic position and pressure P= -0.32 kbar If once again I put all the above parameters in the input file and do the vc-relax calculation, then in first scf step I must get a pressure of -0.32 kbar. But this is not the case. The pressure changes to some other value (-10.38 kbar). So my question is why this happens? Why the pressure again increase or decrease? regards -- 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/20090102/6ca9ba16/attachment.htm From paulatto at sissa.it Fri Jan 2 16:17:22 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Fri, 02 Jan 2009 16:17:22 +0100 Subject: [Pw_forum] question about vc-relax In-Reply-To: <3a749910901020552v3a3cc06bua95b46a7bb5a07a1@mail.gmail.com> References: <3a749910901020552v3a3cc06bua95b46a7bb5a07a1@mail.gmail.com> Message-ID: On Fri, 02 Jan 2009 14:52:10 +0100, Bipul Rakshit wrote: > getting relaxed cell volume, CELL_PARAMETER AND relaxed atomic position > ... > But this is not the case. The pressure changes to some other value > (-10.38 kbar). Dear Bipul, if you had not used ibrav=0 in the initial input file, reconstructiong input cell parameters from the output can be a bit tricky, please double check it. Otherwise provide more detail so we can take a more educated guess. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From bipulrr at gmail.com Fri Jan 2 16:34:16 2009 From: bipulrr at gmail.com (Bipul Rakshit) Date: Fri, 2 Jan 2009 21:04:16 +0530 Subject: [Pw_forum] question about vc-relax In-Reply-To: References: <3a749910901020552v3a3cc06bua95b46a7bb5a07a1@mail.gmail.com> Message-ID: <3a749910901020734v2a04e7d9r64f737b9a7b24f@mail.gmail.com> Dear Paulatto scsb.vc-relax.in is my initial input file and scsb.vc-relax1.in my next input file, in which i use ibrav=0 From zhaohscas at yahoo.com.cn Fri Jan 2 17:38:34 2009 From: zhaohscas at yahoo.com.cn (Hongsheng Zhao) Date: Sat, 3 Jan 2009 00:38:34 +0800 Subject: [Pw_forum] Want to calculate optical properties within pwscf-dfpt. Message-ID: <19510558956.20090103003834@yahoo.com.cn> Dear PWSCF users, I want to konw is it possible to use PWSCF to do some DFPT calculations to obtain the optical properties, such as dielectric functions, electro-optic coefficient, and nonlinear optical coefficient of a system presently? Regards, -- Hongsheng Zhao Xinjiang Technical Institute of Physics and Chemistry Chinese Academy of Sciences GnuPG DSA: 0xD108493 2009-1-3 From giannozz at democritos.it Fri Jan 2 18:34:52 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 2 Jan 2009 18:34:52 +0100 Subject: [Pw_forum] question about vc-relax In-Reply-To: <3a749910901020552v3a3cc06bua95b46a7bb5a07a1@mail.gmail.com> References: <3a749910901020552v3a3cc06bua95b46a7bb5a07a1@mail.gmail.com> Message-ID: <200901021834.52584.giannozz@democritos.it> On Friday 02 January 2009 14:52, Bipul Rakshit wrote: > If once again I put all the above parameters in the input file and do the > vc-relax calculation, then in first scf step I must get a pressure of -0.32 > kbar. But this is not the case. The pressure changes to some other value > (-10.38 kbar). > > So my question is why this happens? most likely, because you did not perform exactly the same calculation in the two cases. Something is different between the two cases: atomic positions, unit cell, k-points, kinetic energy functional, etc.etc. Paolo -- Paolo Giannozzi, Democritos and Udine University From quantumdft at gmail.com Fri Jan 2 18:58:00 2009 From: quantumdft at gmail.com (vega lew) Date: Sat, 3 Jan 2009 01:58:00 +0800 Subject: [Pw_forum] How to understand the intermolecular force between water molecules calculated by PWSCF Message-ID: <412f6c680901020958g2acc1749td3b6469a69a00d20@mail.gmail.com> Dear all, if we calculated the a couple of water molecles in a large enough boxes, we will find the H of water must point to O of another water. We often call this kind of attraction as hydrogen bond. But I want to know except for hydrogen bond, is there van der waals force or other forces included in the results given by PWSCF? thanks 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 ****************************************************************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, Jiangsu, China ****************************************************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090103/2f05aaa3/attachment.htm From akohlmey at cmm.chem.upenn.edu Fri Jan 2 20:04:26 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 2 Jan 2009 14:04:26 -0500 (EST) Subject: [Pw_forum] How to understand the intermolecular force between water molecules calculated by PWSCF In-Reply-To: <412f6c680901020958g2acc1749td3b6469a69a00d20@mail.gmail.com> References: <412f6c680901020958g2acc1749td3b6469a69a00d20@mail.gmail.com> Message-ID: On Sat, 3 Jan 2009, vega lew wrote: VL> Dear all, VL> VL> if we calculated the a couple of water molecles in a large enough boxes, we VL> will find the H of water must point to O of another water. VL> We often call this kind of attraction as hydrogen bond. VL> But I want to know except for hydrogen bond, is there van der waals force or VL> other forces included in the results given by PWSCF? yes and no. PWSCF computes interactions/forces between atoms based on quantum mechanics (using DFT), and thus those forces are the result of _all_ atoms in a non-additive way, i.e. it contains in principle everything. decomposing this into individual (pairwise) contributions (coulomb, non-coulomb, bonded, non-bonded, ...) is entirely empirical (or phenomenological) and the results depend on your definition of the individual contributions. this is what is done in the parametrization of classical force fields and thus explains why those parameters are not very transfereable. hydrogen bonds are a phenomenoligical concept added on top of that and somewhat ill defined at that (there are many different definitions of when there is a hydrogen bond between atoms or not). the most common definition of a hydrogen bond actually involves three atoms, the hydrogen bond donor, the acceptor and the hydrogen that is located in between, and to call the resulting "structural feature" a hydrogen bond the donor and acceptor have to be close enough and the hydrogen must somewhat close to the line connecting those two atoms. in classical force fields hydrogen bonds are favored by a combination of somewhat strong coulombic attraction balanced with vdW repulsion. now, it is similarly difficult to determine the contribution assigned to vdW interactions (except for systems that have bascially no other interactions, e.g. Argon) and on top of that you have the problem that current LDA/GGA functionals underrepresent certain exchange and correlation interactions which corresponds to what described as vdW interactions. you'd have to do prohibitively expensive quantum chemistry calculations (using e.g. CCSD(T) with very large basis sets) to compute those accurately. cheers, axel. VL> VL> thanks for reading VL> VL> vega VL> 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 baroni at sissa.it Fri Jan 2 21:32:16 2009 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 2 Jan 2009 21:32:16 +0100 Subject: [Pw_forum] Want to calculate optical properties within pwscf-dfpt. In-Reply-To: <19510558956.20090103003834@yahoo.com.cn> References: <19510558956.20090103003834@yahoo.com.cn> Message-ID: <711E9A31-43D6-4E1E-BD94-A6C33A12705B@sissa.it> Time dependent DFPT is implemented in a not-yet-public version of QE, following the method explained in JCP 128, 154105 (2008). It will be released in the next few months - stay tuned SB On Jan 2, 2009, at 5:38 PM, Hongsheng Zhao wrote: > Dear PWSCF users, > > I want to konw is it possible to use PWSCF to do some DFPT > calculations to obtain the optical properties, such as dielectric > functions, electro-optic coefficient, and nonlinear optical > coefficient of a system presently? > > Regards, > -- > Hongsheng Zhao > Xinjiang Technical Institute of Physics and Chemistry > Chinese Academy of Sciences > GnuPG DSA: 0xD108493 > 2009-1-3 > > _______________________________________________ > 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/20090102/409f30c3/attachment.htm From carlos.loia.reis at gmail.com Sat Jan 3 17:55:19 2009 From: carlos.loia.reis at gmail.com (Carlos Reis) Date: Sat, 3 Jan 2009 16:55:19 +0000 Subject: [Pw_forum] Advice on Electron-Phonon Calculations Message-ID: <1382facf0901030855j53a2af45raf599c43d51ce2f@mail.gmail.com> Dear pwscf users and developers, I am using pwscf successfully. Both the serial and mpi versions work very well. My setup is a gigabit cluster of quad-core pc's running Linux. I am trying to setup an electron-phonon calculation for a large system with many k-points, and I managed do take advantage of the k-point parallelization implemented in pw.x for the nscf ground state calculation. The calculation of the phonons in parallel also worked reasonably well although I still don't know how to make ph.x to write the large *.dwf and *.bar files in the /tmp of each node (or even if that is possible). The main problem however is that when I activate the electron-phonon calculation (elph=.true.) I get the message: "pools and a2F not implemented" Is there some other way to do this electron-phonon calculation taking advantage of parallelization over the k-points?. Is there a better strategy for distributing the electron-phonon calculation among the nodes? Any advice on this is welcome. I would also like to congratulate the pwscf developers for their excellent work. Best regards, Carlos. -- Carlos Loia Reis Centro de F?sica Te?rica e Computacional Complexo Interdisciplinar da Universidade de Lisboa Av. Prof. Gama Pinto, 2 P-1649-003 Lisboa, Portugal. From w2agz at pacbell.net Sun Jan 4 00:38:16 2009 From: w2agz at pacbell.net (Paul M. Grant) Date: Sat, 3 Jan 2009 15:38:16 -0800 Subject: [Pw_forum] K_POINTS, tpiba or crystal In-Reply-To: <200812301420.47223.giannozz@democritos.it> References: <000501c969e4$eb03bb00$c10b3100$@net> <20081229212836.hut32tqv40csg404@mail.democritos.it> <005901c96a3b$0c9de960$25d9bc20$@net> <200812301420.47223.giannozz@democritos.it> Message-ID: <01a001c96dfc$5a97be10$0fc73a30$@net> Followup on K_POINTS: tpiba vs crystal... As usual, Paolo is right...for sc lattices (ibrav=1), tpiba or crystal give equal nscf outputs. But, for fcc lattices (ibrav=2), it apparently does! See below comparing nscf k-vector outputs for tpiba and crystal, and crystal is very different (and weird...for me anyway). BTW, whether tpiba, etc., should be enclosed in curly braces or parentheses or nothing on the K_POINTS card is confusing. PW_INPUT shows braces, but parenthese work too, and probably no delimiters as well. My agenda here is to generate k-points grids amenable to applications that cannot use simply MP weighted points in the irreducible BZ, such as epsilon.x. I'm finding the algorithm used by Eyvaz Isaev for Fermi Surface plotting which uniformly populate and span the fundamental reciprocal lattice cell works quite well...if inputted with tpiba or "blank," not crystal. TPIBA VS CRYSTAL COMPARISON RE FCC FOLLOWS =================================================== TPIBA nsfc run input (partial) (Al, ibrav=2) K_POINTS {tpiba} 2197 0.00000000 0.00000000 0.00000000 1.00000000 -0.08333333 0.08333333 -0.08333333 1.00000000 -0.16666667 0.16666667 -0.16666667 1.00000000 -0.25000000 0.25000000 -0.25000000 1.00000000 -0.33333333 0.33333333 -0.33333333 1.00000000 -0.41666667 0.41666667 -0.41666667 1.00000000 -0.50000000 0.50000000 -0.50000000 1.00000000 -0.58333333 0.58333333 -0.58333333 1.00000000 -0.66666667 0.66666667 -0.66666667 1.00000000 -0.75000000 0.75000000 -0.75000000 1.00000000 -0.83333333 0.83333333 -0.83333333 1.00000000 -0.91666667 0.91666667 -0.91666667 1.00000000 -1.00000000 1.00000000 -1.00000000 1.00000000 0.08333333 0.08333333 0.08333333 1.00000000 0.00000000 0.16666667 0.00000000 1.00000000 -0.08333333 0.25000000 -0.08333333 1.00000000 -0.16666667 0.33333333 -0.16666667 1.00000000 -0.25000000 0.41666667 -0.25000000 1.00000000 -0.33333333 0.50000000 -0.33333333 1.00000000 .................... nsfc run output (partial) (Al, ibrav=2) reciprocal axes: (cart. coord. in units 2 pi/a_0) b(1) = ( -1.000000 -1.000000 1.000000 ) b(2) = ( 1.000000 1.000000 1.000000 ) b(3) = ( -1.000000 1.000000 -1.000000 ) number of k points= 2197 gaussian broad. (Ry)= 0.0150 ngauss = 0 cart. coord. in units 2pi/a_0 k( 1) = ( 0.0000000 0.0000000 0.0000000), wk = 0.0009103 k( 2) = ( -0.0833333 0.0833333 -0.0833333), wk = 0.0009103 k( 3) = ( -0.1666667 0.1666667 -0.1666667), wk = 0.0009103 k( 4) = ( -0.2500000 0.2500000 -0.2500000), wk = 0.0009103 k( 5) = ( -0.3333333 0.3333333 -0.3333333), wk = 0.0009103 k( 6) = ( -0.4166667 0.4166667 -0.4166667), wk = 0.0009103 k( 7) = ( -0.5000000 0.5000000 -0.5000000), wk = 0.0009103 k( 8) = ( -0.5833333 0.5833333 -0.5833333), wk = 0.0009103 k( 9) = ( -0.6666667 0.6666667 -0.6666667), wk = 0.0009103 k( 10) = ( -0.7500000 0.7500000 -0.7500000), wk = 0.0009103 k( 11) = ( -0.8333333 0.8333333 -0.8333333), wk = 0.0009103 k( 12) = ( -0.9166667 0.9166667 -0.9166667), wk = 0.0009103 k( 13) = ( -1.0000000 1.0000000 -1.0000000), wk = 0.0009103 k( 14) = ( 0.0833333 0.0833333 0.0833333), wk = 0.0009103 k( 15) = ( 0.0000000 0.1666667 0.0000000), wk = 0.0009103 k( 16) = ( -0.0833333 0.2500000 -0.0833333), wk = 0.0009103 ============================================================================ CRYSTAL nsfc run input (partial) (Al, ibrav=2) K_POINTS {crystal} 2197 0.00000000 0.00000000 0.00000000 1.00000000 -0.08333333 0.08333333 -0.08333333 1.00000000 -0.16666667 0.16666667 -0.16666667 1.00000000 -0.25000000 0.25000000 -0.25000000 1.00000000 -0.33333333 0.33333333 -0.33333333 1.00000000 -0.41666667 0.41666667 -0.41666667 1.00000000 -0.50000000 0.50000000 -0.50000000 1.00000000 -0.58333333 0.58333333 -0.58333333 1.00000000 -0.66666667 0.66666667 -0.66666667 1.00000000 -0.75000000 0.75000000 -0.75000000 1.00000000 -0.83333333 0.83333333 -0.83333333 1.00000000 -0.91666667 0.91666667 -0.91666667 1.00000000 -1.00000000 1.00000000 -1.00000000 1.00000000 0.08333333 0.08333333 0.08333333 1.00000000 0.00000000 0.16666667 0.00000000 1.00000000 -0.08333333 0.25000000 -0.08333333 1.00000000 -0.16666667 0.33333333 -0.16666667 1.00000000 -0.25000000 0.41666667 -0.25000000 1.00000000 -0.33333333 0.50000000 -0.33333333 1.00000000 -0.41666667 0.58333333 -0.41666667 1.00000000 ................... nsfc run output (partial) (Al, ibrav=2) reciprocal axes: (cart. coord. in units 2 pi/a_0) b(1) = ( -1.000000 -1.000000 1.000000 ) b(2) = ( 1.000000 1.000000 1.000000 ) b(3) = ( -1.000000 1.000000 -1.000000 ) number of k points= 2197 gaussian broad. (Ry)= 0.0150 ngauss = cart. coord. in units 2pi/a_0 k( 1) = ( 0.0000000 0.0000000 0.0000000), wk = 0.0009103 k( 2) = ( 0.2500000 0.0833333 0.0833333), wk = 0.0009103 k( 3) = ( 0.5000000 0.1666667 0.1666667), wk = 0.0009103 k( 4) = ( 0.7500000 0.2500000 0.2500000), wk = 0.0009103 k( 5) = ( 1.0000000 0.3333333 0.3333333), wk = 0.0009103 k( 6) = ( 1.2500000 0.4166667 0.4166667), wk = 0.0009103 k( 7) = ( 1.5000000 0.5000000 0.5000000), wk = 0.0009103 k( 8) = ( 1.7500000 0.5833333 0.5833333), wk = 0.0009103 k( 9) = ( 2.0000000 0.6666667 0.6666667), wk = 0.0009103 k( 10) = ( 2.2500000 0.7500000 0.7500000), wk = 0.0009103 k( 11) = ( 2.5000000 0.8333333 0.8333333), wk = 0.0009103 k( 12) = ( 2.7500000 0.9166667 0.9166667), wk = 0.0009103 k( 13) = ( 3.0000000 1.0000000 1.0000000), wk = 0.0009103 k( 14) = ( -0.0833333 0.0833333 0.0833333), wk = 0.0009103 k( 15) = ( 0.1666667 0.1666667 0.1666667), wk = 0.0009103 k( 16) = ( 0.4166667 0.2500000 0.2500000), wk = 0.0009103 k( 17) = ( 0.6666667 0.3333333 0.3333333), wk = 0.0009103 k( 18) = ( 0.9166667 0.4166667 0.4166667), wk = 0.0009103 k( 19) = ( 1.1666667 0.5000000 0.5000000), wk = 0.0009103 k( 20) = ( 1.4166667 0.5833333 0.5833333), wk = 0.0009103 k( 21) = ( 1.6666667 0.6666667 0.6666667), wk = 0.0009103 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 ? ? -----Original Message----- From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Paolo Giannozzi Sent: Tuesday, December 30, 2008 5:21 AM To: PWSCF Forum Subject: Re: [Pw_forum] K_POINTS, tpiba or crystal On Tuesday 30 December 2008 05:56, Paul M. Grant wrote: > b(1) = ( 1.000000 0.000000 0.000000 ) > b(2) = ( 0.000000 1.000000 0.000000 ) > b(3) = ( 0.000000 0.000000 1.000000 ) these are the lattice vectors generating a simple cubic lattice > Now, I want to input this grid to an nscf run via K_POINTS...is my grid > "tpiba" or "crystal?" in this particular case, it shouldn't matter, since the crystal and cartesian axis are the same, and the length of crystal axis is 2pi/a . Paolo -- Paolo Giannozzi, Democritos and Udine University _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From jameslipd at gmail.com Sun Jan 4 04:20:21 2009 From: jameslipd at gmail.com (jameslipd at gmail.com) Date: Sun, 04 Jan 2009 03:20:21 +0000 Subject: [Pw_forum] Problem with average.x Message-ID: <00163646ccde7ce575045f9fa70c@google.com> Dear users I want to calculate the planar and macroscopic averages of (100) surface. But I met a problem with average.x. When i type "average.x Dear users I want to calculate the planar and macroscopic averages of (100) surface. But I met a problem with average.x. When i type "average.x References: <000501c969e4$eb03bb00$c10b3100$@net> <20081229212836.hut32tqv40csg404@mail.democritos.it> <005901c96a3b$0c9de960$25d9bc20$@net> <200812301420.47223.giannozz@democritos.it> <01a001c96dfc$5a97be10$0fc73a30$@net> Message-ID: <94EC2CAB-BF97-491C-9B1A-735DF7C22E84@sissa.it> On Jan 4, 2009, at 12:38 AM, Paul M. Grant wrote: > Followup on K_POINTS: tpiba vs crystal... Hi Paul! > As usual, Paolo is right... he always does! > for sc lattices (ibrav=1), tpiba or crystal give > equal nscf outputs. no surprise, here > But, for fcc lattices (ibrav=2), it apparently does! you mean it does NOT .?... no surprise here either ... > See below comparing > nscf k-vector outputs for tpiba and crystal, and crystal is very > different > (and weird...for me anyway). I am not sure I understand what the question really is, but I do not what would weird here. My understanding is that "tpiba" means "cartesian coordinates in units of tipiba". By the way, it could be a good idea to replace "tpiba" (a word that does exist in the webster nor in any other dictionary) with "cartesian": this could help avoid some confusion. "crystal" means that the coordinates that you give are the coefficient that express a k-point as a linear combination of the reciprocal lattice basis g(1), g(2), g(3). In formulas: k("tpiba") = tpiba * (k(1)*x(1)+k(2)*x(2)+k(3)*x(3)) k("crystal") = tpiba* (k(1)*b(1)+k(2)*b(2)+k(3)*b(3)), where x(1)= (1,0,0), x(2)=(0,1,0), x(3)=(0,0,1), and the slightly confusing tpiba in the second line is due to the code convention: g(*)=tpiba*b(*) Let us take for instance the second vector of Paul's list. In the "tpiba" case, the output is just a copy of the input, as it should. In the crystal case, one should have: output(*) = input(1)*b(1) + input(2)*b(2)+input(3)*b(3) = -1/12 (-1,-1,1) +1/12 (1,1,1)-1/12(-1,1,-1) = 1/12 (1+1+1,1+1-1,-1+1+1) = (1/4,1/12,1/12), which is what it is. I have noot checked the other points, but I assume that they are what they should be according to the above argument > BTW, whether tpiba, etc., should be enclosed > in curly braces or parentheses or nothing on the K_POINTS card is > confusing. > PW_INPUT shows braces, but parenthese work too, and probably no > delimiters > as well. sorry, no hints here ... > My agenda here is to generate k-points grids amenable to > applications that > cannot use simply MP weighted points in the irreducible BZ, such as > epsilon.x. I'm finding the algorithm used by Eyvaz Isaev for Fermi > Surface > plotting which uniformly populate and span the fundamental reciprocal > lattice cell works quite well...if inputted with tpiba or "blank," not > crystal. most probably, it give the cartesian representation of the k-points ... hope the above is not (too) wrong and helpful to some extent ... Happy new year to Paul, Paolo, and the whole newsgroup. Stefano > > > > TPIBA VS CRYSTAL COMPARISON RE FCC FOLLOWS > =================================================== > TPIBA > nsfc run input (partial) (Al, ibrav=2) > K_POINTS {tpiba} > 2197 > 0.00000000 0.00000000 0.00000000 1.00000000 > -0.08333333 0.08333333 -0.08333333 1.00000000 > -0.16666667 0.16666667 -0.16666667 1.00000000 > -0.25000000 0.25000000 -0.25000000 1.00000000 > -0.33333333 0.33333333 -0.33333333 1.00000000 > -0.41666667 0.41666667 -0.41666667 1.00000000 > -0.50000000 0.50000000 -0.50000000 1.00000000 > -0.58333333 0.58333333 -0.58333333 1.00000000 > -0.66666667 0.66666667 -0.66666667 1.00000000 > -0.75000000 0.75000000 -0.75000000 1.00000000 > -0.83333333 0.83333333 -0.83333333 1.00000000 > -0.91666667 0.91666667 -0.91666667 1.00000000 > -1.00000000 1.00000000 -1.00000000 1.00000000 > 0.08333333 0.08333333 0.08333333 1.00000000 > 0.00000000 0.16666667 0.00000000 1.00000000 > -0.08333333 0.25000000 -0.08333333 1.00000000 > -0.16666667 0.33333333 -0.16666667 1.00000000 > -0.25000000 0.41666667 -0.25000000 1.00000000 > -0.33333333 0.50000000 -0.33333333 1.00000000 > .................... > nsfc run output (partial) (Al, ibrav=2) > reciprocal axes: (cart. coord. in units 2 pi/a_0) > b(1) = ( -1.000000 -1.000000 1.000000 ) > b(2) = ( 1.000000 1.000000 1.000000 ) > b(3) = ( -1.000000 1.000000 -1.000000 ) > number of k points= 2197 gaussian broad. (Ry)= 0.0150 ngauss > = 0 > cart. coord. in units 2pi/a_0 > k( 1) = ( 0.0000000 0.0000000 0.0000000), wk = 0.0009103 > k( 2) = ( -0.0833333 0.0833333 -0.0833333), wk = 0.0009103 > k( 3) = ( -0.1666667 0.1666667 -0.1666667), wk = 0.0009103 > k( 4) = ( -0.2500000 0.2500000 -0.2500000), wk = 0.0009103 > k( 5) = ( -0.3333333 0.3333333 -0.3333333), wk = 0.0009103 > k( 6) = ( -0.4166667 0.4166667 -0.4166667), wk = 0.0009103 > k( 7) = ( -0.5000000 0.5000000 -0.5000000), wk = 0.0009103 > k( 8) = ( -0.5833333 0.5833333 -0.5833333), wk = 0.0009103 > k( 9) = ( -0.6666667 0.6666667 -0.6666667), wk = 0.0009103 > k( 10) = ( -0.7500000 0.7500000 -0.7500000), wk = 0.0009103 > k( 11) = ( -0.8333333 0.8333333 -0.8333333), wk = 0.0009103 > k( 12) = ( -0.9166667 0.9166667 -0.9166667), wk = 0.0009103 > k( 13) = ( -1.0000000 1.0000000 -1.0000000), wk = 0.0009103 > k( 14) = ( 0.0833333 0.0833333 0.0833333), wk = 0.0009103 > k( 15) = ( 0.0000000 0.1666667 0.0000000), wk = 0.0009103 > k( 16) = ( -0.0833333 0.2500000 -0.0833333), wk = 0.0009103 > = > = > = > = > = > = > ====================================================================== > CRYSTAL > nsfc run input (partial) (Al, ibrav=2) > K_POINTS {crystal} > 2197 > 0.00000000 0.00000000 0.00000000 1.00000000 > -0.08333333 0.08333333 -0.08333333 1.00000000 > -0.16666667 0.16666667 -0.16666667 1.00000000 > -0.25000000 0.25000000 -0.25000000 1.00000000 > -0.33333333 0.33333333 -0.33333333 1.00000000 > -0.41666667 0.41666667 -0.41666667 1.00000000 > -0.50000000 0.50000000 -0.50000000 1.00000000 > -0.58333333 0.58333333 -0.58333333 1.00000000 > -0.66666667 0.66666667 -0.66666667 1.00000000 > -0.75000000 0.75000000 -0.75000000 1.00000000 > -0.83333333 0.83333333 -0.83333333 1.00000000 > -0.91666667 0.91666667 -0.91666667 1.00000000 > -1.00000000 1.00000000 -1.00000000 1.00000000 > 0.08333333 0.08333333 0.08333333 1.00000000 > 0.00000000 0.16666667 0.00000000 1.00000000 > -0.08333333 0.25000000 -0.08333333 1.00000000 > -0.16666667 0.33333333 -0.16666667 1.00000000 > -0.25000000 0.41666667 -0.25000000 1.00000000 > -0.33333333 0.50000000 -0.33333333 1.00000000 > -0.41666667 0.58333333 -0.41666667 1.00000000 > ................... > nsfc run output (partial) (Al, ibrav=2) > reciprocal axes: (cart. coord. in units 2 pi/a_0) > b(1) = ( -1.000000 -1.000000 1.000000 ) > b(2) = ( 1.000000 1.000000 1.000000 ) > b(3) = ( -1.000000 1.000000 -1.000000 ) > number of k points= 2197 gaussian broad. (Ry)= 0.0150 ngauss = > cart. coord. in units 2pi/a_0 > k( 1) = ( 0.0000000 0.0000000 0.0000000), wk = 0.0009103 > k( 2) = ( 0.2500000 0.0833333 0.0833333), wk = 0.0009103 > k( 3) = ( 0.5000000 0.1666667 0.1666667), wk = 0.0009103 > k( 4) = ( 0.7500000 0.2500000 0.2500000), wk = 0.0009103 > k( 5) = ( 1.0000000 0.3333333 0.3333333), wk = 0.0009103 > k( 6) = ( 1.2500000 0.4166667 0.4166667), wk = 0.0009103 > k( 7) = ( 1.5000000 0.5000000 0.5000000), wk = 0.0009103 > k( 8) = ( 1.7500000 0.5833333 0.5833333), wk = 0.0009103 > k( 9) = ( 2.0000000 0.6666667 0.6666667), wk = 0.0009103 > k( 10) = ( 2.2500000 0.7500000 0.7500000), wk = 0.0009103 > k( 11) = ( 2.5000000 0.8333333 0.8333333), wk = 0.0009103 > k( 12) = ( 2.7500000 0.9166667 0.9166667), wk = 0.0009103 > k( 13) = ( 3.0000000 1.0000000 1.0000000), wk = 0.0009103 > k( 14) = ( -0.0833333 0.0833333 0.0833333), wk = 0.0009103 > k( 15) = ( 0.1666667 0.1666667 0.1666667), wk = 0.0009103 > k( 16) = ( 0.4166667 0.2500000 0.2500000), wk = 0.0009103 > k( 17) = ( 0.6666667 0.3333333 0.3333333), wk = 0.0009103 > k( 18) = ( 0.9166667 0.4166667 0.4166667), wk = 0.0009103 > k( 19) = ( 1.1666667 0.5000000 0.5000000), wk = 0.0009103 > k( 20) = ( 1.4166667 0.5833333 0.5833333), wk = 0.0009103 > k( 21) = ( 1.6666667 0.6666667 0.6666667), wk = 0.0009103 > > 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 > > > > > -----Original Message----- > From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] > On > Behalf Of Paolo Giannozzi > Sent: Tuesday, December 30, 2008 5:21 AM > To: PWSCF Forum > Subject: Re: [Pw_forum] K_POINTS, tpiba or crystal > > On Tuesday 30 December 2008 05:56, Paul M. Grant wrote: > >> b(1) = ( 1.000000 0.000000 0.000000 ) >> b(2) = ( 0.000000 1.000000 0.000000 ) >> b(3) = ( 0.000000 0.000000 1.000000 ) > > these are the lattice vectors generating a simple cubic lattice > >> Now, I want to input this grid to an nscf run via K_POINTS...is my >> grid >> "tpiba" or "crystal?" > > in this particular case, it shouldn't matter, since the crystal and > cartesian axis are the same, and the length of crystal axis is 2pi/a . > > Paolo > -- > Paolo Giannozzi, Democritos and Udine University > _______________________________________________ > 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/20090104/c6cb0d6e/attachment-0001.htm From giannozz at democritos.it Sun Jan 4 10:25:22 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sun, 4 Jan 2009 10:25:22 +0100 Subject: [Pw_forum] Advice on Electron-Phonon Calculations In-Reply-To: <1382facf0901030855j53a2af45raf599c43d51ce2f@mail.gmail.com> References: <1382facf0901030855j53a2af45raf599c43d51ce2f@mail.gmail.com> Message-ID: <200901041025.23384.giannozz@democritos.it> On Saturday 03 January 2009 17:55, Carlos Reis wrote: > The calculation of the phonons in parallel also worked reasonably well > although I still don't know how to make ph.x to write the large *.dwf > and *.bar files in the /tmp of each node (or even if that is possible). if you specify a directory that is local to each node, each process should write its slice of data to it. Note however that each process will try to read its slice of data from it as well (unless you specify to pw.x that you want all your data written to a single place, wf_collect=.true.) > Is there some other way to do this electron-phonon calculation taking > advantage of parallelization over the k-points?. unfortunately the parallelization over k-points is not implemented for electron-phonon calculations. if I remember correctly, its implementation together with the current algorithm is not completely trivial. Paolo -- Paolo Giannozzi, Democritos and Udine University From giannozz at democritos.it Sun Jan 4 10:29:01 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sun, 4 Jan 2009 10:29:01 +0100 Subject: [Pw_forum] K_POINTS, tpiba or crystal In-Reply-To: <01a001c96dfc$5a97be10$0fc73a30$@net> References: <000501c969e4$eb03bb00$c10b3100$@net> <200812301420.47223.giannozz@democritos.it> <01a001c96dfc$5a97be10$0fc73a30$@net> Message-ID: <200901041029.01350.giannozz@democritos.it> On Sunday 04 January 2009 00:38, Paul M. Grant wrote: > BTW, whether tpiba, etc., should be enclosed in curly braces or > parentheses or nothing on the K_POINTS card is confusing. it is irrelevant: only the keyword is relevant P. -- Paolo Giannozzi, Democritos and Udine University From giannozz at democritos.it Mon Jan 5 16:19:55 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 05 Jan 2009 16:19:55 +0100 Subject: [Pw_forum] Problem with average.x In-Reply-To: <00163646ccde7ce575045f9fa70c@google.com> References: <00163646ccde7ce575045f9fa70c@google.com> Message-ID: <4962251B.7050406@democritos.it> jameslipd at gmail.com wrote: > from plot_io : error [...] error number is irrelevant > opening file IO-O.pp this is the real error message, and it means that the file is not there, or not readable > I checked my IO.O.pp file you should check if file "IO-O.pp" is where the code expects it, with the proper name. IIRC it should be in the current directory, i.e. the one in which you execute average.x . In parallel execution, however, the current directory may not be where you think it is: some MPI's assume that it is in the directory where the executable is Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From panda.deng.pan at gmail.com Mon Jan 5 17:16:46 2009 From: panda.deng.pan at gmail.com (=?GB2312?B?xcu1xw==?=) Date: Tue, 6 Jan 2009 00:16:46 +0800 Subject: [Pw_forum] Problem about "ibrav=7" running scf calculation Message-ID: Dear All, I have a problem while running the scf calculation with the option "ibrav=7". Here is the error information: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from read_namelists : error # 19 reading namelist electrons %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from read_namelists : error # 19 reading namelist electrons %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... And my input file is : &control calculation = 'scf' restart_mode='from_scratch' prefix='BaNiAs', pseudo_dir = '/mywork' / &system ibrav= 7 , celldm(1) = 7.7702 , celldm(3) = 2.806, nat= 10, ntyp= 3, ecutwfc =18.0 / &electrons mixing_beta = 0.7 conv_thr = 1.0d-8 ATOMIC_SPECIES Ba Ba.pw91-nsp-van.UPF Ni NiUS.RRKJ3.UPF As As.gon.UPF ATOMIC_POSITIONS Ba 0.0000 0.0000 0.0000 Ni 0.5000 0.0000 0.2500 Ni 0.0000 0.5000 0.2500 As 0.0000 0.0000 0.3476 As 0.0000 0.0000 -0.3476 K_POINTS (automatic) 2 2 2 0 0 0 I have tried to remove the option 'celldm(1) = 7.7702', at this time the error information gives: from iosys : error # 1 invalid lattice parameters ( celldm or a ) If it helps.I was trying to calculation the Dos of BaNiAs,which have symmetry of the space group I4_mmm,with the lattice parameter a=7.7702 c=21.81 (bohr). Please tell me the proper option sets, or point in out where I was wrong. Thanks Pan Deng Nanjing University From akohlmey at cmm.chem.upenn.edu Mon Jan 5 17:33:33 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 5 Jan 2009 11:33:33 -0500 (EST) Subject: [Pw_forum] Problem about "ibrav=7" running scf calculation In-Reply-To: References: Message-ID: On Tue, 6 Jan 2009, ???? wrote: PD> Dear All, dear pan deng, PD> I have a problem while running the scf calculation with the option "ibrav=7". PD> Here is the error information: PD> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% PD> from read_namelists : error # 19 PD> reading namelist electrons PD> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% PD> PD> stopping ... ok. so that means that there is something in the &electrons section that is not in accord with the syntax requirements for namelists. PD> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% PD> from read_namelists : error # 19 PD> reading namelist electrons PD> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% PD> PD> stopping ... thank you for giving a good example and providing a _full_ input. this helps a lot. PD> And my input file is : PD> PD> &control PD> calculation = 'scf' PD> restart_mode='from_scratch' PD> prefix='BaNiAs', PD> pseudo_dir = '/mywork' PD> / PD> &system PD> ibrav= 7 , celldm(1) = 7.7702 , celldm(3) = 2.806, nat= 10, ntyp= 3, PD> ecutwfc =18.0 PD> / PD> &electrons PD> mixing_beta = 0.7 PD> conv_thr = 1.0d-8 PD> ATOMIC_SPECIES ...and here is the problem. you don't "close" the namelist. so the code will try to interpret ATOMIC_SPECIES as part of the namelist, which - of course - fails. hence the error. insert a forward slash on the line before it and you should be fine. PD> Ba Ba.pw91-nsp-van.UPF PD> Ni NiUS.RRKJ3.UPF PD> As As.gon.UPF PD> PD> ATOMIC_POSITIONS PD> Ba 0.0000 0.0000 0.0000 PD> Ni 0.5000 0.0000 0.2500 PD> Ni 0.0000 0.5000 0.2500 PD> As 0.0000 0.0000 0.3476 PD> As 0.0000 0.0000 -0.3476 PD> PD> K_POINTS (automatic) PD> PD> 2 2 2 0 0 0 PD> PD> I have tried to remove the option 'celldm(1) = 7.7702', at this time PD> the error information gives: PD> from iosys : error # 1 PD> invalid lattice parameters ( celldm or a ) PD> PD> If it helps.I was trying to calculation the Dos of BaNiAs,which have PD> symmetry of the space group I4_mmm,with the lattice parameter a=7.7702 PD> c=21.81 (bohr). PD> PD> Please tell me the proper option sets, or point in out where I was wrong. the error was about the file syntax, not about about an individual option. cheers, axel. PD> PD> Thanks PD> PD> Pan Deng PD> Nanjing University PD> _______________________________________________ PD> Pw_forum mailing list PD> Pw_forum at pwscf.org PD> http://www.democritos.it/mailman/listinfo/pw_forum PD> -- ======================================================================= 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 Mon Jan 5 17:34:26 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 5 Jan 2009 17:34:26 +0100 Subject: [Pw_forum] Problem about "ibrav=7" running scf calculation In-Reply-To: References: Message-ID: On Jan 5, 2009, at 17:16 , ?? wrote: > from read_namelists : error # 19 > reading namelist electrons there is an error in namelist "electrons" > &electrons > mixing_beta = 0.7 > conv_thr = 1.0d-8 > ATOMIC_SPECIES what happened to the terminator (the character "/") of the namelist? --- Paolo Giannozzi, Democritos and University of Udine, Italy From jlz101 at unlserve.unl.edu Mon Jan 5 17:23:45 2009 From: jlz101 at unlserve.unl.edu (J. D. Burton) Date: Mon, 5 Jan 2009 10:23:45 -0600 Subject: [Pw_forum] Problem about "ibrav=7" running scf calculation In-Reply-To: References: Message-ID: <000c01c96f51$fb18a420$f149ec60$@unl.edu> Dear Pan Deng, The namelist &electrons must be terminated with "/" just as the other namelists are. Cheers, J. D. ************************************ J. D. Burton, Ph.D. jlz101 at unlserve.unl.edu University of Nebraska Lincoln Physics and Astronomy Office Ph. (402) 472 2499 Mobile Ph. (402) 419 9918 213C Ferguson Hall ************************************ "The job of a scientist is to generate wrong ideas as fast as possible." -- Murray Gell-Mann -----Original Message----- From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of ?? Sent: Monday, January 05, 2009 10:17 AM To: pw_forum at pwscf.org Subject: [Pw_forum] Problem about "ibrav=7" running scf calculation Dear All, I have a problem while running the scf calculation with the option "ibrav=7". Here is the error information: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from read_namelists : error # 19 reading namelist electrons %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from read_namelists : error # 19 reading namelist electrons %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... And my input file is : &control calculation = 'scf' restart_mode='from_scratch' prefix='BaNiAs', pseudo_dir = '/mywork' / &system ibrav= 7 , celldm(1) = 7.7702 , celldm(3) = 2.806, nat= 10, ntyp= 3, ecutwfc =18.0 / &electrons mixing_beta = 0.7 conv_thr = 1.0d-8 ATOMIC_SPECIES Ba Ba.pw91-nsp-van.UPF Ni NiUS.RRKJ3.UPF As As.gon.UPF ATOMIC_POSITIONS Ba 0.0000 0.0000 0.0000 Ni 0.5000 0.0000 0.2500 Ni 0.0000 0.5000 0.2500 As 0.0000 0.0000 0.3476 As 0.0000 0.0000 -0.3476 K_POINTS (automatic) 2 2 2 0 0 0 I have tried to remove the option 'celldm(1) = 7.7702', at this time the error information gives: from iosys : error # 1 invalid lattice parameters ( celldm or a ) If it helps.I was trying to calculation the Dos of BaNiAs,which have symmetry of the space group I4_mmm,with the lattice parameter a=7.7702 c=21.81 (bohr). Please tell me the proper option sets, or point in out where I was wrong. Thanks Pan Deng Nanjing University _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From pushpajnc at gmail.com Mon Jan 5 23:50:15 2009 From: pushpajnc at gmail.com (pushpa raghani) Date: Mon, 5 Jan 2009 19:50:15 -0300 Subject: [Pw_forum] relaxation problem on IBM bluegene machine Message-ID: Dear all, I am doing slab relaxation (surface) calculations on IBM bluegene machine. For certain cell sizes, the calculations go very fine but for some others, it stops after doing the first scf run and then calculating forces and the new atomic positions, ie, it stops while 'Writing output data file pwscf.save'. The status shows that the calculation is running but it doesn't do anything. Could somebody tell me why does it this happen and how can it be fixed? I am using 3.2.3 version. Many Thanks, Pushpa Raghani Stanford University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090105/1136eaa8/attachment.htm From manoj at phys.ufl.edu Tue Jan 6 01:31:00 2009 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Mon, 5 Jan 2009 19:31:00 -0500 (EST) Subject: [Pw_forum] complex band Message-ID: Dear PWSCF users, The complex band code of PWSCF is based on Choi & Ihm's paper, which uses Kleinmann Bylandar pseudopotential. But, the code can work for ultra soft pseuodpotential as well. My question is what are the extra steps i need to take to change from KB to USPP? Thanks, Manoj Srivastava Graduate Student Department of Physics University of Florida, gainesville, FL. From thmason24 at yahoo.com Tue Jan 6 04:27:49 2009 From: thmason24 at yahoo.com (Timothy Mason) Date: Mon, 5 Jan 2009 19:27:49 -0800 (PST) Subject: [Pw_forum] vc_relax convergence criteria In-Reply-To: Message-ID: <604029.34454.qm@web81803.mail.mud.yahoo.com> Group, I find that vc_relax appears to have differente convergence criteria depending on which algorithm I use. why is this? both output files reveal the same forces on every atom after the first scf run. the following control inputs and following cell parameters are considered converged for BFGS but not for damp-w. input file is provided below as are the force and stress results of the first scf run. The only difference between the other input file is that ion_Dynanics and cell dynamics are replaced with 'bfgs'. thanks, Tim Mason : graduate student , university of missouri st. louis &CONTROL calculation = "vc-relax", prefix = 'n9_cont', tprnfor = .TRUE. pseudo_dir = '/home/thmmqc/pwscf/psps/US_GGA', outdir='/home/thmmqc/tmp' etot_conv_thr = 1.0D-8 forc_conv_thr = 0.001, dt =30 nstep = 100 / &SYSTEM ibrav = 0, nat = 15, ntyp = 5, ecutwfc = 40 ecutrho = 480 celldm(1) = 1.889725989 / &ELECTRONS mixing_mode = 'plain', mixing_beta = 0.7, conv_thr = 1.0D-10, / &IONS ion_dynamics = 'damp', ion_damping = 0.5, ion_positions = 'from_input', pot_extrapolation = "second_order", wfc_extrapolation = "second_order", upscale = 100, / &CELL cell_dynamics = "damp-w", press = 0.0, / ATOMIC_SPECIES Li 1.0 Li.pbe-n-van.UPF Ca 1.0 Ca.pbe-nsp-van.UPF N 1.0 N.pbe-rrkjus.UPF H 1.0 H.pbe-rrkjus.UPF B 1.0 B.pbe-n-van.UPF CELL_PARAMETERS 6.04984441 0.07947561 0.15763675 -1.73979839 9.57714512 0.75491312 -1.44861278 4.74878467 5.48694513 ATOMIC_POSITIONS {crystal} Li 0.520452827000 -0.010315542000 1.409909233000 Ca 0.695778370000 0.477023050000 -0.049490890000 N 0.431958544000 0.703293048000 0.694429814000 H 0.318686095000 0.666672029000 0.612147697000 H 0.335911859000 0.689088529000 0.810462743000 H 0.360709987000 0.300964023000 0.638729658000 H 0.201482918000 0.495572043000 0.343835757000 H 0.128550349000 0.243802563000 0.512226432000 H 0.079408634000 0.326602576000 0.652481493000 H 0.849854970000 0.835524836000 -0.001662989000 H 0.744851787000 0.657416041000 0.317036490000 H 0.808377359000 0.910829253000 0.149232970000 H 0.571308073000 0.847867969000 0.035708674000 B 0.192474730000 0.341596546000 0.537027123000 B 0.743409823000 0.812482642000 0.125603330000 K_POINTS automatic 4 4 4 0 0 0 ************forces/stresses************** Forces acting on atoms (Ry/au): negative rho (up, down): 0.141E-03 0.000E+00 atom 1 type 1 force = -0.00003810 -0.00012386 0.00011116 atom 2 type 2 force = -0.00001862 -0.00011006 0.00011622 atom 3 type 3 force = -0.00013613 0.00018277 0.00037626 atom 4 type 4 force = -0.00018482 -0.00013005 0.00033089 atom 5 type 4 force = 0.00014399 -0.00018545 0.00010084 atom 6 type 4 force = -0.00033470 -0.00009963 0.00013277 atom 7 type 4 force = -0.00001692 0.00010745 0.00044525 atom 8 type 4 force = 0.00014860 0.00054225 0.00007950 atom 9 type 4 force = 0.00004670 -0.00009161 0.00021222 atom 10 type 4 force = -0.00004822 0.00001277 0.00020842 atom 11 type 4 force = 0.00017014 -0.00020814 0.00004839 atom 12 type 4 force = 0.00007402 0.00019658 0.00009918 atom 13 type 4 force = -0.00036718 -0.00001416 0.00022673 atom 14 type 5 force = 0.00006543 0.00025720 0.00043955 atom 15 type 5 force = 0.00049580 -0.00033607 0.00017440 Total force = 0.001489 Total SCF correction = 0.000015 entering subroutine stress ... negative rho (up, down): 0.141E-03 0.000E+00 total stress (Ry/bohr**3) (kbar) P= 1.59 0.00000309 0.00000387 0.00000120 0.45 0.57 0.18 0.00000387 0.00001595 0.00000723 0.57 2.35 1.06 0.00000120 0.00000723 0.00001333 0.18 1.06 1.96 ? From jameslipd at gmail.com Tue Jan 6 04:50:52 2009 From: jameslipd at gmail.com (Jiaye, Li) Date: Tue, 6 Jan 2009 11:50:52 +0800 Subject: [Pw_forum] Problem with average.x Message-ID: Dear Paolo Thanks for answering me. My IO-O.pp file is readable and rewritable and it is has been placed in the current directory where I performed pw.x and pp.x calculation. The temp dir is the same between pw.x input file and pp.x input file. At the first glance, I doubted that it may be due to the error in plot_io.f90, but the test of the example in WorkFct_example worked well, which really confused me. As you mentioned that the error must be lying in the input file for average.x and the IO-O.pp file may be in the wrong place. I tried to place it in the dir where average.x is placed, but in vain. Could you please help me to fix the problem? Thank you. ps. my souce code version is espresso-4.0.3. -- Sincerely Jiaye Li ============================================== Li, Jiaye M.S. in designing and developing novel materials Research experience: *ALD precursor. ie. Metal complexes *Surface modification. ie. Indium tin oxide, SiO2 *Gas storage and adsorption. ie. Ionic Liquids Skills: *ab-initial method, DFT simulation and Force Field simulation Please contact me at: 1. jameslipd at gmail.com or 2. jameslibd at gmail.com ============================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090106/6a522a02/attachment-0001.htm From jasius_1 at yahoo.com Tue Jan 6 04:53:08 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Mon, 5 Jan 2009 19:53:08 -0800 (PST) Subject: [Pw_forum] from read_namelists : error # 1 Message-ID: <113160.78400.qm@web54402.mail.yahoo.com> Not sure why, but program aborts with the following error. The input script is below, I was trying to run first test example #!/bin/bash #PBS -l nodes=1:ppn=4 #PBS -S /bin/bash #PBS -j eo #PBS -e atom-lsda.err #PBS -V cd $PBS_O_WORKDIR echo "Current working directory is `pwd`" echo "Node file: $PBS_NODEFILE :" echo "---------------------" cat $PBS_NODEFILE echo "---------------------" echo "Starting run at: `date`" export PATH=${PATH}:/share/apps/espresso-4.0.4/bin mpiexec -np 4 -machinefile $PBS_NODEFILE /share/apps/espresso-4.0.4/bin/pw.x -npool 8 < atom-lsda echo "Job finished at: `date`" Error: Current working directory is /home/jbaltrus/atom-lsda Node file: /opt/torque/aux//130.address.edu : --------------------- compute-0-1 compute-0-1 compute-0-1 compute-0-1 --------------------- Starting run at: Mon Jan 5 21:44:50 CST 2009 Program PWSCF v.4.0.4 starts ... Today is 5Jan2009 at 21:44:51 Parallel version (MPI) Number of processors in use: 4 K-points division: npool = 4 For Norm-Conserving or Ultrasoft (Vanderbilt) Pseudopotentials or PAW Current dimensions of program pwscf are: Max number of different atomic species (ntypx) = 10 Max number of k-points (npk) = 40000 Max angular momentum in pseudopotentials (lmaxx) = 3 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from read_namelists : error # 1 reading namelist control %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... From akohlmey at cmm.chem.upenn.edu Tue Jan 6 05:16:30 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 5 Jan 2009 23:16:30 -0500 (EST) Subject: [Pw_forum] from read_namelists : error # 1 In-Reply-To: <113160.78400.qm@web54402.mail.yahoo.com> References: <113160.78400.qm@web54402.mail.yahoo.com> Message-ID: On Mon, 5 Jan 2009, Jonas Baltrusaitis wrote: JB> Not sure why, but program aborts with the following error. The input JB> script is below, I was trying to run first test example jonas, have you checked the respective FAQ on http://www.quantum-espresso.org/ ? it is highly likely that your problem is one of those mentioned there. my guess would be that your MPI startup scheme doesn't replicate redirected input to all nodes. cheers, axel. p.s.: please honor the policy of this mailing list to provide your full name and affiliation. thanks. a. JB> JB> #!/bin/bash JB> #PBS -l nodes=1:ppn=4 JB> #PBS -S /bin/bash JB> #PBS -j eo JB> #PBS -e atom-lsda.err JB> #PBS -V JB> JB> cd $PBS_O_WORKDIR JB> JB> echo "Current working directory is `pwd`" JB> JB> echo "Node file: $PBS_NODEFILE :" JB> echo "---------------------" JB> cat $PBS_NODEFILE JB> echo "---------------------" JB> JB> echo "Starting run at: `date`" JB> JB> export PATH=${PATH}:/share/apps/espresso-4.0.4/bin JB> JB> JB> mpiexec -np 4 -machinefile $PBS_NODEFILE /share/apps/espresso-4.0.4/bin/pw.x -npool 8 < atom-lsda JB> JB> JB> JB> echo "Job finished at: `date`" JB> JB> Error: JB> JB> Current working directory is /home/jbaltrus/atom-lsda JB> Node file: /opt/torque/aux//130.address.edu : JB> --------------------- JB> compute-0-1 JB> compute-0-1 JB> compute-0-1 JB> compute-0-1 JB> --------------------- JB> Starting run at: Mon Jan 5 21:44:50 CST 2009 JB> JB> Program PWSCF v.4.0.4 starts ... JB> Today is 5Jan2009 at 21:44:51 JB> JB> Parallel version (MPI) JB> JB> Number of processors in use: 4 JB> K-points division: npool = 4 JB> JB> For Norm-Conserving or Ultrasoft (Vanderbilt) Pseudopotentials or PAW JB> JB> Current dimensions of program pwscf are: JB> Max number of different atomic species (ntypx) = 10 JB> Max number of k-points (npk) = 40000 JB> Max angular momentum in pseudopotentials (lmaxx) = 3 JB> JB> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% JB> from read_namelists : error # 1 JB> reading namelist control JB> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% JB> JB> stopping ... JB> JB> JB> JB> JB> JB> _______________________________________________ JB> Pw_forum mailing list JB> Pw_forum at pwscf.org JB> http://www.democritos.it/mailman/listinfo/pw_forum JB> -- ======================================================================= 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 lanhaiping at gmail.com Tue Jan 6 06:07:02 2009 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 6 Jan 2009 13:07:02 +0800 Subject: [Pw_forum] from read_namelists : error # 1 In-Reply-To: <113160.78400.qm@web54402.mail.yahoo.com> References: <113160.78400.qm@web54402.mail.yahoo.com> Message-ID: Hi, " reading namelist control " means that there is something wrong with your input file. therefore, you need send your input file to the forum, or please check it thorough by yourself. Regards On Tue, Jan 6, 2009 at 11:53 AM, Jonas Baltrusaitis wrote: > Not sure why, but program aborts with the following error. The input script > is below, I was trying to run first test example > > #!/bin/bash > #PBS -l nodes=1:ppn=4 > #PBS -S /bin/bash > #PBS -j eo > #PBS -e atom-lsda.err > #PBS -V > > cd $PBS_O_WORKDIR > > echo "Current working directory is `pwd`" > > echo "Node file: $PBS_NODEFILE :" > echo "---------------------" > cat $PBS_NODEFILE > echo "---------------------" > > echo "Starting run at: `date`" > > export PATH=${PATH}:/share/apps/espresso-4.0.4/bin > > > mpiexec -np 4 -machinefile $PBS_NODEFILE > /share/apps/espresso-4.0.4/bin/pw.x -npool 8 < atom-lsda > > > > echo "Job finished at: `date`" > > Error: > > Current working directory is /home/jbaltrus/atom-lsda > Node file: /opt/torque/aux//130.address.edu : > --------------------- > compute-0-1 > compute-0-1 > compute-0-1 > compute-0-1 > --------------------- > Starting run at: Mon Jan 5 21:44:50 CST 2009 > > Program PWSCF v.4.0.4 starts ... > Today is 5Jan2009 at 21:44:51 > > Parallel version (MPI) > > Number of processors in use: 4 > K-points division: npool = 4 > > For Norm-Conserving or Ultrasoft (Vanderbilt) Pseudopotentials or PAW > > Current dimensions of program pwscf are: > Max number of different atomic species (ntypx) = 10 > Max number of k-points (npk) = 40000 > Max angular momentum in pseudopotentials (lmaxx) = 3 > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from read_namelists : error # 1 > reading namelist control > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > > > > > _______________________________________________ > 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/20090106/88e91625/attachment.htm From lanhaiping at gmail.com Tue Jan 6 06:10:27 2009 From: lanhaiping at gmail.com (lan haiping) Date: Tue, 6 Jan 2009 13:10:27 +0800 Subject: [Pw_forum] Problem with average.x In-Reply-To: References: Message-ID: Hi, Please donot disrupt the threads of maillist. As far as i know, you always repeat sending mails of the same topics with different threads or subjects. Keep the thread is useful to yourself and others. Regards, On Tue, Jan 6, 2009 at 11:50 AM, Jiaye, Li wrote: > Dear Paolo > > > Thanks for answering me. My IO-O.pp file is readable and rewritable and it > is has been placed in the current directory where I performed pw.x and pp.x > calculation. The temp dir is the same between pw.x input file and pp.x input > file. > > At the first glance, I doubted that it may be due to the error in > plot_io.f90, but the test of the example in WorkFct_example worked well, > which really confused me. As you mentioned that the error must be lying in > the input file for average.x and the IO-O.pp file may be in the wrong place. > I tried to place it in the dir where average.x is placed, but in vain. > > Could you please help me to fix the problem? Thank you. > > ps. my souce code version is espresso-4.0.3. > > -- > Sincerely > > Jiaye Li > > > > ============================================== > Li, Jiaye > > M.S. in designing and developing novel materials > > Research experience: > *ALD precursor. ie. Metal complexes > *Surface modification. ie. Indium tin oxide, SiO2 > *Gas storage and adsorption. ie. Ionic Liquids > > Skills: > *ab-initial method, DFT simulation and Force Field simulation > > Please contact me at: > 1. jameslipd at gmail.com or > 2. jameslibd at gmail.com > > ============================================== > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090106/1cb98fc4/attachment.htm From giannozz at democritos.it Tue Jan 6 09:36:24 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 6 Jan 2009 09:36:24 +0100 Subject: [Pw_forum] Problem with average.x In-Reply-To: References: Message-ID: On Jan 6, 2009, at 4:50 , Jiaye, Li wrote: > My IO-O.pp file is readable and rewritable and it is has been > placed in the current directory where I performed pw.x and pp.x > calculation. The temp dir is the same between pw.x input file and > pp.x input file. [...] Could you please help me to fix the problem? locate the place where the error occurs in the fortran code, then place a command like call shell ('ls -l ') recompile and see what happens. It should list the files in the current directory P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Tue Jan 6 09:38:29 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 6 Jan 2009 09:38:29 +0100 Subject: [Pw_forum] vc_relax convergence criteria In-Reply-To: <604029.34454.qm@web81803.mail.mud.yahoo.com> References: <604029.34454.qm@web81803.mail.mud.yahoo.com> Message-ID: <46B02EAD-2C93-46F1-B7D4-07D5AE98566C@democritos.it> On Jan 6, 2009, at 4:27 , Timothy Mason wrote: > I find that vc_relax appears to have differente convergence criteria > depending on which algorithm I use. why is this? likely because they were written by different people. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Tue Jan 6 09:41:22 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 6 Jan 2009 09:41:22 +0100 Subject: [Pw_forum] complex band In-Reply-To: References: Message-ID: On Jan 6, 2009, at 1:31 , Manoj Srivastava wrote: > The complex band code of PWSCF is based on Choi & Ihm's paper, > which uses > Kleinmann Bylandar pseudopotential. But, the code can work for ultra > soft pseuodpotential as well. My question is what are the extra > steps i > need to take to change from KB to USPP? try this: title = {Ballistic conductance of magnetic Co and Ni nanowires with ultrasoft pseudopotentials}, author = {A. Smogunov and A. Dal Corso and E. Tosatti}, journal = {Phys. Rev. B}, volume = {70}, pages= {045417}, year = 2004} Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Tue Jan 6 09:53:06 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 6 Jan 2009 09:53:06 +0100 Subject: [Pw_forum] relaxation problem on IBM bluegene machine In-Reply-To: References: Message-ID: On Jan 5, 2009, at 23:50 , pushpa raghani wrote: > I am doing slab relaxation (surface) calculations on IBM bluegene > machine. > For certain cell sizes, the calculations go very fine but for some > others, it > stops after doing the first scf run and then calculating forces > and the new > atomic positions, ie, it stops while 'Writing output data file > pwscf.save'. The > status shows that the calculation is running but it doesn't do > anything. Could > somebody tell me why does it this happen and how can it be fixed? short answer: no, nobody can. You have to discover what happens exactly and under which circumstances: if this is blue-gene specific or not, if it is reproducible or erratic, what happens if you change options (try to keep data into memory instead of saving them to file), etc. etc.. It looks to me like a problem in the file system, or in its usage, rather than in quantum espresso itself. > I am using 3.2.3 version. you should use more recent versions. They better support bluegene machines and have a better scalability. Not sure this will do anything to fix your problem, though Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From paulatto at sissa.it Tue Jan 6 10:07:52 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Tue, 06 Jan 2009 10:07:52 +0100 Subject: [Pw_forum] Problem about "ibrav=7" running scf calculation In-Reply-To: References: Message-ID: On Mon, 05 Jan 2009 17:16:46 +0100, ?????? wrote: > I have a problem while running the scf calculation with the option > "ibrav=7". Dear ??, the fact that you are using ibrav=7 is totally unrelated; as the error message says, you have a problem in the electron namelist, actually you have forgotten to terminate it with a line containing single "/" (like other namelists). cheers -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From paulatto at sissa.it Tue Jan 6 10:11:24 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Tue, 06 Jan 2009 10:11:24 +0100 Subject: [Pw_forum] from read_namelists : error # 1 In-Reply-To: <113160.78400.qm@web54402.mail.yahoo.com> References: <113160.78400.qm@web54402.mail.yahoo.com> Message-ID: On Tue, 06 Jan 2009 04:53:08 +0100, Jonas Baltrusaitis wrote: > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from read_namelists : error # 1 > reading namelist control > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Dear Jonas, all we can say is you have a syntax error in the control namelist. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From panda.deng.pan at gmail.com Tue Jan 6 11:03:02 2009 From: panda.deng.pan at gmail.com (=?GB2312?B?xcu1xw==?=) Date: Tue, 6 Jan 2009 18:03:02 +0800 Subject: [Pw_forum] Problem about "ibrav=7" running scf calculation In-Reply-To: References: Message-ID: Dear All I am gratreful for all your help to me. Pan Deng Nanjing University From manoj at phys.ufl.edu Tue Jan 6 15:25:18 2009 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Tue, 6 Jan 2009 09:25:18 -0500 (EST) Subject: [Pw_forum] complex band In-Reply-To: Message-ID: Thanks Paolo. I appreciate it. -manoj On Tue, 6 Jan 2009, Paolo Giannozzi wrote: > > On Jan 6, 2009, at 1:31 , Manoj Srivastava wrote: > > > The complex band code of PWSCF is based on Choi & Ihm's paper, > > which uses > > Kleinmann Bylandar pseudopotential. But, the code can work for ultra > > soft pseuodpotential as well. My question is what are the extra > > steps i > > need to take to change from KB to USPP? > > try this: > title = {Ballistic conductance of magnetic Co and Ni nanowires with > ultrasoft pseudopotentials}, > author = {A. Smogunov and A. Dal Corso and E. Tosatti}, > journal = {Phys. Rev. B}, > volume = {70}, > pages= {045417}, > year = 2004} > > 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 thmason24 at yahoo.com Tue Jan 6 19:30:15 2009 From: thmason24 at yahoo.com (Timothy Mason) Date: Tue, 6 Jan 2009 10:30:15 -0800 (PST) Subject: [Pw_forum] vc_relax convergence criteria Message-ID: <389925.75195.qm@web81802.mail.mud.yahoo.com> On Jan 6, 2009, at 4:27 , Timothy Mason wrote: > I find that vc_relax appears to have differente convergence criteria > depending on which algorithm I use. why is this? >>likely because they were written by different people. >>Paolo I used version 4.03. does anyone see value in having consistent convergence VC-relax criteria for future releases? the documentaion has definitions for what force_conv_thresh and etot_conv_thr mean. Do these definitions not apply to bfgs? From pushpajnc at gmail.com Tue Jan 6 20:28:53 2009 From: pushpajnc at gmail.com (pushpa raghani) Date: Tue, 6 Jan 2009 16:28:53 -0300 Subject: [Pw_forum] relaxation problem on IBM bluegene machine In-Reply-To: References: Message-ID: On Tue, Jan 6, 2009 at 5:53 AM, Paolo Giannozzi wrote: > > On Jan 5, 2009, at 23:50 , pushpa raghani wrote: > > > I am doing slab relaxation (surface) calculations on IBM bluegene > > machine. > > For certain cell sizes, the calculations go very fine but for some > > others, it > > stops after doing the first scf run and then calculating forces > > and the new > > atomic positions, ie, it stops while 'Writing output data file > > pwscf.save'. The > > status shows that the calculation is running but it doesn't do > > anything. Could > > somebody tell me why does it this happen and how can it be fixed? > > short answer: no, nobody can. You have to discover what happens exactly > and under which circumstances: if this is blue-gene specific or not, It is bluegene specific. If I do the same calculation on another cluster, it works fine but it has lesser number of processors, so the calculation is very slow. > > if it is > reproducible or erratic, It is reproducible, I tried to change the number of processors, change the symmetries of the system (which anyway I had to do). Some more info: I am doing calculations on FCC(100) surface with Hubbard U. When I use c(2x3) cell (with U and/or without U), everything goes fine. When I use c(2x4) cell without U (just spin polarized calculation), then also there is no such problem, it writes all the data in the directory (pwscf.save) and finishes smoothly. The problem arises when I do LDA+U for c(2x4) cell. Then for some reason it stops (not really stops) while writing pwscf.save! what happens if you change options (try to keep > data into memory instead of saving them to file), etc. etc.. This could work but how to do that? Should I comment out the line in the program where it writes the data to pwscf.save or there is some simple switch to do that? It looks > to me > like a problem in the file system, or in its usage, rather than in > quantum > espresso itself. As I said, it all works fine for other cell sizes and also without U. Then it doesn't seem to be the problem with file system? > > > > I am using 3.2.3 version. > > you should use more recent versions. They better support bluegene > machines > and have a better scalability. Not sure this will do anything to fix > your problem, > though I will also try with the new version. Many Thanks, Pushpa Stanford Universiy > > _______________________________________________ > 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/20090106/387fecf3/attachment.htm From 041019019 at fudan.edu.cn Wed Jan 7 07:37:35 2009 From: 041019019 at fudan.edu.cn (Wang Qinjing) Date: Wed, 07 Jan 2009 14:37:35 +0800 Subject: [Pw_forum] the efficiency problem of transmission calculation by pwcond.x Message-ID: <001c01c97092$6d490d50$3f7c620a@fduphy778548b0> Hi all I am doing transmission calculation by pwcond.x. I have 2 question about the efficiency of it. (1) The most severe problem was that I felt the transmission calculation by pwcond.x was too slow. As an example, the system contains a 6-atom lead and about 100-atom scattering region. T he kx and ky was 8 and 1 respectively. The input file for pwcond.x is as follows: ---------------- &inputcond outdir='./outdir', prefixl='lead' prefixs='cen', tran_file='trans.0' ikind = 1, energy0=2.0d0, denergy=-0.2d0 ewind=4.d0, epsproj=1.d-7, nz1 = 10, / 5 0.0000000 0.0000000 0.2500000 0.1250000 0.0000000 0.5000000 0.2500000 0.0000000 0.5000000 0.3750000 0.0000000 0.5000000 -0.5000000 0.0000000 0.2500000 21 ------------------ It took about 1 hour to get a transmission value for a certain energy on a 8-node parallel cluster with 1 G memory, 2.5 GHz CPU and infiniband. Is it a reasonable one? (2) I notice that the time needed for the transmission is linear proportional to the number of energy points. That means if the time needed for a single energy point is 1 hour, it will take 20 hours for 20 energy points. This is different with some other scientific softwares. Is it natural for pwcond.x? Thank you in advance Best wishes Wang ------------------------------------ ?????????????? Wang Qinjing, Physics Department, Fudan University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090107/cb7fd264/attachment-0001.htm From giannozz at democritos.it Wed Jan 7 09:11:20 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 07 Jan 2009 09:11:20 +0100 Subject: [Pw_forum] relaxation problem on IBM bluegene machine In-Reply-To: References: Message-ID: <496463A8.8020409@democritos.it> pushpa raghani wrote: > When I use c(2x3) cell (with U and/or without U), everything goes > fine. When I use c(2x4) cell without U (just spin polarized > calculation), then also there is no such problem [...]. > The problem arises when I do LDA+U for c(2x4) cell. then it might also be a new version of an old problem. Excerpt from http://www.quantum-espresso.org/wiki/index.php/Developer_Manual#Parallelization: --- Beware: replicated calculations may either be performed independently on each processor, or performed on one processor and broadcast to all others. The first approach requires less programming, but it is unsafe: in principle all processors should yield exactly the same results, if they work on the same data, but sometimes they don't (depending on the machine, compiler, and libraries). Even a tiny difference in the last significant digit can eventually cause serious trouble if allowed to build up, especially when a replicated check is performed (in which case the code may hang if the check yields different results on different processors). Never assume that the value of a variable produced by replicated calculations is exactly the same on all processors: when in doubt, broadcast the value calculated on a specific processor (the root processor) to all others. --- Guess what: IBM bluegene is one of those machine where different processors may yield slightly different results starting with the same data. Some time ago it took me a week to figure out why the parallel diagonalization was hanging on a BG and not on other machines. > what happens if you change options (try to keep > data into memory instead of saving them to file), etc. etc.. > > This could work but how to do that? option "disk_io", in namelist &control In any case: try the latest stable version first, and maybe also the cvs version (a snapshot is available from the download page) Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From smogunov at sissa.it Wed Jan 7 09:38:38 2009 From: smogunov at sissa.it (alexandr smogunov) Date: Wed, 7 Jan 2009 09:38:38 +0100 Subject: [Pw_forum] the efficiency problem of transmission calculation by pwcond.x In-Reply-To: <001c01c97092$6d490d50$3f7c620a@fduphy778548b0> References: <001c01c97092$6d490d50$3f7c620a@fduphy778548b0> Message-ID: <200901070938.38909.smogunov@sissa.it> Dear Wang Could you please provide PW input files for the scattering and lead regions and also PWCOND output file for one k point and one energy to see what is going on? Then, how many CPU do you use? Regards, Alexander On Wednesday 07 January 2009 07:37, Wang Qinjing wrote: > Hi all > > I am doing transmission calculation by pwcond.x. I have 2 question about > the efficiency of it. > > (1) The most severe problem was that I felt the transmission calculation by > pwcond.x was too slow. As an example, the system contains a 6-atom lead and > about 100-atom scattering region. T he kx and ky was 8 and 1 respectively. > The input file for pwcond.x is as follows: > > ---------------- > &inputcond > outdir='./outdir', > prefixl='lead' > prefixs='cen', > tran_file='trans.0' > ikind = 1, > energy0=2.0d0, > denergy=-0.2d0 > ewind=4.d0, > epsproj=1.d-7, > nz1 = 10, > / > 5 > 0.0000000 0.0000000 0.2500000 > 0.1250000 0.0000000 0.5000000 > 0.2500000 0.0000000 0.5000000 > 0.3750000 0.0000000 0.5000000 > -0.5000000 0.0000000 0.2500000 > 21 > ------------------ > > It took about 1 hour to get a transmission value for a certain energy on a > 8-node parallel cluster with 1 G memory, 2.5 GHz CPU and infiniband. Is it > a reasonable one? > > (2) I notice that the time needed for the transmission is linear > proportional to the number of energy points. That means if the time needed > for a single energy point is 1 hour, it will take 20 hours for 20 energy > points. This is different with some other scientific softwares. Is it > natural for pwcond.x? > > Thank you in advance > > Best wishes > > Wang > > > ------------------------------------ > ?????????????? > Wang Qinjing, Physics Department, > Fudan University From giannozz at democritos.it Wed Jan 7 09:44:24 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 07 Jan 2009 09:44:24 +0100 Subject: [Pw_forum] vc_relax convergence criteria In-Reply-To: <389925.75195.qm@web81802.mail.mud.yahoo.com> References: <389925.75195.qm@web81802.mail.mud.yahoo.com> Message-ID: <49646B68.6070704@democritos.it> Timothy Mason wrote: > does anyone see value in having consistent convergence > VC-relax criteria for future releases? of course everybody sees value in it. The relevant question is however a different one: "is anybody 1) interested in implementing consistent criteria for convergence in VC-relax, and 2) able to do it?" Note that the intersection of set 1) and set 2) might be the empty set. > the documentation has definitions for what force_conv_thresh "forc_conv_thr" > and etot_conv_thr mean. Do these definitions not apply to bfgs? they should, since both are passed to the bfgs routine (see Modules/bfgs_module.f90). You can check this quite easily: locate the lines after this one ! ... convergence is checked here It seems to me that they are used for damped dynamics as well (routine PW/vcsmd.f90; locate line ! ... check if convergence for structural minimization is achieved ). Note that there are other "damped dynamics" in the code Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From panda.deng.pan at gmail.com Wed Jan 7 09:47:22 2009 From: panda.deng.pan at gmail.com (=?GB2312?B?xcu1xw==?=) Date: Wed, 7 Jan 2009 16:47:22 +0800 Subject: [Pw_forum] What are 'forrtl: severe (59):' means Message-ID: Dear All, Thanks for your help before but I still couldn't work yet.There is something wrong in my input and I tried to find out but failed. Here is the error message: forrtl: severe (59): list-directed I/O syntax error, unit -5, file Internal List-Directed Read Image PC Routine Line Source pw.x 4000000000CEFA50 Unknown Unknown Unknown pw.x 4000000000CED560 Unknown Unknown Unknown pw.x 4000000000BD1610 Unknown Unknown Unknown pw.x 4000000000B125D0 Unknown Unknown Unknown pw.x 4000000000B11520 Unknown Unknown Unknown pw.x 4000000000B50980 Unknown Unknown Unknown pw.x 4000000000168910 Unknown Unknown Unknown pw.x 4000000000165970 Unknown Unknown Unknown pw.x 4000000000253F90 Unknown Unknown Unknown pw.x 400000000024C280 Unknown Unknown Unknown pw.x 4000000000006A50 Unknown Unknown Unknown pw.x 40000000000068D0 Unknown Unknown Unknown libc.so.6.1 200000000588BC20 Unknown Unknown Unknown pw.x 4000000000006700 Unknown Unknown Unknown And also my input is &control calculation = 'scf' restart_mode='from_scratch' prefix='BaNiAs', pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/', outdir='/disk2/xgwan/tmp/' / &system ibrav= 7 ,celldm(1) = 7.7702 , celldm(3) = 2.806, nat= 10, ntyp= 3, ecutwfc =18.0 / &electrons mixing_beta = 0.7 conv_thr = 1.0d-8 / ATOMIC_SPECIES Ba Ba.pw91-nsp-van.UPF Ni NiUS.RRKJ3.UPF As As.gon.UPF ATOMIC_POSITIONS Ba 0.0000 0.0000 0.0000 Ni 0.5000 0.0000 0.2500 Ni 0.0000 0.5000 0.2500 As 0.0000 0.0000 0.3476 As 0.0000 0.0000 -0.3476 K_POINTS (automatic) 2 2 2 0 0 0 I have no idea what to do.Is that possible the problem is I used improper psudopotential? Thanks Pan Deng Nanjing University From Giovanni.Cantele at na.infn.it Wed Jan 7 09:54:08 2009 From: Giovanni.Cantele at na.infn.it (Giovanni Cantele) Date: Wed, 07 Jan 2009 09:54:08 +0100 Subject: [Pw_forum] What are 'forrtl: severe (59):' means In-Reply-To: References: Message-ID: <49646DB0.20200@na.infn.it> ?? wrote: > Dear All, > Thanks for your help before but I still couldn't work yet.There is > something wrong in my input and I tried to find out but failed. > Here is the error message: > > forrtl: severe (59): list-directed I/O syntax error, unit -5, file > Internal List-Directed Read > Image PC Routine Line > Source > pw.x 4000000000CEFA50 Unknown Unknown Unknown > > ATOMIC_SPECIES > Ba Ba.pw91-nsp-van.UPF > Ni NiUS.RRKJ3.UPF > As As.gon.UPF > > > maybe, the card "ATOMIC_SPECIES" is uncorrectly specified, because atomic masses are missing. From Doc/INPUT_PW : ATOMIC_SPECIES X Mass_X PseudoPot_X Y Mass_Y PseudoPot_Y Z Mass_Z PseudoPot_Z Mass_X,Y,Z,... *MUST* be specified even though they are not needed in your calculation, I think. Giovanni -- Dr. Giovanni Cantele Coherentia CNR-INFM and Dipartimento di Scienze Fisiche Universita' di Napoli "Federico II" Complesso Universitario di Monte S. Angelo - Ed. 6 Via Cintia, I-80126, Napoli, Italy Phone: +39 081 676910 Fax: +39 081 676346 E-mail: giovanni.cantele at cnr.it giovanni.cantele at na.infn.it Web: http://people.na.infn.it/~cantele Research Group: http://www.nanomat.unina.it From giannozz at democritos.it Wed Jan 7 09:59:51 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 07 Jan 2009 09:59:51 +0100 Subject: [Pw_forum] What are 'forrtl: severe (59):' means In-Reply-To: References: Message-ID: <49646F07.2090806@democritos.it> ?? wrote: > forrtl: severe (59): it means that error messages for run-time libraries are even more obscure than quantum-espresso error messages list-directed I/O syntax error, unit -5, file > Internal List-Directed Read it is an error while reading from a character variable > ATOMIC_SPECIES > Ba Ba.pw91-nsp-van.UPF wrong: atomic mass is missing -- Paolo Giannozzi, Democritos and University of Udine, Italy From bipulrr at gmail.com Wed Jan 7 10:06:11 2009 From: bipulrr at gmail.com (Bipul Rakshit) Date: Wed, 7 Jan 2009 14:36:11 +0530 Subject: [Pw_forum] about vc-relax Message-ID: <3a749910901070106p6c093649uf441bb97d3c2dcb3@mail.gmail.com> Hello Pwscf users, I am using vc-relax calculation. At the end of the output I am getting the new, atomic positions and the cell parameters. But there is not much change in the atomic positions while the cell parameters changed, that means the angle of the crystal is changing. So my question is what parameters I use, so that I able to vary the atomic-positions and keep the angle of the crystal fix? Any advice is greatly helpful for my work. regards -- 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/20090107/bf345d29/attachment.htm From panda.deng.pan at gmail.com Wed Jan 7 10:10:55 2009 From: panda.deng.pan at gmail.com (=?GB2312?B?xcu1xw==?=) Date: Wed, 7 Jan 2009 17:10:55 +0800 Subject: [Pw_forum] What are 'forrtl: severe (59):' means In-Reply-To: <49646F07.2090806@democritos.it> References: <49646F07.2090806@democritos.it> Message-ID: thanks a lot for answering me for my naive mistake. From sebastijan.peljhan at ijs.si Wed Jan 7 10:34:13 2009 From: sebastijan.peljhan at ijs.si (Sebastijan Peljhan) Date: Wed, 07 Jan 2009 10:34:13 +0100 Subject: [Pw_forum] about vc-relax In-Reply-To: <3a749910901070106p6c093649uf441bb97d3c2dcb3@mail.gmail.com> References: <3a749910901070106p6c093649uf441bb97d3c2dcb3@mail.gmail.com> Message-ID: <1231320853.16625.40.camel@yunnan.ijs.si> If you would like to vary just atomic positions you should try to perform "relax" calculation instead of "vc-relax". Otherwise you can specify a new flag in input file namelist CELL &CELL cell_dofree xyz / if you would like to change only the length of lattice vectors and not angles. On Wed, 2009-01-07 at 14:36 +0530, Bipul Rakshit wrote: > Hello Pwscf users, > I am using vc-relax calculation. At the end of the output I am getting > the new, atomic positions and the cell parameters. But there is not > much change in the atomic positions while the cell parameters changed, > that means the angle of the crystal is changing. So my question is > what parameters I use, so that I able to vary the atomic-positions and > keep the angle of the crystal fix? -- Sebastijan Peljhan J. Stefan Institute Jamova 39 1000 Ljubljana Slovenia From chuchun.fu at cea.fr Wed Jan 7 10:49:39 2009 From: chuchun.fu at cea.fr (chu-chun fu) Date: Wed, 07 Jan 2009 10:49:39 +0100 Subject: [Pw_forum] generalized Bloch theorem Message-ID: <49647AB3.3060900@cea.fr> Dear PWscf developers and users, I would like to know whether the generalized bloch theorem to consider spin waves is implemented or not in the present PWscf version ( say espresso-4.0.3) ? If I am not wrong, I believed it was implemented in some old versions of PWscf ? Thank you for your response Chu Chun FU -- Chu Chun FU DEN/DMN/SRMP CEA-Saclay 91191 Gif-sur-yvette Cedex France Tel: +33 1 69 08 29 32 Fax: +33 1 69 08 68 67 E-mail: chuchun.fu at cea.fr From panda.deng.pan at gmail.com Wed Jan 7 11:09:27 2009 From: panda.deng.pan at gmail.com (=?GB2312?B?xcu1xw==?=) Date: Wed, 7 Jan 2009 18:09:27 +0800 Subject: [Pw_forum] "ibrav=7" Message-ID: Dear All. Although I have corrected some stupid mistake by the help of All, the error appears again. Here is the error message: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from read_namelists : error # 19 reading namelist system %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... And also my inputs: &control calculation = 'scf' restart_mode='from_scratch' prefix='BaNiAs', pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/', outdir='/disk2/xgwan/tmp/' / &system ibrav=7, cellem(1)=7.7702 , celldm(3)=2.806, nat=10, ntyp=3, ecutwfc =18.0 / &electrons mixing_beta = 0.7 conv_thr = 1.0d-8 / ATOMIC_SPECIES Ba 56.00 Ba.pw91-nsp-van.UPF Ni 28.00 NiUS.RRKJ3.UPF As 33.00 As.gon.UPF ATOMIC_POSITIONS Ba 0.0000 0.0000 0.0000 Ni 0.5000 0.0000 0.2500 Ni 0.0000 0.5000 0.2500 As 0.0000 0.0000 0.3476 As 0.0000 0.0000 -0.3476 K_POINTS (automatic) 2 2 2 0 0 0 I thought I might make some mistake about the option 'ibrav=7'.Also I not quite sure about the option 'ecutwfc=18.0' has the proper value. Thaks again! Pan Deng Nanjing University From giannozz at democritos.it Wed Jan 7 11:17:42 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 07 Jan 2009 11:17:42 +0100 Subject: [Pw_forum] "ibrav=7" In-Reply-To: References: Message-ID: <49648146.3070302@democritos.it> ?? wrote: > Although I have corrected some stupid mistake by the help of All, > the error appears again. you haven't corrected all stupid mistakes... > cellem(1)=7.7702 like this one -- Paolo Giannozzi, Democritos and University of Udine, Italy From panda.deng.pan at gmail.com Wed Jan 7 11:37:47 2009 From: panda.deng.pan at gmail.com (=?GB2312?B?xcu1xw==?=) Date: Wed, 7 Jan 2009 18:37:47 +0800 Subject: [Pw_forum] "ibrav=7" In-Reply-To: <49648146.3070302@democritos.it> References: <49648146.3070302@democritos.it> Message-ID: Dear Paolo, Would you tell more detail about my problem. > > cellem(1)=7.7702 Is this one nosense in the namelist ?Or it needs for other settings? I have tried run without this option 'cellem(1)=7.7702', it will give the error message 'Invalid lattice parametre'.Please tell me how to set it properly. Have a nice day. From giannozz at democritos.it Wed Jan 7 11:41:31 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 07 Jan 2009 11:41:31 +0100 Subject: [Pw_forum] "ibrav=7" In-Reply-To: References: <49648146.3070302@democritos.it> Message-ID: <496486DB.6060203@democritos.it> ?? wrote: >> > cellem(1)=7.7702 > > Is this one nosense in the namelist ? "celldm", not "cellem" -- Paolo Giannozzi, Democritos and University of Udine, Italy From lanhaiping at gmail.com Wed Jan 7 11:47:54 2009 From: lanhaiping at gmail.com (lan haiping) Date: Wed, 7 Jan 2009 18:47:54 +0800 Subject: [Pw_forum] "ibrav=7" In-Reply-To: References: <49648146.3070302@democritos.it> Message-ID: Dear Deng Pan, a suggestion is that you can try to use xcrysden to examine your input file. a simple command just likes xcrysden --pwi your.inp good luck. On Wed, Jan 7, 2009 at 6:37 PM, ?? wrote: > Dear Paolo, > > Would you tell more detail about my problem. > > > > cellem(1)=7.7702 > > Is this one nosense in the namelist ?Or it needs for other settings? > I have tried run without this option 'cellem(1)=7.7702', it will give > the error message 'Invalid lattice parametre'.Please tell me how to > set it properly. > > Have a nice day. > _______________________________________________ > 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/20090107/60095aed/attachment.htm From baroni at sissa.it Wed Jan 7 11:51:01 2009 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 7 Jan 2009 11:51:01 +0100 Subject: [Pw_forum] generalized Bloch theorem In-Reply-To: <49647AB3.3060900@cea.fr> References: <49647AB3.3060900@cea.fr> Message-ID: You are right. Spiral waves used to be implemented long ago in pwscf. I am almost sure they are not any longer, but you may want to check with the original author of the implementation (ralph gebauer, rgebauer at ictp.itO or Andrea Dal Corso (dalcorso at sissa.it) who has implemented most of recent spin-related stuff. Stefano Baroni On Jan 7, 2009, at 10:49 AM, chu-chun fu wrote: > Dear PWscf developers and users, > > I would like to know whether the generalized bloch theorem to consider > spin waves is implemented or not in the present PWscf version > ( say > espresso-4.0.3) ? If I am not wrong, I believed it was implemented in > some old versions of PWscf ? > > Thank you for your response > > Chu Chun FU > > -- > Chu Chun FU > DEN/DMN/SRMP CEA-Saclay > 91191 Gif-sur-yvette Cedex > France > Tel: +33 1 69 08 29 32 > Fax: +33 1 69 08 68 67 > E-mail: chuchun.fu at cea.fr > > _______________________________________________ > 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/20090107/06f312a1/attachment.htm From mambom1902 at yahoo.com Wed Jan 7 13:06:21 2009 From: mambom1902 at yahoo.com (loc duong ding) Date: Wed, 7 Jan 2009 04:06:21 -0800 (PST) Subject: [Pw_forum] problem with vc-relax Message-ID: <415768.12230.qm@web38808.mail.mud.yahoo.com> Dear all, I have a problem when I run vc-relax input file. The force is increasing when the energy is decreasing. I wonder this is the normal case or not. Total force = 0.008423 Total SCF correction = 0.001451 Total force = 0.250397 Total SCF correction = 0.001413 Total force = 0.050557 Total SCF correction = 0.000845 Total force = 0.023345 Total SCF correction = 0.000806 Total force = 0.023806 Total SCF correction = 0.000490 Total force = 0.012668 Total SCF correction = 0.000976 Total force = 0.018448 Total SCF correction = 0.000564 Total force = 0.007273 Total SCF correction = 0.000217 Total force = 0.009299 Total SCF correction = 0.000773 Total force = 0.020024 Total SCF correction = 0.000250 Total force = 0.020090 Total SCF correction = 0.000275 Total force = 0.019091 Total SCF correction = 0.000327 Total force = 0.015802 Total SCF correction = 0.000400 Total force = 0.015637 Total SCF correction = 0.000278 Total force = 0.016418 Total SCF correction = 0.000350 Total force = 0.015572 Total SCF correction = 0.000426 Total force = 0.012873 Total SCF correction = 0.001104 Total force = 0.010242 Total SCF correction = 0.000505 Total force = 0.011922 Total SCF correction = 0.000188 Total force = 0.010932 Total SCF correction = 0.000262 Total force = 0.009932 Total SCF correction = 0.000551 Total force = 0.011343 Total SCF correction = 0.000600 Total force = 0.014450 Total SCF correction = 0.000690 Total force = 0.014524 Total SCF correction = 0.000966 Total force = 0.011143 Total SCF correction = 0.000587 Total force = 0.009504 Total SCF correction = 0.000221 Total force = 0.011470 Total SCF correction = 0.000335 Total force = 0.014026 Total SCF correction = 0.000562 Total force = 0.014418 Total SCF correction = 0.000610 Total force = 0.014141 Total SCF correction = 0.000876 Total force = 0.014633 Total SCF correction = 0.000997 Total force = 0.015387 Total SCF correction = 0.001111 Total force = 0.015936 Total SCF correction = 0.001014 Total force = 0.015588 Total SCF correction = 0.000677 Total force = 0.014659 Total SCF correction = 0.000955 Total force = 0.013135 Total SCF correction = 0.000591 Total force = 0.011533 Total SCF correction = 0.000887 Total force = 0.011797 Total SCF correction = 0.000345 Total force = 0.011249 Total SCF correction = 0.000261 Total force = 0.009558 Total SCF correction = 0.000150 Total force = 0.009430 Total SCF correction = 0.000226 Total force = 0.009257 Total SCF correction = 0.000241 Total force = 0.010596 Total SCF correction = 0.000144 Total force = 0.011993 Total SCF correction = 0.000211 Total force = 0.012208 Total SCF correction = 0.000626 Total force = 0.011316 Total SCF correction = 0.000538 Total force = 0.012541 Total SCF correction = 0.000283 Total force = 0.013752 Total SCF correction = 0.000313 Total force = 0.014861 Total SCF correction = 0.000345 Total force = 0.014663 Total SCF correction = 0.000492 ! total energy = -311.65667063 Ry ! total energy = -311.53467416 Ry ! total energy = -311.65476738 Ry ! total energy = -311.65809859 Ry ! total energy = -311.65746738 Ry ! total energy = -311.65818908 Ry ! total energy = -311.65836015 Ry ! total energy = -311.65850034 Ry ! total energy = -311.65862191 Ry ! total energy = -311.65888214 Ry ! total energy = -311.65922406 Ry ! total energy = -311.65963798 Ry ! total energy = -311.66003012 Ry ! total energy = -311.66040611 Ry ! total energy = -311.66078796 Ry ! total energy = -311.66117923 Ry ! total energy = -311.66141965 Ry ! total energy = -311.66161990 Ry ! total energy = -311.66177274 Ry ! total energy = -311.66190991 Ry ! total energy = -311.66205510 Ry ! total energy = -311.66222069 Ry ! total energy = -311.66243963 Ry ! total energy = -311.66268794 Ry ! total energy = -311.66289592 Ry ! total energy = -311.66304908 Ry ! total energy = -311.66320915 Ry ! total energy = -311.66340482 Ry ! total energy = -311.66361063 Ry ! total energy = -311.66386806 Ry ! total energy = -311.66413998 Ry ! total energy = -311.66442184 Ry ! total energy = -311.66467976 Ry ! total energy = -311.66497799 Ry ! total energy = -311.66518668 Ry ! total energy = -311.66545797 Ry ! total energy = -311.66563977 Ry ! total energy = -311.66584671 Ry ! total energy = -311.66601473 Ry ! total energy = -311.66615866 Ry ! total energy = -311.66626437 Ry ! total energy = -311.66638311 Ry ! total energy = -311.66651248 Ry ! total energy = -311.66665949 Ry ! total energy = -311.66687447 Ry ! total energy = -311.66704465 Ry ! total energy = -311.66725797 Ry ! total energy = -311.66744622 Ry ! total energy = -311.66766939 Ry ! total energy = -311.66791148 Ry Another problem is related c_bands eigenvalue. I often get the notice: Davidson diagonalization with overlap c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged But It until runs well. I wonder the result for this calculation is believable or not. Best regards, ----------------------------------------------- Loc Duong Dinh Ms-Ph.D Student Sungkyunkwan Advanced Institute of Nanotechnology, Sungkyunkwan University, Suwon, 440-746, Korea Email: mambom1902 at yahoo.com From degironc at sissa.it Wed Jan 7 13:19:53 2009 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 07 Jan 2009 13:19:53 +0100 Subject: [Pw_forum] problem with vc-relax In-Reply-To: <415768.12230.qm@web38808.mail.mud.yahoo.com> References: <415768.12230.qm@web38808.mail.mud.yahoo.com> Message-ID: <49649DE9.6090904@sissa.it> loc duong ding wrote: > Dear all, > > I have a problem when I run vc-relax input file. > The force is increasing when the energy is decreasing. > I wonder this is the normal case or not. > this may indicate that you starter from (close to ) a saddle point and so you are moving away from a local maximum. as long as the energy decreases it should be ok. stefano > > Total force = 0.008423 Total SCF correction = 0.001451 > Total force = 0.250397 Total SCF correction = 0.001413 > Total force = 0.050557 Total SCF correction = 0.000845 > Total force = 0.023345 Total SCF correction = 0.000806 > Total force = 0.023806 Total SCF correction = 0.000490 > Total force = 0.012668 Total SCF correction = 0.000976 > Total force = 0.018448 Total SCF correction = 0.000564 > Total force = 0.007273 Total SCF correction = 0.000217 > Total force = 0.009299 Total SCF correction = 0.000773 > Total force = 0.020024 Total SCF correction = 0.000250 > Total force = 0.020090 Total SCF correction = 0.000275 > Total force = 0.019091 Total SCF correction = 0.000327 > Total force = 0.015802 Total SCF correction = 0.000400 > Total force = 0.015637 Total SCF correction = 0.000278 > Total force = 0.016418 Total SCF correction = 0.000350 > Total force = 0.015572 Total SCF correction = 0.000426 > Total force = 0.012873 Total SCF correction = 0.001104 > Total force = 0.010242 Total SCF correction = 0.000505 > Total force = 0.011922 Total SCF correction = 0.000188 > Total force = 0.010932 Total SCF correction = 0.000262 > Total force = 0.009932 Total SCF correction = 0.000551 > Total force = 0.011343 Total SCF correction = 0.000600 > Total force = 0.014450 Total SCF correction = 0.000690 > Total force = 0.014524 Total SCF correction = 0.000966 > Total force = 0.011143 Total SCF correction = 0.000587 > Total force = 0.009504 Total SCF correction = 0.000221 > Total force = 0.011470 Total SCF correction = 0.000335 > Total force = 0.014026 Total SCF correction = 0.000562 > Total force = 0.014418 Total SCF correction = 0.000610 > Total force = 0.014141 Total SCF correction = 0.000876 > Total force = 0.014633 Total SCF correction = 0.000997 > Total force = 0.015387 Total SCF correction = 0.001111 > Total force = 0.015936 Total SCF correction = 0.001014 > Total force = 0.015588 Total SCF correction = 0.000677 > Total force = 0.014659 Total SCF correction = 0.000955 > Total force = 0.013135 Total SCF correction = 0.000591 > Total force = 0.011533 Total SCF correction = 0.000887 > Total force = 0.011797 Total SCF correction = 0.000345 > Total force = 0.011249 Total SCF correction = 0.000261 > Total force = 0.009558 Total SCF correction = 0.000150 > Total force = 0.009430 Total SCF correction = 0.000226 > Total force = 0.009257 Total SCF correction = 0.000241 > Total force = 0.010596 Total SCF correction = 0.000144 > Total force = 0.011993 Total SCF correction = 0.000211 > Total force = 0.012208 Total SCF correction = 0.000626 > Total force = 0.011316 Total SCF correction = 0.000538 > Total force = 0.012541 Total SCF correction = 0.000283 > Total force = 0.013752 Total SCF correction = 0.000313 > Total force = 0.014861 Total SCF correction = 0.000345 > Total force = 0.014663 Total SCF correction = 0.000492 > > > ! total energy = -311.65667063 Ry > ! total energy = -311.53467416 Ry > ! total energy = -311.65476738 Ry > ! total energy = -311.65809859 Ry > ! total energy = -311.65746738 Ry > ! total energy = -311.65818908 Ry > ! total energy = -311.65836015 Ry > ! total energy = -311.65850034 Ry > ! total energy = -311.65862191 Ry > ! total energy = -311.65888214 Ry > ! total energy = -311.65922406 Ry > ! total energy = -311.65963798 Ry > ! total energy = -311.66003012 Ry > ! total energy = -311.66040611 Ry > ! total energy = -311.66078796 Ry > ! total energy = -311.66117923 Ry > ! total energy = -311.66141965 Ry > ! total energy = -311.66161990 Ry > ! total energy = -311.66177274 Ry > ! total energy = -311.66190991 Ry > ! total energy = -311.66205510 Ry > ! total energy = -311.66222069 Ry > ! total energy = -311.66243963 Ry > ! total energy = -311.66268794 Ry > ! total energy = -311.66289592 Ry > ! total energy = -311.66304908 Ry > ! total energy = -311.66320915 Ry > ! total energy = -311.66340482 Ry > ! total energy = -311.66361063 Ry > ! total energy = -311.66386806 Ry > ! total energy = -311.66413998 Ry > ! total energy = -311.66442184 Ry > ! total energy = -311.66467976 Ry > ! total energy = -311.66497799 Ry > ! total energy = -311.66518668 Ry > ! total energy = -311.66545797 Ry > ! total energy = -311.66563977 Ry > ! total energy = -311.66584671 Ry > ! total energy = -311.66601473 Ry > ! total energy = -311.66615866 Ry > ! total energy = -311.66626437 Ry > ! total energy = -311.66638311 Ry > ! total energy = -311.66651248 Ry > ! total energy = -311.66665949 Ry > ! total energy = -311.66687447 Ry > ! total energy = -311.66704465 Ry > ! total energy = -311.66725797 Ry > ! total energy = -311.66744622 Ry > ! total energy = -311.66766939 Ry > ! total energy = -311.66791148 Ry > > Another problem is related c_bands eigenvalue. I often get the notice: > > Davidson diagonalization with overlap > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > > > But It until runs well. I wonder the result for this calculation is believable or not. > > Best regards, > ----------------------------------------------- > Loc Duong Dinh > Ms-Ph.D Student > Sungkyunkwan Advanced Institute of Nanotechnology, > Sungkyunkwan University, > Suwon, 440-746, Korea > Email: mambom1902 at yahoo.com > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From baroni at sissa.it Wed Jan 7 13:24:21 2009 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 7 Jan 2009 13:24:21 +0100 Subject: [Pw_forum] problem with vc-relax In-Reply-To: <49649DE9.6090904@sissa.it> References: <415768.12230.qm@web38808.mail.mud.yahoo.com> <49649DE9.6090904@sissa.it> Message-ID: On Jan 7, 2009, at 1:19 PM, Stefano de Gironcoli wrote: > loc duong ding wrote: >> Dear all, >> >> I have a problem when I run vc-relax input file. >> The force is increasing when the energy is decreasing. >> I wonder this is the normal case or not. >> > this may indicate that you starter from (close to ) a saddle point and > so you are moving away from a local maximum. > as long as the energy decreases it should be ok. another more exotic possibility is that the energy gain due to a decreasing stress is larger than the increase due to an increasing force. I am not sure that the algorithm implemented guarantees that the individual components of the energy gradients decrease during along the minimization process. Stefano B > > stefano > >> >> Total force = 0.008423 Total SCF correction = >> 0.001451 >> Total force = 0.250397 Total SCF correction = >> 0.001413 >> Total force = 0.050557 Total SCF correction = >> 0.000845 >> Total force = 0.023345 Total SCF correction = >> 0.000806 >> Total force = 0.023806 Total SCF correction = >> 0.000490 >> Total force = 0.012668 Total SCF correction = >> 0.000976 >> Total force = 0.018448 Total SCF correction = >> 0.000564 >> Total force = 0.007273 Total SCF correction = >> 0.000217 >> Total force = 0.009299 Total SCF correction = >> 0.000773 >> Total force = 0.020024 Total SCF correction = >> 0.000250 >> Total force = 0.020090 Total SCF correction = >> 0.000275 >> Total force = 0.019091 Total SCF correction = >> 0.000327 >> Total force = 0.015802 Total SCF correction = >> 0.000400 >> Total force = 0.015637 Total SCF correction = >> 0.000278 >> Total force = 0.016418 Total SCF correction = >> 0.000350 >> Total force = 0.015572 Total SCF correction = >> 0.000426 >> Total force = 0.012873 Total SCF correction = >> 0.001104 >> Total force = 0.010242 Total SCF correction = >> 0.000505 >> Total force = 0.011922 Total SCF correction = >> 0.000188 >> Total force = 0.010932 Total SCF correction = >> 0.000262 >> Total force = 0.009932 Total SCF correction = >> 0.000551 >> Total force = 0.011343 Total SCF correction = >> 0.000600 >> Total force = 0.014450 Total SCF correction = >> 0.000690 >> Total force = 0.014524 Total SCF correction = >> 0.000966 >> Total force = 0.011143 Total SCF correction = >> 0.000587 >> Total force = 0.009504 Total SCF correction = >> 0.000221 >> Total force = 0.011470 Total SCF correction = >> 0.000335 >> Total force = 0.014026 Total SCF correction = >> 0.000562 >> Total force = 0.014418 Total SCF correction = >> 0.000610 >> Total force = 0.014141 Total SCF correction = >> 0.000876 >> Total force = 0.014633 Total SCF correction = >> 0.000997 >> Total force = 0.015387 Total SCF correction = >> 0.001111 >> Total force = 0.015936 Total SCF correction = >> 0.001014 >> Total force = 0.015588 Total SCF correction = >> 0.000677 >> Total force = 0.014659 Total SCF correction = >> 0.000955 >> Total force = 0.013135 Total SCF correction = >> 0.000591 >> Total force = 0.011533 Total SCF correction = >> 0.000887 >> Total force = 0.011797 Total SCF correction = >> 0.000345 >> Total force = 0.011249 Total SCF correction = >> 0.000261 >> Total force = 0.009558 Total SCF correction = >> 0.000150 >> Total force = 0.009430 Total SCF correction = >> 0.000226 >> Total force = 0.009257 Total SCF correction = >> 0.000241 >> Total force = 0.010596 Total SCF correction = >> 0.000144 >> Total force = 0.011993 Total SCF correction = >> 0.000211 >> Total force = 0.012208 Total SCF correction = >> 0.000626 >> Total force = 0.011316 Total SCF correction = >> 0.000538 >> Total force = 0.012541 Total SCF correction = >> 0.000283 >> Total force = 0.013752 Total SCF correction = >> 0.000313 >> Total force = 0.014861 Total SCF correction = >> 0.000345 >> Total force = 0.014663 Total SCF correction = >> 0.000492 >> >> >> ! total energy = -311.65667063 Ry >> ! total energy = -311.53467416 Ry >> ! total energy = -311.65476738 Ry >> ! total energy = -311.65809859 Ry >> ! total energy = -311.65746738 Ry >> ! total energy = -311.65818908 Ry >> ! total energy = -311.65836015 Ry >> ! total energy = -311.65850034 Ry >> ! total energy = -311.65862191 Ry >> ! total energy = -311.65888214 Ry >> ! total energy = -311.65922406 Ry >> ! total energy = -311.65963798 Ry >> ! total energy = -311.66003012 Ry >> ! total energy = -311.66040611 Ry >> ! total energy = -311.66078796 Ry >> ! total energy = -311.66117923 Ry >> ! total energy = -311.66141965 Ry >> ! total energy = -311.66161990 Ry >> ! total energy = -311.66177274 Ry >> ! total energy = -311.66190991 Ry >> ! total energy = -311.66205510 Ry >> ! total energy = -311.66222069 Ry >> ! total energy = -311.66243963 Ry >> ! total energy = -311.66268794 Ry >> ! total energy = -311.66289592 Ry >> ! total energy = -311.66304908 Ry >> ! total energy = -311.66320915 Ry >> ! total energy = -311.66340482 Ry >> ! total energy = -311.66361063 Ry >> ! total energy = -311.66386806 Ry >> ! total energy = -311.66413998 Ry >> ! total energy = -311.66442184 Ry >> ! total energy = -311.66467976 Ry >> ! total energy = -311.66497799 Ry >> ! total energy = -311.66518668 Ry >> ! total energy = -311.66545797 Ry >> ! total energy = -311.66563977 Ry >> ! total energy = -311.66584671 Ry >> ! total energy = -311.66601473 Ry >> ! total energy = -311.66615866 Ry >> ! total energy = -311.66626437 Ry >> ! total energy = -311.66638311 Ry >> ! total energy = -311.66651248 Ry >> ! total energy = -311.66665949 Ry >> ! total energy = -311.66687447 Ry >> ! total energy = -311.66704465 Ry >> ! total energy = -311.66725797 Ry >> ! total energy = -311.66744622 Ry >> ! total energy = -311.66766939 Ry >> ! total energy = -311.66791148 Ry >> >> Another problem is related c_bands eigenvalue. I often get the >> notice: >> >> Davidson diagonalization with overlap >> c_bands: 1 eigenvalues not converged >> c_bands: 1 eigenvalues not converged >> >> >> But It until runs well. I wonder the result for this calculation is >> believable or not. >> >> Best regards, >> ----------------------------------------------- >> Loc Duong Dinh >> Ms-Ph.D Student >> Sungkyunkwan Advanced Institute of Nanotechnology, >> Sungkyunkwan University, >> Suwon, 440-746, Korea >> Email: mambom1902 at yahoo.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 From panda.deng.pan at gmail.com Wed Jan 7 13:54:09 2009 From: panda.deng.pan at gmail.com (=?GB2312?B?xcu1xw==?=) Date: Wed, 7 Jan 2009 20:54:09 +0800 Subject: [Pw_forum] inconsistent DFT read Message-ID: Dear All, I have another problem here. the error message: from readpp : error # 2 inconsistent DFT read and input &control calculation = 'scf' restart_mode='from_scratch' prefix='BaNiAs', pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/', outdir='/disk2/xgwan/tmp/' / &system ibrav=7, celldm(1)=7.7702 , celldm(3)=2.806, nat=5, ntyp=3, ecutwfc =18.0 / &electrons mixing_beta = 0.7 conv_thr = 1.0d-8 / ATOMIC_SPECIES Ba 56.00 Ba.pw91-nsp-van.UPF Ni 28.00 NiUS.RRKJ3.UPF As 33.00 As.gon.UPF ATOMIC_POSITIONS Ba 0.0000 0.0000 0.0000 Ni 0.5000 0.0000 0.2500 Ni 0.0000 0.5000 0.2500 As 0.0000 0.0000 0.3476 As 0.0000 0.0000 0.6524 K_POINTS (automatic) 2 2 2 0 0 0 I do something stupig again.thanks for all the help. Have a nice day. Pan Deng Nanjing University From vormar at gmail.com Wed Jan 7 14:17:15 2009 From: vormar at gmail.com (Marci) Date: Wed, 7 Jan 2009 14:17:15 +0100 Subject: [Pw_forum] inconsistent DFT read In-Reply-To: References: Message-ID: <34ad5960901070517y3935c932x9384d7bfbad5e548@mail.gmail.com> Dear Pan, I think you simply didn't take care of your pseudopotentials. NiUS.RRKJ3.UPF and As.gon.UPF corresponds to a PZ (LDA) functional but Ba.pw91-nsp-van.UPF sets PW91 (GGA) as a functional. Also take care that you're mixing ultrasoft and norm-conserving pseudopotentials, so you might need a higher ecutwc and ecutrho. Yours, Marton ----------------------------------------------------- M\'arton V\"or\"os, physicist student Department of Atomic Physics Budapest University of Technology and Economics (BUTE) Budafoki ?t 8., H-1111, Budapest, Hungary vormar_at_gmail_dot_com, vm776_at_hszk_dot_bme_dot_hu ----------------------------------------------------- 2009/1/7 ?? : > Dear All, > > I have another problem here. > the error message: > > from readpp : error # 2 > inconsistent DFT read > > and input > > &control > calculation = 'scf' > restart_mode='from_scratch' > prefix='BaNiAs', > pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/', > outdir='/disk2/xgwan/tmp/' > / > &system > ibrav=7, celldm(1)=7.7702 , celldm(3)=2.806, nat=5, ntyp=3, > ecutwfc =18.0 > / > &electrons > mixing_beta = 0.7 > conv_thr = 1.0d-8 > / > ATOMIC_SPECIES > Ba 56.00 Ba.pw91-nsp-van.UPF > Ni 28.00 NiUS.RRKJ3.UPF > As 33.00 As.gon.UPF > > ATOMIC_POSITIONS > Ba 0.0000 0.0000 0.0000 > Ni 0.5000 0.0000 0.2500 > Ni 0.0000 0.5000 0.2500 > As 0.0000 0.0000 0.3476 > As 0.0000 0.0000 0.6524 > > K_POINTS (automatic) > > 2 2 2 0 0 0 > > > I do something stupig again.thanks for all the help. > > Have a nice day. > > Pan Deng > Nanjing University > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From degironc at sissa.it Wed Jan 7 14:18:43 2009 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 07 Jan 2009 14:18:43 +0100 Subject: [Pw_forum] inconsistent DFT read In-Reply-To: References: Message-ID: <4964ABB3.2030800@sissa.it> This means that the pseudopotentials used have been generated with different XC functionals. you can see which XC is used looking inside the pseudopotential file. stefano ?? wrote: > Dear All, > > I have another problem here. > the error message: > > from readpp : error # 2 > inconsistent DFT read > > and input > > &control > calculation = 'scf' > restart_mode='from_scratch' > prefix='BaNiAs', > pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/', > outdir='/disk2/xgwan/tmp/' > / > &system > ibrav=7, celldm(1)=7.7702 , celldm(3)=2.806, nat=5, ntyp=3, > ecutwfc =18.0 > / > &electrons > mixing_beta = 0.7 > conv_thr = 1.0d-8 > / > ATOMIC_SPECIES > Ba 56.00 Ba.pw91-nsp-van.UPF > Ni 28.00 NiUS.RRKJ3.UPF > As 33.00 As.gon.UPF > > ATOMIC_POSITIONS > Ba 0.0000 0.0000 0.0000 > Ni 0.5000 0.0000 0.2500 > Ni 0.0000 0.5000 0.2500 > As 0.0000 0.0000 0.3476 > As 0.0000 0.0000 0.6524 > > K_POINTS (automatic) > > 2 2 2 0 0 0 > > > I do something stupig again.thanks for all the help. > > Have a nice day. > > Pan Deng > Nanjing University > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From Giovanni.Cantele at na.infn.it Wed Jan 7 14:16:11 2009 From: Giovanni.Cantele at na.infn.it (Giovanni Cantele) Date: Wed, 07 Jan 2009 14:16:11 +0100 Subject: [Pw_forum] inconsistent DFT read In-Reply-To: References: Message-ID: <4964AB1B.4090405@na.infn.it> ?? wrote: > Dear All, > > I have another problem here. > the error message: > > from readpp : error # 2 > inconsistent DFT read > > and input > > &control > calculation = 'scf' > restart_mode='from_scratch' > prefix='BaNiAs', > pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/', > outdir='/disk2/xgwan/tmp/' > / > &system > ibrav=7, celldm(1)=7.7702 , celldm(3)=2.806, nat=5, ntyp=3, > ecutwfc =18.0 > / > &electrons > mixing_beta = 0.7 > conv_thr = 1.0d-8 > / > ATOMIC_SPECIES > Ba 56.00 Ba.pw91-nsp-van.UPF > Ni 28.00 NiUS.RRKJ3.UPF > As 33.00 As.gon.UPF > > ATOMIC_POSITIONS > Ba 0.0000 0.0000 0.0000 > Ni 0.5000 0.0000 0.2500 > Ni 0.0000 0.5000 0.2500 > As 0.0000 0.0000 0.3476 > As 0.0000 0.0000 0.6524 > > K_POINTS (automatic) > > 2 2 2 0 0 0 > > > I do something stupig again.thanks for all the help. > > Have a nice day. > > Pan Deng > Nanjing University > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > maybe, it could be helpful both using the QE Wiki and browsing the archive of this forum, where this kind of questions have been answered several times: http://www.quantum-espresso.org/wiki/index.php/Troubleshooting_(PWscf)#pw.x_stops_with_.27inconsistent_DFT.27_error http://www.democritos.it/pipermail/pw_forum/2004-September/001392.html http://www.democritos.it/pipermail/pw_forum/2005-February/002148.html .... giovanni -- Dr. Giovanni Cantele Coherentia CNR-INFM and Dipartimento di Scienze Fisiche Universita' di Napoli "Federico II" Complesso Universitario di Monte S. Angelo - Ed. 6 Via Cintia, I-80126, Napoli, Italy Phone: +39 081 676910 Fax: +39 081 676346 E-mail: giovanni.cantele at cnr.it giovanni.cantele at na.infn.it Web: http://people.na.infn.it/~cantele Research Group: http://www.nanomat.unina.it From mambom1902 at yahoo.com Wed Jan 7 14:28:22 2009 From: mambom1902 at yahoo.com (loc duong ding) Date: Wed, 7 Jan 2009 05:28:22 -0800 (PST) Subject: [Pw_forum] problem in vc-relax with c_bands Message-ID: <123073.74110.qm@web38806.mail.mud.yahoo.com> Thank you for all your replies. I have another problem that is related c_bands eigenvalue. I often get the notice: Davidson diagonalization with overlap c_bands: 1 eigenvalues not converged c_bands: 1 eigenvalues not converged But It until runs well. I wonder the result for this calculation is believable or not. Best regards, ----------------------------------------------- Loc Duong Dinh Ms-Ph.D Student Sungkyunkwan Advanced Institute of Nanotechnology, Sungkyunkwan University, Suwon, 440-746, Korea Email: mambom1902 at yahoo.com From lanhaiping at gmail.com Wed Jan 7 16:16:42 2009 From: lanhaiping at gmail.com (lan haiping) Date: Wed, 7 Jan 2009 23:16:42 +0800 Subject: [Pw_forum] problem in vc-relax with c_bands In-Reply-To: <123073.74110.qm@web38806.mail.mud.yahoo.com> References: <123073.74110.qm@web38806.mail.mud.yahoo.com> Message-ID: hi, it has been answered many times take a look at http://www.quantum-espresso.org/wiki/index.php/Troubleshooting_% 28PWscf%29 item 1.12 On Wed, Jan 7, 2009 at 9:28 PM, loc duong ding wrote: > Thank you for all your replies. > > > I have another problem that is related c_bands eigenvalue. I often get the > notice: > > Davidson diagonalization with overlap > c_bands: 1 eigenvalues not converged > c_bands: 1 eigenvalues not converged > > > But It until runs well. I wonder the result for this calculation is > believable > or not. > > Best regards, > ----------------------------------------------- > Loc Duong Dinh > Ms-Ph.D Student > Sungkyunkwan Advanced Institute of Nanotechnology, > Sungkyunkwan University, > Suwon, 440-746, Korea > Email: mambom1902 at yahoo.com > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090107/2207929f/attachment.htm From rezaebraahimi at yahoo.com Wed Jan 7 16:59:39 2009 From: rezaebraahimi at yahoo.com (Reza Ebraahimi) Date: Wed, 7 Jan 2009 07:59:39 -0800 (PST) Subject: [Pw_forum] 2D electron gas Message-ID: <439522.88224.qm@web59804.mail.ac4.yahoo.com> Hello everyone As i understand, the use of the different approximations for XC energy in the code is defined?by the kind of psedopotentials we use.. I want to know that if someone tries to study 2D systems, is it possible in the code?to generate PP's which the XC energy is calculated by the mean of 2D electron gas database (For example the work done by Tanatar & Ceperley, PRB 39, 5005(1989))? Thank You R.Ebraahimi -- graduate student Tehran University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090107/3ac05409/attachment.htm From baroni at sissa.it Wed Jan 7 17:12:58 2009 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 7 Jan 2009 17:12:58 +0100 Subject: [Pw_forum] 2D electron gas In-Reply-To: <439522.88224.qm@web59804.mail.ac4.yahoo.com> References: <439522.88224.qm@web59804.mail.ac4.yahoo.com> Message-ID: <9E478D96-38E0-4E67-9886-B7C8769FA1E6@sissa.it> Hi, R. Ebrahaami: I am puzzled by the question. You mean an atomic pseudpotential generated with a 2D XC energy? Real 2D systems are made by 3D atoms, which obviously require XC ata from 3D simulations. 2D models, of course, also exist (such as the 2D electron gas), but they do not require any pseudopotential. Or am I wrong? SB On Jan 7, 2009, at 4:59 PM, Reza Ebraahimi wrote: > Hello everyone > As i understand, the use of the different approximations for XC > energy in the code is defined by the kind of psedopotentials we use. > I want to know that if someone tries to study 2D systems, is it > possible in the code to generate PP's which the XC energy is > calculated by the mean of 2D electron gas database (For example the > work done by Tanatar & Ceperley, PRB 39, 5005(1989))? > Thank You > R.Ebraahimi -- graduate student > Tehran University > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste 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/20090107/49d2aef3/attachment.htm From akohlmey at cmm.chem.upenn.edu Wed Jan 7 17:34:28 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 7 Jan 2009 11:34:28 -0500 (EST) Subject: [Pw_forum] problem with vc-relax In-Reply-To: <415768.12230.qm@web38808.mail.mud.yahoo.com> References: <415768.12230.qm@web38808.mail.mud.yahoo.com> Message-ID: On Wed, 7 Jan 2009, loc duong ding wrote: LDD> Dear all, LDD> LDD> I have a problem when I run vc-relax input file. LDD> The force is increasing when the energy is decreasing. LDD> I wonder this is the normal case or not. in addition to what else was mentioned, you should check whether your stress tensor is well converged with respect to wavefunction and density cutoff. if this is not the case your system can behave very strange and you have factor in that your "effective" cutoff is increasing with a shrinking cell and decreasing with an expanding cell. [...] LDD> But It until runs well. I wonder the result for this calculation is LDD> believable or not. another general remark. the fact that a calculation completes only means that the computations could be completed without errors in the input and without numerical instabilities. it says very little about whether you calculation has physical meaning. this is something _you_ have to find out and that can require some rather indirect tests and checks. scientific codes - and quantum espresso is no exception - do not "know" physics but only apply the math that has been programmed. parameters and "symptoms" that may be acceptable for one calculation, can indicate big problems from some other. the best way is to start by trying to experiment with calculations where you know the result of and learning how to get it right and how different parameters affect the correctness of the result. cheers, axel. LDD> LDD> Best regards, LDD> ----------------------------------------------- LDD> Loc Duong Dinh LDD> Ms-Ph.D Student LDD> Sungkyunkwan Advanced Institute of Nanotechnology, LDD> Sungkyunkwan University, LDD> Suwon, 440-746, Korea LDD> Email: mambom1902 at yahoo.com LDD> LDD> LDD> LDD> LDD> _______________________________________________ LDD> Pw_forum mailing list LDD> Pw_forum at pwscf.org LDD> http://www.democritos.it/mailman/listinfo/pw_forum LDD> -- ======================================================================= 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 reboredofa at ornl.gov Wed Jan 7 17:33:19 2009 From: reboredofa at ornl.gov (Reboredo, Fernando A.) Date: Wed, 07 Jan 2009 11:33:19 -0500 Subject: [Pw_forum] 2D electron gas In-Reply-To: <439522.88224.qm@web59804.mail.ac4.yahoo.com> References: <439522.88224.qm@web59804.mail.ac4.yahoo.com> Message-ID: <3E4D6911DB2CA94DA31AC5E1CB522DBB03979DE7@ORNLEXCHANGE.ornl.gov> Dear Ebraahimi, I imagine that you want to use real atoms the system is not 2D but just quasi-2D. In that case one just uses the standard 3D approximations. In local and quasi-local approximations of the energy functional (LDA PBE) the long range dimensionality of the system is not taken into account. The calculations by Tanatar are for strictly two dimensional systems. Fernando Reboredo ________________________________ From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Reza Ebraahimi Sent: Wednesday, January 07, 2009 11:00 AM To: pw_forum at pwscf.org Subject: [Pw_forum] 2D electron gas Hello everyone As i understand, the use of the different approximations for XC energy in the code is defined by the kind of psedopotentials we use. I want to know that if someone tries to study 2D systems, is it possible in the code to generate PP's which the XC energy is calculated by the mean of 2D electron gas database (For example the work done by Tanatar & Ceperley, PRB 39, 5005(1989))? Thank You R.Ebraahimi -- graduate student Tehran University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090107/0abdb14b/attachment-0001.htm From vparkash at mtu.edu Wed Jan 7 23:57:26 2009 From: vparkash at mtu.edu (Vidur Parkash) Date: Wed, 7 Jan 2009 14:57:26 -0800 (PST) Subject: [Pw_forum] WFC file record format Message-ID: <25364.78850.qm@web35403.mail.mud.yahoo.com> Hello Everyone I need to extract wavefunctions dumped into the wfc file by espresso. Can anybody tell me how records in the wfc file are organized. A structure declaration from the source code would be useful. Vidur Parkash, MultiScale Technologies Institute Electrical & Computer Engineering Michigan Technological University 716 EERC, 1400 Townsend Drive Houghton, MI-49931, USA Tel: 1-906-487-0023 From akohlmey at cmm.chem.upenn.edu Thu Jan 8 00:33:27 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 7 Jan 2009 18:33:27 -0500 (EST) Subject: [Pw_forum] WFC file record format In-Reply-To: <25364.78850.qm@web35403.mail.mud.yahoo.com> References: <25364.78850.qm@web35403.mail.mud.yahoo.com> Message-ID: On Wed, 7 Jan 2009, Vidur Parkash wrote: VP> Hello Everyone VP> vidur, VP> I need to extract wavefunctions dumped into the wfc file by VP> espresso. Can anybody tell me how records in the wfc file are VP> organized. A structure declaration from the source code would be VP> useful. wouldn't it be easier to just add whatever manipulation/analysis you want to do to the already existing postprocessing software (directory PP)? this way you don't have to worry about the format on disk and avoid code duplication and having to play catchup whenever the format changes... cheers, axel. VP> VP> Vidur Parkash, VP> VP> MultiScale Technologies Institute VP> Electrical & Computer Engineering VP> Michigan Technological University VP> 716 EERC, 1400 Townsend Drive VP> Houghton, MI-49931, USA VP> Tel: 1-906-487-0023 VP> VP> _______________________________________________ VP> Pw_forum mailing list VP> Pw_forum at pwscf.org VP> http://www.democritos.it/mailman/listinfo/pw_forum VP> -- ======================================================================= 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 mambom1902 at yahoo.com Thu Jan 8 02:58:50 2009 From: mambom1902 at yahoo.com (loc duong ding) Date: Wed, 7 Jan 2009 17:58:50 -0800 (PST) Subject: [Pw_forum] problem in vc-relax with c_bands References: Message-ID: <656439.78770.qm@web38801.mail.mud.yahoo.com> Dear Ian Haiping, > From: "lan haiping" > Subject: Re: [Pw_forum] problem in vc-relax with c_bands > To: "PWSCF Forum" > Message-ID: > > Content-Type: text/plain; charset="iso-8859-1" > > hi, it has been answered many times > take a look at > > http://www.quantum-espresso.org/wiki/index.php/Troubleshooting_% > 28PWscf%29 > item 1.12 Thank you for your instructions. However, It very hard for me to go to the quantum-espresso web pages. I don't know this problem is caused of my computer setting or the website page. If possible, could you send me the content of this problem? I greatly appreciate your help. > > > On Wed, Jan 7, 2009 at 9:28 PM, loc duong ding wrote: > > > Thank you for all your replies. > > > > > > I have another problem that is related c_bands eigenvalue. I often get the > > notice: > > > > Davidson diagonalization with overlap > > c_bands: 1 eigenvalues not converged > > c_bands: 1 eigenvalues not converged > > > > > > But It until runs well. I wonder the result for this calculation is > > believable > > or not. > > > > Best regards, > > ----------------------------------------------- > > Loc Duong Dinh > > Ms-Ph.D Student > > Sungkyunkwan Advanced Institute of Nanotechnology, > > Sungkyunkwan University, > > Suwon, 440-746, Korea > > Email: mambom1902 at yahoo.com ----------------------------------------------- Loc Duong Dinh Ms-Ph.D Student Sungkyunkwan Advanced Institute of Nanotechnology, Sungkyunkwan University, Suwon, 440-746, Korea Email: mambom1902 at yahoo.com From mambom1902 at yahoo.com Thu Jan 8 03:04:12 2009 From: mambom1902 at yahoo.com (loc duong ding) Date: Wed, 7 Jan 2009 18:04:12 -0800 (PST) Subject: [Pw_forum] problem with vc-relax References: Message-ID: <177592.50503.qm@web38802.mail.mud.yahoo.com> Dear all, Could you introduce to me some references for cell relaxation? I think I don't have enough knowledge for such kind of calculation. Thanks for all. ----------------------------------------------- Loc Duong Dinh Ms-Ph.D Student Sungkyunkwan Advanced Institute of Nanotechnology, Sungkyunkwan University, Suwon, 440-746, Korea Email: mambom1902 at yahoo.com From giannozz at democritos.it Thu Jan 8 09:20:39 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 8 Jan 2009 09:20:39 +0100 Subject: [Pw_forum] problem in vc-relax with c_bands In-Reply-To: <656439.78770.qm@web38801.mail.mud.yahoo.com> References: <656439.78770.qm@web38801.mail.mud.yahoo.com> Message-ID: On Jan 8, 2009, at 2:58 , loc duong ding wrote: > Thank you for your instructions. However, It very hard for me to go > to the > quantum-espresso web pages. I don't know this problem is caused of my > computer setting or the website page. there is nothing wrong in the website page as far as I know > If possible, could you send me the content of this problem? > recent quantum-espresso releases should contain a copy of most of the contents of the web pages in Doc/ Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Thu Jan 8 09:25:02 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 8 Jan 2009 09:25:02 +0100 Subject: [Pw_forum] WFC file record format In-Reply-To: <25364.78850.qm@web35403.mail.mud.yahoo.com> References: <25364.78850.qm@web35403.mail.mud.yahoo.com> Message-ID: On Jan 7, 2009, at 23:57 , Vidur Parkash wrote: > I need to extract wavefunctions dumped into the wfc file by espresso. > Can anybody tell me how records in the wfc file are organized. http://www.quantum-espresso.org/wiki/index.php/ Developer_Manual#File_Formats My advice is to follow Axel's advice, though Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From vcuteym at 126.com Thu Jan 8 09:27:50 2009 From: vcuteym at 126.com (=?gbk?B?0e7Drw==?=) Date: Thu, 8 Jan 2009 16:27:50 +0800 (CST) Subject: [Pw_forum] How to configure "MPI_LIB=" in make.sys ? Message-ID: <10536982.621161231403270723.JavaMail.coremail@bj126app84.126.com> Hi, everyone I want to compile PWSCF in parallel using mpich. I do not konw what is needed in "MPI_LIB". I configure it "MPI_LIB=-L/..../mpich/lib -lmpi -lmpichf90 ". Is there any example to tell me about "MPI_LIB"? I use PGI compiler and AMD64 CPU. thanks myang Tel:?86-010-6275-7594 Email: vcuteym at 126.com School of Physics Peking University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090108/44d06663/attachment.htm From yukihiro_okuno at fujifilm.co.jp Thu Jan 8 11:08:04 2009 From: yukihiro_okuno at fujifilm.co.jp (yukihiro_okuno at fujifilm.co.jp) Date: Thu, 8 Jan 2009 19:08:04 +0900 Subject: [Pw_forum] Definition of qfunc in UPF file and the source code realus, f90 and init_us_1.f90 Message-ID: Dear PWSCF users and developer. I want to know the detail about UPF file. In the UPF format, in tab, the definition of qfunc is r**2 Q_nm(r) ? Or the value of upf(nt)%qfunc(ir,nm) in the source code is r**2 * Q_nm(r) ? In the manual of UPF , it is written that qfunc is q_{ij}(r) for r > rinner, but the source codes of PWSCF/PW/realus,f90 and init_us_1.f90, it seems that qfunc is consistent that it is defined by r**2 * Q_nm(r). And the values of qfunc in UPF file, is it also written the values of (r < rinner) Q^{rad}_nm(r) though it is not used when qfcoef is defined ? Sincerely, Yukihiro Okuno From giannozz at democritos.it Thu Jan 8 12:00:11 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 8 Jan 2009 12:00:11 +0100 Subject: [Pw_forum] How to configure "MPI_LIB=" in make.sys ? In-Reply-To: <10536982.621161231403270723.JavaMail.coremail@bj126app84.126.com> References: <10536982.621161231403270723.JavaMail.coremail@bj126app84.126.com> Message-ID: <78E645FB-40F4-477F-8A2D-EA5BCC4E5E7F@democritos.it> On Jan 8, 2009, at 9:27 , ?? wrote: > I want to compile PWSCF in parallel using mpich. > I do not konw what is needed in "MPI_LIB". use "configure" with no optiobs first. If it doesn't find the MPI libraries, they are in some nonstandard place, in which case, YOU have to know where they are. > I configure it "MPI_LIB=-L/..../mpich/lib -lmpi -lmpichf90 " Maybe. -L/directory/where/mpich/libraries/are -lxyz will find /directory/where/mpich/libraries/are/libxyz.a (or .so if it is a shared library) Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From greatnaol at gmail.com Thu Jan 8 13:45:00 2009 From: greatnaol at gmail.com (Naol Regassa) Date: Thu, 8 Jan 2009 15:45:00 +0300 Subject: [Pw_forum] calculating U Message-ID: <97e4b88e0901080445m5ada4eads39620b9f9e5562b0@mail.gmail.com> Dear all I was calculating Hubbard U for strongly correlated perovskite like SrMnO3. I can run scf and scf+u, perturbed for different Hubbard_alpha. My question is 1- Do we check the convergence of alpha? If yes with what parameter? 2- How do we find the value of U? What are the important parameter there(in the perturbed outputs)? 3 - Is it needed to construct super cells so that we can check the convergence with them? great thanks for your response Thesis student From penghua8503 at 163.com Thu Jan 8 14:32:37 2009 From: penghua8503 at 163.com (penghua) Date: Thu, 8 Jan 2009 21:32:37 +0800 (CST) Subject: [Pw_forum] electron-phonon coefficients Message-ID: <16047701.244461231421557750.JavaMail.coremail@app160.163.com> Dear all, When I use lambda.x to calculate electron-phonon coefficients in example07, the value of g and wk should be given. q, wk : q-points and weights But, how can I get the two values, using kpoints.x or finding them in some files, like q2r.out, which have got? When I running the lambda.x in the example07, I got the fowling information forrtl: severe (29): file not found, unit 4, file /espresso-3.2.3/examples/example07/results/fort.4 Thanks in advance. Hua Peng ------------------------------------------------------------------- Hua Peng Penghua8503 at 163.com ShandongUniversity, Jinan, Shandong Province, China, 250100 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090108/6a2bae42/attachment.htm From matteo at umn.edu Thu Jan 8 16:22:49 2009 From: matteo at umn.edu (Matteo Cococcioni) Date: Thu, 08 Jan 2009 09:22:49 -0600 Subject: [Pw_forum] calculating U In-Reply-To: <97e4b88e0901080445m5ada4eads39620b9f9e5562b0@mail.gmail.com> References: <97e4b88e0901080445m5ada4eads39620b9f9e5562b0@mail.gmail.com> Message-ID: <49661A49.4060600@umn.edu> Dear Naol. you can find many details about this in this paper: PRB 71, 035105 (2005). I will try to answer to your questions below. Naol Regassa wrote: > Dear all > > I was calculating Hubbard U for strongly correlated perovskite like > SrMnO3. I can run scf and scf+u, perturbed for different > Hubbard_alpha. My question is > those with alpha are just "fake" lda+U calculations: lda_plus_u needs to be set to .true. (so that the code will write occupations on the output file) but U is very small (practically 0; normally a value of 1.d-10 eV is chosen) because we usually perturb LDA or GGA ground states. > 1- Do we check the convergence of alpha? If yes with what parameter? > I'm not sure what you mean by convergence of alpha. It's a good habit to make sure that the total occupation of each site has a linear dependence on alpha around alpha=0. > 2- How do we find the value of U? What are the important parameter > there(in the perturbed outputs)? > > This requires constructing response matrices and a little bit of processing of the results you obtain from the code. U is not something you directly read from any ourput file of the code. > 3 - Is it needed to construct super cells so that we can check the > convergence with them? > > Yes. again please check the reference above. keep in mind that in the perturbed run diago_thr_init needs to be set equal to the threshold of the iterative (davidson) diagonalization of the last electronic iteration of the unperturbed run. see the example input attached for details. > great thanks for your response > Thesis student > would you please add your affiliation? regards, Matteo > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: run_Fe_ucalc_example Url: http://www.democritos.it/pipermail/pw_forum/attachments/20090108/7b4b867f/attachment-0001.txt -------------- 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/20090108/7b4b867f/attachment-0001.vcf From akohlmey at cmm.chem.upenn.edu Thu Jan 8 16:48:17 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 8 Jan 2009 10:48:17 -0500 (EST) Subject: [Pw_forum] problem with vc-relax In-Reply-To: <177592.50503.qm@web38802.mail.mud.yahoo.com> References: <177592.50503.qm@web38802.mail.mud.yahoo.com> Message-ID: On Wed, 7 Jan 2009, loc duong ding wrote: LDD> Dear all, LDD> LDD> Could you introduce to me some references for cell relaxation? I LDD> think I don't have enough knowledge for such kind of calculation. how about reading some of the recommended literature? and _studying the available tutorial material. check out http://www.quantum-espresso.org/learn.php axel LDD> Thanks for all. LDD> LDD> ----------------------------------------------- LDD> Loc Duong Dinh LDD> Ms-Ph.D Student LDD> Sungkyunkwan Advanced Institute of Nanotechnology, LDD> Sungkyunkwan University, LDD> Suwon, 440-746, Korea LDD> Email: mambom1902 at yahoo.com LDD> LDD> LDD> LDD> _______________________________________________ LDD> Pw_forum mailing list LDD> Pw_forum at pwscf.org LDD> http://www.democritos.it/mailman/listinfo/pw_forum LDD> -- ======================================================================= 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 Thu Jan 8 17:37:24 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 8 Jan 2009 17:37:24 +0100 Subject: [Pw_forum] Definition of qfunc in UPF file and the source code realus, f90 and init_us_1.f90 In-Reply-To: References: Message-ID: On Jan 8, 2009, at 11:08 , yukihiro_okuno at fujifilm.co.jp wrote: > I want to know the detail about UPF file. the "official" (so to speak) document describing the UPF file format is here: http://www.quantum-espresso.org/wiki/index.php/ Unified_pseudopotential_format > In the UPF format, in tab, the definition of qfunc is > r**2 Q_nm(r) ? > Or the value of upf(nt)%qfunc(ir,nm) in the source code is r**2 * > Q_nm(r) ? > In the manual of UPF , it is written that qfunc is q_{ij}(r) for r > > rinner, > but the source codes of PWSCF/PW/realus,f90 and init_us_1.f90, it > seems > that qfunc is consistent that it is defined by r**2 * Q_nm(r). > And the values of qfunc in UPF file, is it also written the > values of (r< rinner) > Q^{rad}_nm(r) though it is not used when qfcoef is defined ? This requires some investigation P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From ptanhphys at gmail.com Thu Jan 8 21:30:51 2009 From: ptanhphys at gmail.com (TuanAnh Pham) Date: Thu, 8 Jan 2009 12:30:51 -0800 Subject: [Pw_forum] Dielectric constant with Lyddane-Sachs-Teller relation Message-ID: Dear Pwscf users, I have a question about calculation of the static dielectric constant. In the phonon code the high frequency dielectric constant is given and in order to find \epsilon_0 we need to add the ionic contribution. The ionic contribution can be defined as in this link of Quantum-espresso: http://www.quantum-espresso.org/wiki/index.php/Dielectric_constant as well as in several papers (Gonze et al, PRB 55, 10355 (1997)): \epsilon_0 - \epsilon_{\infty} = 4*pi*\sum_{m}{\frac{S_m}{\omega^2_m}} [1] in which omega_m is the phonon frequency and S_m is the mode-oscillator strength. On another hand, \epsilon_0 can also be calculated by Lyddane-Sachs-Teller (LST) relation using LO-TO splitting provided in the code: \frac{\epsilon_0}{\epsilon_{\infty}} = \prod{\frac{\omega^2_LO}{\omega^2_TO}} [2] Could you please let me know what is the different between these two methods ? Is that correct that the LST relation only works for the system with sufficient symmetry? Thank you very much and best regards, Tuan Anh. -- Tuan Anh Pham Graduate Student University of California, Davis Phone: 530-752-0957 Homepage: http://angstrom.ucdavis.edu/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090108/e3ae9c18/attachment.htm From jasius_1 at yahoo.com Fri Jan 9 01:21:40 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Thu, 8 Jan 2009 16:21:40 -0800 (PST) Subject: [Pw_forum] wrong path to the pseudo folder In-Reply-To: <78E645FB-40F4-477F-8A2D-EA5BCC4E5E7F@democritos.it> Message-ID: <670497.34738.qm@web54406.mail.yahoo.com> After reading the manual carefully I failed to find a cause for this error. I specified pseudo_dir and an erro is: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from readpp : error # 1 file /home/jbaltrus/espresso/pseudo/O.pz-rrkjus.UPF not found %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% My input echo script points to the non-existent pseudo folder in $HOME folder, however it's not there Cannot match namelist object name ?/share/apps/espresso-4.0.4/pseudo?, Cannot match namelist object name ?/home/jbaltrus/atom-lsda? &control calculation='scf', pseudo_dir=?/share/apps/espresso-4.0.4/pseudo?, outdir=?/home/jbaltrus/atom-lsda? / &system ibrav=1, celldm(1)=10.0, nat=1, ntyp=1, nbnd=6, ecutwfc=25.0, ecutrho=200.0, occupations='from_input', nspin=2 / &electrons mixing_beta=0.25, / ATOMIC_SPECIES O 15.99994 O.pz-rrkjus.UPF ATOMIC_POSITIONS O 0.000000000 0.000000000 0.000000000 K_POINTS {gamma} OCCUPATIONS 1.0 1.0000000000 1.0000000000 1.0000000000 0.0 0.0 1.0 0.3333333333 0.3333333333 0.3333333333 0.0 0.0 Jonas Baltrusaitis University of Iowa From akohlmey at cmm.chem.upenn.edu Fri Jan 9 01:38:43 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 8 Jan 2009 19:38:43 -0500 (EST) Subject: [Pw_forum] wrong path to the pseudo folder In-Reply-To: <670497.34738.qm@web54406.mail.yahoo.com> References: <670497.34738.qm@web54406.mail.yahoo.com> Message-ID: On Thu, 8 Jan 2009, Jonas Baltrusaitis wrote: jonas, JB> After reading the manual carefully I failed to find a cause for this JB> error. I specified pseudo_dir and an erro is: JB> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% JB> from readpp : error # 1 JB> file /home/jbaltrus/espresso/pseudo/O.pz-rrkjus.UPF not found JB> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% JB> My input echo script points to the non-existent pseudo folder in JB> $HOME folder, however it's not there JB> JB> Cannot match namelist object name ??/share/apps/espresso-4.0.4/pseudo??, JB> JB> Cannot match namelist object name ??/home/jbaltrus/atom-lsda?? these two lines are suspicious. obviously the namelist parser could not parse the two strings. please check whether the single quotation marks are proper ASCII characters. particularly when transferring files from windows machines or but also when using "smart" editors that use non-ASCII compatible encodings, bad things can happen. the code falls back to pseudo_dir="$HOME/espresso/pseudo" if pseudo_dir is not set. see Modules/read_namelists.f90 for details. cheers, axel. JB> JB> JB> &control JB> calculation='scf', JB> pseudo_dir=??/share/apps/espresso-4.0.4/pseudo??, JB> outdir=??/home/jbaltrus/atom-lsda?? JB> / JB> &system JB> ibrav=1, JB> celldm(1)=10.0, JB> nat=1, JB> ntyp=1, JB> nbnd=6, JB> ecutwfc=25.0, JB> ecutrho=200.0, JB> occupations='from_input', JB> nspin=2 JB> / JB> &electrons JB> mixing_beta=0.25, JB> / JB> ATOMIC_SPECIES JB> O 15.99994 O.pz-rrkjus.UPF JB> ATOMIC_POSITIONS JB> O 0.000000000 0.000000000 0.000000000 JB> K_POINTS {gamma} JB> OCCUPATIONS JB> 1.0 1.0000000000 1.0000000000 1.0000000000 0.0 0.0 JB> 1.0 0.3333333333 0.3333333333 0.3333333333 0.0 0.0 JB> JB> Jonas Baltrusaitis JB> University of Iowa JB> JB> JB> JB> JB> _______________________________________________ JB> Pw_forum mailing list JB> Pw_forum at pwscf.org JB> http://www.democritos.it/mailman/listinfo/pw_forum JB> -- ======================================================================= 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 yukihiro_okuno at fujifilm.co.jp Fri Jan 9 01:51:31 2009 From: yukihiro_okuno at fujifilm.co.jp (yukihiro_okuno at fujifilm.co.jp) Date: Fri, 9 Jan 2009 09:51:31 +0900 Subject: [Pw_forum] Definition of qfunc in UPF file and the source code realus, f90 and init_us_1.f90 Message-ID: Dear?Prof. Giannozzi. Thank you for your reply. >the "official" (so to speak) document describing the UPF file format >is here: http://www.quantum-espresso.org/wiki/index.php/ Unified_pseudopotential_format I have see this page and it is simply written that qfunc = q_{ij}(r) But, in the source code (realus.f90, init_us_1.f90) , it seems that qfunc is treated as r**2 * q_{ij}(r). I want to make atomic charge density in each orbital chi_{l}(r) which is USPP pseudo atomic orbital from UPF pseudo potential file, I need the correct definition of UPF. Is it possible to constract atomic orbital charge by qfunc in the region r < rinner instead of construct the pseudo polynominal form of pseudo Q_{ij} from qfcoef ? Sincerely. Yukihiro Okuno. From paulatto at sissa.it Fri Jan 9 09:31:05 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Fri, 9 Jan 2009 09:31:05 +0100 (CET) Subject: [Pw_forum] wrong path to the pseudo folder In-Reply-To: References: <670497.34738.qm@web54406.mail.yahoo.com> Message-ID: <1286.147.122.3.89.1231489865.squirrel@webmail.sissa.it> On Ven, Gennaio 9, 2009 01:38, Axel Kohlmeyer wrote: > please check whether the single > quotation marks are proper ASCII characters. Furthermore, the line-termination character could be DOS-style, or mac-style; this is often a source of troubles. If you are on a linux machine you usually have the dos2unix command you can use to fix this problem 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 Fri Jan 9 13:51:53 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 9 Jan 2009 13:51:53 +0100 Subject: [Pw_forum] Definition of qfunc in UPF file and the source code realus, f90 and init_us_1.f90 In-Reply-To: References: Message-ID: <4CC42C5B-B97B-4CA7-95E7-4F47167E9F1E@democritos.it> On Jan 9, 2009, at 1:51 , yukihiro_okuno at fujifilm.co.jp wrote: > it seems that qfunc is treated as r**2 * q_{ij}(r). they are. For the Q(r), we used the same convention as in the ultrasoft pseudopotential generation code by David Vanderbilt. Thank you for remarking this. > Is it possible to constract atomic orbital charge by qfunc in the > region r < rinner instead of construct the pseudo polynominal > form of pseudo Q_{ij} from qfcoef ? The Q(r) should be good also for r Dear Forum, I have been trying to calculate the energy of a CO crystal using the coordinates given in cst-www.nrl.navy.mul/lattice/struk.xmol/b21.pos. This has 14 CO molecules. I used the format shown in example03 with the proper changes for the number of atoms and crystal structure. Unfortunately the cal- culation does not converge. Increasing the number of iterations and ecutwfc and ecutrho improved things a little but the convergence is still about 10^-4 Ry. I am particularly interested in the interactions between molecules so this is not satis- factory. Is there anything I can do to improve the calculation to reach convergence to about 10^-7 Ry? Incidentally the calculated triple bond energy for a single CO molecule is about 43 Ry while the literature value given in texts is about 54 Ry. Is this to be expected? Hopefully someone has ideas about how to improve the convergence. Amos Leffler -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090109/95c3d712/attachment.htm From akohlmey at cmm.chem.upenn.edu Fri Jan 9 21:13:41 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 9 Jan 2009 15:13:41 -0500 (EST) Subject: [Pw_forum] Energy of CO Crystal In-Reply-To: <61b264610901091033n64865b9hbff420f824ed4fe3@mail.gmail.com> References: <61b264610901091033n64865b9hbff420f824ed4fe3@mail.gmail.com> Message-ID: On Fri, 9 Jan 2009, Amos Leffler wrote: AL> Dear Forum, dear amos, AL> I have been trying to calculate the energy of a CO crystal AL> using the coordinates given in AL> cst-www.nrl.navy.mul/lattice/struk.xmol/b21.pos. This has 14 CO AL> molecules. I used the format shown in example03 with the proper no it doesn't! there are 4 CO molecules per unit cell. one in the corner and 3 in the faces. AL> changes for the number of atoms and crystal structure. AL> Unfortunately the cal- culation does not converge. Increasing the converge for self-consistency or for geometry? AL> number of iterations and ecutwfc and ecutrho improved things a AL> little but the convergence is still about 10^-4 Ry. I am AL> particularly interested in the interactions between molecules so AL> this is not satis- factory. Is there anything I can do to improve AL> the calculation to reach convergence to about 10^-7 Ry? impossible to say without seeing your input file. i just did a quick test and it converges rather easily. AL> Incidentally the calculated triple bond energy for a AL> single CO molecule is about 43 Ry while the literature value given AL> in texts is about 54 Ry. Is this to be expected? how did you determine this 43Ry value (or the 54Ry)? this would be equivalent to 27000 (or 33900) kcal/mol. hardly realistic. cheers, axel. AL> Hopefully someone has ideas about how to improve the convergence. AL> Amos AL> Leffler AL> -- ======================================================================= 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 thmason24 at yahoo.com Sat Jan 10 03:46:37 2009 From: thmason24 at yahoo.com (Timothy Mason) Date: Fri, 9 Jan 2009 18:46:37 -0800 (PST) Subject: [Pw_forum] problem with vc-relax In-Reply-To: <49131688.3040102@sissa.it> Message-ID: <866595.30786.qm@web81801.mail.mud.yahoo.com> I see this all the time in my vc-relax attempts. I've wondered if its actually necessary that the gradients always decrease? If the potential is not quite quadratic around the minimum then i suppose there is no reason to assume that the forces would aways be decreasing. Tim Mason university of missouri St. Louis From baroni at sissa.it Sat Jan 10 07:19:41 2009 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 10 Jan 2009 07:19:41 +0100 Subject: [Pw_forum] problem with vc-relax In-Reply-To: <866595.30786.qm@web81801.mail.mud.yahoo.com> References: <866595.30786.qm@web81801.mail.mud.yahoo.com> Message-ID: <2A21EB2C-2F3B-4C9A-B104-8CC8BD1A2350@sissa.it> You are right, Tim. One can easily imagine non-quadratic functions such that the gradients are smaller far from the minimum that near it. A stupid example is a double well close to the maximum. Stefano On Jan 10, 2009, at 3:46 AM, Timothy Mason wrote: > > I see this all the time in my vc-relax attempts. I've wondered if > its actually necessary that the gradients always decrease? > > If the potential is not quite quadratic around the minimum then i > suppose there is no reason to assume that the forces would aways be > decreasing. > > Tim Mason > university of missouri St. Louis > _______________________________________________ > 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/20090110/0dc359bd/attachment.htm From sclauzer at sissa.it Sat Jan 10 15:17:45 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Sat, 10 Jan 2009 15:17:45 +0100 Subject: [Pw_forum] Different TRC will be obtained for the same system? In-Reply-To: <000e01c96903$f4075770$3f7c620a@fduphy778548b0> References: <000e01c96903$f4075770$3f7c620a@fduphy778548b0> Message-ID: <4968AE09.80307@sissa.it> Dear Wang, Wang Qinjing wrote: > Dear all: > > I am a new PWSCF user and I was confused by two things. > > One thing is different TRC (Transimission coefficient) at the same > energy point will be obtained for the same system. After get the wave > functions by the scf calculation, I tried to get the TRC. First I just > calculate the TRC at the fermi energy, i.e., the inputfile cond.in for > the pwcond.x fiis as follows: > > ========= > energy0=0.0d0, > denergy=0.0d0, > / > 1 > 0.0 0.0 1.0 > 1 > 0.0 > ========= > in this way the TRC at fermi energy is about 0.106 > > But if I calculate the TRCs for different energies, i.e., the input file > for the pwcond.x is as: > > ======== > energy0=2.0d0, > denergy=-0.02d0, > / > 1 > 0.0 0.0 1.0 > 201 > ======== > in this way a series of TRC is obtained and the TRC at fermi energy is > 5*10^-3. There must be something wrong. Is it a bug? or related to the > complie problem? None of the two. In fact the code performs a basis set reduction of the G_perp vectors based on two parameters: ewind and epsproj (see also Doc/INPUT_PWCOND.txt). The first one selects an energy window starting from the bottom of the band structure up to energy0+ewind that will be used to construct the basis set. As you can notice, the upper limit of this window depends on energy0, i.e. the higher scattering energy you want to compute. Since this basis set can change as you change energy0, also transmission coefficients can. Of course you want to avoid this, and in order to do this you have to check convergence of your complex band structure w.r.t. both ewind and epsproj before computing conductance. I think that the default values are not enough to achieve convergence: reasonable values may be ewind ~ 5 and epsproj ~ 10^-5, but the accuracy and the efficiency in the reduction may depend quite a lot on the details of the sistem and of the calculation (so please check on your specific system!). If you want to use the full basis, you can specify ewind=1.d2 or higher (in this case the basis reduction will not be used). Hope this helps, GS > > Thank you in advance. > > Yours Wang > > > ------------------------------------ > ?????????????? > Wang Qinjing, Physics Department, > Fudan University > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From jasius_1 at yahoo.com Sat Jan 10 17:53:31 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Sat, 10 Jan 2009 08:53:31 -0800 (PST) Subject: [Pw_forum] vdW-df functional Message-ID: <783753.99691.qm@web54405.mail.yahoo.com> What keywords do I use to treat the system vdW-df functional? thanks JOnas Baltrusaitis University of Iowa From akohlmey at cmm.chem.upenn.edu Sat Jan 10 18:00:32 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 10 Jan 2009 12:00:32 -0500 (EST) Subject: [Pw_forum] vdW-df functional In-Reply-To: <783753.99691.qm@web54405.mail.yahoo.com> References: <783753.99691.qm@web54405.mail.yahoo.com> Message-ID: On Sat, 10 Jan 2009, Jonas Baltrusaitis wrote: JB> What keywords do I use to treat the system vdW-df functional? do you mean DFT with empirical van-der-waals correction? axel. JB> JB> thanks JB> JB> JOnas Baltrusaitis JB> University of Iowa JB> JB> JB> JB> JB> _______________________________________________ JB> Pw_forum mailing list JB> Pw_forum at pwscf.org JB> http://www.democritos.it/mailman/listinfo/pw_forum JB> -- ======================================================================= 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 jasius_1 at yahoo.com Sat Jan 10 22:00:13 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Sat, 10 Jan 2009 13:00:13 -0800 (PST) Subject: [Pw_forum] vdW-df functional In-Reply-To: Message-ID: <755183.70661.qm@web54405.mail.yahoo.com> yes, that's what I mean. come to think of it I couldn't find in manual any keywords to specify any functional. what's the best source for quantum espresso keywords? do you by any chance know if pwgui can import coordinates from .cif? thanks Jonas --- On Sat, 1/10/09, Axel Kohlmeyer wrote: > From: Axel Kohlmeyer > Subject: Re: [Pw_forum] vdW-df functional > To: "Jonas Baltrusaitis" > Cc: "PWSCF Forum" > Date: Saturday, January 10, 2009, 9:00 AM > On Sat, 10 Jan 2009, Jonas Baltrusaitis wrote: > > JB> What keywords do I use to treat the system vdW-df > functional? > > do you mean DFT with empirical van-der-waals correction? > > axel. > > JB> > JB> thanks > JB> > JB> JOnas Baltrusaitis > JB> University of Iowa > JB> > JB> > JB> > JB> > JB> _______________________________________________ > JB> Pw_forum mailing list > JB> Pw_forum at pwscf.org > JB> http://www.democritos.it/mailman/listinfo/pw_forum > JB> > > -- > ======================================================================= > 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 Sat Jan 10 22:12:04 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Sat, 10 Jan 2009 22:12:04 +0100 Subject: [Pw_forum] vdW-df functional In-Reply-To: <755183.70661.qm@web54405.mail.yahoo.com> References: <755183.70661.qm@web54405.mail.yahoo.com> Message-ID: <1DD51C87-8816-4E30-86BC-99AD591CE8E8@democritos.it> On Jan 10, 2009, at 22:00 , Jonas Baltrusaitis wrote: > yes, that's what I mean. empirical vdW corrections used to be implemented in some version of FPMD once upon a time, but they have slowly faded away into oblivion, I think. The code is apparently still there (see espresso/CPV/vanderwaals.f90), and it shouldn't be difficult to put it back to work, though. More recently, somebody claimed that he was interested in implementing it, but I never heard any news since one year or so. True (nonlocal) vdW is a hot topic; there is some very preliminary implementation, I just heard at a conference an algorithmic improvement that may open the way to serious calculations, there is a lot of interest in its implementation, but it is not going to materialize very soon > what's the best source for quantum espresso keywords? for functionals? espresso/Modules/functionals.f90 > do you by any chance know if pwgui can import coordinates from .cif? from where? :-) Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From eyvaz_isaev at yahoo.com Sun Jan 11 00:46:36 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sat, 10 Jan 2009 15:46:36 -0800 (PST) Subject: [Pw_forum] electron-phonon coefficients In-Reply-To: <16047701.244461231421557750.JavaMail.coremail@app160.163.com> Message-ID: <50188.79341.qm@web65702.mail.ac4.yahoo.com> Dear Penghua, > forrtl: severe (29): file not found, unit 4, file > /espresso-3.2.3/examples/example07/results/fort.4 lambda.x reads a file where you should list elph-files (with q-points before, and additional information in the first line). Most likely, it could not find these files, as they are in a different directory or they have names different those you specified. q-points and their weights you can find using kpoints.x and answering the questions it posts. A *.ph.out file (*.dyn0 as well) does not the weights, but it is a good idea to compare q-points you got by means of kpoints.x and that contained in *.dyn0 file. 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 Thu, 1/8/09, penghua wrote: > From: penghua > Subject: [Pw_forum] electron-phonon coefficients > To: "pw_forum at pwscf.org" > Date: Thursday, January 8, 2009, 4:32 PM > Dear all, > > When I use lambda.x to calculate electron-phonon > coefficients in example07, the value of g and wk should be > given. > > q, wk : q-points and weights > > But, how can I get the two values, using kpoints.x or > finding them in some files, like q2r.out, which have got? > > When I running the lambda.x in the example07, I got the > fowling information > > forrtl: severe (29): file not found, unit 4, file > /espresso-3.2.3/examples/example07/results/fort.4 > > > > Thanks in advance. > > Hua Peng > > ------------------------------------------------------------------- > > Hua Peng Penghua8503 at 163.com > > ShandongUniversity, > > Jinan, Shandong Province, China, 250100 > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From naromero at gmail.com Sun Jan 11 03:36:34 2009 From: naromero at gmail.com (Nichols A. Romero) Date: Sat, 10 Jan 2009 20:36:34 -0600 Subject: [Pw_forum] vdW-df functional In-Reply-To: <1DD51C87-8816-4E30-86BC-99AD591CE8E8@democritos.it> References: <755183.70661.qm@web54405.mail.yahoo.com> <1DD51C87-8816-4E30-86BC-99AD591CE8E8@democritos.it> Message-ID: <6ac064b60901101836p3afb07c9yacd633ae25f98e9a@mail.gmail.com> There is nonlocal vdW DFT in GPAW https://wiki.fysik.dtu.dk/gpaw/ It is also available in a fully self-consistent. Don't know how well it is tested. On Sat, Jan 10, 2009 at 3:12 PM, Paolo Giannozzi wrote: > > On Jan 10, 2009, at 22:00 , Jonas Baltrusaitis wrote: > > > yes, that's what I mean. > > empirical vdW corrections used to be implemented in some version > of FPMD once upon a time, but they have slowly faded away into > oblivion, I think. The code is apparently still there (see > espresso/CPV/vanderwaals.f90), and it shouldn't be difficult to put > it back to work, though. More recently, somebody claimed that he > was interested in implementing it, but I never heard any news since > one year or so. > > True (nonlocal) vdW is a hot topic; there is some very preliminary > implementation, I just heard at a conference an algorithmic improvement > that may open the way to serious calculations, there is a lot of > interest > in its implementation, but it is not going to materialize very soon > > > what's the best source for quantum espresso keywords? > > for functionals? espresso/Modules/functionals.f90 > > > do you by any chance know if pwgui can import coordinates from .cif? > > from where? :-) > > 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 > -- Nichols A. Romero, Ph.D. Argonne Leadership Computing Facility Argonne, IL 60490 (630) 252-3441 (O) (630) 470-0462 (C) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090110/add11980/attachment.htm From jasius_1 at yahoo.com Sun Jan 11 04:12:09 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Sat, 10 Jan 2009 19:12:09 -0800 (PST) Subject: [Pw_forum] input for a polymer crystal In-Reply-To: <6ac064b60901101836p3afb07c9yacd633ae25f98e9a@mail.gmail.com> Message-ID: <858684.26099.qm@web54410.mail.yahoo.com> I am trying to optimize a hydrogen bonded polymer crystal structure. I have experimental a,b,c and all three angles of the cell, as well as the point group. Also fractional coordinates what I don't have is an idea how to prepare an input. I read Doc/Input_pw closely and couldn't find how to utilize any of those (except atomic positions). CELL_PARAMETERS should accommodate those somwehow but it's only 3x3 matrix input in the manual instead. also, specifying ibrav, celldm, nat, ntyp and nbnd are not clear as it's nothing like a simple Al or Si structure. also, what do I put for K_POINTS? Gamma only? thanks for the help Jonas Baltrusaitis University of Iowa From akohlmey at cmm.chem.upenn.edu Sun Jan 11 05:12:49 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 10 Jan 2009 23:12:49 -0500 Subject: [Pw_forum] input for a polymer crystal In-Reply-To: <858684.26099.qm@web54410.mail.yahoo.com> References: <6ac064b60901101836p3afb07c9yacd633ae25f98e9a@mail.gmail.com> <858684.26099.qm@web54410.mail.yahoo.com> Message-ID: <7b6913e90901102012g784a951p2b97f60c3de0dd6@mail.gmail.com> On Sat, Jan 10, 2009 at 10:12 PM, Jonas Baltrusaitis wrote: > I am trying to optimize a hydrogen bonded polymer crystal structure. I have experimental a,b,c and all three angles of the cell, as well as the point group. Also fractional coordinates > > what I don't have is an idea how to prepare an input. I read Doc/Input_pw closely and couldn't find how to utilize any of those (except atomic positions). > > CELL_PARAMETERS should accommodate those somwehow but it's only 3x3 matrix input in the manual instead. also, specifying ibrav, celldm, nat, ntyp and nbnd are not clear as it's nothing like a simple Al or Si structure. also, what do I put for K_POINTS? Gamma only? if you have all you this information, ibrav, celldm, nat, ntyp are all defined. CELL_PARAMETERS is only needed if you set ibrav=0. nbnd depends on your pseudopotentials and whether you use smearing or not. probably the default set by pw.x is sufficient and you can probably get away with fixed occupations and gamma, but that just needs to be tested. there is a wealth of material on how to set up proper jobs with pw.x and other codes, also discussing convergence issues (which cannot be predefined) at: http://www.quantum-espresso.org/wiki/index.php/Tutorials_and_Examples perhaps you shoud study this more closely. if all these rather elementary parameters tell you nothing, than you should work on your understanding of the methods you use _before_ you use them for a production calculation... cheers, axel. > > thanks for the help > > Jonas Baltrusaitis > University of Iowa > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From penghua8503 at 163.com Sun Jan 11 07:58:08 2009 From: penghua8503 at 163.com (penghua) Date: Sun, 11 Jan 2009 14:58:08 +0800 (CST) Subject: [Pw_forum] electron-phonon coefficients In-Reply-To: <50188.79341.qm@web65702.mail.ac4.yahoo.com> References: <50188.79341.qm@web65702.mail.ac4.yahoo.com> Message-ID: <3436250.77761231657088853.JavaMail.coremail@bj163app21.163.com> Dear Prof. Eyvaz Isaev Thank you for your replay. Since elph file is needed in the calculation of lambda, I rerunning ph.x, adding filelph='elphmatrx', I didn?t find any elph matrix file in directory. I don?t know the reason, how can I get the files? ------------------------------------------ Hua Peng Penghua8503 at 163.com ShandongUniversity , Jinan, Shandong Province, China, 250100 2009-01-11?"Eyvaz Isaev" ? >Dear Penghua, > >> forrtl: severe (29): file not found, unit 4, file >> /espresso-3.2.3/examples/example07/results/fort.4 > >lambda.x reads a file where you should list elph-files (with q-points before, and additional information in the first line). Most likely, it could not find these files, as they are in a different directory or they have names different those you specified. > >q-points and their weights you can find using kpoints.x and answering the questions it posts. A *.ph.out file (*.dyn0 as well) does not the weights, but it is a good idea to compare q-points you got by means of kpoints.x and that contained in *.dyn0 file. > >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 > > >_______________________________________________ >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/20090111/d68f8c3e/attachment.htm From eyvaz_isaev at yahoo.com Sun Jan 11 16:27:52 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sun, 11 Jan 2009 07:27:52 -0800 (PST) Subject: [Pw_forum] electron-phonon coefficients In-Reply-To: <3436250.77761231657088853.JavaMail.coremail@bj163app21.163.com> Message-ID: <970969.60624.qm@web65705.mail.ac4.yahoo.com> Well, I see. You need a revised elphon.90 to get these files. I will send you these files (including those for "artificial parallel" calculations) later, hopefully today. Bests, Eyvaz. --- On Sun, 1/11/09, penghua wrote: > From: penghua > Subject: Re: [Pw_forum] electron-phonon coefficients > To: eyvaz_isaev at yahoo.com, "PWSCF Forum" > Date: Sunday, January 11, 2009, 9:58 AM > Dear Prof. Eyvaz Isaev > > Thank you for your replay. Since elph file is needed in the > calculation of lambda, I rerunning ph.x, adding > filelph='elphmatrx', I didn?t find any elph matrix > file in directory. I don?t know the reason, how can I get > the files? > > ------------------------------------------ > > Hua Peng Penghua8503 at 163.com > > ShandongUniversity > > , > > Jinan, Shandong Province, China, 250100 > > > > 2009-01-11?"Eyvaz Isaev" > ? > >Dear Penghua, > > > >> forrtl: severe (29): file not found, unit 4, file > >> /espresso-3.2.3/examples/example07/results/fort.4 > > > >lambda.x reads a file where you should list elph-files > (with q-points before, and additional information in the > first line). Most likely, it could not find these files, as > they are in a different directory or they have names > different those you specified. > > > >q-points and their weights you can find using kpoints.x > and answering the questions it posts. A *.ph.out file > (*.dyn0 as well) does not the weights, but it is a good idea > to compare q-points you got by means of kpoints.x and that > contained in *.dyn0 file. > > > >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 > > > > > >_______________________________________________ > >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 chaohao.mse at gmail.com Mon Jan 12 02:06:29 2009 From: chaohao.mse at gmail.com (C.H. Hu) Date: Mon, 12 Jan 2009 09:06:29 +0800 Subject: [Pw_forum] electron-phonon coefficients In-Reply-To: <970969.60624.qm@web65705.mail.ac4.yahoo.com> References: <3436250.77761231657088853.JavaMail.coremail@bj163app21.163.com> <970969.60624.qm@web65705.mail.ac4.yahoo.com> Message-ID: <6502ee650901111706q67c3941dt20c8f44d93a274f3@mail.gmail.com> Dear Prof. Eyvaz Isaev, Can you also send me this revised file? It is helpful to me since I am also investigating the electron-phonon interaction in a hydride system using PWSCF. Thanks in advance for your help. With my best regards, C.H. On Sun, Jan 11, 2009 at 11:27 PM, Eyvaz Isaev wrote: > Well, I see. You need a revised elphon.90 to get these files. I will send you these files (including those for "artificial parallel" calculations) later, hopefully today. > > Bests, > Eyvaz. > > --- On Sun, 1/11/09, penghua wrote: > >> From: penghua >> Subject: Re: [Pw_forum] electron-phonon coefficients >> To: eyvaz_isaev at yahoo.com, "PWSCF Forum" >> Date: Sunday, January 11, 2009, 9:58 AM >> Dear Prof. Eyvaz Isaev >> >> Thank you for your replay. Since elph file is needed in the >> calculation of lambda, I rerunning ph.x, adding >> filelph='elphmatrx', I didn't find any elph matrix >> file in directory. I don't know the reason, how can I get >> the files? >> >> ------------------------------------------ >> >> Hua Peng Penghua8503 at 163.com >> >> ShandongUniversity >> >> , >> >> Jinan, Shandong Province, China, 250100 >> >> >> >> 2009-01-11?"Eyvaz Isaev" >> ? >> >Dear Penghua, >> > >> >> forrtl: severe (29): file not found, unit 4, file >> >> /espresso-3.2.3/examples/example07/results/fort.4 >> > >> >lambda.x reads a file where you should list elph-files >> (with q-points before, and additional information in the >> first line). Most likely, it could not find these files, as >> they are in a different directory or they have names >> different those you specified. >> > >> >q-points and their weights you can find using kpoints.x >> and answering the questions it posts. A *.ph.out file >> (*.dyn0 as well) does not the weights, but it is a good idea >> to compare q-points you got by means of kpoints.x and that >> contained in *.dyn0 file. >> > >> >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 >> > >> > >> >_______________________________________________ >> >Pw_forum mailing list >> >Pw_forum at pwscf.org >> >http://www.democritos.it/mailman/listinfo/pw_forum >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From lzh-ming at 163.com Mon Jan 12 04:42:33 2009 From: lzh-ming at 163.com (lzh-ming) Date: Mon, 12 Jan 2009 11:42:33 +0800 (CST) Subject: [Pw_forum] cell_dofree = volume not yet implemented Message-ID: <6412755.495371231731753225.JavaMail.coremail@bj163app113.163.com> Hello everyone, I have tried to run pw.x to make a relax. I set &ions / &cell cell_dofree='volume', / I got from metric_setup : error # 1 cell_dofree = volume not yet implemented So, what's the problem? How can I implement volumn? Thank you. li-ming-fan ----------------------- Student for MoS, Institute of Theoretical Physics, Lanzhou University, Gansu Province, China, -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090112/979da1a7/attachment.htm From padmaja_patnaik at yahoo.co.uk Mon Jan 12 05:56:50 2009 From: padmaja_patnaik at yahoo.co.uk (Padmaja Patnaik) Date: Mon, 12 Jan 2009 04:56:50 +0000 (GMT) Subject: [Pw_forum] Doubt in occupations Message-ID: <770135.63659.qm@web25405.mail.ukl.yahoo.com> Dear all I have a doubt in the 'occupations'. As mentioned in the INPUT_PW document the occupations = tetrahedra for DOS calculations. Is it compulsory for all materials? Occupations = tetrahedra is not accepting K_POINTs = Gamma. If i want to do Gamma point calculations then can i change the occupations argument? Say for my case its an semiconductor so can i give 'occupations= fixed' in the DOS calculations and take K_POINTS as Gamma? Please suggest. Regards and thanks, Padmaja Patnaik Research Scholar Dept of Physics IIT Bombay Mumbai, India -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090112/896f217c/attachment.htm From giannozz at democritos.it Mon Jan 12 09:04:03 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 12 Jan 2009 09:04:03 +0100 Subject: [Pw_forum] cell_dofree = volume not yet implemented In-Reply-To: <6412755.495371231731753225.JavaMail.coremail@bj163app113.163.com> References: <6412755.495371231731753225.JavaMail.coremail@bj163app113.163.com> Message-ID: <496AF973.6020502@democritos.it> lzh-ming wrote: > cell_dofree = volume not yet implemented > So, what's the problem? the problem is, I think, that keeping the volume fixed requires a slightly different algorithm from all other cases. Unless somebody volunteers to implement it, that case should not be mentioned in the documentation (unfortunately it is) Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Mon Jan 12 09:27:21 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 12 Jan 2009 09:27:21 +0100 Subject: [Pw_forum] Doubt in occupations In-Reply-To: <770135.63659.qm@web25405.mail.ukl.yahoo.com> References: <770135.63659.qm@web25405.mail.ukl.yahoo.com> Message-ID: <496AFEE9.5030403@democritos.it> Padmaja Patnaik wrote: > I have a doubt in the 'occupations'. As mentioned in the INPUT_PW > document the occupations = tetrahedra for DOS calculations. Is it > compulsory for all materials? the tetrahedron method is compulsory if you want a nice DOS that looks like those in textbooks; if not, you can use a gaussian broadening Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From wangjunjie1981_0 at 163.com Mon Jan 12 10:00:20 2009 From: wangjunjie1981_0 at 163.com (wangjunjie1981_0) Date: Mon, 12 Jan 2009 17:00:20 +0800 (CST) Subject: [Pw_forum] a question of el-ph calculation Message-ID: <29165382.140941231750820241.JavaMail.coremail@bj163app89.163.com> Dear pwscf users, When i calculate el-ph interaction, a question happens. I want to know what are the reasons that output file (name.elph.out) do not wirte completely, and not give error information. Because of calculated need, i increased the npk in Module/parameter.f90, and recomplied the program. The normal calculations of scf, nscf, and nscf2 are done, but elph.out is not given completely. I want to know if larger k points in the part of nscf2 results in the output file wirting completely. Best regards, Junjie Wang -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090112/539e5c66/attachment.htm From sclauzer at sissa.it Mon Jan 12 10:15:35 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Mon, 12 Jan 2009 10:15:35 +0100 Subject: [Pw_forum] K_POINTS, tpiba or crystal In-Reply-To: <01a001c96dfc$5a97be10$0fc73a30$@net> References: <000501c969e4$eb03bb00$c10b3100$@net> <20081229212836.hut32tqv40csg404@mail.democritos.it> <005901c96a3b$0c9de960$25d9bc20$@net> <200812301420.47223.giannozz@democritos.it> <01a001c96dfc$5a97be10$0fc73a30$@net> Message-ID: <496B0A37.6020604@sissa.it> Dear Paul, I add just a small note that maybe will help you further in checking k-points coordinates. If you put verbosity='high' in the control namelist, the program will show you on output the set of k-points coordinates used for that calculation expressed in both 'tpiba' and 'crystal' reference frames, so that you can better check the difference (if any, depending on the crystal itself). GS Paul M. Grant wrote: > Followup on K_POINTS: tpiba vs crystal... > > As usual, Paolo is right...for sc lattices (ibrav=1), tpiba or crystal give > equal nscf outputs. > > But, for fcc lattices (ibrav=2), it apparently does! See below comparing > nscf k-vector outputs for tpiba and crystal, and crystal is very different > (and weird...for me anyway). BTW, whether tpiba, etc., should be enclosed > in curly braces or parentheses or nothing on the K_POINTS card is confusing. > PW_INPUT shows braces, but parenthese work too, and probably no delimiters > as well. > > My agenda here is to generate k-points grids amenable to applications that > cannot use simply MP weighted points in the irreducible BZ, such as > epsilon.x. I'm finding the algorithm used by Eyvaz Isaev for Fermi Surface > plotting which uniformly populate and span the fundamental reciprocal > lattice cell works quite well...if inputted with tpiba or "blank," not > crystal. > > > TPIBA VS CRYSTAL COMPARISON RE FCC FOLLOWS > =================================================== -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From lzh-ming at 163.com Mon Jan 12 11:06:43 2009 From: lzh-ming at 163.com (lzh-ming) Date: Mon, 12 Jan 2009 18:06:43 +0800 (CST) Subject: [Pw_forum] cell_dofree = volume not yet implemented In-Reply-To: <496AF973.6020502@democritos.it> References: <496AF973.6020502@democritos.it> <6412755.495371231731753225.JavaMail.coremail@bj163app113.163.com> Message-ID: <7999089.578591231754803527.JavaMail.coremail@bj163app19.163.com> Dear Paolo Giannozzi, I was running a relax for simple cubic crystal. I want to find the best lattice parametre a_0. In the file INPUT_PW.html in the directory Doc, the annotation for cell_dofree says 'CHARACTER','volume = the cell is simply rescaled, without changing the shape'. So this is right what I want. But I got the problem. And how can I do if I want to relax the crystal for that purpose? Thanks! li-ming-fan ?2009-01-12?"Paolo Giannozzi" ??? >lzh-ming wrote: > >> cell_dofree = volume not yet implemented >> So, what's the problem? > >the problem is, I think, that keeping the volume fixed >requires a slightly different algorithm from all other >cases. Unless somebody volunteers to implement it, >that case should not be mentioned in the documentation >(unfortunately it is) > >Paolo >-- >Paolo Giannozzi, Democritos and University of Udine, Italy -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090112/94d45e8b/attachment.htm From giannozz at democritos.it Mon Jan 12 11:13:50 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 12 Jan 2009 11:13:50 +0100 Subject: [Pw_forum] cell_dofree = volume not yet implemented In-Reply-To: <7999089.578591231754803527.JavaMail.coremail@bj163app19.163.com> References: <496AF973.6020502@democritos.it> <6412755.495371231731753225.JavaMail.coremail@bj163app113.163.com> <7999089.578591231754803527.JavaMail.coremail@bj163app19.163.com> Message-ID: <496B17DE.7010001@democritos.it> lzh-ming wrote: > In the file INPUT_PW.html in the directory Doc, > the annotation for cell_dofree says > 'CHARACTER','volume = the cell is simply rescaled, without changing the > shape'. > So this is right what I want. But I got the problem. > And how can I do if I want to relax the crystal for that purpose? either you implement it or you do it the good old way, by calculating the energy-vs-volume curve at different volumes Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Mon Jan 12 11:34:45 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Mon, 12 Jan 2009 11:34:45 +0100 Subject: [Pw_forum] a question of el-ph calculation In-Reply-To: <29165382.140941231750820241.JavaMail.coremail@bj163app89.163.com> References: <29165382.140941231750820241.JavaMail.coremail@bj163app89.163.com> Message-ID: <496B1CC5.5050400@democritos.it> wangjunjie1981_0 wrote: > I want to know if larger k points in the part of nscf2 results > in the output file wirting completely. if I understand what you mean, you do not get any output from the electron-phonon calculation. It is impossible to say why on the basis of the information you provide. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From J.Kendrick at Bradford.ac.uk Mon Jan 12 12:41:46 2009 From: J.Kendrick at Bradford.ac.uk (John Kendrick) Date: Mon, 12 Jan 2009 11:41:46 +0000 Subject: [Pw_forum] Projection, convergence and imaginary phonon frequencies Message-ID: <496B2C7A.30206@bradford.ac.uk> Hi, I have been doing some calculations recently on molecular crystals. Because of our interest in the lowest phonon frequencies I have been concerned that translational invariance is maintained as far as possible. To this end I have increased the accuracy of the optimisation and the scf convergence, to a point where I cannot see how to converge the calculation any tighter. I noticed that the norm conserving pseudopotentials were having greater difficulties than the others. In particular for one molecule I have seen significant differences between the different projection schemes that ensure translational invariance. Particularly the intensities in the infrared a very different. Below is some output from a representative calculation. Crystal Projection # mode [cm-1] [THz] IR Raman depol 1 0.00 0.0000 0.0000 0.0011 0.7500 2 0.00 0.0000 0.0000 0.0043 0.7500 3 0.00 0.0000 0.0000 0.0047 0.7500 4 22.00 0.6596 0.0000 6.7060 0.7500 5 56.93 1.7067 0.0033 0.1326 0.7500 6 56.93 1.7067 0.0033 0.1326 0.7500 7 73.28 2.1969 0.0000 9.8083 0.6361 8 82.51 2.4737 0.0000 0.0000 0.7422 9 83.26 2.4960 0.2348 0.0334 0.7500 10 83.26 2.4960 0.2348 0.0334 0.7500 NOASR No Projection # mode [cm-1] [THz] IR Raman depol 1 -52.77 -1.5820 0.0016 0.7753 0.7500 2 22.00 0.6596 0.0000 6.7060 0.7500 3 44.67 1.3392 0.0147 0.4140 0.7500 4 44.67 1.3392 0.0147 0.4140 0.7500 5 72.20 2.1646 0.0211 0.0034 0.7500 6 72.20 2.1646 0.0211 0.0034 0.7500 7 73.28 2.1969 0.0000 9.8083 0.6361 8 82.51 2.4737 0.0000 0.0000 0.7425 9 83.79 2.5120 0.0000 0.1142 0.7500 10 84.69 2.5390 0.2889 0.6525 0.7500 SIMPLE Projection # mode [cm-1] [THz] IR Raman depol 1 -1.31 -0.0394 0.0000 0.0120 0.7500 2 -1.31 -0.0394 0.0000 0.0120 0.7500 3 -0.92 -0.0277 0.0000 0.0010 0.7500 4 39.67 1.1893 0.0001 0.4896 0.7500 5 39.67 1.1893 0.0001 0.4896 0.7500 6 51.30 1.5380 0.0000 2.9853 0.7120 7 60.39 1.8104 0.0000 6.5707 0.7500 8 61.02 1.8293 0.0040 2.8566 0.7500 9 61.02 1.8293 0.0040 2.8566 0.7500 10 61.94 1.8570 0.0000 0.0000 0.7499 11 68.41 2.0507 0.0000 7.7313 0.7500 12 76.98 2.3078 0.1318 0.0930 0.7500 13 76.98 2.3078 0.1318 0.0930 0.7500 Should I be looking at the energy cutoff - 80Ryd? the kspace integration - 1 1 2? The unit cell is approximately 9X9X6 angstrom Other convergence thresholds? There maybe some meshes used for integration that could still help improve the calculation or perhaps an approach to further tightening the optimisation. Or this maybe as good as it gets? I was wondering if anyone had any experience they could offer. Input to the optimisation step looks like this. &CONTROL restart_mode = 'restart', calculation = 'relax' , outdir = '.' , pseudo_dir = '/home/jkendric/Data/PseudoPotentials' , disk_io = 'minimal' , nstep = 200, etot_conv_thr = 1.0D-6, forc_conv_thr = 1.0D-5, / &SYSTEM ibrav = 0, celldm(1) = 1.8897262, nat = 58, ntyp = 4, ecutwfc = 80.D0 , occupations = 'fixed' , / &ELECTRONS conv_thr = 1.D-11 , mixing_beta = 0.3D0 , / &IONS ion_dynamics = 'bfgs' pot_extrapolation = 'second_order' , wfc_extrapolation = 'second_order' , / CELL_PARAMETERS cubic 9.3027 0.0 0.0 0.0 9.3027 0.0 0.0 0.0 6.6403 ATOMIC_SPECIES C 12.01070 C.blyp-mt.UPF N 14.00674 N.blyp-mt.UPF O 15.99940 O.blyp-mt.UPF H 1.00794 H.blyp-vbc.UPF ATOMIC_POSITIONS crystal C 0.500000000 0.500000000 0.000000000 C 0.623684469 0.560707616 0.129462381 H 0.717493110 0.584367933 0.037673510 H 0.590256782 0.657115895 0.209561631 ..... K_POINTS automatic 1 1 2 0 0 0 Energies in the output file look like this ! total energy = -1055.84030152 Ry ! total energy = -1055.83327033 Ry ! total energy = -1055.84030145 Ry ! total energy = -1055.84030152 Ry ! total energy = -1055.84030153 Ry ! total energy = -1055.84030153 Ry ! total energy = -1055.84030154 Ry ! total energy = -1055.84030154 Ry ! total energy = -1055.84030154 Ry Total force = 0.000129 Total SCF correction = 0.000006 Input to the phonon calculations looks like this Phonons at Gamma Point with raman &inputph tr2_ph=1.0d-18, epsil=.true., lraman=.true., outdir='.', fildyn='petn.dynG', / 0.0 0.0 0.0 Thanks. John Kendrick University of Bradford Bradford BD13 1DP UK From akohlmey at cmm.chem.upenn.edu Mon Jan 12 15:44:00 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 12 Jan 2009 09:44:00 -0500 (EST) Subject: [Pw_forum] cell_dofree = volume not yet implemented In-Reply-To: <6412755.495371231731753225.JavaMail.coremail@bj163app113.163.com> References: <6412755.495371231731753225.JavaMail.coremail@bj163app113.163.com> Message-ID: On Mon, 12 Jan 2009, lzh-ming wrote: LMF> Hello everyone, hello li-ming-fang, LMF> I have tried to run pw.x to make a relax. LMF> I set LMF> &ions LMF> / LMF> &cell LMF> cell_dofree='volume', LMF> / LMF> I got LMF> from metric_setup : error # 1 LMF> cell_dofree = volume not yet implemented LMF> LMF> So, what's the problem? the code to do this is obviously missing and somebody has actually been so nice to put a not to this extent there. LMF> How can I implement volumn? you can take your favorate text editor and write the code to do it. it is as simple as that. this is how packages like Q-E come together: people add to it what they need. ;) cheers, axel. LMF> Thank you. LMF> LMF> li-ming-fan LMF> ----------------------- LMF> Student for MoS, LMF> Institute of Theoretical Physics, LMF> Lanzhou University, Gansu Province, LMF> China, -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From marzari at MIT.EDU Mon Jan 12 16:09:52 2009 From: marzari at MIT.EDU (Nicola Marzari) Date: Mon, 12 Jan 2009 18:09:52 +0300 Subject: [Pw_forum] Projection, convergence and imaginary phonon frequencies In-Reply-To: <496B2C7A.30206@bradford.ac.uk> References: <496B2C7A.30206@bradford.ac.uk> Message-ID: <496B5D40.6010907@mit.edu> Dear John, if it is a crystal, the most critical parameter should be k-point sampling. Increase that, maybe using shifted meshes of the form 2n 2n 3n 1 1 1 , with n=1,2,3... Norm-conserving and ultrasoft should matter little (unless there is a problem with one pseudo), while GGAs should be a bit tougher to converge than LDA. Do "crystal" and "simple" really lead to a different number of modes (10 vs 13) ?? nicola John Kendrick wrote: > Hi, > > I have been doing some calculations recently on molecular crystals. > Because of our interest in the lowest phonon frequencies I have been > concerned that translational invariance is maintained as far as > possible. To this end I have increased the accuracy of the optimisation > and the scf convergence, to a point where I cannot see how to converge > the calculation any tighter. I noticed that the norm conserving > pseudopotentials were having greater difficulties than the others. In > particular for one molecule I have seen significant differences between > the different projection schemes that ensure translational invariance. > Particularly the intensities in the infrared a very different. Below is > some output from a representative calculation. --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From akohlmey at cmm.chem.upenn.edu Mon Jan 12 16:54:18 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 12 Jan 2009 10:54:18 -0500 (EST) Subject: [Pw_forum] Projection, convergence and imaginary phonon frequencies In-Reply-To: <496B5D40.6010907@mit.edu> References: <496B2C7A.30206@bradford.ac.uk> <496B5D40.6010907@mit.edu> Message-ID: On Mon, 12 Jan 2009, Nicola Marzari wrote: NM> Dear John, dear nicola, happy new year from cold philly. NM> if it is a crystal, the most critical parameter should be k-point NM> sampling. Increase that, maybe using shifted meshes of the form NM> 2n 2n 3n 1 1 1 , with n=1,2,3... but it is a _molecular_ crystal, so i guess there are some areas with comparatively low electron density and quite weak interactions, as seen by the long list of rather soft modes. i would suspect that will need to be converged very well wrt to the basis set to not see any ripples on the potential surface. for TM oxygen pseudos i would go to 120ry or higher to get tight convergence. for MD 80-85ry is usually sufficient (the more traditional 70ry has been shown to be very sloppy), but if you want a tight geometry convergence you will have to be more careful. BFGS can be quite sensitive and will essentially refuse to converge if you have "ripples" due to a very flat potential surface unless you crank up the cutoff significantly. NM> Norm-conserving and ultrasoft should matter little (unless NM> there is a problem with one pseudo), while GGAs should be a bit NM> tougher to converge than LDA. NM> NM> Do "crystal" and "simple" really lead to a different number of NM> modes (10 vs 13) ?? those are obviously only the lowest few modes. it is a 9x9x6 angstrom cell! cheers, axel. NM> NM> nicola NM> NM> NM> John Kendrick wrote: NM> > Hi, NM> > NM> > I have been doing some calculations recently on molecular crystals. NM> > Because of our interest in the lowest phonon frequencies I have been NM> > concerned that translational invariance is maintained as far as NM> > possible. To this end I have increased the accuracy of the optimisation NM> > and the scf convergence, to a point where I cannot see how to converge NM> > the calculation any tighter. I noticed that the norm conserving NM> > pseudopotentials were having greater difficulties than the others. In NM> > particular for one molecule I have seen significant differences between NM> > the different projection schemes that ensure translational invariance. NM> > Particularly the intensities in the infrared a very different. Below is NM> > some output from a representative calculation. NM> NM> --------------------------------------------------------------------- NM> Prof Nicola Marzari Department of Materials Science and Engineering NM> 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA NM> tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu NM> _______________________________________________ NM> Pw_forum mailing list NM> Pw_forum at pwscf.org NM> http://www.democritos.it/mailman/listinfo/pw_forum NM> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From J.Kendrick at Bradford.ac.uk Mon Jan 12 17:31:29 2009 From: J.Kendrick at Bradford.ac.uk (John Kendrick) Date: Mon, 12 Jan 2009 16:31:29 +0000 Subject: [Pw_forum] Projection, convergence and imaginary phonon frequencies In-Reply-To: Message-ID: <496B7061.2070104@bradford.ac.uk> Hi, Yes Happy New Year. I only gave the first few soft modes for each case. It is these soft modes I am interested in. There are 174 modes (58 atoms) altogether. I listed 13 in one case as there was more of a match with the intensities from the other projection methods. The comment about cutoff is interesting as I had found that the optimiser (BFGS) was having problems. This would also be consistent with getting more consistent results with the USP potentials. Thanks John Axel Kohlmeyer wrote: > On Mon, 12 Jan 2009, Nicola Marzari wrote: > > NM> Dear John, > > dear nicola, > > happy new year from cold philly. > > NM> if it is a crystal, the most critical parameter should be k-point > NM> sampling. Increase that, maybe using shifted meshes of the form > NM> 2n 2n 3n 1 1 1 , with n=1,2,3... > > but it is a _molecular_ crystal, so i guess there are some > areas with comparatively low electron density and quite weak > interactions, as seen by the long list of rather soft modes. > i would suspect that will need to be converged very well wrt > to the basis set to not see any ripples on the potential surface. > > for TM oxygen pseudos i would go to 120ry or higher to get > tight convergence. for MD 80-85ry is usually sufficient > (the more traditional 70ry has been shown to be very sloppy), > but if you want a tight geometry convergence you will have > to be more careful. BFGS can be quite sensitive and will > essentially refuse to converge if you have "ripples" due > to a very flat potential surface unless you crank up the > cutoff significantly. > > NM> Norm-conserving and ultrasoft should matter little (unless > NM> there is a problem with one pseudo), while GGAs should be a bit > NM> tougher to converge than LDA. > NM> > NM> Do "crystal" and "simple" really lead to a different number of > NM> modes (10 vs 13) ?? > > those are obviously only the lowest few modes. > it is a 9x9x6 angstrom cell! > > cheers, > axel. > > NM> > NM> nicola > NM> > NM> > NM> John Kendrick wrote: > NM> > Hi, > NM> > > NM> > I have been doing some calculations recently on molecular crystals. > NM> > Because of our interest in the lowest phonon frequencies I have been > NM> > concerned that translational invariance is maintained as far as > NM> > possible. To this end I have increased the accuracy of the optimisation > NM> > and the scf convergence, to a point where I cannot see how to converge > NM> > the calculation any tighter. I noticed that the norm conserving > NM> > pseudopotentials were having greater difficulties than the others. In > NM> > particular for one molecule I have seen significant differences between > NM> > the different projection schemes that ensure translational invariance. > NM> > Particularly the intensities in the infrared a very different. Below is > NM> > some output from a representative calculation. > NM> > NM> --------------------------------------------------------------------- > NM> Prof Nicola Marzari Department of Materials Science and Engineering > NM> 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA > NM> tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu > NM> _______________________________________________ > NM> Pw_forum mailing list > NM> Pw_forum at pwscf.org > NM> http://www.democritos.it/mailman/listinfo/pw_forum > NM> > From akohlmey at cmm.chem.upenn.edu Mon Jan 12 18:27:03 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 12 Jan 2009 12:27:03 -0500 (EST) Subject: [Pw_forum] Projection, convergence and imaginary phonon frequencies In-Reply-To: <496B7061.2070104@bradford.ac.uk> References: <496B7061.2070104@bradford.ac.uk> Message-ID: On Mon, 12 Jan 2009, John Kendrick wrote: JK> Hi, hi, JK> Yes Happy New Year. JK> JK> I only gave the first few soft modes for each case. It is these soft JK> modes I am interested in. There are 174 modes (58 atoms) altogether. I hmmmm... since the topic of treating dispersion interactions came up the other day, this would be a case where it may be worth investigating how much a dispersion correction, even an empirical one, would impact those modes. i know of somebody who did this successfully on an alanine crystal with some other plane wave pseudopotential code (i can give you references, if needed). JK> listed 13 in one case as there was more of a match with the intensities JK> from the other projection methods. JK> JK> The comment about cutoff is interesting as I had found that the JK> optimiser (BFGS) was having problems. This would also be consistent JK> with getting more consistent results with the USP potentials. depending on the settings. with ultra-soft psps it is important that you have a high enough density cutoff. e.g. with a wfc cutoff at 30ry, you may need 300ry for rho. if you had been mix-and-matching potentials or kept the 80ry, then you would have been in the right ball park (80ry*4 = 320ry which would be the default rho cutoff). cheers, axel. JK> JK> Thanks JK> JK> John JK> JK> Axel Kohlmeyer wrote: JK> > On Mon, 12 Jan 2009, Nicola Marzari wrote: JK> > JK> > NM> Dear John, JK> > JK> > dear nicola, JK> > JK> > happy new year from cold philly. JK> > JK> > NM> if it is a crystal, the most critical parameter should be k-point JK> > NM> sampling. Increase that, maybe using shifted meshes of the form JK> > NM> 2n 2n 3n 1 1 1 , with n=1,2,3... JK> > JK> > but it is a _molecular_ crystal, so i guess there are some JK> > areas with comparatively low electron density and quite weak JK> > interactions, as seen by the long list of rather soft modes. JK> > i would suspect that will need to be converged very well wrt JK> > to the basis set to not see any ripples on the potential surface. JK> > JK> > for TM oxygen pseudos i would go to 120ry or higher to get JK> > tight convergence. for MD 80-85ry is usually sufficient JK> > (the more traditional 70ry has been shown to be very sloppy), JK> > but if you want a tight geometry convergence you will have JK> > to be more careful. BFGS can be quite sensitive and will JK> > essentially refuse to converge if you have "ripples" due JK> > to a very flat potential surface unless you crank up the JK> > cutoff significantly. JK> > JK> > NM> Norm-conserving and ultrasoft should matter little (unless JK> > NM> there is a problem with one pseudo), while GGAs should be a bit JK> > NM> tougher to converge than LDA. JK> > NM> JK> > NM> Do "crystal" and "simple" really lead to a different number of JK> > NM> modes (10 vs 13) ?? JK> > JK> > those are obviously only the lowest few modes. JK> > it is a 9x9x6 angstrom cell! JK> > JK> > cheers, JK> > axel. JK> > JK> > NM> JK> > NM> nicola JK> > NM> JK> > NM> JK> > NM> John Kendrick wrote: JK> > NM> > Hi, JK> > NM> > JK> > NM> > I have been doing some calculations recently on molecular crystals. JK> > NM> > Because of our interest in the lowest phonon frequencies I have been JK> > NM> > concerned that translational invariance is maintained as far as JK> > NM> > possible. To this end I have increased the accuracy of the optimisation JK> > NM> > and the scf convergence, to a point where I cannot see how to converge JK> > NM> > the calculation any tighter. I noticed that the norm conserving JK> > NM> > pseudopotentials were having greater difficulties than the others. In JK> > NM> > particular for one molecule I have seen significant differences between JK> > NM> > the different projection schemes that ensure translational invariance. JK> > NM> > Particularly the intensities in the infrared a very different. Below is JK> > NM> > some output from a representative calculation. JK> > NM> JK> > NM> --------------------------------------------------------------------- JK> > NM> Prof Nicola Marzari Department of Materials Science and Engineering JK> > NM> 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA JK> > NM> tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu JK> > NM> _______________________________________________ JK> > NM> Pw_forum mailing list JK> > NM> Pw_forum at pwscf.org JK> > NM> http://www.democritos.it/mailman/listinfo/pw_forum JK> > NM> JK> > JK> _______________________________________________ JK> Pw_forum mailing list JK> Pw_forum at pwscf.org JK> http://www.democritos.it/mailman/listinfo/pw_forum JK> -- ======================================================================= 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 baro at democritos.it Wed Jan 14 14:24:30 2009 From: baro at democritos.it (Moreno Baricevic) Date: 14 Jan 2009 14:24:30 +0100 Subject: [Pw_forum] mailman test - please ignore this mail Message-ID: <1231939470.28609.162.camel@balrog.sissa.it> M. From giannozz at democritos.it Wed Jan 14 17:00:12 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 14 Jan 2009 17:00:12 +0100 Subject: [Pw_forum] Mailing list & web site malfunctioning Message-ID: <496E0C0C.5080909@democritos.it> If you notice problems in the www.quantum-espresso.org web site, please be patient: we have just moved it to a different server. An unintended consequence of such move is that the mailing list pw_forum at pwscf.org is presently not working (I am using an alias). It has already been fixed but it will take some time before the needed DNS change propagates. If a message you sent bounces, just sent it again (if it doesn't bounce, don't send it again: it should go through sooner or later). Sorry for the inconvenience. Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From mansourehp at gmail.com Wed Jan 14 13:23:55 2009 From: mansourehp at gmail.com (Mansoureh Pashangpour) Date: Wed, 14 Jan 2009 15:53:55 +0330 Subject: [Pw_forum] Molecular dynamics with change of temperature Message-ID: Dear all I would like to perform a MD simulation with increasing temperature for it .I am doing this simulation by change of T manually .Is there any instruction to do this automaticlly? Thank you in advance for your help and attention. Pashangpour -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090114/0c318d24/attachment.htm From giannozz at democritos.it Tue Jan 13 18:15:02 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Tue, 13 Jan 2009 18:15:02 +0100 Subject: [Pw_forum] Dielectric constant with Lyddane-Sachs-Teller relation In-Reply-To: References: Message-ID: <496CCC16.8000501@democritos.it> TuanAnh Pham wrote: > Could you please let me know what is the different between > these two methods ? they should give the same results, when both methods apply > Is that correct that the LST relation only works for the system with > sufficient symmetry? typically the LST relation is derived for high-symmetry crystals with two atoms in the unit cell, in which all you have is one TO and one LO nonzero frequency at Gamma. The LST relation can be generalized, though. See for instance this old paper: W. Cochran and R. A. Cowley, J. Phys. Chem. Solids 23, 447 (1962) Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From bipulrr at gmail.com Wed Jan 14 18:30:49 2009 From: bipulrr at gmail.com (Bipul Rakshit) Date: Wed, 14 Jan 2009 23:00:49 +0530 Subject: [Pw_forum] band gap problem Message-ID: <3a749910901140930k7027ea58ufcee04d1ee60b24@mail.gmail.com> Dear Pwscf users, I am doing electronic calculation of SnO2 in rutile structure. I am using Sn.pw91-n-van.UPF and O.pw91-van_ak.UPF PP. With this i am getting a band gap of 7.2 eV, while LDA calculations (by other authors) shows band gap of 3.36 eV and experimental is 1.1 eV. So is the PP which I am using gives this much band gap? or there is some error in my cal.? I am sending the scf file. Please check it and suggest me. thanks -- 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/20090114/b3481c08/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: sno2.scf.in Type: application/octet-stream Size: 2683 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20090114/b3481c08/attachment.obj From giannozz at democritos.it Wed Jan 14 18:23:46 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 14 Jan 2009 18:23:46 +0100 Subject: [Pw_forum] Molecular dynamics with change of temperature In-Reply-To: References: Message-ID: <496E1FA2.9030107@democritos.it> Mansoureh Pashangpour wrote: > I would like to perform a MD simulation with increasing temperature for > it .I am doing this simulation by change of T manually .Is there any > instruction to do this automaticlly? not in CP, I think. In PWscf, see variable "nraise" and "deltaT". Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Wed Jan 14 11:39:03 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 14 Jan 2009 11:39:03 +0100 Subject: [Pw_forum] Dielectric constant with Lyddane-Sachs-Teller relation Message-ID: <496DC0C7.4050902@democritos.it> TuanAnh Pham wrote: > Could you please let me know what is the different between these two methods ? they should give the same results, when both methods apply > Is that correct that the LST relation only works for the system with sufficient symmetry? typically the LST relation is derived for high-symmetry crystals with two atoms in the unit cell, in which all you have is one TO and one LO nonzero frequency at Gamma. The LST relation can be generalized, though. See for instance this old paper: W. Cochran and R. A. Cowley, J. Phys. Chem. Solids 23, 447 (1962) Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy From whitley3 at llnl.gov Wed Jan 14 19:20:58 2009 From: whitley3 at llnl.gov (Heather Whitley) Date: Wed, 14 Jan 2009 10:20:58 -0800 Subject: [Pw_forum] Increase in number of diagonalization errors In-Reply-To: References: Message-ID: <496E2D0A.8080803@llnl.gov> Hi, I recently started using the new version of pwscf from QE 4.0.4; I had previously been using QE 3.2.3. I've noticed that it seems like I get these Cholesky diagonalization errors much more frequently with the new version of the code. I've read the advice on how to avoid these errors (running with the "-ndiag 1" flag or changing the option for diagonalization method). I am, however, curious if there is some reason why I would see these errors more frequently with the new version? Maybe it's something in my input? I've attached the text from the error (typical of what other people have asked about in previous threads), as well as a sample input file. Thank you, Heather %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% task # 63 from pdpotf : error # 422 problems computing cholesky decomposition %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% &control calculation='nscf' restart_mode='restart' tstress=.true. tprnfor=.true. etot_conv_thr=1.0d-5 forc_conv_thr=1.0d-4 pseudo_dir = './' outdir='./' title='cd15se15' prefix='cd15se15' wf_collect = .false. max_seconds=1.43e+04 nstep=2500 tefield=.false. dipfield=.false. / &system ibrav=8 celldm(1) =60.261212118 celldm(2) =1.0 celldm(3) =1.0 nat=36 ntyp=5 nelec=286 ecutwfc =35.0 emaxpos = .45 eopreg = 0.1 eamp = 0.001 edir = 3 occupations='smearing' degauss=0.001 smearing='gaussian' / &electrons diagonalization='david' electron_maxstep=500 conv_thr = 1.0d-10 mixing_beta=0.5 startingwfc='random' -- *********************************************** Heather D. Whitley, Quantum Simulations Group Lawrence Livermore National Laboratory, L-415 Livermore, CA 94551 http://qsg.llnl.gov From yuminqian at gmail.com Wed Jan 14 02:00:34 2009 From: yuminqian at gmail.com (yumin qian) Date: Wed, 14 Jan 2009 09:00:34 +0800 Subject: [Pw_forum] spin glass calculation by first principle method Message-ID: Dear all PWSCF user Right now I meet a problem in solid crystal calculation , if I think the crystal is at the state of spin glass, i.e. the transition metal in the crystal have a local moment but does not have a long range order , how should I use PWSCF to investigate electronic structure of this kind of ground state . Many thanks -- 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/20090114/b5678b5b/attachment.htm From baroni at sissa.it Wed Jan 14 20:22:16 2009 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 14 Jan 2009 20:22:16 +0100 Subject: [Pw_forum] spin glass calculation by first principle method In-Reply-To: References: Message-ID: <1D2AB9CC-F439-4FCD-8464-C9503F1062B1@sissa.it> using supercells and a lot of intuition/common sense SB On Jan 14, 2009, at 2:00 AM, yumin qian wrote: > Dear all PWSCF user > Right now I meet a problem in solid crystal calculation , if I > think the crystal is at the state of spin glass, i.e. the transition > metal in the crystal have a local moment but does not have a long > range order , how should I use PWSCF to investigate > electronic structure of this kind of ground state . > Many thanks > > -- > 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 --- 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/20090114/b763ce29/attachment-0001.htm From chenhanghuipwscf at gmail.com Tue Jan 13 10:29:15 2009 From: chenhanghuipwscf at gmail.com (hanghui chen) Date: Tue, 13 Jan 2009 17:29:15 +0800 Subject: [Pw_forum] Hubbard_U and Hubbard_alpha Message-ID: <22ae3ca40901130129w310cce22q990e0104f73bcb23@mail.gmail.com> Dear PWSCF users, I have been recently working on the magnetic properties of transitional metal oxides and need to turn on Hubbard U. I found that in PWSCF there are two variables Hubbard_U and Hubbard_alpha. In other packages like VASP, there are also two variables U (Hubbard U) and J (Hund's exchange or Stoner parameter). I would like to know what is Hubbard_alpha. Does it have the same meaning as J in VASP? I read the menu but could not find the meaning of Hubbard_alpha. Thank you very much. Hanghui Chen Department of Physics Yale University From matteo at umn.edu Wed Jan 14 21:15:54 2009 From: matteo at umn.edu (Matteo Cococcioni) Date: Wed, 14 Jan 2009 14:15:54 -0600 Subject: [Pw_forum] Hubbard_U and Hubbard_alpha In-Reply-To: <22ae3ca40901130129w310cce22q990e0104f73bcb23@mail.gmail.com> References: <22ae3ca40901130129w310cce22q990e0104f73bcb23@mail.gmail.com> Message-ID: <496E47FA.6030507@umn.edu> Dear Hanghui, Hubbard_alpha is an input parameter used to compute the value of U (PRB 71 35105 (2005)). So you don't need to use it for a simple DFT+U calculation if you already know the value of U. J is not implemented in pwscf; U in pwscf (and it's the same on most of papers using DFT+U) corresponds to an effective U_eff = U - J. I have never seen a paper in literature where the effect of having J treated separately was important so I think this the one implemented is a good enough functional in most of cases. regards, Matteo hanghui chen wrote: > Dear PWSCF users, > I have been recently working on the magnetic properties of > transitional metal oxides and need to turn on Hubbard U. I found that > in PWSCF there are two variables Hubbard_U and Hubbard_alpha. In other > packages like VASP, there are also two variables U (Hubbard U) and J > (Hund's exchange or Stoner parameter). I would like to know what is > Hubbard_alpha. Does it have the same meaning as J in VASP? I read the > menu but could not find the meaning of Hubbard_alpha. > Thank you very much. > > > Hanghui Chen > Department of Physics > Yale University > _______________________________________________ > 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/20090114/beeceae3/attachment.vcf From giannozz at democritos.it Wed Jan 14 21:40:37 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 14 Jan 2009 21:40:37 +0100 Subject: [Pw_forum] Increase in number of diagonalization errors In-Reply-To: <496E2D0A.8080803@llnl.gov> References: <496E2D0A.8080803@llnl.gov> Message-ID: <2D415020-F72E-4719-9B03-D32D8B07E6F8@democritos.it> On Jan 14, 2009, at 19:20 , Heather Whitley wrote: > I am, however, curious if there is some reason why I would see these > errors more frequently with the new version? not that I know, but since the reason for these errors is usually hard to find, it is also hard to give a more definite answer. At a given time the algorithm for subspace diagonalization has been replaced by a better performing one. Maybe it is also less stable. > from pdpotf : error # 422 > problems computing cholesky decomposition this is the parallel version of subspace diagonalization. In the CVS version there is another one available, by the way, based on SCALAPACK > startingwfc='random' there is usually no need to start with random wavefunctions Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Wed Jan 14 21:42:13 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Wed, 14 Jan 2009 21:42:13 +0100 Subject: [Pw_forum] band gap problem In-Reply-To: <3a749910901140930k7027ea58ufcee04d1ee60b24@mail.gmail.com> References: <3a749910901140930k7027ea58ufcee04d1ee60b24@mail.gmail.com> Message-ID: On Jan 14, 2009, at 18:30 , Bipul Rakshit wrote: > With this i am getting a band gap of 7.2 eV, while LDA calculations > (by other > authors) shows band gap of 3.36 eV and experimental is 1.1 eV. I have never heard of a case where the LDA band gap is 3 times the experimental one! Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From panda.deng.pan at gmail.com Wed Jan 14 09:46:12 2009 From: panda.deng.pan at gmail.com (=?GB2312?B?xcu1xw==?=) Date: Wed, 14 Jan 2009 16:46:12 +0800 Subject: [Pw_forum] Problem about the inputs of 'dos.x' Message-ID: Dear All, I was trying to calculate the dos of BaNiAs, when I was running the dos.x it gives the error message : from dos : error # 1 not implemented I do not know what mistake I have made in my inputs.Would someone point it out Please. &inputpp outdir='/disk2/xgwan/tmp/' prefix='BaNiAs' fildos='BaNiAs.dos', Emin=0.0, Emax=25.0, DeltaE=0.1 / And still I ran the projwfc.x to get some projected dos after I ran the dos.x.Does the results of the projwfc.x be affected by the failing of dos.x? Thanks all. Pan Deng Nanjing University From kondor.jess at gmail.com Wed Jan 14 22:42:24 2009 From: kondor.jess at gmail.com (Jess Kondor) Date: Wed, 14 Jan 2009 16:42:24 -0500 Subject: [Pw_forum] Hubbard_U and Hubbard_alpha In-Reply-To: <496E47FA.6030507@umn.edu> References: <22ae3ca40901130129w310cce22q990e0104f73bcb23@mail.gmail.com> <496E47FA.6030507@umn.edu> Message-ID: <1d9d5d9d0901141342p24b2ff19vce67eb45a094cae7@mail.gmail.com> Dear Matteo, > J is not implemented in pwscf; U in pwscf (and it's the same on most of > papers using DFT+U) > corresponds to an effective U_eff = U - J. I have never seen a paper in > literature where the effect of having J treated separately was important so > I think this the one implemented is > a good enough functional in most of cases. I've seen at least one paper, Claude Ederer * and Nicola A. Spaldin, "Influence of strain and oxygen vacancies on the magnetoelectric properties of multiferroic bismuth ferrite", Phys. Rev. B 71, 224103 (2005) where authors said 'In contrast, the magnetization of BiFeO3, which is due to a small canting of the mainly antiferromagnetically coupled magnetic moments of the Fe cations, is strongly J dependent.' - ===================================== Jess Kondor PICYT, PhD student ===================================== From mansourehp at gmail.com Tue Jan 13 10:39:55 2009 From: mansourehp at gmail.com (Mansoureh Pashangpour) Date: Tue, 13 Jan 2009 13:09:55 +0330 Subject: [Pw_forum] Molecular dynamics with change of temperature Message-ID: Dear all I would like to perform a MD simulation with increasing temperature for it .I am doing this simulation by change of T manually .Is there any instruction to do this automaticlly? Thank you in advance for your help and attention. Mansoureh Pashangpour -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090113/96954091/attachment.htm From hqzhou at nju.edu.cn Thu Jan 15 05:11:52 2009 From: hqzhou at nju.edu.cn (Huiqun Zhou) Date: Thu, 15 Jan 2009 12:11:52 +0800 Subject: [Pw_forum] band gap problem References: <3a749910901140930k7027ea58ufcee04d1ee60b24@mail.gmail.com> Message-ID: Have you tested convergency against ecutwfc? I remember that because of the pseudo potential for oxigen you're using, ecutwfc can go as high as 100 rye if your criterion for convergency of total energy is 1 mRy. Although it sounds rediculous for USPP, you may need to see if the ecutwfc influences the band gap. Huiqun Zhou @Nanjing University, China ----- Original Message ----- From: Bipul Rakshit To: PWSCF Forum Sent: Thursday, January 15, 2009 1:30 AM Subject: [Pw_forum] band gap problem Dear Pwscf users, I am doing electronic calculation of SnO2 in rutile structure. I am using Sn.pw91-n-van.UPF and O.pw91-van_ak.UPF PP. With this i am getting a band gap of 7.2 eV, while LDA calculations (by other authors) shows band gap of 3.36 eV and experimental is 1.1 eV. So is the PP which I am using gives this much band gap? or there is some error in my cal.? I am sending the scf file. Please check it and suggest me. thanks -- 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 -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090115/52a32888/attachment.htm From lanhaiping at gmail.com Thu Jan 15 05:28:20 2009 From: lanhaiping at gmail.com (lan haiping) Date: Thu, 15 Jan 2009 12:28:20 +0800 Subject: [Pw_forum] Molecular dynamics with change of temperature In-Reply-To: References: Message-ID: Hi, a simple method is to follow the script in example01. you can write some script for a range of temperature set. Regards On Tue, Jan 13, 2009 at 5:39 PM, Mansoureh Pashangpour wrote: > Dear all > I would like to perform a MD simulation with increasing temperature for it > .I am doing > this simulation by change of T manually .Is there any instruction to do > this automaticlly? > Thank you in advance for your help and attention. > > Mansoureh Pashangpour > > _______________________________________________ > 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/20090115/1df3f0d3/attachment.htm From giannozz at democritos.it Thu Jan 15 09:26:57 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 15 Jan 2009 09:26:57 +0100 Subject: [Pw_forum] Problem about the inputs of 'dos.x' In-Reply-To: References: Message-ID: <99E6AB2D-9A38-4548-97FC-051FE229C6EC@democritos.it> On Jan 14, 2009, at 9:46 , ?? wrote: > from dos : error # 1 > not implemented you cannot use k-point parallelization ("pools") with dos.x > Does the results of the projwfc.x be affected by the failing of dos.x? no Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at democritos.it Thu Jan 15 09:51:53 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 15 Jan 2009 09:51:53 +0100 Subject: [Pw_forum] Molecular dynamics with change of temperature In-Reply-To: References: Message-ID: <4C29640F-A1C7-4469-B5EF-333083BE93D9@democritos.it> For CP there is also the "autopilot" feature that might be appropriate for this case. Documentation in Doc/README.AUTOPILOT Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From sclauzer at sissa.it Wed Jan 14 09:47:52 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Wed, 14 Jan 2009 09:47:52 +0100 Subject: [Pw_forum] question on optimization In-Reply-To: <49554733.7020806@ciw.edu> References: <793818.1251.qm@web65712.mail.ac4.yahoo.com> <49554733.7020806@ciw.edu> Message-ID: <496DA6B8.9000602@sissa.it> I noticed your mail has not been replied yet. I'm not an expert in variable cell calculations, but I would suggest you to try vc-relax with the option cell_dofree='z' in the cell namelist (see also Doc/INPUT_PW.txt, if you haven't yet). It should do what you need, if I have well understood and if the option is correctly working. If you send me an input file for your (possibly small) system I can try if it works. GS P. Ganesh wrote: > Hi, > > I was wondering if it possible to do a constrained structural > optimization using 'pwscf' of a monoclinic structure whereby > the 'a' and 'b' in-plane lattices are kept fixed, while the 'c' axis > (i.e. mod(c), and angles alpha and beta) alone is allowed to relax. > > Wish you all happy holidays. > > Thanks, > Ganesh > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From sclauzer at sissa.it Thu Jan 15 09:35:25 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Thu, 15 Jan 2009 09:35:25 +0100 Subject: [Pw_forum] Problem about the inputs of 'dos.x' In-Reply-To: References: Message-ID: <496EF54D.5000907@sissa.it> Dear Pan Deng, are you doing a parallel calculation? If so, most probably you started dos.x with a different number of pools or processors with respect to the pw.x run. When running any of the postprocessing tools which read output files of pw.x in the parallel version, you must use the same number of processors AND the same number of pools used for the pw.x run. In principle the dos.x should not modify pw.x output files (i.e. those in prefix.save and the wavefunctions), since it simply needs to read them (you can checkd the time stamps of those files to see if they have been modified when dos.x was running). Regards GS ?? wrote: > Dear All, > > I was trying to calculate the dos of BaNiAs, when I was running > the dos.x it gives the error message : > from dos : error # 1 > not implemented > I do not know what mistake I have made in my inputs.Would someone > point it out Please. > &inputpp > outdir='/disk2/xgwan/tmp/' > prefix='BaNiAs' > fildos='BaNiAs.dos', > Emin=0.0, Emax=25.0, DeltaE=0.1 > / > And still I ran the projwfc.x to get some projected dos after I ran > the dos.x.Does the results of the projwfc.x be affected by the failing > of dos.x? > > Thanks all. > > Pan Deng > Nanjing University > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From sclauzer at sissa.it Wed Jan 14 10:12:04 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Wed, 14 Jan 2009 10:12:04 +0100 Subject: [Pw_forum] error :kpoint 7 ibnd 17 solve_linter: root not converged NaN In-Reply-To: References: Message-ID: <496DAC64.5030404@sissa.it> Hi, maybe try to reduce the mixing parameter alpha_mix, perhaps the default is too high for your system and causes problems in the potential-mixing subroutine. You can also try to play with the nmix_ph parameter. GS xu yuehua wrote: > error in the phonon calculation! (the 19th phonon problem)???? > hi all : > my system is a excited state, my aim is to get the phonon phonon at G > point at this state,but i face > the error is following in the 19 phonon calculation ? > > Representation # 19 mode # 19 > Self-consistent Calculation > iter # 1 total cpu time : 3876.6 secs av.it > .: 3.9 > thresh= 0.100E-01 alpha_mix = 0.700 |ddv_scf|^2 = 0.412E-07 > iter # 2 total cpu time : 3896.1 secs av.it .: > 10.3 > thresh= 0.203E-04 alpha_mix = 0.700 |ddv_scf|^2 = 0.891E-09 > iter # 3 total cpu time : 3918.1 secs av.it .: > 12.9 > thresh= 0.298E-05 alpha_mix = 0.700 |ddv_scf|^2 = 0.155E-08 > iter # 4 total cpu time : 3939.5 secs av.it .: > 11.9 > thresh= 0.394E-05 alpha_mix = 0.700 |ddv_scf|^2 = 0.415E-11 > kpoint 7 ibnd 17 solve_linter: root not converged NaN > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from broyden : error # 3 > factorization > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from broyden : error # 3 > factorization > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from broyden : error # 3 > factorization > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > the error is in the 19th phonon calculation ,before is OK. > SO i do not know what happened. i need your help. > > -- > Xu Yuehua > physics Department of Nanjing university > China > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From paulatto at sissa.it Wed Jan 14 12:09:15 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Wed, 14 Jan 2009 12:09:15 +0100 Subject: [Pw_forum] Test In-Reply-To: <496DC65F.6060409@democritos.it> References: <496D96EE.1070805@uniud.it> <1231928480.28609.1.camel@balrog.sissa.it> <496DBEEF.4080203@uniud.it> <1231929248.28610.4.camel@balrog.sissa.it> <496DC127.3080105@democritos.it> <1231930297.28609.28.camel@balrog.sissa.it> <496DC535.4070506@democritos.it> <1231930820.28511.49.camel@balrog.sissa.it> <496DC65F.6060409@democritos.it> Message-ID: test email, nothing to worry about (It looks like some people are not receiving pw_forum emails.) -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From giannozz at democritos.it Thu Jan 15 10:36:27 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 15 Jan 2009 10:36:27 +0100 Subject: [Pw_forum] band gap problem In-Reply-To: <3a749910901140930k7027ea58ufcee04d1ee60b24@mail.gmail.com> References: <3a749910901140930k7027ea58ufcee04d1ee60b24@mail.gmail.com> Message-ID: <6BBC68C9-4864-4134-9359-A5CC11293950@democritos.it> On Jan 14, 2009, at 18:30 , Bipul Rakshit wrote: > With this i am getting a band gap of 7.2 eV really? I get 2.2 eV with your data Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From nemeth at anl.gov Wed Jan 14 21:20:42 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Wed, 14 Jan 2009 14:20:42 -0600 (CST) Subject: [Pw_forum] PDOS In-Reply-To: References: Message-ID: Hi, please ignore my previous request as I have found the answer on a previous PWSCF forum site (http://www.democritos.it/pipermail/pw_forum/2004-May/001081.html). Thanks: Karoly On Tue, 13 Jan 2009, Karoly Nemeth wrote: > > Hi, > > I recently calculated PDOS of a system using the pseudopotential file > Ni.pbe-nd-rrkjus.UPF. The output does not contain any PDOS for 4P of Ni, only > PDOS for atomic orbitals occupied in the Ni 3d9 4s1 > electron configuration (used in the pseudopotential file) are given. Is there > any way to set up the PROJWFC calculation such that it delivers the missing > PDOS (for 4P of Ni) as well? > Thanks: > > Karoly > > > From nemeth at anl.gov Wed Jan 14 01:48:05 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Tue, 13 Jan 2009 18:48:05 -0600 (CST) Subject: [Pw_forum] PDOS In-Reply-To: References: Message-ID: Hi, I recently calculated PDOS of a system using the pseudopotential file Ni.pbe-nd-rrkjus.UPF. The output does not contain any PDOS for 4P of Ni, only PDOS for atomic orbitals occupied in the Ni 3d9 4s1 electron configuration (used in the pseudopotential file) are given. Is there any way to set up the PROJWFC calculation such that it delivers the missing PDOS (for 4P of Ni) as well? Thanks: Karoly From yuminqian at gmail.com Thu Jan 15 11:45:15 2009 From: yuminqian at gmail.com (yumin qian) Date: Thu, 15 Jan 2009 18:45:15 +0800 Subject: [Pw_forum] spin glass calculation by first principle method In-Reply-To: <1D2AB9CC-F439-4FCD-8464-C9503F1062B1@sissa.it> References: <1D2AB9CC-F439-4FCD-8464-C9503F1062B1@sissa.it> Message-ID: thanks for your reply -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090115/0d32cba3/attachment.htm From mambom1902 at yahoo.com Wed Jan 14 12:01:08 2009 From: mambom1902 at yahoo.com (loc duong ding) Date: Wed, 14 Jan 2009 03:01:08 -0800 (PST) Subject: [Pw_forum] Relax structure after getting the cell relaxation result Message-ID: <393913.35245.qm@web38804.mail.mud.yahoo.com> Dear all, I relax the cell of structure. After that, I use this result to relax ion positions. The input file for cell relaxation and ion relax is the same (only change 'vc-relax' to 'relax'). These are the result I got. For Vc-relax: total stress (Ry/bohr**3) (kbar) P= 0.02 0.00000094 -0.00000009 -0.00000014 0.14 -0.01 -0.02 -0.00000009 0.00000082 -0.00000024 -0.01 0.12 -0.04 -0.00000014 -0.00000024 -0.00000142 -0.02 -0.04 -0.21 bfgs converged in 62 scf cycles and 52 bfgs steps End of BFGS Geometry Optimization Final enthalpy = -441.4525306199 Ry CELL_PARAMETERS (alat) 0.984130693 0.001683023 0.030237080 -0.490603832 0.848326583 -0.013643812 0.066169871 0.004483599 2.350435692 For ion relax: total stress (Ry/bohr**3) (kbar) P= 0.28 0.00000250 -0.00000002 -0.00000113 0.37 0.00 -0.17 -0.00000002 0.00000191 -0.00000041 0.00 0.28 -0.06 -0.00000113 -0.00000041 0.00000128 -0.17 -0.06 0.19 BFGS Geometry Optimization bfgs converged in 1 scf cycles and 0 bfgs steps End of BFGS Geometry Optimization Final energy = -441.4515856032 Ry CELL_PARAMETERS (alat) 0.984130693 0.001683023 0.030237080 -0.490603832 0.848326583 -0.013643812 0.066169871 0.004483599 2.350435692 I don't understand why the pressures in two case are different. I also checked the position of ions and nothing is different. And another thing is Final energy and Final enthalpy. I know enthalpy is the energy of system at constant pressure. What is difference between Final energy and Final enthalpy for this case? Best regards, ----------------------------------------------- Loc Duong Dinh Ms-Ph.D Student Sungkyunkwan Advanced Institute of Nanotechnology, Sungkyunkwan University, Suwon, 440-746, Korea Email: mambom1902 at yahoo.com From paulatto at sissa.it Thu Jan 15 13:23:23 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Thu, 15 Jan 2009 13:23:23 +0100 Subject: [Pw_forum] Relax structure after getting the cell relaxation result In-Reply-To: <393913.35245.qm@web38804.mail.mud.yahoo.com> References: <393913.35245.qm@web38804.mail.mud.yahoo.com> Message-ID: On Wed, 14 Jan 2009 12:01:08 +0100, loc duong ding wrote: > I relax the cell of structure. After that, I use > this result to relax ion positions. The input file for cell relaxation > and ion relax is the same (only change 'vc-relax' to 'relax'). As far as I know, vc-relax relaxes the atomic positions as well, unless you specify ion_dynamic='none', which is not possible with bfgs. > I don't understand why the pressures in two case are different. They are zero in both case, where zero means "less than the convergence threshold". I cannot be sure that when you are saying "I use this result" you are actually doing the right thing, but if you do then your 2 calculations are the same, and will give you the same results, inside the convergence threshold. regards -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From akohlmey at cmm.chem.upenn.edu Wed Jan 14 17:16:36 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 14 Jan 2009 11:16:36 -0500 (EST) Subject: [Pw_forum] Molecular dynamics with change of temperature In-Reply-To: References: Message-ID: On Wed, 14 Jan 2009, Mansoureh Pashangpour wrote: MP> Dear all MP> I would like to perform a MD simulation with increasing temperature for it MP> .I am doing MP> this simulation by change of T manually. Is there any instruction to do this MP> automaticlly? well, you can take an editor, program it, recompile and be happy. ;-) however, i don't quite see how much of a difference that would make for the kind and particularly the size of system that you can handle with quantum espresso. it would be just as easy to simply write a small shell script to run a chunk of trajectory and then restart with ramped up temperature. as usual, it is hard to give any specific advice without knowing about the purpose of your temperature ramping. axel. MP> Thank you in advance for your help and attention. MP> Pashangpour 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 giannozz at democritos.it Thu Jan 15 14:42:36 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Thu, 15 Jan 2009 14:42:36 +0100 Subject: [Pw_forum] Relax structure after getting the cell relaxation result In-Reply-To: <393913.35245.qm@web38804.mail.mud.yahoo.com> References: <393913.35245.qm@web38804.mail.mud.yahoo.com> Message-ID: <774D962B-6588-47F9-A317-BCA4889D5853@democritos.it> On Jan 14, 2009, at 12:01 , loc duong ding wrote: > I know enthalpy is the energy of system at constant pressure. really? I knew that H=E+PV P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From 041019019 at fudan.edu.cn Thu Jan 15 13:51:00 2009 From: 041019019 at fudan.edu.cn (Wang Qinjing) Date: Thu, 15 Jan 2009 20:51:00 +0800 Subject: [Pw_forum] Fw: How to use the parameters orbj_in and orbj_fin in calculation in pwcond.x Message-ID: <002b01c9770f$eb1f4680$3f7c620a@fduphy778548b0> Dear all: I'd like to calculate the transmission of some special orbital. In the INPUT_PWCOND.html I was told there are two parameters orbj_in and orbj_fin as follows: ================ orbj_in REAL the initial orbital for projecting the transmission orbj_fin REAL the final orbital for projecting the transmission ============ But what value can I set for these two parameters? What do they mean? Are they related to the transmission of some special orbital? Thank you in advance Cheers Wang ------------------------------------ ?????????????? Wang Qinjing, Physics Department, Fudan University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090115/91d19f2a/attachment.htm From nicvok at freenet.de Tue Jan 13 15:09:43 2009 From: nicvok at freenet.de (Nicki F. Hinsche) Date: Tue, 13 Jan 2009 15:09:43 +0100 Subject: [Pw_forum] Transport with pw_cond copper Message-ID: <496CA0A7.3090007@freenet.de> Hi, I want to make some test calculations of transport through 'bulk' copper. Left and right lead shall be equal and the scattering region should be the same material as the lead itself. My problem is, that I cant get any results and I can't recognize any mistakes in my input cards -> my inputcards are: # self-consistent calculation for Cu bulk &control calculation='scf' restart_mode='from_scratch', pseudo_dir = '$PSEUDO_DIR/', outdir='$TMP_DIR/' prefix='cu' / &system ibrav = 1 celldm(1) = 6.719 nat= 1, ntyp= 1, ecutwfc = 40.0, occupations='smearing', smearing='methfessel-paxton', degauss=0.01 / &electrons conv_thr = 1.0e-12 mixing_beta = 0.5 / ATOMIC_SPECIES Cu 63.55 Cu.pz-d-rrkjus.UPF ATOMIC_POSITIONS Cu 0.0 0.0 0.0 K_POINTS (automatic) 16 16 16 0 0 0 # complex bands of Copper along the 001 direction K_perp=0 cat > cu.cond.in << EOF &inputcond oudir='$TMP_DIR/' prefixl='cu' band_file ='bands.cu' ikind=0 energy0=10.d0 denergy=-0.1d0 ewind=1.d0 epsproj=1.d-3 delgep = 1.d-12 cutplot = 2.d0 / 1 0.0 0.0 1.0 100 # transmission calculation for the perfect copper contact cat > cutrans.cond.in << EOF &inputcond outdir='$TMP_DIR/', prefixl='cu', prefixs='cu', tran_file='trans.cu', ikind=1, energy0=2.95d0, denergy=-0.1d0, ewind=1.d0, epsproj=1.d-3, nz1 = 1 / 1 0.0 0.0 1.0 100 I get the error message: task # 1 from in init_cond : error # 172992232 wrong nrztot I would be thankful for every hint that you might give me... ------------------------------------------------------- Nicki F. Hinsche Martin-Luther-Universit?t Halle-Wittenberg Naturwissenschaftliche Fakult?t II - Chemie und Physik Institut f?r Physik, Theoretische Physik D-06099 Halle (Saale) Telefon : +49-345/55-25460 Forschungszentrum Dresden-Rossendorf Institut f?r Ionenstrahlphysik und Materialforschung Postfach 51 01 19 D-01314 Dresden, Germany E-Mail : n.hinsche at fzd.de -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090113/70b77b03/attachment.htm From 041019019 at fudan.edu.cn Wed Jan 14 14:48:50 2009 From: 041019019 at fudan.edu.cn (Wang Qinjing) Date: Wed, 14 Jan 2009 21:48:50 +0800 Subject: [Pw_forum] How to use the parameters orbj_in and orbj_fin in calculation in pwcond.x Message-ID: <000801c9764e$d4a30570$3f7c620a@fduphy778548b0> Dear all: I'd like to calculate the transmission of some special orbital. In the INPUT_PWCOND.html I was told there are two parameters orbj_in and orbj_fin as follows: ================ orbj_in REAL the initial orbital for projecting the transmission orbj_fin REAL the final orbital for projecting the transmission ============ But what value can I set for these two parameters? What do they mean? Are they related to the transmission of some special orbital? Thank you in advance Cheers Wang ------------------------------------ ?????????????? Wang Qinjing, Physics Department, Fudan University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090114/c4c67b98/attachment-0001.htm From manoj at phys.ufl.edu Thu Jan 15 16:36:38 2009 From: manoj at phys.ufl.edu (Manoj Srivastava) Date: Thu, 15 Jan 2009 10:36:38 -0500 (EST) Subject: [Pw_forum] Transport with pw_cond copper In-Reply-To: <496CA0A7.3090007@freenet.de> Message-ID: Hi Nicki, I think your unit cell is wrong. The unit cell for Cu is FCC, but you have specfied it as simple cubic. Try to change ibrav to 2 and see if it works. -Manoj On Tue, 13 Jan 2009, Nicki F. Hinsche wrote: > Hi, I want to make some test calculations of transport through 'bulk' > copper. Left and right lead shall be equal and the scattering region > should be the same material as the lead itself. My problem is, that I > cant get any results and I can't recognize any mistakes in my input > cards -> my inputcards are: > > # self-consistent calculation for Cu bulk > > &control > calculation='scf' > restart_mode='from_scratch', > pseudo_dir = '$PSEUDO_DIR/', > outdir='$TMP_DIR/' > prefix='cu' > / > &system > ibrav = 1 > celldm(1) = 6.719 > nat= 1, > ntyp= 1, > ecutwfc = 40.0, > occupations='smearing', > smearing='methfessel-paxton', > degauss=0.01 > / > &electrons > conv_thr = 1.0e-12 > mixing_beta = 0.5 > / > ATOMIC_SPECIES > Cu 63.55 Cu.pz-d-rrkjus.UPF > ATOMIC_POSITIONS > Cu 0.0 0.0 0.0 > K_POINTS (automatic) > 16 16 16 0 0 0 > > > # complex bands of Copper along the 001 direction K_perp=0 > > cat > cu.cond.in << EOF > &inputcond > oudir='$TMP_DIR/' > prefixl='cu' > band_file ='bands.cu' > ikind=0 > energy0=10.d0 > denergy=-0.1d0 > ewind=1.d0 > epsproj=1.d-3 > delgep = 1.d-12 > cutplot = 2.d0 > / > 1 > 0.0 0.0 1.0 > 100 > > > # transmission calculation for the perfect copper contact > > cat > cutrans.cond.in << EOF > &inputcond > outdir='$TMP_DIR/', > prefixl='cu', > prefixs='cu', > tran_file='trans.cu', > ikind=1, > energy0=2.95d0, > denergy=-0.1d0, > ewind=1.d0, > epsproj=1.d-3, > nz1 = 1 > / > 1 > 0.0 0.0 1.0 > 100 > > > I get the error message: > > task # 1 > from in init_cond : error # 172992232 > wrong nrztot > > > I would be thankful for every hint that you might give me... > > > > ------------------------------------------------------- > > Nicki F. Hinsche > > Martin-Luther-Universit?t Halle-Wittenberg > > Naturwissenschaftliche Fakult?t II - Chemie und Physik > > Institut f?r Physik, Theoretische Physik > > D-06099 Halle (Saale) > > Telefon : +49-345/55-25460 > > Forschungszentrum Dresden-Rossendorf > > Institut f?r Ionenstrahlphysik und Materialforschung > > Postfach 51 01 19 > > D-01314 Dresden, Germany > E-Mail : n.hinsche at fzd.de > > > > From jlz101 at unlserve.unl.edu Thu Jan 15 17:46:54 2009 From: jlz101 at unlserve.unl.edu (J. D. Burton) Date: Thu, 15 Jan 2009 10:46:54 -0600 Subject: [Pw_forum] Transport with pw_cond copper In-Reply-To: <496CA0A7.3090007@freenet.de> References: <496CA0A7.3090007@freenet.de> Message-ID: <003c01c97730$df9be040$9ed3a0c0$@unl.edu> Nicki, PWCOND only works with tetragonal cells: either ibrav = 6 or use ibrav = 0 with tetragonal CELL_PARAMETERS in your scf calculation. Notice all the calculations in example12 use ibrav = 6. Why do you want simple cubic Cu anyway? Cu is fcc, but you can?t use ibrav = 2 for PWCOND. You can set up your scf calculation like this: ibrav = 0 celldm(1) = cubic lattice parameter for Cu in a.u. ATOMIC_POSITONS (crystal) Cu 0.0 0.0 0.0 Cu 0.5 0.5 0.5 CELL_PARAMETERS sqrt(2)/2 0.00 0.00 0.00 sqrt(2)/2 0.00 0.00 0.00 1.00 Anything useful that you might want to calculate (i.e. a junction) will be have to be at least tetragonal anyway. (Although I don't think even orthorhombic cells are implemented yet.) Cheers, J. D. ************************************ J. D. Burton, Ph.D. jlz101 at unlserve.unl.edu University of Nebraska Lincoln Physics and Astronomy Office Ph. (402) 472 2499 Mobile Ph. (402) 419 9918 213C Ferguson Hall ************************************ "The job of a scientist is to generate wrong ideas as fast as possible." -- Murray Gell-Mann From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On Behalf Of Nicki F. Hinsche Sent: Tuesday, January 13, 2009 8:10 AM To: pw_forum at pwscf.org Subject: [Pw_forum] Transport with pw_cond copper Hi, I want to make some test calculations of transport through 'bulk' copper. Left and right lead shall be equal and the scattering region should be the same material as the lead itself. My problem is, that I cant get any results and I can't recognize any mistakes in my input cards -> my inputcards are: # self-consistent calculation for Cu bulk ?&control ??? calculation='scf' ??? restart_mode='from_scratch', ??? pseudo_dir = '$PSEUDO_DIR/', ??? outdir='$TMP_DIR/' ??? prefix='cu' ?/ ?&system ??? ibrav = 1 ??? celldm(1) = 6.719 ??? nat= 1, ??? ntyp= 1, ??? ecutwfc = 40.0, ??? occupations='smearing', ??? smearing='methfessel-paxton', ??? degauss=0.01 ?/ ?&electrons ??? conv_thr = 1.0e-12 ??? mixing_beta = 0.5 ?/ ATOMIC_SPECIES ?Cu 63.55 Cu.pz-d-rrkjus.UPF ATOMIC_POSITIONS ?Cu 0.0 0.0 0.0 K_POINTS (automatic) ?16 16 16 0 0 0 # complex bands of Copper along the 001 direction K_perp=0 cat > cu.cond.in << EOF ?&inputcond ??? oudir='$TMP_DIR/' ??? prefixl='cu' ??? band_file ='bands.cu' ??? ikind=0 ??? energy0=10.d0 ??? denergy=-0.1d0 ??? ewind=1.d0 ??? epsproj=1.d-3 ??? delgep = 1.d-12 ??? cutplot = 2.d0 ?/ ??? 1 ??? 0.0 0.0 1.0 ??? 100 # transmission calculation for the perfect copper contact cat > cutrans.cond.in << EOF ?&inputcond ??? outdir='$TMP_DIR/', ??? prefixl='cu', ??? prefixs='cu', ??? tran_file='trans.cu', ??? ikind=1, ??? energy0=2.95d0, ??? denergy=-0.1d0, ??? ewind=1.d0, ??? epsproj=1.d-3, ??? nz1 = 1 ?/ ?? 1 ??? 0.0? 0.0? 1.0 ??? 100 ?? I get the error message: ? ?? task #???????? 1 ???? from in init_cond : error # 172992232 ???? wrong nrztot I would be thankful for every hint that you might give me... ------------------------------------------------------- Nicki F. Hinsche Martin-Luther-Universit?t Halle-Wittenberg Naturwissenschaftliche Fakult?t II - Chemie und Physik Institut f?r Physik, Theoretische Physik D-06099 Halle (Saale) Telefon : +49-345/55-25460 Forschungszentrum Dresden-Rossendorf Institut f?r Ionenstrahlphysik und Materialforschung Postfach 51 01 19 D-01314 Dresden, Germany E-Mail : n.hinsche at fzd.de From mambom1902 at yahoo.com Fri Jan 16 01:52:01 2009 From: mambom1902 at yahoo.com (loc duong ding) Date: Thu, 15 Jan 2009 16:52:01 -0800 (PST) Subject: [Pw_forum] Relax structure after getting the cell References: Message-ID: <131926.13075.qm@web38804.mail.mud.yahoo.com> > > I know enthalpy is the energy of system at constant pressure. > > really? I knew that H=E+PV I am sorry for my stupid confusion.? Anyway, what value (energy or enthalpy) should I use to compare the stability of ground state of two systems? ?----------------------------------------------- Loc Duong Dinh Ms-Ph.D Student Sungkyunkwan Advanced Institute of Nanotechnology, Sungkyunkwan University, Suwon, 440-746, Korea Email: mambom1902 at yahoo.com From ptanhphys at gmail.com Fri Jan 16 02:07:13 2009 From: ptanhphys at gmail.com (TuanAnh Pham) Date: Thu, 15 Jan 2009 17:07:13 -0800 Subject: [Pw_forum] Dielectric constant with Lyddane-Sachs-Teller relation In-Reply-To: <496CCC16.8000501@democritos.it> References: <496CCC16.8000501@democritos.it> Message-ID: Dear Paolo, Thank you very much for your suggestion. Best regards, Tuan Anh. On Tue, Jan 13, 2009 at 9:15 AM, Paolo Giannozzi wrote: > TuanAnh Pham wrote: > > > Could you please let me know what is the different between > > these two methods ? > > they should give the same results, when both methods apply > > > Is that correct that the LST relation only works for the system with > > sufficient symmetry? > > typically the LST relation is derived for high-symmetry crystals > with two atoms in the unit cell, in which all you have is one TO > and one LO nonzero frequency at Gamma. The LST relation can be > generalized, though. See for instance this old paper: > W. Cochran and R. A. Cowley, J. Phys. Chem. Solids 23, 447 (1962) > > 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 > -- Tuan Anh Pham Graduate Student University of California, Davis Phone: 530-752-0957 Homepage: http://angstrom.ucdavis.edu/ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090115/891557e6/attachment.htm From j.antonio.montoya at gmail.com Fri Jan 16 03:44:03 2009 From: j.antonio.montoya at gmail.com (j.antonio.montoya at gmail.com) Date: Fri, 16 Jan 2009 02:44:03 +0000 Subject: [Pw_forum] Relax structure after getting the cell Message-ID: <452190409-1232073859-cardhu_decombobulator_blackberry.rim.net-1186566870-@bxe249.bisx.prod.on.blackberry> Dear Loc, only at zero pressure H reduces to E. Therefore is clear that, for any pressure different than zero in your two systems, what you need to consider is enthalpy, i.e. compare E+PV for the two systems at the same pressure for both... And also check Wikipedia for more hints on thermodynamics. Javier Montoya PhD CIW, DC 20015 USA ------Original Message------ From: loc duong ding Sender: pw_forum-bounces at pwscf.org To: pw_forum at pwscf.org ReplyTo: PWSCF Forum Sent: Jan 15, 2009 7:52 PM Subject: Re: [Pw_forum] Relax structure after getting the cell > > I know enthalpy is the energy of system at constant pressure. > > really? I knew that H=E+PV I am sorry for my stupid confusion.? Anyway, what value (energy or enthalpy) should I use to compare the stability of ground state of two systems? ?----------------------------------------------- Loc Duong Dinh Ms-Ph.D Student Sungkyunkwan Advanced Institute of Nanotechnology, Sungkyunkwan University, Suwon, 440-746, Korea Email: mambom1902 at yahoo.com _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum Sent via BlackBerry by AT&T From njuxuyuehua at gmail.com Fri Jan 16 07:17:28 2009 From: njuxuyuehua at gmail.com (xu yuehua) Date: Fri, 16 Jan 2009 14:17:28 +0800 Subject: [Pw_forum] error :kpoint 7 ibnd 17 solve_linter: root not converged NaN In-Reply-To: <496DAC64.5030404@sissa.it> References: <496DAC64.5030404@sissa.it> Message-ID: thank you. i shall try it later. 2009/1/14 Gabriele Sclauzero > Hi, > > maybe try to reduce the mixing parameter alpha_mix, perhaps the default > is too high > for your system and causes problems in the potential-mixing subroutine. > You can also try to play with the nmix_ph parameter. > > GS > > xu yuehua wrote: > > error in the phonon calculation! (the 19th phonon problem)???? > > hi all : > > my system is a excited state, my aim is to get the phonon phonon at G > > point at this state,but i face > > the error is following in the 19 phonon calculation ? > > > > Representation # 19 mode # 19 > > Self-consistent Calculation > > iter # 1 total cpu time : 3876.6 secs av.it > > .: 3.9 > > thresh= 0.100E-01 alpha_mix = 0.700 |ddv_scf|^2 = 0.412E-07 > > iter # 2 total cpu time : 3896.1 secs av.it .: > > 10.3 > > thresh= 0.203E-04 alpha_mix = 0.700 |ddv_scf|^2 = 0.891E-09 > > iter # 3 total cpu time : 3918.1 secs av.it .: > > 12.9 > > thresh= 0.298E-05 alpha_mix = 0.700 |ddv_scf|^2 = 0.155E-08 > > iter # 4 total cpu time : 3939.5 secs av.it .: > > 11.9 > > thresh= 0.394E-05 alpha_mix = 0.700 |ddv_scf|^2 = 0.415E-11 > > kpoint 7 ibnd 17 solve_linter: root not converged NaN > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > from broyden : error # 3 > > factorization > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > from broyden : error # 3 > > factorization > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > from broyden : error # 3 > > factorization > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > the error is in the 19th phonon calculation ,before is OK. > > SO i do not know what happened. i need your help. > > > > -- > > Xu Yuehua > > physics Department of Nanjing university > > China > > > > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > -- > > > o ------------------------------------------------ o > | Gabriele Sclauzero, PhD Student | > | c/o: SISSA & CNR-INFM Democritos, | > | via Beirut 2-4, 34014 Trieste (Italy) | > | email: sclauzer at sissa.it | > | phone: +39 040 3787 511 | > | skype: gurlonotturno | > o ------------------------------------------------ o > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Xu Yuehua physics Department of Nanjing university China -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090116/c100c49d/attachment-0001.htm From mambom1902 at yahoo.com Fri Jan 16 07:38:12 2009 From: mambom1902 at yahoo.com (loc duong ding) Date: Thu, 15 Jan 2009 22:38:12 -0800 (PST) Subject: [Pw_forum] press_conv_thr value Message-ID: <697647.3915.qm@web38802.mail.mud.yahoo.com> Dear all, From my previous problem, I?think I have to reduce the press_conv_thr. Could you give me some?hints to chose this value? I use the default value (0.5 kbar) for my previous calculation. I?great appreciate?your instructions. Best regards, ?----------------------------------------------- Loc Duong Dinh Ms-Ph.D Student Sungkyunkwan Advanced Institute of Nanotechnology, Sungkyunkwan University, Suwon, 440-746, Korea Email: mambom1902 at yahoo.com From jasius_1 at yahoo.com Fri Jan 16 07:56:42 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Thu, 15 Jan 2009 22:56:42 -0800 (PST) Subject: [Pw_forum] specifyint angle in monoclinic cell In-Reply-To: <7b6913e90901102012g784a951p2b97f60c3de0dd6@mail.gmail.com> Message-ID: <674797.62952.qm@web54408.mail.yahoo.com> I am not sure hot to specify the angle in my monoclinic cell which is 111.668, should I input this value or the cos of it (which I did)? Specifying cos is somewhat odd. Other two angles are 90: &system ibrav=12, nat=62, ntyp=4, a=11.367, b=9.749, c=14.273, cosab=-0.36922777, ecutwfc=25.0 Also, is there a way to utilize crystal symmetry as this is an experimental structure and I know the space group of it Lastly, is there a way to setup a compounds jobs where after optimization band structure is run from the same input and optimized coordinates are read? I know I will be referred to the examples but I can't quite figure it out from example 1 thanks Jonas From smogunov at sissa.it Fri Jan 16 09:08:37 2009 From: smogunov at sissa.it (Alexander Smogunov) Date: Fri, 16 Jan 2009 09:08:37 +0100 Subject: [Pw_forum] Transport with pw_cond copper In-Reply-To: <003c01c97730$df9be040$9ed3a0c0$@unl.edu> References: <496CA0A7.3090007@freenet.de> <003c01c97730$df9be040$9ed3a0c0$@unl.edu> Message-ID: <200901160908.37658.smogunov@sissa.it> Dear Nicki, John On Thursday 15 January 2009 17:46, J. D. Burton wrote: > Nicki, > > PWCOND only works with tetragonal cells: either ibrav = 6 or use ibrav = 0 > with tetragonal CELL_PARAMETERS in your scf calculation. Notice all the > calculations in example12 use ibrav = 6. Why do you want simple cubic Cu > anyway? Cu is fcc, but you can?t use ibrav = 2 for PWCOND. You can set up > your scf calculation like this: Some time ago there was a discussion on the systems where PWCOND works, It works not only for tetragonal unit cells but for general MONOCLINIC systems where the transport direction (the z axis) is perpendicular to the xy plane. This is the only restriction on the geometry. It is right that some check on the allowed geometry would be useful to add before the code starts running... > > ibrav = 0 > celldm(1) = cubic lattice parameter for Cu in a.u. > > ATOMIC_POSITONS (crystal) > Cu 0.0 0.0 0.0 > Cu 0.5 0.5 0.5 > > CELL_PARAMETERS > sqrt(2)/2 0.00 0.00 > 0.00 sqrt(2)/2 0.00 > 0.00 0.00 1.00 > I think that should work. But you can use a simple tetragonal unit cell with also 2 atoms/cell as it is shown in the example12 for bulk Al and Ni. Best regards, Alexander > Anything useful that you might want to calculate (i.e. a junction) will be > have to be at least tetragonal anyway. (Although I don't think even > orthorhombic cells are implemented yet.) > > Cheers, > J. D. > > ************************************ > J. D. Burton, Ph.D. > jlz101 at unlserve.unl.edu > University of Nebraska Lincoln > Physics and Astronomy > Office Ph. (402) 472 2499 > Mobile Ph. (402) 419 9918 > 213C Ferguson Hall > ************************************ > "The job of a scientist is to generate wrong ideas as fast as possible." > -- Murray Gell-Mann > > From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On > Behalf Of Nicki F. Hinsche > Sent: Tuesday, January 13, 2009 8:10 AM > To: pw_forum at pwscf.org > Subject: [Pw_forum] Transport with pw_cond copper > > Hi, I want to make some test calculations of transport through 'bulk' > copper. Left and right lead shall be equal and the scattering region should > be the same material as the lead itself. My problem is, that I cant get any > results and I can't recognize any mistakes in my input cards -> my > inputcards are: > > # self-consistent calculation for Cu bulk > > ?&control > ??? calculation='scf' > ??? restart_mode='from_scratch', > ??? pseudo_dir = '$PSEUDO_DIR/', > ??? outdir='$TMP_DIR/' > ??? prefix='cu' > ?/ > ?&system > ??? ibrav = 1 > ??? celldm(1) = 6.719 > ??? nat= 1, > ??? ntyp= 1, > ??? ecutwfc = 40.0, > ??? occupations='smearing', > ??? smearing='methfessel-paxton', > ??? degauss=0.01 > ?/ > ?&electrons > ??? conv_thr = 1.0e-12 > ??? mixing_beta = 0.5 > ?/ > ATOMIC_SPECIES > ?Cu 63.55 Cu.pz-d-rrkjus.UPF > ATOMIC_POSITIONS > ?Cu 0.0 0.0 0.0 > K_POINTS (automatic) > ?16 16 16 0 0 0 > > > # complex bands of Copper along the 001 direction K_perp=0 > > cat > cu.cond.in << EOF > ?&inputcond > ??? oudir='$TMP_DIR/' > ??? prefixl='cu' > ??? band_file ='bands.cu' > ??? ikind=0 > ??? energy0=10.d0 > ??? denergy=-0.1d0 > ??? ewind=1.d0 > ??? epsproj=1.d-3 > ??? delgep = 1.d-12 > ??? cutplot = 2.d0 > ?/ > ??? 1 > ??? 0.0 0.0 1.0 > ??? 100 > > > # transmission calculation for the perfect copper contact > > cat > cutrans.cond.in << EOF > ?&inputcond > ??? outdir='$TMP_DIR/', > ??? prefixl='cu', > ??? prefixs='cu', > ??? tran_file='trans.cu', > ??? ikind=1, > ??? energy0=2.95d0, > ??? denergy=-0.1d0, > ??? ewind=1.d0, > ??? epsproj=1.d-3, > ??? nz1 = 1 > ?/ > ?? 1 > ??? 0.0? 0.0? 1.0 > ??? 100 > > ?? > I get the error message: > > ? ?? task #???????? 1 > ???? from in init_cond : error # 172992232 > ???? wrong nrztot > > > I would be thankful for every hint that you might give me... > > > > > ------------------------------------------------------- > Nicki F. Hinsche > Martin-Luther-Universit?t Halle-Wittenberg > Naturwissenschaftliche Fakult?t II - Chemie und Physik > Institut f?r Physik, Theoretische Physik > D-06099 Halle (Saale) > Telefon : +49-345/55-25460 > Forschungszentrum Dresden-Rossendorf > Institut f?r Ionenstrahlphysik und Materialforschung > Postfach 51 01 19 > D-01314 Dresden, Germany > E-Mail : n.hinsche at fzd.de > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From giannozz at democritos.it Fri Jan 16 09:28:17 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 16 Jan 2009 09:28:17 +0100 Subject: [Pw_forum] specifyint angle in monoclinic cell In-Reply-To: <674797.62952.qm@web54408.mail.yahoo.com> References: <674797.62952.qm@web54408.mail.yahoo.com> Message-ID: <49704521.8010902@democritos.it> Jonas Baltrusaitis wrote: > I am not sure hot to specify the angle in my monoclinic cell from the documentation: cosab = cosine of the angle between axis a and b > Also, is there a way to utilize crystal symmetry as this is an > experimental structure and I know the space group of it currently under development > Lastly, is there a way to setup a compounds jobs where after > optimization band structure is run from the same input and > optimized coordinates are read? not yet. You need to get the optimized coordinates from the output and put them into a new input (this is why the new coordinates are printed in a format that is compatible with input format) P. -- Paolo Giannozzi, Democritos and University of Udine, Italy From kazempoor2000 at yahoo.com Fri Jan 16 09:40:02 2009 From: kazempoor2000 at yahoo.com (ali kazempoor) Date: Fri, 16 Jan 2009 00:40:02 -0800 (PST) Subject: [Pw_forum] k grid Message-ID: <878258.76316.qm@web33107.mail.mud.yahoo.com> Dear all I am a user of espresso and also sax. For calculation of ?quasiparticle band structure,I need to homogenious k-mesh with equal weigth. My question is : can we define different k grid for 3 3 3 kpoint and 0 0 0 shifts? thanks a lot ? Ali Kazempour Isfahan university of technology,Phys.Dept ,Iran -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090116/69816c33/attachment.htm From smogunov at sissa.it Fri Jan 16 09:34:07 2009 From: smogunov at sissa.it (Alexander Smogunov) Date: Fri, 16 Jan 2009 09:34:07 +0100 Subject: [Pw_forum] How to use the parameters orbj_in and orbj_fin in calculation in pwcond.x In-Reply-To: <000801c9764e$d4a30570$3f7c620a@fduphy778548b0> References: <000801c9764e$d4a30570$3f7c620a@fduphy778548b0> Message-ID: <200901160934.07693.smogunov@sissa.it> Dear Wang On Wednesday 14 January 2009 14:48, Wang Qinjing wrote: > Dear all: > > I'd like to calculate the transmission of some special orbital. In the > INPUT_PWCOND.html I was told there are two parameters orbj_in and orbj_fin > as follows: I do not know what you mean by "the transmission of some special orbital". What the code calculates are the scalar products of transmission eigenchannel wave functions with projector wave functions centered at different atoms and provided by pseudo potentials. If you specify orbj_in and orbj_fin, the code will print out these products for projector functions in the range [orbj_in,orbj_fin]. It is sometimes useful to analize the symmetry of transmission eigenchannels, there is no, however, specific meaning in these numbers. Regards, Alexander > > ================ > orbj_in REAL > the initial orbital for projecting the transmission > > > orbj_fin REAL > the final orbital for projecting the transmission > ============ > > But what value can I set for these two parameters? What do they mean? Are > they related to the transmission of some special orbital? > > Thank you in advance > > Cheers > Wang > > ------------------------------------ > ???????????????????????????? > Wang Qinjing, Physics Department, > Fudan University From akohlmey at cmm.chem.upenn.edu Fri Jan 16 16:50:47 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 16 Jan 2009 10:50:47 -0500 (EST) Subject: [Pw_forum] k grid In-Reply-To: <878258.76316.qm@web33107.mail.mud.yahoo.com> References: <878258.76316.qm@web33107.mail.mud.yahoo.com> Message-ID: On Fri, 16 Jan 2009, ali kazempoor wrote: AK> Dear all AK> I am a user of espresso and also sax. For calculation of AK> ?quasiparticle band structure,I need to homogenious k-mesh with AK> equal weigth. My question is : can we define different k grid for 3 AK> 3 3 kpoint and 0 0 0 shifts? have you thought about entering the k-points manually? if you want the full grid, they are trivial to determine... axel. AK> thanks a lot AK> ? AK> Ali Kazempour AK> Isfahan university of technology,Phys.Dept ,Iran -- ======================================================================= 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 sclauzer at sissa.it Sat Jan 17 09:55:15 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Sat, 17 Jan 2009 09:55:15 +0100 Subject: [Pw_forum] How to use the parameters orbj_in and orbj_fin in calculation in pwcond.x In-Reply-To: <200901160934.07693.smogunov@sissa.it> References: <000801c9764e$d4a30570$3f7c620a@fduphy778548b0> <200901160934.07693.smogunov@sissa.it> Message-ID: <49719CF3.20000@sissa.it> Dear Wang, Alexander Smogunov wrote: > Dear Wang > > On Wednesday 14 January 2009 14:48, Wang Qinjing wrote: >> Dear all: >> >> I'd like to calculate the transmission of some special orbital. In the >> INPUT_PWCOND.html I was told there are two parameters orbj_in and orbj_fin >> as follows: > I do not know what you mean by "the transmission of some special orbital". > What the code calculates are the scalar products of transmission eigenchannel > wave functions with projector wave functions centered at different atoms and > provided by pseudo potentials. The orbitals onto which the scattering states are projected are the beta functions centered on atoms belonging to the scattering region only. They are listed after the lines: ----- Information about scattering region ----- noins = ... norb = ... norbf = ... nrz = ... iorb type ibeta ang. mom. m position (a_0) 1 1 1 1 1 taunew( 1)=( 0.0000 0.0000 0.0000) ... You can select the values for orbj_in,orbj_fin in the range [1,orbj_max], where orbj_max is the maximum value reached in the iorb column (and obviously orbj_in < orbj_fin). As Alexander pointed out, they do not give you any transmission, but they allow you to identify the symmetry of eigenchannels and eventually to separate the transmission according to the symmetry of the system by summing up the transmissions of all eigenchannels of a given symmetry. Hope this helps, GS > If you specify orbj_in and orbj_fin, the code > will print out these products for projector functions in the range > [orbj_in,orbj_fin]. It is sometimes useful to analize the symmetry of > transmission eigenchannels, there is no, however, specific meaning in these > numbers. > Regards, > Alexander > > >> ================ >> orbj_in REAL >> the initial orbital for projecting the transmission >> >> >> orbj_fin REAL >> the final orbital for projecting the transmission >> ============ >> >> But what value can I set for these two parameters? What do they mean? Are >> they related to the transmission of some special orbital? >> >> Thank you in advance >> >> Cheers >> Wang >> >> ------------------------------------ >> ???????????????????????????? >> Wang Qinjing, Physics Department, >> Fudan University > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From suzuki.furitsu at fs3.ecs.kyoto-u.ac.jp Sat Jan 17 11:05:17 2009 From: suzuki.furitsu at fs3.ecs.kyoto-u.ac.jp (=?ISO-2022-JP?B?GyRCTmtMWiEhSVROJxsoQg==?=) Date: Sat, 17 Jan 2009 19:05:17 +0900 Subject: [Pw_forum] Generating pseudo potentials containing GIPAW datas Message-ID: <123218673033067.490989626@mail3.ecs.kyoto-u.ac.jp> Dear PWSCF users and developer, I apologize if this has come up before, but I hope someone will have the advice for me about generating GIPAW pseudo potentials. I want to know generate pseudo potentials containing GIPAW datas similar to the ones used in the GIPAW_example directory. I found it needs lgipaw_reconstruction tag in input file, but the created UPF file by ld1.x doesn't have similar headers, and it can't be used with the gipaw.x (QE v.4.0.3). Would you please give me some advice? Sincerely, Furitsu Suzuki Department of Molecular Engineering, Graduate School of Engineering, Kyoto University, Japan E-mail: suzuki.furitsu at fs3.ecs.kyoto-u.ac.jp From meheut at impmc.jussieu.fr Sun Jan 18 03:45:37 2009 From: meheut at impmc.jussieu.fr (merlin meheut) Date: Sun, 18 Jan 2009 03:45:37 +0100 Subject: [Pw_forum] supercell calculations with matdyn Message-ID: <20090118034537.14326o4naulse381@www-ext.impmc.jussieu.fr> Dear all, I would like to use the supercell calculation with matdyn.x, however it does not seem to work. My input is: ------------------------------------------------ &input dos=.false., flfrc = 'matifc', flvec='', asr = 'no', l1=2, l2=1, l3=1, flfrq = 'frqbruc.pur', fltau='pos' /&end 1 0.333333 0.577350 -0.330831 ------------------------------------------------- When I put l1=1, this works perfectly fine, but with this input it gives me strange errors such as: *** glibc detected *** /home/meheut/program_files/espresso-3.2.3/pwtools//matdyn.x: malloc(): memory corruption: 0x0000000001c0f060 *** ======= Backtrace: ========= /lib/libc.so.6[0x7f7ec5b8ea14] /lib/libc.so.6(__libc_malloc+0x90)[0x7f7ec5b90360] etc.... It seems to me related to memory allocation, but I am not very good at compilation troubles so I do not know how to handle that. I tried versions 4.0, 3.2.3 on my computer (gfortran compiler, intel dual core) and version 3.0 on a rocks cluster. All show some trouble. I used previously this option with success on an old version (1.3.0), that I could not compile on my new computer, and I recall having customized the Makefile with something like LDFLAGS= (...) -bmaxdata:1024000 but I am not sure if it is related to this problem. However I tried something alike but this option was not recognized by gfortran (the compiler I used for the 1.3.0 was mpif77) . Thank you in advance for any hints, ----Merlin -- Merlin M?heut professional adress: 595 Charles Young Drive East 3806 Geology Building Los Angeles, California 90095-1597 United States of America tel: 310 825 7934 cell phone: 310 893 4253 ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From jasius_1 at yahoo.com Sun Jan 18 06:36:45 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Sat, 17 Jan 2009 21:36:45 -0800 (PST) Subject: [Pw_forum] vibrational frequencies In-Reply-To: <1DD51C87-8816-4E30-86BC-99AD591CE8E8@democritos.it> Message-ID: <324518.41123.qm@web54405.mail.yahoo.com> Can pwscf calculate vibrational frequencies (IR)? From lanhaiping at gmail.com Sun Jan 18 06:40:06 2009 From: lanhaiping at gmail.com (lan haiping) Date: Sun, 18 Jan 2009 13:40:06 +0800 Subject: [Pw_forum] vibrational frequencies In-Reply-To: <324518.41123.qm@web54405.mail.yahoo.com> References: <1DD51C87-8816-4E30-86BC-99AD591CE8E8@democritos.it> <324518.41123.qm@web54405.mail.yahoo.com> Message-ID: Hi, Please take a look at www.quantum-espresso.org or the userguide along with the QE package. A simple answer is "Yes"! And please supply your affiliation! best On Sun, Jan 18, 2009 at 1:36 PM, Jonas Baltrusaitis wrote: > Can pwscf calculate vibrational frequencies (IR)? > > > > _______________________________________________ > 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/20090118/6f979117/attachment.htm From anna.ferrari at unito.it Sun Jan 18 17:04:39 2009 From: anna.ferrari at unito.it (anna.ferrari at unito.it) Date: Sun, 18 Jan 2009 17:04:39 +0100 (CET) Subject: [Pw_forum] TD-DFT Message-ID: <11338.79.6.168.51.1232294679.squirrel@mail.unito.it> dear all, TD-DFT is implemented in PWSCF? In which codes is TDDFT implemented for solids? thanks a lot anna ferrari dip. chimica IFM univ. Torino From baroni at sissa.it Sun Jan 18 21:25:19 2009 From: baroni at sissa.it (Stefano Baroni) Date: Sun, 18 Jan 2009 21:25:19 +0100 Subject: [Pw_forum] TD-DFT In-Reply-To: <11338.79.6.168.51.1232294679.squirrel@mail.unito.it> References: <11338.79.6.168.51.1232294679.squirrel@mail.unito.it> Message-ID: <79CA99D6-A6B0-4CA2-A7E2-2A358FE68A1F@sissa.it> Anna: we do have an experimental code implementing TDDFT in QE, see: J. Chem. Phys. 128, 154105 (2008). It is not yet public, though. If you want to give it a try, please contact me or Baris Malcioglu who is reading me in cc. Cheers - Stefano On Jan 18, 2009, at 5:04 PM, anna.ferrari at unito.it wrote: > dear all, > TD-DFT is implemented in PWSCF? > In which codes is TDDFT implemented for solids? > > thanks a lot > > anna ferrari > dip. chimica IFM > univ. Torino > _______________________________________________ > 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/20090118/15a8d561/attachment.htm From jasius_1 at yahoo.com Sun Jan 18 23:27:29 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Sun, 18 Jan 2009 14:27:29 -0800 (PST) Subject: [Pw_forum] vibrational frequencies In-Reply-To: Message-ID: <274832.87529.qm@web54408.mail.yahoo.com> Well, I did my homework but couldn't find anything about calculating vibrational frequencies (infrared spectrum). Only phonon frequencies... IAny clues? JOnas Baltrusaitis University of IOwa --- On Sat, 1/17/09, lan haiping wrote: > From: lan haiping > Subject: Re: [Pw_forum] vibrational frequencies > To: jasius_1 at yahoo.com, "PWSCF Forum" > Date: Saturday, January 17, 2009, 9:40 PM > Hi, > Please take a look at www.quantum-espresso.org or the > userguide along with > the QE package. > A simple answer is "Yes"! > > And please supply your affiliation! > > best > > On Sun, Jan 18, 2009 at 1:36 PM, Jonas Baltrusaitis > wrote: > > > Can pwscf calculate vibrational frequencies (IR)? > > > > > > > > _______________________________________________ > > 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 From j.antonio.montoya at gmail.com Mon Jan 19 00:23:54 2009 From: j.antonio.montoya at gmail.com (j.antonio.montoya at gmail.com) Date: Sun, 18 Jan 2009 23:23:54 +0000 Subject: [Pw_forum] vibrational frequencies In-Reply-To: <274832.87529.qm@web54408.mail.yahoo.com> References: <274832.87529.qm@web54408.mail.yahoo.com> Message-ID: <1102475624-1232321057-cardhu_decombobulator_blackberry.rim.net-536706383-@bxe249.bisx.prod.on.blackberry> Dear Jonas, From the phonon calculation you can automatically get the IR and Raman active frequencies, and with a bit more of effort also the intensities. If you don't know the relation between phonons and the frequencies you are looking for, please read a book on solid state physics, or in IR spectroscopy. Javier Montoya CIW, DC 20015 USA Sent via BlackBerry by AT&T -----Original Message----- From: Jonas Baltrusaitis Date: Sun, 18 Jan 2009 14:27:29 To: PWSCF Forum; lan haiping Subject: Re: [Pw_forum] vibrational frequencies Well, I did my homework but couldn't find anything about calculating vibrational frequencies (infrared spectrum). Only phonon frequencies... IAny clues? JOnas Baltrusaitis University of IOwa --- On Sat, 1/17/09, lan haiping wrote: > From: lan haiping > Subject: Re: [Pw_forum] vibrational frequencies > To: jasius_1 at yahoo.com, "PWSCF Forum" > Date: Saturday, January 17, 2009, 9:40 PM > Hi, > Please take a look at www.quantum-espresso.org or the > userguide along with > the QE package. > A simple answer is "Yes"! > > And please supply your affiliation! > > best > > On Sun, Jan 18, 2009 at 1:36 PM, Jonas Baltrusaitis > wrote: > > > Can pwscf calculate vibrational frequencies (IR)? > > > > > > > > _______________________________________________ > > 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 From jasius_1 at yahoo.com Mon Jan 19 01:25:31 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Sun, 18 Jan 2009 16:25:31 -0800 (PST) Subject: [Pw_forum] vibrational frequencies In-Reply-To: <1102475624-1232321057-cardhu_decombobulator_blackberry.rim.net-536706383-@bxe249.bisx.prod.on.blackberry> Message-ID: <184126.77730.qm@web54405.mail.yahoo.com> I see the problem now. I am so used to calculating Hessian upon each atom displacement in non-plane wave type calculations. Are there any examples on how to setup calculations to eventually get IR frequencies? JOnas --- On Sun, 1/18/09, j.antonio.montoya at gmail.com wrote: > From: j.antonio.montoya at gmail.com > Subject: Re: [Pw_forum] vibrational frequencies > To: jasius_1 at yahoo.com, "PWSCF Forum" > Date: Sunday, January 18, 2009, 3:23 PM > Dear Jonas, > > From the phonon calculation you can automatically get the > IR and Raman active frequencies, and with a bit more of > effort also the intensities. > > If you don't know the relation between phonons and the > frequencies you are looking for, please read a book on solid > state physics, or in IR spectroscopy. > > Javier Montoya > CIW, DC 20015 > USA > Sent via BlackBerry by AT&T > > -----Original Message----- > From: Jonas Baltrusaitis > > Date: Sun, 18 Jan 2009 14:27:29 > To: PWSCF Forum; lan > haiping > Subject: Re: [Pw_forum] vibrational frequencies > > > Well, I did my homework but couldn't find anything > about calculating vibrational frequencies (infrared > spectrum). Only phonon frequencies... IAny clues? > > JOnas Baltrusaitis > University of IOwa > > > --- On Sat, 1/17/09, lan haiping > wrote: > > > From: lan haiping > > Subject: Re: [Pw_forum] vibrational frequencies > > To: jasius_1 at yahoo.com, "PWSCF Forum" > > > Date: Saturday, January 17, 2009, 9:40 PM > > Hi, > > Please take a look at www.quantum-espresso.org or the > > userguide along with > > the QE package. > > A simple answer is "Yes"! > > > > And please supply your affiliation! > > > > best > > > > On Sun, Jan 18, 2009 at 1:36 PM, Jonas Baltrusaitis > > wrote: > > > > > Can pwscf calculate vibrational frequencies > (IR)? > > > > > > > > > > > > _______________________________________________ > > > 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 From akohlmey at cmm.chem.upenn.edu Mon Jan 19 02:15:21 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sun, 18 Jan 2009 20:15:21 -0500 Subject: [Pw_forum] vibrational frequencies In-Reply-To: <184126.77730.qm@web54405.mail.yahoo.com> References: <1102475624-1232321057-cardhu_decombobulator_blackberry.rim.net-536706383-@bxe249.bisx.prod.on.blackberry> <184126.77730.qm@web54405.mail.yahoo.com> Message-ID: <7b6913e90901181715o57574b9s6939a5c8ead609ab@mail.gmail.com> for all practical purposes, there are no differences between phonons and infrared frequencies. axel. On 1/18/09, Jonas Baltrusaitis wrote: > I see the problem now. I am so used to calculating Hessian upon each atom > displacement in non-plane wave type calculations. Are there any examples on > how to setup calculations to eventually get IR frequencies? > > JOnas > > > --- On Sun, 1/18/09, j.antonio.montoya at gmail.com > wrote: > >> From: j.antonio.montoya at gmail.com >> Subject: Re: [Pw_forum] vibrational frequencies >> To: jasius_1 at yahoo.com, "PWSCF Forum" >> Date: Sunday, January 18, 2009, 3:23 PM >> Dear Jonas, >> >> From the phonon calculation you can automatically get the >> IR and Raman active frequencies, and with a bit more of >> effort also the intensities. >> >> If you don't know the relation between phonons and the >> frequencies you are looking for, please read a book on solid >> state physics, or in IR spectroscopy. >> >> Javier Montoya >> CIW, DC 20015 >> USA >> Sent via BlackBerry by AT&T >> >> -----Original Message----- >> From: Jonas Baltrusaitis >> >> Date: Sun, 18 Jan 2009 14:27:29 >> To: PWSCF Forum; lan >> haiping >> Subject: Re: [Pw_forum] vibrational frequencies >> >> >> Well, I did my homework but couldn't find anything >> about calculating vibrational frequencies (infrared >> spectrum). Only phonon frequencies... IAny clues? >> >> JOnas Baltrusaitis >> University of IOwa >> >> >> --- On Sat, 1/17/09, lan haiping >> wrote: >> >> > From: lan haiping >> > Subject: Re: [Pw_forum] vibrational frequencies >> > To: jasius_1 at yahoo.com, "PWSCF Forum" >> >> > Date: Saturday, January 17, 2009, 9:40 PM >> > Hi, >> > Please take a look at www.quantum-espresso.org or the >> > userguide along with >> > the QE package. >> > A simple answer is "Yes"! >> > >> > And please supply your affiliation! >> > >> > best >> > >> > On Sun, Jan 18, 2009 at 1:36 PM, Jonas Baltrusaitis >> > wrote: >> > >> > > Can pwscf calculate vibrational frequencies >> (IR)? >> > > >> > > >> > > >> > > _______________________________________________ >> > > 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 > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From hqzhou at nju.edu.cn Mon Jan 19 05:27:04 2009 From: hqzhou at nju.edu.cn (Huiqun Zhou) Date: Mon, 19 Jan 2009 12:27:04 +0800 Subject: [Pw_forum] TD-DFT References: <11338.79.6.168.51.1232294679.squirrel@mail.unito.it> Message-ID: <3A3D2F431ED14E0689BF194C7DF7F6C9@solarflare> Anna, You can try octopus: http://www.tddft.org/ Huiqun Zhou @Nanjing University, China ----- Original Message ----- From: To: Sent: Monday, January 19, 2009 12:04 AM Subject: [Pw_forum] TD-DFT > dear all, > TD-DFT is implemented in PWSCF? > In which codes is TDDFT implemented for solids? > > thanks a lot > > anna ferrari > dip. chimica IFM > univ. Torino > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From jasius_1 at yahoo.com Mon Jan 19 06:51:19 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Sun, 18 Jan 2009 21:51:19 -0800 (PST) Subject: [Pw_forum] compound jobs In-Reply-To: Message-ID: <996084.91666.qm@web54406.mail.yahoo.com> Is there a way to setup one input for a compound job? Say optimize geometry, then read coordinates and calculate frequency and finally produce several plots (band structure and DOS) I tried to accomplish first two steps but the second one just got neglected when job was run Jonas Baltrusaitis University of Iowa # self-consistent calculation &control calculation='relax', restart_mode='from_scratch', prefix='h2o' pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', outdir='/home/jbaltrus/h2o/' / &SYSTEM ibrav = 0, celldm(1) = 25.d0, nat = 3, ntyp = 2, ecutwfc = 30.0d0, ecutrho = 300.0d0 / &ELECTRONS conv_thr = 1.0d-8, mixing_mode = 'plain', mixing_beta = 0.3d0 / &ions ion_dynamics = ?none?, / ATOMIC_SPECIES O 15.9994 O.pbe-van_ak.UPF H 1.00794 H.pbe-van_ak.UPF ATOMIC_POSITIONS O -2.405637886 0.000000000 -1.440078763 H -2.405637886 0.759337000 -0.844035763 H -2.405637886 -0.759337000 -0.844035763 K_POINTS { gamma } CELL_PARAMETERS { cubic } 1.00000 0.00000 0.00000 0.00000 1.00000 0.00000 0.00000 0.00000 1.00000 # phonon calculation at Gamma &inputph tr2_ph=1.0d-14, prefix='h2o', epsil=.true., amass(1)=28.08, outdir='/home/jbaltrus/h2o/', fildyn='h2o_qe.dynG', / From roccad at gmail.com Mon Jan 19 08:05:27 2009 From: roccad at gmail.com (dario rocca) Date: Sun, 18 Jan 2009 23:05:27 -0800 Subject: [Pw_forum] TD-DFT In-Reply-To: <3A3D2F431ED14E0689BF194C7DF7F6C9@solarflare> References: <11338.79.6.168.51.1232294679.squirrel@mail.unito.it> <3A3D2F431ED14E0689BF194C7DF7F6C9@solarflare> Message-ID: For what concern linear response, real time codes (such as Octopus) are more computationally demanding with respect to the new TDDFT implementation in quantum espresso. In case Baris is not available, if Stefano still agrees, I can also provide you with a version. Dario On Sun, Jan 18, 2009 at 8:27 PM, Huiqun Zhou wrote: > Anna, > > You can try octopus: http://www.tddft.org/ > > Huiqun Zhou > @Nanjing University, China > > ----- Original Message ----- > From: > To: > Sent: Monday, January 19, 2009 12:04 AM > Subject: [Pw_forum] TD-DFT > > > > dear all, > > TD-DFT is implemented in PWSCF? > > In which codes is TDDFT implemented for solids? > > > > thanks a lot > > > > anna ferrari > > dip. chimica IFM > > univ. Torino > > _______________________________________________ > > 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/20090118/bbba0cfc/attachment.htm From anna.ferrari at unito.it Mon Jan 19 10:03:23 2009 From: anna.ferrari at unito.it (anna.ferrari at unito.it) Date: Mon, 19 Jan 2009 10:03:23 +0100 (CET) Subject: [Pw_forum] TD-DFT In-Reply-To: <79CA99D6-A6B0-4CA2-A7E2-2A358FE68A1F@sissa.it> References: <11338.79.6.168.51.1232294679.squirrel@mail.unito.it> <79CA99D6-A6B0-4CA2-A7E2-2A358FE68A1F@sissa.it> Message-ID: <12177.79.6.168.51.1232355803.squirrel@mail.unito.it> thanks a lot i'v already received some indications from baris anna > > > ------------------ Messaggio originale ------------------- > Oggetto: Re: [Pw_forum] TD-DFT > Da: "Stefano Baroni" > Data: Dom, 18 Gennaio 2009, 9:25 pm > A: "PWSCF Forum" > Cc: "O. Baris Malcioglu" > ---------------------------------------------------------- > > Anna: we do have an experimental code implementing TDDFT > in QE, see: > J. Chem. Phys. 128, 154105 (2008). It is not yet public, > though. If > you want to give it a try, please contact me or Baris > Malcioglu who is > reading me in cc. Cheers - Stefano > > > On Jan 18, 2009, at 5:04 PM, anna.ferrari at unito.it wrote: > >> dear all, >> TD-DFT is implemented in PWSCF? >> In which codes is TDDFT implemented for solids? >> >> thanks a lot >> >> anna ferrari >> dip. chimica IFM >> univ. Torino >> _______________________________________________ >> 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 > From hqzhou at nju.edu.cn Mon Jan 19 10:39:25 2009 From: hqzhou at nju.edu.cn (Huiqun Zhou) Date: Mon, 19 Jan 2009 17:39:25 +0800 Subject: [Pw_forum] TD-DFT References: <11338.79.6.168.51.1232294679.squirrel@mail.unito.it><3A3D2F431ED14E0689BF194C7DF7F6C9@solarflare> Message-ID: I indeed have been forbearing from complaining its computationally demanding character, so I'd like to try qe's experimental tddft function (^o^) Huiqun ----- Original Message ----- From: dario rocca To: PWSCF Forum Sent: Monday, January 19, 2009 3:05 PM Subject: Re: [Pw_forum] TD-DFT For what concern linear response, real time codes (such as Octopus) are more computationally demanding with respect to the new TDDFT implementation in quantum espresso. In case Baris is not available, if Stefano still agrees, I can also provide you with a version. Dario On Sun, Jan 18, 2009 at 8:27 PM, Huiqun Zhou wrote: Anna, You can try octopus: http://www.tddft.org/ Huiqun Zhou @Nanjing University, China ----- Original Message ----- From: To: Sent: Monday, January 19, 2009 12:04 AM Subject: [Pw_forum] TD-DFT > dear all, > TD-DFT is implemented in PWSCF? > In which codes is TDDFT implemented for solids? > > thanks a lot > > anna ferrari > dip. chimica IFM > univ. Torino > _______________________________________________ > 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/20090119/9d901f2a/attachment.htm From eyvaz_isaev at yahoo.com Mon Jan 19 10:48:39 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 19 Jan 2009 01:48:39 -0800 (PST) Subject: [Pw_forum] compound jobs In-Reply-To: <996084.91666.qm@web54406.mail.yahoo.com> Message-ID: <397574.24151.qm@web65702.mail.ac4.yahoo.com> Hi, I do not think so, they have to be done separately. As concerns your input file you have to split it into two scripts for self-consistent calculaculations, and then for phonons. They use different codes, as first, self-consistent step, usues pw.x, the second one uses ph.x. Of course, lines from "> # phonon calculation at Gamma" to "/" will be neglected as, most likely, they are not read, or ignored. Bests, P.S. What you suggest, should be expanded - then write a good paper and send it to PRL, at least . ------------------------------------------------------------------- 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, 1/19/09, Jonas Baltrusaitis wrote: > From: Jonas Baltrusaitis > Subject: [Pw_forum] compound jobs > To: "PWSCF Forum" > Date: Monday, January 19, 2009, 8:51 AM > Is there a way to setup one input for a compound job? Say > optimize geometry, then read coordinates and calculate > frequency and finally produce several plots (band structure > and DOS) > > I tried to accomplish first two steps but the second one > just got neglected when job was run > > Jonas Baltrusaitis > University of Iowa > > # self-consistent calculation > &control > calculation='relax', > restart_mode='from_scratch', > prefix='h2o' > > pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', > outdir='/home/jbaltrus/h2o/' > / > &SYSTEM > ibrav = 0, > celldm(1) = 25.d0, > nat = 3, > ntyp = 2, > ecutwfc = 30.0d0, > ecutrho = 300.0d0 > / > &ELECTRONS > conv_thr = 1.0d-8, > mixing_mode = 'plain', > mixing_beta = 0.3d0 > / > &ions > ion_dynamics = ?none?, > / > ATOMIC_SPECIES > O 15.9994 O.pbe-van_ak.UPF > H 1.00794 H.pbe-van_ak.UPF > ATOMIC_POSITIONS > O -2.405637886 0.000000000 -1.440078763 > H -2.405637886 0.759337000 -0.844035763 > H -2.405637886 -0.759337000 -0.844035763 > K_POINTS { gamma } > CELL_PARAMETERS { cubic } > 1.00000 0.00000 0.00000 > 0.00000 1.00000 0.00000 > 0.00000 0.00000 1.00000 > # phonon calculation at Gamma > &inputph > tr2_ph=1.0d-14, > prefix='h2o', > epsil=.true., > amass(1)=28.08, > outdir='/home/jbaltrus/h2o/', > fildyn='h2o_qe.dynG', > / > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From ammar112 at yahoo.com Mon Jan 19 12:44:12 2009 From: ammar112 at yahoo.com (benamrani ammar) Date: Mon, 19 Jan 2009 03:44:12 -0800 (PST) Subject: [Pw_forum] notification Message-ID: <124818.45723.qm@web51605.mail.re2.yahoo.com> Hi... my inbox get too heavy and charged with daily messages. please remove my subscription from your forum.or diminish your messages. greetings Ammar -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090119/8ddc533b/attachment.htm From Giovanni.Cantele at na.infn.it Mon Jan 19 13:19:27 2009 From: Giovanni.Cantele at na.infn.it (Giovanni Cantele) Date: Mon, 19 Jan 2009 13:19:27 +0100 Subject: [Pw_forum] notification In-Reply-To: <124818.45723.qm@web51605.mail.re2.yahoo.com> References: <124818.45723.qm@web51605.mail.re2.yahoo.com> Message-ID: <49746FCF.1080900@na.infn.it> benamrani ammar wrote: > Hi... > > my inbox get too heavy and charged with daily messages. > please remove my subscription from your forum. > maybe you should enter here: http://www.democritos.it/mailman/listinfo/pw_forum and follow "To unsubscribe from Pw_forum, get a password reminder, or change your subscription options enter your subscription email address:" > or diminish your messages. > maybe the above solution is better, it would accomplish both your inbox needs and those of all the other people who like discussing in this forum!!! ;-) giovanni -- Dr. Giovanni Cantele Coherentia CNR-INFM and Dipartimento di Scienze Fisiche Universita' di Napoli "Federico II" Complesso Universitario di Monte S. Angelo - Ed. 6 Via Cintia, I-80126, Napoli, Italy Phone: +39 081 676910 Fax: +39 081 676346 E-mail: giovanni.cantele at cnr.it giovanni.cantele at na.infn.it Web: http://people.na.infn.it/~cantele Research Group: http://www.nanomat.unina.it From quantumdft at gmail.com Mon Jan 19 13:50:33 2009 From: quantumdft at gmail.com (vega lew) Date: Mon, 19 Jan 2009 20:50:33 +0800 Subject: [Pw_forum] a question for projwfc.x Message-ID: <412f6c680901190450l1a995e4eib7e44c90583487f5@mail.gmail.com> Dear all, in input file description of projwfc.x, it's said that the code could identify the different contribution of Px, Py and Pz, or D sub orbitals. I wonder if there is a literature describe the algorithm and the physical meaning in detail. 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 ****************************************************************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, Jiangsu, China ****************************************************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090119/bc74109e/attachment.htm From zucco at dipteris.unige.it Mon Jan 19 15:46:22 2009 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Mon, 19 Jan 2009 15:46:22 +0100 Subject: [Pw_forum] Using wfc with scf after vc-relax Message-ID: <87EA57CF-EEF2-4139-9C3D-C235A44A2B37@dipteris.unige.it> Hello to all users, I'm using QE 4.0.4 with mpi on my macosx. I relaxed successfully a structure with variable-cell method obtaining a minimum with a diagonalized matrix of stresses. Interested in the postprocessing of wfc I used the files written at the end of the relaxation with projwfc or pp, but it seems that the wfc are not readable by those tools with the well known error messages found in the FAQ when cutoff of mesh nodes are not of the same size (dimensions do not match ..). Then, computing a single scf point with the equilibium (!) geometries of the converged vc-relax job, this geometries is not recognized as minimum, with a mean of the trace of stress eclose to 87 kbars, appearing mechanically unstable. Then the question is why vc-relax generate an odd wfc useless to other tools and the geometries of vc-relax are not computed as equilibrium geometries by scf run? ..or why I'm in this strange situation... I apologize if the questions hides lack of knowledge of mine. Thanks to all in advance 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 akohlmey at cmm.chem.upenn.edu Mon Jan 19 16:48:52 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 19 Jan 2009 10:48:52 -0500 (EST) Subject: [Pw_forum] compound jobs In-Reply-To: <996084.91666.qm@web54406.mail.yahoo.com> References: <996084.91666.qm@web54406.mail.yahoo.com> Message-ID: On Sun, 18 Jan 2009, Jonas Baltrusaitis wrote: JB> Is there a way to setup one input for a compound job? Say optimize JB> geometry, then read coordinates and calculate frequency and finally JB> produce several plots (band structure and DOS) have a look at the scripts in the example directory. this is how you'd set up multi-step jobs. optimized geometries are also printed to the output in a format that can be given directly to the input, so you could build input files using a template and some shell script magic to insert optimized geometries. JB> I tried to accomplish first two steps but the second one just got JB> neglected when job was run in general, it is a bad idea to build "one-size-fits-all-purposes" executables. the overhead to maintain them, is far too high. the system applied in Q-E is that you have a set of different executables, each of them performing one kind of task and then it is up to you to combine them. this way you are not limited but whatever combinations of tasks the programmers allow you to do. cheers, axel. JB> JB> Jonas Baltrusaitis JB> University of Iowa JB> JB> # self-consistent calculation JB> &control JB> calculation='relax', JB> restart_mode='from_scratch', JB> prefix='h2o' JB> pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', JB> outdir='/home/jbaltrus/h2o/' JB> / JB> &SYSTEM JB> ibrav = 0, JB> celldm(1) = 25.d0, JB> nat = 3, JB> ntyp = 2, JB> ecutwfc = 30.0d0, JB> ecutrho = 300.0d0 JB> / JB> &ELECTRONS JB> conv_thr = 1.0d-8, JB> mixing_mode = 'plain', JB> mixing_beta = 0.3d0 JB> / JB> &ions JB> ion_dynamics = ??none??, JB> / JB> ATOMIC_SPECIES JB> O 15.9994 O.pbe-van_ak.UPF JB> H 1.00794 H.pbe-van_ak.UPF JB> ATOMIC_POSITIONS JB> O -2.405637886 0.000000000 -1.440078763 JB> H -2.405637886 0.759337000 -0.844035763 JB> H -2.405637886 -0.759337000 -0.844035763 JB> K_POINTS { gamma } JB> CELL_PARAMETERS { cubic } JB> 1.00000 0.00000 0.00000 JB> 0.00000 1.00000 0.00000 JB> 0.00000 0.00000 1.00000 JB> # phonon calculation at Gamma JB> &inputph JB> tr2_ph=1.0d-14, JB> prefix='h2o', JB> epsil=.true., JB> amass(1)=28.08, JB> outdir='/home/jbaltrus/h2o/', JB> fildyn='h2o_qe.dynG', JB> / JB> JB> JB> JB> JB> JB> JB> _______________________________________________ JB> Pw_forum mailing list JB> Pw_forum at pwscf.org JB> http://www.democritos.it/mailman/listinfo/pw_forum JB> -- ======================================================================= 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 Jan 19 16:57:16 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 19 Jan 2009 10:57:16 -0500 (EST) Subject: [Pw_forum] a question for projwfc.x In-Reply-To: <412f6c680901190450l1a995e4eib7e44c90583487f5@mail.gmail.com> References: <412f6c680901190450l1a995e4eib7e44c90583487f5@mail.gmail.com> Message-ID: On Mon, 19 Jan 2009, vega lew wrote: VL> Dear all, VL> VL> in input file description of projwfc.x, it's said that the code could VL> identify the different contribution of Px, Py and Pz, or D sub orbitals. VL> VL> I wonder if there is a literature describe the algorithm and the physical VL> meaning in detail. have a look into a text book on linear algebra and look up projections. pw.x uses a plane wave basis set, projwfc.x projects the wavefunction expressed in the plane wave basis upon the atomic pseudowavefunctions stored in the pseudopotentials (which is not perfect, btw. some part of the wavefunction will remain unassigned, since you are using an incomplete basis). the physical meaning is that one tries to attribute the contributions from the individual atoms to the individual states. the process is similar to, say, expanding an irrational function through a polynomial and then try to describe the original function through the contributions of the individual "power of x" coefficients. cheers, axel. VL> VL> thank you for reading VL> VL> vega 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 akohlmey at cmm.chem.upenn.edu Mon Jan 19 17:05:35 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 19 Jan 2009 11:05:35 -0500 (EST) Subject: [Pw_forum] Using wfc with scf after vc-relax In-Reply-To: <87EA57CF-EEF2-4139-9C3D-C235A44A2B37@dipteris.unige.it> References: <87EA57CF-EEF2-4139-9C3D-C235A44A2B37@dipteris.unige.it> Message-ID: On Mon, 19 Jan 2009, Marino Vetuschi Zuccolini wrote: MVZ> Hello to all users, hello marino, MVZ> I'm using QE 4.0.4 with mpi on my macosx. MVZ> I relaxed successfully a structure with variable-cell method obtaining MVZ> a minimum with a diagonalized matrix of stresses. MVZ> MVZ> Interested in the postprocessing of wfc I used the files written at MVZ> the end of the relaxation with projwfc or pp, but it seems that the MVZ> wfc are not readable by those tools with the well known error messages MVZ> found in the FAQ when cutoff of mesh nodes are not of the same size MVZ> (dimensions do not match ..). that is most likely due to mistakes in your input files. hard to tell without seeing them, though. MVZ> Then, computing a single scf point with the equilibium (!) geometries MVZ> of the converged vc-relax job, this geometries is not recognized as MVZ> minimum, with a mean of the trace of stress eclose to 87 kbars, MVZ> appearing mechanically unstable. again, this is most likely due to a mistake in the input. have you checked for convergence of the stress tensor wrt. cutoffs in your vc-relax calculation. please note, that when the box size changes this also changes the "effective" cutoff, so you have to be _well_ converged or you get significant pulay stress. using vc-relax is convenient, but has one serious drawback: you run at a constant number of plane waves with a fixed grid and not at a constant cutoff (there is a method available to work around it to some degree). but in general, the requirements for convergence are much higher than for regular relaxation or single point calculations. also, when you do a single point calculation after vc-relax and your cell has changed, so that a different FFT grid would fit into the cell, it will be used by default and if you are not well converged wrt stree, it will show as a significant difference. cheers, axel. MVZ> Then the question is why vc-relax generate an odd wfc useless to other MVZ> tools and the geometries of vc-relax are not computed as equilibrium MVZ> geometries by scf run? ..or why I'm in this strange situation... MVZ> MVZ> I apologize if the questions hides lack of knowledge of mine. MVZ> MVZ> Thanks to all in advance MVZ> MVZ> m. MVZ> MVZ> MVZ> ******************************************************* MVZ> Marino Vetuschi Zuccolini MVZ> zucco at dipteris.unige.it MVZ> Researcher / Geochemist MVZ> Laboratory of Geochemistry MVZ> MVZ> DIPartimento per lo studio della TErra e delle sue RISorse - MVZ> Universit? di Genova MVZ> Tel. ++39 010 3538136 Fax. ++39 010 352169 MVZ> Corso Europa 26, 16132 - Genova - Italy MVZ> MVZ> MVZ> _______________________________________________ MVZ> Pw_forum mailing list MVZ> Pw_forum at pwscf.org MVZ> http://www.democritos.it/mailman/listinfo/pw_forum MVZ> -- ======================================================================= 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 quantumdft at gmail.com Mon Jan 19 19:20:28 2009 From: quantumdft at gmail.com (vega lew) Date: Tue, 20 Jan 2009 02:20:28 +0800 Subject: [Pw_forum] a question for projwfc.x In-Reply-To: References: <412f6c680901190450l1a995e4eib7e44c90583487f5@mail.gmail.com> Message-ID: <412f6c680901191020k1f606b4chdff3acc7c41ec121@mail.gmail.com> Dear sir, I'm very grateful to your help and time. Now, my understanding about this issue improved greatly with your help. best wishes, vega On Mon, Jan 19, 2009 at 11:57 PM, Axel Kohlmeyer < akohlmey at cmm.chem.upenn.edu> wrote: > On Mon, 19 Jan 2009, vega lew wrote: > > VL> Dear all, > VL> > VL> in input file description of projwfc.x, it's said that the code could > VL> identify the different contribution of Px, Py and Pz, or D sub > orbitals. > VL> > VL> I wonder if there is a literature describe the algorithm and the > physical > VL> meaning in detail. > > have a look into a text book on linear algebra and look up projections. > > pw.x uses a plane wave basis set, projwfc.x projects the wavefunction > expressed in the plane wave basis upon the atomic pseudowavefunctions > stored in the pseudopotentials (which is not perfect, btw. some part > of the wavefunction will remain unassigned, since you are using an > incomplete basis). the physical meaning is that one tries to attribute > the contributions from the individual atoms to the individual states. > > the process is similar to, say, expanding an irrational function > through a polynomial and then try to describe the original function > through the contributions of the individual "power of x" coefficients. > > cheers, > axel. > > VL> > VL> thank you for reading > VL> > VL> vega > 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. > -- ================================================================================== 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 ****************************************************************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, Jiangsu, China ****************************************************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090120/7d4ba853/attachment.htm From davidfarrell2008 at u.northwestern.edu Mon Jan 19 20:06:36 2009 From: davidfarrell2008 at u.northwestern.edu (David Farrell) Date: Mon, 19 Jan 2009 13:06:36 -0600 Subject: [Pw_forum] Compile error on OS X PPC w/ gfortran Message-ID: <5EF76550-B1A7-4F6C-B18D-3A8218F8FEBB@u.northwestern.edu> Hi all, I am trying to build QE on my laptop (an older G4 powerbook, OS X 10.5.6) with gfortran & gcc v 4.4.0 and came across the following errors when trying to build 'tools': fortran -O3 -x f95-cpp-input -D__XLF -D__GFORTRAN -D__FFTW - D__USE_INTERNAL_FFTW -I../include -I./ -I../Modules -I../iotk/src - I../PW -I../PH -c parser.f90 parser.f90:66.21: CHARACTER(LEN=1), AUTOMATIC :: sep1, sep2 1 Error: Invalid character in name at (1) parser.f90:67.12: INTEGER, AUTOMATIC :: j 1 Error: Invalid character in name at (1) parser.f90:81.11: DO j = 2, MAX( LEN( line ), 256 ) 1 Error: Symbol 'j' at (1) has no IMPLICIT type parser.f90:78.11: sep1 = char(32) ! ... blank character 1 Error: Symbol 'sep1' at (1) has no IMPLICIT type parser.f90:79.11: sep2 = char(9) ! ... tab character 1 Error: Symbol 'sep2' at (1) has no IMPLICIT type make[1]: *** [parser.o] Error 1 make: *** [mods] Error 2 Any ideas? seems like my compiler is choking on the formatting, but I am not really sure why. Thanks, Dave David E. Farrell Post-Doctoral Fellow Department of Materials Science and Engineering Northwestern University email: d-farrell2 at northwestern.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090119/d3bf373b/attachment.htm From giannozz at democritos.it Mon Jan 19 22:37:47 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Mon, 19 Jan 2009 22:37:47 +0100 Subject: [Pw_forum] Compile error on OS X PPC w/ gfortran In-Reply-To: <5EF76550-B1A7-4F6C-B18D-3A8218F8FEBB@u.northwestern.edu> References: <5EF76550-B1A7-4F6C-B18D-3A8218F8FEBB@u.northwestern.edu> Message-ID: <20090119223747.tzvkrbqfc4gss0gc@mail.democritos.it> Quoting David Farrell : > I am trying to build QE on my laptop (an older G4 powerbook, OS X > 10.5.6) with gfortran & gcc v 4.4.0 and came across the following > errors when trying to build 'tools': > > fortran -O3 -x f95-cpp-input -D__XLF -D__GFORTRAN -D__FFTW - > D__USE_INTERNAL_FFTW -I../include -I./ -I../Modules -I../iotk/src > - I../PW -I../PH -c parser.f90 remove "-D__XLF" Paolo ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From giannozz at democritos.it Mon Jan 19 22:46:15 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Mon, 19 Jan 2009 22:46:15 +0100 Subject: [Pw_forum] supercell calculations with matdyn In-Reply-To: <20090118034537.14326o4naulse381@www-ext.impmc.jussieu.fr> References: <20090118034537.14326o4naulse381@www-ext.impmc.jussieu.fr> Message-ID: <20090119224615.8sbly1h5kw8ossk4@mail.democritos.it> Quoting merlin meheut : > I would like to use the supercell calculation with matdyn.x I am afraid it hasn't been used since years. Anyway: gfortran is not very stable and cannot be trusted for bug reports. I'll try to figure out if it is a bug. > LDFLAGS= (...) -bmaxdata:1024000 this is for IBM xlf and has nothing to do with your problem Paolo ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From baroni at sissa.it Mon Jan 19 23:39:06 2009 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 19 Jan 2009 23:39:06 +0100 Subject: [Pw_forum] vibrational frequencies In-Reply-To: <7b6913e90901181715o57574b9s6939a5c8ead609ab@mail.gmail.com> References: <1102475624-1232321057-cardhu_decombobulator_blackberry.rim.net-536706383-@bxe249.bisx.prod.on.blackberry> <184126.77730.qm@web54405.mail.yahoo.com> <7b6913e90901181715o57574b9s6939a5c8ead609ab@mail.gmail.com> Message-ID: to add my 5 cents to the thread: not all vibrational mode are IR active. The IR activity is what is usually nicknamed as "oscillator strength" of the vibrational mode, which is a linear combination of the ionic effective (Born) charges, the linear coefficient being suitable components of the vibrational eigenmode. all the necessary ingredients are calculated by QE. the theory can be learnt in many textbooks. Stefano On Jan 19, 2009, at 2:15 AM, Axel Kohlmeyer wrote: > for all practical purposes, there are no differences between phonons > and infrared frequencies. axel. > > On 1/18/09, Jonas Baltrusaitis wrote: >> I see the problem now. I am so used to calculating Hessian upon >> each atom >> displacement in non-plane wave type calculations. Are there any >> examples on >> how to setup calculations to eventually get IR frequencies? >> >> JOnas >> >> >> --- On Sun, 1/18/09, j.antonio.montoya at gmail.com >> wrote: >> >>> From: j.antonio.montoya at gmail.com >>> Subject: Re: [Pw_forum] vibrational frequencies >>> To: jasius_1 at yahoo.com, "PWSCF Forum" >>> Date: Sunday, January 18, 2009, 3:23 PM >>> Dear Jonas, >>> >>> From the phonon calculation you can automatically get the >>> IR and Raman active frequencies, and with a bit more of >>> effort also the intensities. >>> >>> If you don't know the relation between phonons and the >>> frequencies you are looking for, please read a book on solid >>> state physics, or in IR spectroscopy. >>> >>> Javier Montoya >>> CIW, DC 20015 >>> USA >>> Sent via BlackBerry by AT&T >>> >>> -----Original Message----- >>> From: Jonas Baltrusaitis >>> >>> Date: Sun, 18 Jan 2009 14:27:29 >>> To: PWSCF Forum; lan >>> haiping >>> Subject: Re: [Pw_forum] vibrational frequencies >>> >>> >>> Well, I did my homework but couldn't find anything >>> about calculating vibrational frequencies (infrared >>> spectrum). Only phonon frequencies... IAny clues? >>> >>> JOnas Baltrusaitis >>> University of IOwa >>> >>> >>> --- On Sat, 1/17/09, lan haiping >>> wrote: >>> >>>> From: lan haiping >>>> Subject: Re: [Pw_forum] vibrational frequencies >>>> To: jasius_1 at yahoo.com, "PWSCF Forum" >>> >>>> Date: Saturday, January 17, 2009, 9:40 PM >>>> Hi, >>>> Please take a look at www.quantum-espresso.org or the >>>> userguide along with >>>> the QE package. >>>> A simple answer is "Yes"! >>>> >>>> And please supply your affiliation! >>>> >>>> best >>>> >>>> On Sun, Jan 18, 2009 at 1:36 PM, Jonas Baltrusaitis >>>> wrote: >>>> >>>>> Can pwscf calculate vibrational frequencies >>> (IR)? >>>>> >>>>> >>>>> >>>>> _______________________________________________ >>>>> 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 >> >> >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum >> >> > > > -- > = > ====================================================================== > Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http:// > www.cmm.upenn.edu > Center for Molecular Modeling -- University of Pennsylvania > Department of Chemistry, 231 S.34th Street, Philadelphia, PA > 19104-6323 > tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 > = > ====================================================================== > If you make something idiot-proof, the universe creates a better > idiot. > _______________________________________________ > 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/20090119/d26362ae/attachment-0001.htm From baroni at sissa.it Mon Jan 19 23:40:13 2009 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 19 Jan 2009 23:40:13 +0100 Subject: [Pw_forum] notification In-Reply-To: <124818.45723.qm@web51605.mail.re2.yahoo.com> References: <124818.45723.qm@web51605.mail.re2.yahoo.com> Message-ID: <58559199-E2BA-485C-B5D8-30E0723B69A7@sissa.it> please, do it by yourself greetings, Stefano Baroni On Jan 19, 2009, at 12:44 PM, benamrani ammar wrote: > Hi... > > my inbox get too heavy and charged with daily messages. > please remove my subscription from your forum.or diminish your > messages. > greetings > Ammar > > _______________________________________________ > 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/20090119/c69d219e/attachment.htm From davidfarrell2008 at u.northwestern.edu Tue Jan 20 00:41:52 2009 From: davidfarrell2008 at u.northwestern.edu (David Farrell) Date: Mon, 19 Jan 2009 17:41:52 -0600 Subject: [Pw_forum] Compile error on OS X PPC w/ gfortran In-Reply-To: <20090119223747.tzvkrbqfc4gss0gc@mail.democritos.it> References: <5EF76550-B1A7-4F6C-B18D-3A8218F8FEBB@u.northwestern.edu> <20090119223747.tzvkrbqfc4gss0gc@mail.democritos.it> Message-ID: <65504C46-71B2-43E6-B844-4828438DFB87@u.northwestern.edu> Wow - guess I should have seen that one. Thanks On Jan 19, 2009, at 3:37 PM, giannozz at democritos.it wrote: > Quoting David Farrell : > >> I am trying to build QE on my laptop (an older G4 powerbook, OS X >> 10.5.6) with gfortran & gcc v 4.4.0 and came across the following >> errors when trying to build 'tools': >> >> fortran -O3 -x f95-cpp-input -D__XLF -D__GFORTRAN -D__FFTW - >> D__USE_INTERNAL_FFTW -I../include -I./ -I../Modules -I../iotk/src >> - I../PW -I../PH -c parser.f90 > > remove "-D__XLF" > > Paolo > > > ---------------------------------------------------------------- > 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 David E. Farrell Post-Doctoral Fellow Department of Materials Science and Engineering Northwestern University email: d-farrell2 at northwestern.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090119/ec8e774d/attachment.htm From giannozz at democritos.it Tue Jan 20 15:57:26 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Tue, 20 Jan 2009 15:57:26 +0100 Subject: [Pw_forum] compound jobs In-Reply-To: References: <996084.91666.qm@web54406.mail.yahoo.com> Message-ID: <20090120155726.84mbhx2edcwww4o8@mail.democritos.it> Quoting Axel Kohlmeyer : > in general, it is a bad idea to build "one-size-fits-all-purposes" > executables. the overhead to maintain them, is far too high. true, but Jonas has a point. Making a band structure calculation should be easier. The problem with the current procedure is that it is too easy to make a mistake. Since I am not sure if my previous message went through (I am in Chile right now): for IR/Raman vibrational mode, see Doc/INPUT_Gamma Paolo (and please don't tell me to add my affiliation!) ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From giannozz at democritos.it Tue Jan 20 16:01:32 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Tue, 20 Jan 2009 16:01:32 +0100 Subject: [Pw_forum] notification In-Reply-To: <58559199-E2BA-485C-B5D8-30E0723B69A7@sissa.it> References: <124818.45723.qm@web51605.mail.re2.yahoo.com> <58559199-E2BA-485C-B5D8-30E0723B69A7@sissa.it> Message-ID: <20090120160132.uvkfw1jk00kwksos@mail.democritos.it> On Jan 19, 2009, at 12:44 PM, benamrani ammar wrote: > my inbox get too heavy and charged with daily messages. > please remove my subscription from your forum.or diminish your messages. from the same page used to subscribe, one can unsubscribe or set option 'digest" that sends just one message a day Paolo -- Democritos and Udine University, Italy ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From meheut at impmc.jussieu.fr Tue Jan 20 21:58:28 2009 From: meheut at impmc.jussieu.fr (merlin meheut) Date: Tue, 20 Jan 2009 21:58:28 +0100 Subject: [Pw_forum] supercell calculations with matdyn In-Reply-To: <20090119224615.8sbly1h5kw8ossk4@mail.democritos.it> References: <20090118034537.14326o4naulse381@www-ext.impmc.jussieu.fr> <20090119224615.8sbly1h5kw8ossk4@mail.democritos.it> Message-ID: <20090120215828.64486y5kroj6nr3o@www-ext.impmc.jussieu.fr> Hi Paolo, thanks a lot for looking at it! I am trying to look at it myself right now. Actually, I was not very accurate in my first mail: the output I mentionned was obtained with dos=.true. with dos=.false., which I can use, the error is different: -------------------------------------------- At line 289 of file matdyn.f90 Fortran runtime error: End of file ------------------------------------------ At first sight, it seems less important. Best regards, Merlin Meheut ESS Department at UCLA Quoting giannozz at democritos.it: > Quoting merlin meheut : > >> I would like to use the supercell calculation with matdyn.x > > I am afraid it hasn't been used since years. Anyway: gfortran > is not very stable and cannot be trusted for bug reports. I'll > try to figure out if it is a bug. > >> LDFLAGS= (...) -bmaxdata:1024000 > > this is for IBM xlf and has nothing to do with your problem > > Paolo > > ---------------------------------------------------------------- > 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 > -- Merlin M?heut professional adress: 595 Charles Young Drive East 3806 Geology Building Los Angeles, California 90095-1597 United States of America tel: 310 825 7934 cell phone: 310 893 4253 ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From giannozz at democritos.it Tue Jan 20 22:04:53 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Tue, 20 Jan 2009 22:04:53 +0100 Subject: [Pw_forum] supercell calculations with matdyn In-Reply-To: <20090120215828.64486y5kroj6nr3o@www-ext.impmc.jussieu.fr> References: <20090118034537.14326o4naulse381@www-ext.impmc.jussieu.fr> <20090119224615.8sbly1h5kw8ossk4@mail.democritos.it> <20090120215828.64486y5kroj6nr3o@www-ext.impmc.jussieu.fr> Message-ID: <20090120220453.i5jxez58joc0ocgg@mail.democritos.it> Quoting merlin meheut : > I am trying to look at it myself right now. excellent initiative. Something that was wrong for sure is the following: line 243 (of the cvs version) of matdyn.f90 was ALLOCATE ( tau (3, nat), ityp(nat), itau_blk(nat_blk) ) while it should be instead ALLOCATE ( tau (3, nat), ityp(nat), itau_blk(nat) ) > At line 289 of file matdyn.f90 > Fortran runtime error: End of file this is quite self-explanatory Paolo --- Democritos and University of Udine ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From meheut at impmc.jussieu.fr Tue Jan 20 22:15:30 2009 From: meheut at impmc.jussieu.fr (merlin meheut) Date: Tue, 20 Jan 2009 22:15:30 +0100 Subject: [Pw_forum] supercell calculations with matdyn In-Reply-To: <20090120220453.i5jxez58joc0ocgg@mail.democritos.it> References: <20090118034537.14326o4naulse381@www-ext.impmc.jussieu.fr> <20090119224615.8sbly1h5kw8ossk4@mail.democritos.it> <20090120215828.64486y5kroj6nr3o@www-ext.impmc.jussieu.fr> <20090120220453.i5jxez58joc0ocgg@mail.democritos.it> Message-ID: <20090120221530.9027164q9o90o44y@www-ext.impmc.jussieu.fr> > excellent initiative. Something that was wrong for sure is the > following: line 243 (of the cvs version) of matdyn.f90 was > > ALLOCATE ( tau (3, nat), ityp(nat), itau_blk(nat_blk) ) > > while it should be instead > > ALLOCATE ( tau (3, nat), ityp(nat), itau_blk(nat) ) > This seems to work! I am going to look at it in details. >> At line 289 of file matdyn.f90 >> Fortran runtime error: End of file > > this is quite self-explanatory > Sorry that had nothing to do with the problem! Merlin Meheut ESS department at UCLA -- Merlin M?heut professional adress: 595 Charles Young Drive East 3806 Geology Building Los Angeles, California 90095-1597 United States of America tel: 310 825 7934 cell phone: 310 893 4253 ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From lataprem29 at gmail.com Wed Jan 21 09:55:33 2009 From: lataprem29 at gmail.com (premlata pandit) Date: Wed, 21 Jan 2009 14:25:33 +0530 Subject: [Pw_forum] generating pseudopotential problem Message-ID: <6f289440901210055m7d3444c0xc55950bd7370c174@mail.gmail.com> Dear Pwscf User We have generated ultrasoft pseudopotential for Pm by using following configuration 4F 4 3 5.00 0.00 2.10 2.10 2.50 4F 4 3 0.00 -0.10 2.10 2.10 3.50 6S 1 0 2.00 0.00 2.40 2.40 0.50 By using above configuration, we have got output. But when we used split 4F configuration ( i.e. more than 2 ), we found some error. The input file Pm2.in, which is used to generated pp, is also attached with this mail. There are some question - 1. is input file and generated pseudopotential is correct? 2. Why we did not get results, when we used split 4F configuration ? anyone can help me , i would be greatful. 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/20090121/eaafb9fd/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: pm2.in Type: application/octet-stream Size: 436 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20090121/eaafb9fd/attachment.obj From zucco at dipteris.unige.it Wed Jan 21 11:38:43 2009 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Wed, 21 Jan 2009 11:38:43 +0100 Subject: [Pw_forum] LDA+U and spilling parameter Message-ID: <31E06CF8-3C51-476C-9AC7-E51966F18C5F@dipteris.unige.it> Hello to all, stemming from a De Gironcoli statement about the "spilling param" in this forum , I would like to know if LDA+U theory can be an exception to realize a spilling parameter far to 1.D-2. Thanks 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 sclauzer at sissa.it Wed Jan 21 16:22:07 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Wed, 21 Jan 2009 16:22:07 +0100 Subject: [Pw_forum] generating pseudopotential problem In-Reply-To: <6f289440901210055m7d3444c0xc55950bd7370c174@mail.gmail.com> References: <6f289440901210055m7d3444c0xc55950bd7370c174@mail.gmail.com> Message-ID: <49773D9F.9080501@sissa.it> I've not much experience in producing PP, but I recognize some problems in your input file. I can point them out, while you wait for a more definitive reply... premlata pandit wrote: > Dear Pwscf User > We have generated ultrasoft pseudopotential for Pm by using following > configuration > 4F 4 3 5.00 0.00 2.10 2.10 2.50 > 4F 4 3 0.00 -0.10 2.10 2.10 3.50 These two rows look strange to me. If you want to generate US PP, you need at least 2 energies for each channels and specify the 2nd cutoff radius bigger than the first. They are the same for the channels you want pseudize within the norm-conserving recipe. I don't understand if your input above is for US with rel=1 or for NC with rel=2. In the relativistic case (rel=2), for each l, the number of projectors doubles (you have one with j=l+1/2 and another with j=l-1/2, where j is the number you specify in the last column). So, your input for relativistic USPP should look more like: 4F 4 3 5.00 0.00 2.10 2.40 2.50 4F 4 3 0.00 -0.10 2.10 2.40 2.50 4F 4 3 0.00 0.00 2.10 2.40 3.50 4F 4 3 0.00 -0.10 2.10 2.40 3.50 Of course this is not sufficient to get a relativistic US-PP, you need to play a lot with cut-off radius and energies as well as maybe adding p or d projectors... it is usually a difficult task to obtain a working PP for rare earth elements. > 6S 1 0 2.00 0.00 2.40 2.40 0.50 > > By using above configuration, we have got output. This does not mean that the PP will work correctly elsewhere than the atom in the configuration used for generating the PP... My suggestion is to look at the input files for PP generation in espresso/atomic_doc/pseudo_library/LDA/REL/ and understand what you need to specify. Regards GS >But when we used > split 4F configuration ( i.e. more than 2 ), we found some error. > The input file Pm2.in, which is used to generated pp, is also attached > with this mail. > There are some question - > 1. is input file and generated pseudopotential is correct? > 2. Why we did not get results, when we used split 4F configuration ? > anyone can help me , i would be greatful. > > Premlata Pandit > Ph.d. student, > Barkatullah university, > Bhopal 462026, > MP, India > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From chenweiguang82 at gmail.com Wed Jan 21 17:23:46 2009 From: chenweiguang82 at gmail.com (Weiguang Chen) Date: Thu, 22 Jan 2009 00:23:46 +0800 Subject: [Pw_forum] error got by using program pp.x to deal with the input file Message-ID: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> Hi, I used the postproceeding program pp.x to deal with the input file as follows, but get the error result: pp.out: Program POST-PROC v.4.0.4 starts ... Today is 22Jan2009 at 13:58:56 negative rho (up, down): 0.313E-03 0.353E-03 Calling punch_plot, plot_num = 10 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from local_dos : error # 1 gaussian broadening needed %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... and the pp.in: pp.in: &inputpp prefix = "E", outdir = '/home/name/E/tmp' , filplot = 'spinup-fermi' plot_num = 10 emin =-0.5 emax = 0.5 spin_component = 1 / &plot nfile = 1 filepp(1) = 'spinup-fermi' weight(1) = 10.0 iflag = 3 output_format = 5 fileout = 'fermi.rho.xsf' e1(1) =5.0, e1(2)=0.0, e1(3) = 0.0, e2(1) =0.0, e2(2)=5.0, e2(3) = 0.0, e3(1) =0.0, e3(2)=0.0, e3(3) = 5.0, nx=50, ny=50, nz=50 / Thanks And Happy Spring Festival (Chinese New Year, i.e. Cow Year for this year) -- Best Wishes ChenWeiguang ************************************************ # Chen, Weiguang # # Postgraduate, Ph. D # 75 University Road, Physics Buliding # 218 # School of Physics & Engineering # Zhengzhou University # Zhengzhou, Henan 450052 CHINA # # Tel: 86-13203730117; # E-mail:chenweiguang82 at gmail.com; # chenweiguang82 at qq.com #********************************************** From akohlmey at cmm.chem.upenn.edu Wed Jan 21 17:25:23 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 21 Jan 2009 11:25:23 -0500 (EST) Subject: [Pw_forum] error got by using program pp.x to deal with the input file In-Reply-To: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> References: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> Message-ID: On Thu, 22 Jan 2009, Weiguang Chen wrote: WC> Hi, WC> I used the postproceeding program pp.x to deal with the input file as WC> follows, but get the error result: ...and what is your problem with that? do you disagree with why it refuses to run? and if yes, please explain? the code already tells you what it thinks is wrong. cheers, axel. WC> pp.out: WC> WC> Program POST-PROC v.4.0.4 starts ... WC> Today is 22Jan2009 at 13:58:56 WC> WC> negative rho (up, down): 0.313E-03 0.353E-03 WC> WC> Calling punch_plot, plot_num = 10 WC> WC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% WC> from local_dos : error # 1 WC> gaussian broadening needed WC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% WC> WC> stopping ... WC> WC> WC> and the pp.in: WC> WC> pp.in: WC> WC> &inputpp WC> prefix = "E", WC> outdir = '/home/name/E/tmp' , WC> filplot = 'spinup-fermi' WC> plot_num = 10 WC> emin =-0.5 WC> emax = 0.5 WC> spin_component = 1 WC> / WC> &plot WC> nfile = 1 WC> filepp(1) = 'spinup-fermi' WC> weight(1) = 10.0 WC> iflag = 3 WC> output_format = 5 WC> fileout = 'fermi.rho.xsf' WC> e1(1) =5.0, e1(2)=0.0, e1(3) = 0.0, WC> e2(1) =0.0, e2(2)=5.0, e2(3) = 0.0, WC> e3(1) =0.0, e3(2)=0.0, e3(3) = 5.0, WC> nx=50, ny=50, nz=50 WC> / WC> WC> Thanks WC> And Happy Spring Festival (Chinese New Year, i.e. Cow Year for this year) WC> -- ======================================================================= 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 chenweiguang82 at gmail.com Wed Jan 21 17:46:53 2009 From: chenweiguang82 at gmail.com (Weiguang Chen) Date: Thu, 22 Jan 2009 00:46:53 +0800 Subject: [Pw_forum] error got by using program pp.x to deal with the input file In-Reply-To: References: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> Message-ID: <9e6d976a0901210846y688ffea7s6c539fee26ca0aae@mail.gmail.com> Hi, Thank you for your fast reply. I am a newbie about pwscf and i want to know what parameter should been modified. Could you tell me more detail why this problem happen? Thanks ChenWeiguang On Thu, Jan 22, 2009 at 12:25 AM, Axel Kohlmeyer wrote: > On Thu, 22 Jan 2009, Weiguang Chen wrote: > > WC> Hi, > WC> I used the postproceeding program pp.x to deal with the input file as > WC> follows, but get the error result: > > ...and what is your problem with that? do you disagree > with why it refuses to run? and if yes, please explain? > > the code already tells you what it thinks is wrong. > > cheers, > axel. > > WC> pp.out: > WC> > WC> Program POST-PROC v.4.0.4 starts ... > WC> Today is 22Jan2009 at 13:58:56 > WC> > WC> negative rho (up, down): 0.313E-03 0.353E-03 > WC> > WC> Calling punch_plot, plot_num = 10 > WC> > WC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > WC> from local_dos : error # 1 > WC> gaussian broadening needed > WC> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > WC> > WC> stopping ... > WC> > WC> > WC> and the pp.in: > WC> > WC> pp.in: > WC> > WC> &inputpp > WC> prefix = "E", > WC> outdir = '/home/name/E/tmp' , > WC> filplot = 'spinup-fermi' > WC> plot_num = 10 > WC> emin =-0.5 > WC> emax = 0.5 > WC> spin_component = 1 > WC> / > WC> &plot > WC> nfile = 1 > WC> filepp(1) = 'spinup-fermi' > WC> weight(1) = 10.0 > WC> iflag = 3 > WC> output_format = 5 > WC> fileout = 'fermi.rho.xsf' > WC> e1(1) =5.0, e1(2)=0.0, e1(3) = 0.0, > WC> e2(1) =0.0, e2(2)=5.0, e2(3) = 0.0, > WC> e3(1) =0.0, e3(2)=0.0, e3(3) = 5.0, > WC> nx=50, ny=50, nz=50 > WC> / > WC> > WC> Thanks > WC> And Happy Spring Festival (Chinese New Year, i.e. Cow Year for this year) > WC> > > -- > ======================================================================= > 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. > -- Best Wishes ChenWeiguang ************************************************ # Chen, Weiguang # # Postgraduate, Ph. D # 75 University Road, Physics Buliding # 218 # School of Physics & Engineering # Zhengzhou University # Zhengzhou, Henan 450052 CHINA # # Tel: 86-13203730117; # E-mail:chenweiguang82 at gmail.com; # chenweiguang82 at qq.com #********************************************** From akohlmey at cmm.chem.upenn.edu Wed Jan 21 19:05:53 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 21 Jan 2009 13:05:53 -0500 (EST) Subject: [Pw_forum] error got by using program pp.x to deal with the input file In-Reply-To: <9e6d976a0901210846y688ffea7s6c539fee26ca0aae@mail.gmail.com> References: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> <9e6d976a0901210846y688ffea7s6c539fee26ca0aae@mail.gmail.com> Message-ID: On Thu, 22 Jan 2009, Weiguang Chen wrote: WC> Hi, WC> Thank you for your fast reply. WC> I am a newbie about pwscf and i want to know what parameter should WC> been modified. Could you tell me more detail why this problem happen? i don't understand your question: this is actually an error message that is not kryptic and tells you exactly what is wrong: for the kind of postprocessing that you want to do, the use gaussian broadening in the preceding calculation is required. as simple as that. ;-) axel. WC> WC> Thanks WC> WC> ChenWeiguang WC> -- ======================================================================= 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 Wed Jan 21 19:55:12 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Wed, 21 Jan 2009 19:55:12 +0100 Subject: [Pw_forum] error got by using program pp.x to deal with the input file In-Reply-To: <9e6d976a0901210846y688ffea7s6c539fee26ca0aae@mail.gmail.com> References: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> <9e6d976a0901210846y688ffea7s6c539fee26ca0aae@mail.gmail.com> Message-ID: <20090121195512.pl1arvsm8kow840k@mail.democritos.it> Quoting Weiguang Chen : > I am a newbie about pwscf and i want to know what parameter should > been modified. Could you tell me more detail why this problem happen? documentation in Doc/INPUT_PP.* Paolo ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From mazouz_moulay at yahoo.fr Wed Jan 21 20:08:01 2009 From: mazouz_moulay at yahoo.fr (mazouz moulay) Date: Wed, 21 Jan 2009 19:08:01 +0000 (GMT) Subject: [Pw_forum] Request QHA code Message-ID: <696653.1196.qm@web25507.mail.ukl.yahoo.com> Dear Eyvaz, ? I would like to have the?QHA code distribution?(with Debye.x) if it is ready, if not yet send me a copy of the code. Thank you in avance ? Best regards Mazouz Moulay University of Sciences and Technology (Oran-Algeria) Physics department? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090121/dae45dce/attachment-0001.htm From chenweiguang82 at gmail.com Wed Jan 21 23:45:45 2009 From: chenweiguang82 at gmail.com (Weiguang Chen) Date: Thu, 22 Jan 2009 06:45:45 +0800 Subject: [Pw_forum] error got by using program pp.x to deal with the input file In-Reply-To: <20090121195512.pl1arvsm8kow840k@mail.democritos.it> References: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> <9e6d976a0901210846y688ffea7s6c539fee26ca0aae@mail.gmail.com> <20090121195512.pl1arvsm8kow840k@mail.democritos.it> Message-ID: <9e6d976a0901211445n7277ca56hd7e9d0f035b0c365@mail.gmail.com> Hi, Thanks for your respondence quickly. But there is no tag to set gaussian broadening in the input-file of pp.x . Is this mean that I should set this value in scf calculation? In fact, I have used gaussian broadening in scf calculation. I set the gaussian broadening in scf calculation as follows: occupations='smearing', smearing='gauss', degauss=0.01, Thanks Chen Weiguang On Thu, Jan 22, 2009 at 2:55 AM, wrote: > Quoting Weiguang Chen : > >> I am a newbie about pwscf and i want to know what parameter should >> been modified. Could you tell me more detail why this problem happen? > > documentation in Doc/INPUT_PP.* > > Paolo > > > ---------------------------------------------------------------- > 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 > -- Best Wishes ChenWeiguang ************************************************ # Chen, Weiguang # # Postgraduate, Ph. D # 75 University Road, Physics Buliding # 218 # School of Physics & Engineering # Zhengzhou University # Zhengzhou, Henan 450052 CHINA # # Tel: 86-13203730117; # E-mail:chenweiguang82 at gmail.com; # chenweiguang82 at qq.com #********************************************** From nemeth at anl.gov Thu Jan 22 04:10:57 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Wed, 21 Jan 2009 21:10:57 -0600 (CST) Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: Hi, I would like to obtain the wavefunction coefficients u(i,k,G), where "i" is the band-index, "k" is the k-space vector used in the discretization of the Brillouin-zone in a PW calculation and "G" is the plane-wave basis function's wave-vector (assuming a complex form as exp[i(k+G)r] ). Is there any utility program that is independent from the rest of PWSCF and would provide some sort of tabulation on u(i,k,G) as a function of i,k and G , for a parallel calculation? Perhaps a quick quide to the format of this information in the *.wfc* files written by a parallel run? I would appreciate any hint that could save me some time from going through the source code for this info. Thanks a lot in advance: Karoly From akohlmey at cmm.chem.upenn.edu Thu Jan 22 04:40:51 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 21 Jan 2009 22:40:51 -0500 (EST) Subject: [Pw_forum] error got by using program pp.x to deal with the input file In-Reply-To: <9e6d976a0901211445n7277ca56hd7e9d0f035b0c365@mail.gmail.com> References: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> <9e6d976a0901210846y688ffea7s6c539fee26ca0aae@mail.gmail.com> <20090121195512.pl1arvsm8kow840k@mail.democritos.it> <9e6d976a0901211445n7277ca56hd7e9d0f035b0c365@mail.gmail.com> Message-ID: On Thu, 22 Jan 2009, Weiguang Chen wrote: WC> Hi, WC> Thanks for your respondence quickly. WC> But there is no tag to set gaussian broadening in the input-file of WC> pp.x . Is this mean that I should set this value in scf calculation? where else? WC> In fact, I have used gaussian broadening in scf calculation. WC> WC> I set the gaussian broadening in scf calculation as follows: WC> occupations='smearing', WC> smearing='gauss', WC> degauss=0.01, please provide a complete script or set of inputs that reproduces this behavior, without requiring access to a supercomputer. perhaps then somebody can tell you what went wrong. i just slapped together a stupid example and "it works for me(tm)" have you run some other calculation in between that would modify the saved wavefunction? axel. WC> WC> Thanks WC> Chen Weiguang WC> On Thu, Jan 22, 2009 at 2:55 AM, wrote: WC> > Quoting Weiguang Chen : WC> > WC> >> I am a newbie about pwscf and i want to know what parameter should WC> >> been modified. Could you tell me more detail why this problem happen? WC> > WC> > documentation in Doc/INPUT_PP.* WC> > WC> > Paolo WC> > WC> > WC> > ---------------------------------------------------------------- WC> > This message was sent using IMP, the Internet Messaging Program. WC> > WC> > _______________________________________________ WC> > Pw_forum mailing list WC> > Pw_forum at pwscf.org WC> > http://www.democritos.it/mailman/listinfo/pw_forum WC> > WC> WC> WC> WC> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From baroni at sissa.it Thu Jan 22 08:05:05 2009 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 22 Jan 2009 08:05:05 +0100 Subject: [Pw_forum] error got by using program pp.x to deal with the input file In-Reply-To: References: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> <9e6d976a0901210846y688ffea7s6c539fee26ca0aae@mail.gmail.com> <20090121195512.pl1arvsm8kow840k@mail.democritos.it> <9e6d976a0901211445n7277ca56hd7e9d0f035b0c365@mail.gmail.com> Message-ID: <53588269-F7A5-462F-9D2E-A500AFBCA4FD@sissa.it> In addition to Axel's suggestions, I would also recommend to use a minimum critical mind. Much to my blame, I know very little of the functionality of pp.x, but from your post I can guess that you are trying to calculate some density of states (DOS): ain't you? Have you aver esked yourself HOW a continuous DOS is calculated from a number of DISCRETE tates? (yes, QE as well as virtually any electronic- structure codes approximate continuous spectra with a number of discrete states) Have you ever asked yourself what Gaussian smearing is meant for? (answer: to broaden a discreet peak into a pseduo- continuous spectrum) If not, do it now. From the answer, you will get a hint for your problem. Stefano PS: broadening is not the only way to obtain DOS's. In a periodic system the tetrahedron method does it better, but I am not sure that this method is used by pp.x in you case. In any event, google "tetrahedron" and "density of states" and get an idea of what the method is/does. On Jan 22, 2009, at 4:40 AM, Axel Kohlmeyer wrote: > On Thu, 22 Jan 2009, Weiguang Chen wrote: > > WC> Hi, > WC> Thanks for your respondence quickly. > WC> But there is no tag to set gaussian broadening in the input-file > of > WC> pp.x . Is this mean that I should set this value in scf > calculation? > > where else? > > WC> In fact, I have used gaussian broadening in scf calculation. > WC> > WC> I set the gaussian broadening in scf calculation as follows: > WC> occupations='smearing', > WC> smearing='gauss', > WC> degauss=0.01, > > please provide a complete script or set of inputs that > reproduces this behavior, without requiring access to > a supercomputer. perhaps then somebody can tell you > what went wrong. > > i just slapped together a stupid example and "it works for me(tm)" > > have you run some other calculation in between that > would modify the saved wavefunction? > > axel. > > > WC> > WC> Thanks > WC> Chen Weiguang > WC> On Thu, Jan 22, 2009 at 2:55 AM, wrote: > WC> > Quoting Weiguang Chen : > WC> > > WC> >> I am a newbie about pwscf and i want to know what parameter > should > WC> >> been modified. Could you tell me more detail why this problem > happen? > WC> > > WC> > documentation in Doc/INPUT_PP.* > WC> > > WC> > Paolo > WC> > > WC> > > WC> > ---------------------------------------------------------------- > WC> > This message was sent using IMP, the Internet Messaging Program. > WC> > > WC> > _______________________________________________ > WC> > Pw_forum mailing list > WC> > Pw_forum at pwscf.org > WC> > http://www.democritos.it/mailman/listinfo/pw_forum > WC> > > WC> > WC> > WC> > WC> > > -- > = > ====================================================================== > Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http:// > www.cmm.upenn.edu > Center for Molecular Modeling -- University of Pennsylvania > Department of Chemistry, 231 S.34th Street, Philadelphia, PA > 19104-6323 > tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 > = > ====================================================================== > If you make something idiot-proof, the universe creates a better > idiot. > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste 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/20090122/f8794391/attachment.htm From quantumdft at gmail.com Thu Jan 22 08:44:30 2009 From: quantumdft at gmail.com (vega lew) Date: Thu, 22 Jan 2009 15:44:30 +0800 Subject: [Pw_forum] questions about population analysis and bonding charge density Message-ID: <412f6c680901212344i3681548as3b05d5ca00133bf0@mail.gmail.com> Dear all I have some questions about the post processing the results from pw.x first, i want to know the nature of bond (ionic or covalent). i think i should do overlap population analysis to identify the bond type. As Segall, M. D. reported in *Phys. Rev. B*, *54*, 16317, there is a criterion for bonding between atoms. A high value of the bond population indicates a covalent bond, while a low value indicates an ionic interaction. I know projwfc.x do Lowdin population analysis instead of Mulliken. I only could find Lowdin charges directly in output file. and i wonder if projwfc.x could give us the information about overlap population. how to extract that information? second, I want to plot the bonding charge density of water adsorbed on TiO2 surfaces as the Exercise 4 of postproc.pdf in 'hands on tutorial of QE'. I think i should relax the water/TiO2 system and then run 3 scf calculations based on the relaxed systems, and extract the charge density for the water/TiO2 system, the water alone and the TiO2 surface alone. But when water dissociate on the surface, the scf calculation of dissociated water is not converged. how could we plot the bonding charge density for this case. At last i still have one more question, if I want to calculation the noble metal on TiO2 surfaces, such as Pt, Au or Pd etc., how could I deal with the 'nbnd' options? should we add 20% more states to the system? and do you think whether it's necessary to add empty bands for bare TiO2 suraface or water adsorbed on TiO2 surfaces calculation? If no empty band added, do you think the results are creditable? Finally, I want to report something abnormal with neb restart process. When I restart the calculation of water dissociation barrier using NEB. I found all the coordinate of fixed atoms change to zero in the .path files under /tmp/ and output folder when I triggered the restart command, and the restart calculation crashed after a couple of minute. So I added the all the coordinate of fixed atoms manually before the calculation was down. Then the calculation could restart normally. My QE is 4.0.3 version. Does any one encounter the same problem? thank you for reading. any hints will be deeply 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 ****************************************************************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, Jiangsu, China ****************************************************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090122/72c39b83/attachment.htm From quantumdft at gmail.com Thu Jan 22 09:02:34 2009 From: quantumdft at gmail.com (vega lew) Date: Thu, 22 Jan 2009 16:02:34 +0800 Subject: [Pw_forum] Problem about the inputs of 'dos.x' In-Reply-To: <496EF54D.5000907@sissa.it> References: <496EF54D.5000907@sissa.it> Message-ID: <412f6c680901220002la63ecbud27c9df1ed7802e0@mail.gmail.com> I solve this problem by using the command 'wf_collect = .true.' in scf and nscf calculation vega On Thu, Jan 15, 2009 at 4:35 PM, Gabriele Sclauzero wrote: > Dear Pan Deng, > > are you doing a parallel calculation? If so, most probably you started > dos.x with a > different number of pools or processors with respect to the pw.x run. > When running any of the postprocessing tools which read output files of > pw.x in the > parallel version, you must use the same number of processors AND the same > number of pools > used for the pw.x run. > > In principle the dos.x should not modify pw.x output files (i.e. those in > prefix.save > and the wavefunctions), since it simply needs to read them (you can checkd > the time stamps > of those files to see if they have been modified when dos.x was running). > > Regards > > GS > > ?? wrote: > > Dear All, > > > > I was trying to calculate the dos of BaNiAs, when I was running > > the dos.x it gives the error message : > > from dos : error # 1 > > not implemented > > I do not know what mistake I have made in my inputs.Would someone > > point it out Please. > > &inputpp > > outdir='/disk2/xgwan/tmp/' > > prefix='BaNiAs' > > fildos='BaNiAs.dos', > > Emin=0.0, Emax=25.0, DeltaE=0.1 > > / > > And still I ran the projwfc.x to get some projected dos after I ran > > the dos.x.Does the results of the projwfc.x be affected by the failing > > of dos.x? > > > > Thanks all. > > > > Pan Deng > > Nanjing University > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > -- > > > o ------------------------------------------------ o > | Gabriele Sclauzero, PhD Student | > | c/o: SISSA & CNR-INFM Democritos, | > | via Beirut 2-4, 34014 Trieste (Italy) | > | email: sclauzer at sissa.it | > | phone: +39 040 3787 511 | > | skype: gurlonotturno | > o ------------------------------------------------ o > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- ================================================================================== 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 ****************************************************************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, Jiangsu, China ****************************************************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090122/9f982e72/attachment.htm From sclauzer at sissa.it Thu Jan 22 09:54:17 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Thu, 22 Jan 2009 09:54:17 +0100 Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: <49783439.40104@sissa.it> Hi, of course Paolo and Axel could help you much more than me on this issue, but inthe meanwhile: Karoly Nemeth wrote: > Hi, > > I would like to obtain the wavefunction coefficients u(i,k,G), where "i" > is the band-index, "k" is the k-space vector used in the discretization of > the Brillouin-zone in a PW calculation and "G" is the plane-wave basis > function's wave-vector (assuming a complex form as exp[i(k+G)r] ). I think that the quantities you are interested in are store in the variable "evc" of the pw program (but only in certains parts of the program and only for one k-point at a time; the wavefunctions for all the k-points are stored on the .wfc files). > > Is there any utility program that is independent from the rest of PWSCF > and would provide some sort of tabulation on u(i,k,G) as a function > of i,k and G , for a parallel calculation? Perhaps a quick quide to the > format of this information in the *.wfc* files written by a parallel run? Be careful that in parallel runs the G vectors (i.e. plane waves) are split across processor, so that each processor may be working only with a slab of the full grid. The option wfc_collect=.TRUE. collects all wavefunctions on a single file after the parallel run (maybe it can help you understand how to gather all the coefficients). GS PS: The forum community would be grateful if you specify your affiliation... thanks. > > I would appreciate any hint that could save me some time from going > through the source code for this info. > Thanks a lot in advance: > > Karoly > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From akohlmey at cmm.chem.upenn.edu Thu Jan 22 16:18:43 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Jan 2009 10:18:43 -0500 (EST) Subject: [Pw_forum] error got by using program pp.x to deal with the input file In-Reply-To: <53588269-F7A5-462F-9D2E-A500AFBCA4FD@sissa.it> References: <9e6d976a0901210823y56694d88w872c927379216b98@mail.gmail.com> <9e6d976a0901210846y688ffea7s6c539fee26ca0aae@mail.gmail.com> <20090121195512.pl1arvsm8kow840k@mail.democritos.it> <9e6d976a0901211445n7277ca56hd7e9d0f035b0c365@mail.gmail.com> <53588269-F7A5-462F-9D2E-A500AFBCA4FD@sissa.it> Message-ID: On Thu, 22 Jan 2009, Stefano Baroni wrote: SB> PS: broadening is not the only way to obtain DOS's. In a periodic system the SB> tetrahedron method does it better, but I am not sure that this method is SB> used by pp.x in you case. In any event, google "tetrahedron" and "density of for the sake of completeness: what stefano describes here is implemented in dos.x and an example for using it can be found in examples/example08 cheers, axel. SB> states" and get an idea of what the method is/does. SB> SB> On Jan 22, 2009, at 4:40 AM, Axel Kohlmeyer wrote: SB> SB> >On Thu, 22 Jan 2009, Weiguang Chen wrote: SB> > SB> > WC> Hi, SB> > WC> Thanks for your respondence quickly. SB> > WC> But there is no tag to set gaussian broadening in the input-file of SB> > WC> pp.x . Is this mean that I should set this value in scf calculation? SB> > SB> >where else? SB> > SB> > WC> In fact, I have used gaussian broadening in scf calculation. SB> > WC> SB> > WC> I set the gaussian broadening in scf calculation as follows: SB> > WC> occupations='smearing', SB> > WC> smearing='gauss', SB> > WC> degauss=0.01, SB> > SB> >please provide a complete script or set of inputs that SB> >reproduces this behavior, without requiring access to SB> >a supercomputer. perhaps then somebody can tell you SB> >what went wrong. SB> > SB> >i just slapped together a stupid example and "it works for me(tm)" SB> > SB> >have you run some other calculation in between that SB> >would modify the saved wavefunction? SB> > SB> >axel. SB> > SB> > SB> > WC> SB> > WC> Thanks SB> > WC> Chen Weiguang SB> > WC> On Thu, Jan 22, 2009 at 2:55 AM, wrote: SB> > WC> > Quoting Weiguang Chen : SB> > WC> > SB> > WC> > > I am a newbie about pwscf and i want to know what parameter should SB> > WC> > > been modified. Could you tell me more detail why this problem SB> > WC> > > happen? SB> > WC> > SB> > WC> > documentation in Doc/INPUT_PP.* SB> > WC> > SB> > WC> > Paolo SB> > WC> > SB> > WC> > SB> > WC> > ---------------------------------------------------------------- SB> > WC> > This message was sent using IMP, the Internet Messaging Program. SB> > WC> > SB> > WC>> _______________________________________________ SB> > WC> > Pw_forum mailing list SB> > WC> > Pw_forum at pwscf.org SB> > WC> > http://www.democritos.it/mailman/listinfo/pw_forum SB> > WC> > SB> > WC> SB> > WC> SB> > WC> SB> > WC> SB> > SB> >-- SB> >======================================================================= SB> >Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu SB> > Center for Molecular Modeling -- University of Pennsylvania SB> >Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 SB> >tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 SB> >======================================================================= SB> >If you make something idiot-proof, the universe creates a better idiot. SB> >_______________________________________________ SB> >Pw_forum mailing list SB> >Pw_forum at pwscf.org SB> >http://www.democritos.it/mailman/listinfo/pw_forum SB> SB> --- SB> Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste SB> http://www.sissa.it/~baroni / [+39] 040 3787 406 (tel) -528 (fax) / SB> stefanobaroni (skype) SB> SB> La morale est une logique de l'action comme la logique est une morale de la SB> pens?e - Jean Piaget SB> SB> Please, if possible, don't send me MS Word or PowerPoint attachments SB> Why? See: http://www.gnu.org/philosophy/no-word-attachments.html SB> SB> SB> SB> SB> SB> SB> -- ======================================================================= 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 jjr19 at uakron.edu Thu Jan 22 18:01:09 2009 From: jjr19 at uakron.edu (J. J. Ramsey) Date: Thu, 22 Jan 2009 09:01:09 -0800 (PST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters Message-ID: <448661.52977.qm@web50904.mail.re2.yahoo.com> The cluster I'm running on uses Infiniband for MPI communications, and the queueing software by default limits me to 1280 processor cores for a single job, 1.8 GB per core, and a wall time of 96 hours. As warned about in the user guide, I'm running into problems with running out of memory when using pw.x, so I need to find appropriate values for npool, ntg, and ndiag. (I don't have an NEB calculation, so I don't thing that the # of images is an issue.) I'm a bit unsure about sane values of these, since as far as I can tell, my simulation isn't a "very large" job but isn't exactly small, either: 213 atoms, 727 electrons, and I'm using 4x4x1 k-point sampling. (It's a slab calculation.) How many cores, roughly, is reasonable for a simulation like this, and what's a reasonable way to divide it up into pools, ortho groups, etc.? From nemeth at anl.gov Thu Jan 22 18:33:44 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Thu, 22 Jan 2009 11:33:44 -0600 (CST) Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: Thank you Gabriele! Since most of the time the *.wfc* are split among a few files, a utility that uses as input only the number of processors, nspin, and perhaps the number of basis functions could be useful. If the wfc files would be in hdf5 format it would be probably easier to figure out all their structure, but since this is not the case, a utility program could be quite useful for this purpose. Karoly *********************************************************************** Karoly Nemeth, Ph.D. Argonne National Laboratory 9700 S. Cass Ave. Argonne, IL 60439. Advanced Photon Source Accelerator Physics and Operations Group Building 401, Room B2200 Tel. 630-252-5813 Fax: 630-252-4732 *********************************************************************** > Message: 3 > Date: Thu, 22 Jan 2009 09:54:17 +0100 > From: Gabriele Sclauzero > Subject: Re: [Pw_forum] wavefunction coefficients > To: PWSCF Forum > Message-ID: <49783439.40104 at sissa.it> > Content-Type: text/plain; charset=us-ascii; format=flowed > > Hi, > > of course Paolo and Axel could help you much more than me on this issue, but inthe > meanwhile: > > Karoly Nemeth wrote: >> Hi, >> >> I would like to obtain the wavefunction coefficients u(i,k,G), where "i" >> is the band-index, "k" is the k-space vector used in the discretization of >> the Brillouin-zone in a PW calculation and "G" is the plane-wave basis >> function's wave-vector (assuming a complex form as exp[i(k+G)r] ). > > I think that the quantities you are interested in are store in the variable "evc" of the > pw program (but only in certains parts of the program and only for one k-point at a time; > the wavefunctions for all the k-points are stored on the .wfc files). > >> >> Is there any utility program that is independent from the rest of PWSCF >> and would provide some sort of tabulation on u(i,k,G) as a function >> of i,k and G , for a parallel calculation? Perhaps a quick quide to the >> format of this information in the *.wfc* files written by a parallel run? > > Be careful that in parallel runs the G vectors (i.e. plane waves) are split across > processor, so that each processor may be working only with a slab of the full grid. > The option wfc_collect=.TRUE. collects all wavefunctions on a single file after the > parallel run (maybe it can help you understand how to gather all the coefficients). > > GS > > PS: The forum community would be grateful if you specify your affiliation... thanks. > >> >> I would appreciate any hint that could save me some time from going >> through the source code for this info. >> Thanks a lot in advance: >> >> Karoly >> From lshulenburger at ciw.edu Thu Jan 22 18:57:32 2009 From: lshulenburger at ciw.edu (lshulenburger at ciw.edu) Date: Thu, 22 Jan 2009 12:57:32 -0500 (EST) Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: <59239.10.1.20.105.1232647052.squirrel@mail.ciw.edu> The pw2casino utility may be of use for this case. Run as a postprocessing step after a scf or nscf calculation, it writes a plaintext file which first lists the G vectors in order, then for each k-point the coefficients of those G vectors for each band. It's output is rather verbose, but it is reasonably self documenting so it might be worth a look. As for hdf output, I know the group developing the qmcpack code at the University of Illinois does exactly this, but for the moment they use pw2casino as an intermediate step and I'm not sure whether the code is released. Hope this helps. Luke Shulenburger Geophysical Laboratory Carnegie Institute of Washington > > Thank you Gabriele! Since most of the time the *.wfc* are split among a > few files, a utility that uses as input only the number of processors, > nspin, and perhaps the number of basis functions could be useful. If the > wfc files would be in hdf5 format it would be probably easier to figure > out all their structure, but since this is not the case, a utility program > could be quite useful for this purpose. > > Karoly > > *********************************************************************** > Karoly Nemeth, Ph.D. > Argonne National Laboratory > 9700 S. Cass Ave. Argonne, IL 60439. > Advanced Photon Source > Accelerator Physics and Operations Group > Building 401, Room B2200 > Tel. 630-252-5813 > Fax: 630-252-4732 > *********************************************************************** > > > >> Message: 3 >> Date: Thu, 22 Jan 2009 09:54:17 +0100 >> From: Gabriele Sclauzero >> Subject: Re: [Pw_forum] wavefunction coefficients >> To: PWSCF Forum >> Message-ID: <49783439.40104 at sissa.it> >> Content-Type: text/plain; charset=us-ascii; format=flowed >> >> Hi, >> >> of course Paolo and Axel could help you much more than me on this >> issue, but inthe >> meanwhile: >> >> Karoly Nemeth wrote: >>> Hi, >>> >>> I would like to obtain the wavefunction coefficients u(i,k,G), where >>> "i" >>> is the band-index, "k" is the k-space vector used in the discretization >>> of >>> the Brillouin-zone in a PW calculation and "G" is the plane-wave basis >>> function's wave-vector (assuming a complex form as exp[i(k+G)r] ). >> >> I think that the quantities you are interested in are store in the >> variable "evc" of the >> pw program (but only in certains parts of the program and only for one >> k-point at a time; >> the wavefunctions for all the k-points are stored on the .wfc files). >> >>> >>> Is there any utility program that is independent from the rest of PWSCF >>> and would provide some sort of tabulation on u(i,k,G) as a function >>> of i,k and G , for a parallel calculation? Perhaps a quick quide to the >>> format of this information in the *.wfc* files written by a parallel >>> run? >> >> Be careful that in parallel runs the G vectors (i.e. plane waves) are >> split across >> processor, so that each processor may be working only with a slab of the >> full grid. >> The option wfc_collect=.TRUE. collects all wavefunctions on a single >> file after the >> parallel run (maybe it can help you understand how to gather all the >> coefficients). >> >> GS >> >> PS: The forum community would be grateful if you specify your >> affiliation... thanks. >> >>> >>> I would appreciate any hint that could save me some time from going >>> through the source code for this info. >>> Thanks a lot in advance: >>> >>> Karoly >>> > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From akohlmey at cmm.chem.upenn.edu Thu Jan 22 19:37:00 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Jan 2009 13:37:00 -0500 (EST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters In-Reply-To: <448661.52977.qm@web50904.mail.re2.yahoo.com> References: <448661.52977.qm@web50904.mail.re2.yahoo.com> Message-ID: On Thu, 22 Jan 2009, J. J. Ramsey wrote: JJR> The cluster I'm running on uses Infiniband for MPI communications, JJR> and the queueing software by default limits me to 1280 processor JJR> cores for a single job, 1.8 GB per core, and a wall time of 96 JJR> hours. As warned about in the user guide, I'm running into problems JJR> with running out of memory when using pw.x, so I need to find JJR> appropriate values for npool, ntg, and ndiag. (I don't have an NEB JJR> calculation, so I don't thing that the # of images is an issue.) JJR> I'm a bit unsure about sane values of these, since as far as I can JJR> tell, my simulation isn't a "very large" job but isn't exactly JJR> small, either: 213 atoms, 727 electrons, and I'm using 4x4x1 JJR> k-point sampling. (It's a slab calculation.) JJR> How many cores, roughly, is reasonable for a simulation like this, JJR> and what's a reasonable way to divide it up into pools, ortho JJR> groups, etc.? jj, very simple. -npools can be maximally the actual number of k-points that you are using, and that should be exploited to the maximum. a second option to pay attention to, is the disk_io flag. setting this to 'low' can have a very positive effect on scaling at the expense of additional memory use. for the rest you have to make a sequence of benchmarks to see how far you can push it. there is no way to predict this. too many factors beyond the contol of the source code have some impact. for some of the advanced options, the code tries to measure its efficiency at the beginning of a run. cheers, axel. JJR> JJR> JJR> JJR> JJR> _______________________________________________ JJR> Pw_forum mailing list JJR> Pw_forum at pwscf.org JJR> http://www.democritos.it/mailman/listinfo/pw_forum JJR> -- ======================================================================= 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 jjr19 at uakron.edu Thu Jan 22 23:12:07 2009 From: jjr19 at uakron.edu (J. J. Ramsey) Date: Thu, 22 Jan 2009 14:12:07 -0800 (PST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters Message-ID: <422993.54661.qm@web50901.mail.re2.yahoo.com> ----- Original Message ---- > From: Axel Kohlmeyer > To: J. J. Ramsey > Cc: PWSCF > Sent: Thursday, January 22, 2009 1:37:00 PM > Subject: Re: [Pw_forum] Trying to find reasonable parallelization parameters > very simple. -npools can be maximally the actual number of k-points that > you are using, and that should be exploited to the maximum. a second > option to pay attention to, is the disk_io flag. setting this to 'low' > can have a very positive effect on scaling at the expense of additional > memory use. Well, I'm trying to *reduce* memory usage per processor, so I'm leery of lowering the disk-IO. What I'd like to do is try to distribute the memory for the job as much as possible. From akohlmey at cmm.chem.upenn.edu Thu Jan 22 23:38:47 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Jan 2009 17:38:47 -0500 (EST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters In-Reply-To: <422993.54661.qm@web50901.mail.re2.yahoo.com> References: <422993.54661.qm@web50901.mail.re2.yahoo.com> Message-ID: On Thu, 22 Jan 2009, J. J. Ramsey wrote: JJR> ----- Original Message ---- JJR> JJR> > From: Axel Kohlmeyer JJR> > To: J. J. Ramsey JJR> > Cc: PWSCF JJR> > Sent: Thursday, January 22, 2009 1:37:00 PM JJR> > Subject: Re: [Pw_forum] Trying to find reasonable parallelization parameters JJR> JJR> > very simple. -npools can be maximally the actual number of k-points that JJR> > you are using, and that should be exploited to the maximum. a second JJR> > option to pay attention to, is the disk_io flag. setting this to 'low' JJR> > can have a very positive effect on scaling at the expense of additional JJR> > memory use. JJR> JJR> Well, I'm trying to *reduce* memory usage per processor, so I'm JJR> leery of lowering the disk-IO. What I'd like to do is try to JJR> distribute the memory for the job as much as possible. perhaps i didn't explain it too well, but reducing disk i/o will _significantly_ enhance scaling and that would mean that you can use many more processors for the g-space parallelization, which in turn will significantly reduce the per MPI-task memory requirements. basically, you want to get the most performance out of your machine, right? and this is how you get it. another option that frequently helps, particularly with current intel quad-core cpus, is to use only half the available cpus. due to the architecture of that intel platform, this reduces the load on the memory bus and effectively doubles the cpu cache (and per MPI-task memory, of course). the effect can be quite drastoc, as shown, e.g. in this scaling test on our local machine (using a different DFT code that is much more demanding in terms of memory bandwidth than Q-E). http://cp2k.googlegroups.com/web/cp2k-water-bench-cmm.png so less is more in this case, and the only way to find this out is to run systematic tests (this example is gamma only). cheers, axel. JJR> JJR> JJR> JJR> JJR> _______________________________________________ JJR> Pw_forum mailing list JJR> Pw_forum at pwscf.org JJR> http://www.democritos.it/mailman/listinfo/pw_forum JJR> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From akohlmey at cmm.chem.upenn.edu Thu Jan 22 23:49:36 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Jan 2009 17:49:36 -0500 Subject: [Pw_forum] Trying to find reasonable parallelization parameters In-Reply-To: References: <422993.54661.qm@web50901.mail.re2.yahoo.com> Message-ID: <7b6913e90901221449j288d2e37i6f7a3f9f9e087f9@mail.gmail.com> oh... and one more thing. _what_ MPI are you using? most of the time, the infiniband MPI implementations are not optimized for large problems and allocate too many non-swappable RDMA buffers. now, this can create the impression of running out of memory, but in reality all you'd need to do is to change the settings of your MPI lib to not use fixed buffers, but rather shared request queues. have a chat with the sysadmin folks of the machine. in any case, since there is no way to make any predictions, you should first do some systematic scaling tests (perhaps with gamma only) and see whether it would run better with less processors (at the numbers you are mentioning, this is what i would definitely expect) altogether, and then less processors pernode (i.e. allocate all cores of a node and then use only half of them). cheers, axel. On Thu, Jan 22, 2009 at 5:38 PM, Axel Kohlmeyer wrote: > On Thu, 22 Jan 2009, J. J. Ramsey wrote: > -- ======================================================================= 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 jjr19 at uakron.edu Fri Jan 23 00:34:24 2009 From: jjr19 at uakron.edu (J. J. Ramsey) Date: Thu, 22 Jan 2009 15:34:24 -0800 (PST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters References: <422993.54661.qm@web50901.mail.re2.yahoo.com> <7b6913e90901221449j288d2e37i6f7a3f9f9e087f9@mail.gmail.com> Message-ID: <293629.84871.qm@web50902.mail.re2.yahoo.com> ----- Original Message ---- > From: Axel Kohlmeyer > To: J. J. Ramsey > Cc: pw_forum at pwscf.org > Sent: Thursday, January 22, 2009 5:49:36 PM > Subject: Re: [Pw_forum] Trying to find reasonable parallelization parameters > > oh... and one more thing. > > _what_ MPI are you using? OpenMPI 1.2.3. Unfortunately, I don't really have a whole lot of control over the MPI implementation. It's on a cluster where I have an account but not much if any say as to how it's administered. FWIW, Linux Networx is the provider of the cluster. From akohlmey at cmm.chem.upenn.edu Fri Jan 23 00:39:59 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Jan 2009 18:39:59 -0500 (EST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters In-Reply-To: <293629.84871.qm@web50902.mail.re2.yahoo.com> References: <422993.54661.qm@web50901.mail.re2.yahoo.com> <7b6913e90901221449j288d2e37i6f7a3f9f9e087f9@mail.gmail.com> <293629.84871.qm@web50902.mail.re2.yahoo.com> Message-ID: On Thu, 22 Jan 2009, J. J. Ramsey wrote: JJR> ----- Original Message ---- JJR> > From: Axel Kohlmeyer JJR> > To: J. J. Ramsey JJR> > Cc: pw_forum at pwscf.org JJR> > Sent: Thursday, January 22, 2009 5:49:36 PM JJR> > Subject: Re: [Pw_forum] Trying to find reasonable parallelization parameters JJR> > JJR> > oh... and one more thing. JJR> > JJR> > _what_ MPI are you using? JJR> JJR> OpenMPI 1.2.3. Unfortunately, I don't really have a whole lot of JJR> control over the MPI implementation. It's on a cluster where I have oh. you can compile your own. compiling MPI does not require administrator privilege. i do it all the time when i am not happy with how it is being set up (even in some of the supercomputing centers). in this case it would probably be beneficial to upgrade to 1.2.8, but please try adding the following settings to your mpirun commandline as it is. --mca btl_openib_use_srq 1 and --mca mpi_leave_pinned 1 the former is the memory saver (and performance enhancer), the second may be needed if you get crashes (if not, don't use it). ...and as i said before, i get _much_ better performance with -npernode 4 on nodes with 8 cores. cheers, axel. JJR> an account but not much if any say as to how it's administered. JJR> FWIW, Linux Networx is the provider of the cluster. JJR> JJR> JJR> JJR> JJR> _______________________________________________ JJR> Pw_forum mailing list JJR> Pw_forum at pwscf.org JJR> http://www.democritos.it/mailman/listinfo/pw_forum JJR> -- ======================================================================= 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 nemeth at anl.gov Fri Jan 23 00:53:45 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Thu, 22 Jan 2009 17:53:45 -0600 (CST) Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: Thank you Luke! Karoly *********************************************************************** Karoly Nemeth, Ph.D. Argonne National Laboratory 9700 S. Cass Ave. Argonne, IL 60439. Advanced Photon Source Accelerator Physics and Operations Group Building 401, Room B2200 Tel. 630-252-5813 Fax: 630-252-4732 *********************************************************************** > Message: 2 > Date: Thu, 22 Jan 2009 12:57:32 -0500 (EST) > From: lshulenburger at ciw.edu > Subject: Re: [Pw_forum] wavefunction coefficients > To: "PWSCF Forum" > Message-ID: <59239.10.1.20.105.1232647052.squirrel at mail.ciw.edu> > Content-Type: text/plain;charset=iso-8859-1 > > The pw2casino utility may be of use for this case. Run as a > postprocessing step after a scf or nscf calculation, it writes a plaintext > file which first lists the G vectors in order, then for each k-point the > coefficients of those G vectors for each band. It's output is rather > verbose, but it is reasonably self documenting so it might be worth a > look. As for hdf output, I know the group developing the qmcpack code at > the University of Illinois does exactly this, but for the moment they use > pw2casino as an intermediate step and I'm not sure whether the code is > released. Hope this helps. > > Luke Shulenburger > Geophysical Laboratory > Carnegie Institute of Washington From jjr19 at uakron.edu Fri Jan 23 01:03:09 2009 From: jjr19 at uakron.edu (J. J. Ramsey) Date: Thu, 22 Jan 2009 16:03:09 -0800 (PST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters References: <422993.54661.qm@web50901.mail.re2.yahoo.com> <7b6913e90901221449j288d2e37i6f7a3f9f9e087f9@mail.gmail.com> <293629.84871.qm@web50902.mail.re2.yahoo.com> Message-ID: <850650.89060.qm@web50910.mail.re2.yahoo.com> ----- Original Message ---- > JJR> OpenMPI 1.2.3. Unfortunately, I don't really have a whole lot of > JJR> control over the MPI implementation. It's on a cluster where I have > > oh. you can compile your own. compiling MPI does not require > administrator privilege. Compiling is one thing. The cluster, however, has its own mpirun.lsf script, and the number of CPUs to request is handled through the queueing system. Also, I'm not sure how I'd make sure that OpenMPI was set up so that mpiexec could access all the compute nodes in the cluster. It's not like I'm compiling it on a multi-core workstation. > please try adding > the following settings to your mpirun commandline as it is. > > --mca btl_openib_use_srq 1 > > and > > --mca mpi_leave_pinned 1 Might I be able to use those with the stock MPI implementation on the cluster? From akohlmey at cmm.chem.upenn.edu Fri Jan 23 01:28:46 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Jan 2009 19:28:46 -0500 (EST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters In-Reply-To: <850650.89060.qm@web50910.mail.re2.yahoo.com> References: <422993.54661.qm@web50901.mail.re2.yahoo.com> <7b6913e90901221449j288d2e37i6f7a3f9f9e087f9@mail.gmail.com> <293629.84871.qm@web50902.mail.re2.yahoo.com> <850650.89060.qm@web50910.mail.re2.yahoo.com> Message-ID: On Thu, 22 Jan 2009, J. J. Ramsey wrote: JJR> Compiling is one thing. The cluster, however, has its own JJR> mpirun.lsf script, and the number of CPUs to request is handled JJR> through the queueing system. Also, I'm not sure how I'd make sure same here, only that we don't waste money on overkill like LSF. i would just make a copy of the script and adapt it to my needs. scripts - unlike compiled binaries - have to be readable to be executed, ;-) JJR> that OpenMPI was set up so that mpiexec could access all the JJR> compute nodes in the cluster. It's not like I'm compiling it on a JJR> multi-core workstation. as i was saying, i've rolled my own even on machines that were set up by very paranoid people. every once in a while i was even fixing bugs in the respective mpirun scripts (or mpicc/mpif77). you basically have to tighten a machine to the point where it would become unusable (an adminstrators dream and a user's nightmare) to not have any options to replace mpi libs. JJR> > please try adding JJR> > the following settings to your mpirun commandline as it is. JJR> > JJR> > --mca btl_openib_use_srq 1 JJR> > JJR> > and JJR> > JJR> > --mca mpi_leave_pinned 1 JJR> JJR> Might I be able to use those with the stock MPI implementation on the cluster? by all means give it a try. i just don't know when they got added. they are for certain available in version 1.2.6. it also depends on the openfabrics distribution, but it looks like the one you have should be recent enough. and you should also give -npernode 4 or -npernode 2 a try and see if it results in faster or slower execution. it will most certainly limit the load on the infiniband and that can only help. cheers, axel. p.s.: please note that it is customary on this mailing list to sign e-mails with full name and affiliation. JJR> _______________________________________________ JJR> Pw_forum mailing list JJR> Pw_forum at pwscf.org JJR> http://www.democritos.it/mailman/listinfo/pw_forum JJR> -- ======================================================================= 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 thmason24 at yahoo.com Fri Jan 23 03:20:20 2009 From: thmason24 at yahoo.com (Timothy Mason) Date: Thu, 22 Jan 2009 18:20:20 -0800 (PST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters Message-ID: <957255.10890.qm@web81802.mail.mud.yahoo.com> I have a follow up question on this issue. if I get the following message at the beginning of the output file "Iterative solution of the eigenvalue problem. Too few procs for parallel algorith we need 4 procs per pool a serial algorithm will be used" does this suggest that if I assign 2 procs per pool then I'm wasting a processor? Tim Mason University of Missouri St. Louis From akohlmey at cmm.chem.upenn.edu Fri Jan 23 03:40:39 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 22 Jan 2009 21:40:39 -0500 (EST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters In-Reply-To: <957255.10890.qm@web81802.mail.mud.yahoo.com> References: <957255.10890.qm@web81802.mail.mud.yahoo.com> Message-ID: On Thu, 22 Jan 2009, Timothy Mason wrote: TM> I have a follow up question on this issue. TM> TM> if I get the following message at the beginning of the output file TM> TM> "Iterative solution of the eigenvalue problem. TM> Too few procs for parallel algorith we need 4 procs per pool TM> a serial algorithm will be used" TM> TM> does this suggest that if I assign 2 procs per pool then I'm wasting a processor? tim, only a benchmark can tell for certain. in most cases the k-point parallelization is by far the most efficient. thus you should first try to use as many pools as possible. even if you are not using a parallel diagonalization algorithm, you are still using the g-space parallelization and distribute the wavefunction and density (see summary at the beginning), which coincides with parallelization of several time consuming operations. HTH, axel. TM> TM> Tim Mason TM> University of Missouri St. Louis TM> _______________________________________________ TM> Pw_forum mailing list TM> Pw_forum at pwscf.org TM> http://www.democritos.it/mailman/listinfo/pw_forum TM> -- ======================================================================= 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 nemeth at anl.gov Fri Jan 23 05:04:57 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Thu, 22 Jan 2009 22:04:57 -0600 (CST) Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: Hi, I have tried out to generate the list of wavefunction coefficients by running pw2casino.x, on the same number of processors (Nproc), the pw.x calculation was running with. It generated Nproc pieces of files of the Cu111.pwfn.data* kind, each containing a list of wavefunction coefficients for a certain fraction of the G vectors. Each of these files indicates the number of G vectors contained, under "Number of G-vectors". Surprisingly, these numbers are not the same as what are listed on the top of the output of the pw.x calculation (under "columns G" and "(wavefct grid)"), for each processor. What is the reason for this difference? I have used version 4.0.4 of QE. Thanks: Karoly >> Message: 2 >> Date: Thu, 22 Jan 2009 12:57:32 -0500 (EST) >> From: lshulenburger at ciw.edu >> Subject: Re: [Pw_forum] wavefunction coefficients >> To: "PWSCF Forum" >> Message-ID: <59239.10.1.20.105.1232647052.squirrel at mail.ciw.edu> >> Content-Type: text/plain;charset=iso-8859-1 >> >> The pw2casino utility may be of use for this case. Run as a >> postprocessing step after a scf or nscf calculation, it writes a plaintext >> file which first lists the G vectors in order, then for each k-point the >> coefficients of those G vectors for each band. It's output is rather >> verbose, but it is reasonably self documenting so it might be worth a >> look. As for hdf output, I know the group developing the qmcpack code at >> the University of Illinois does exactly this, but for the moment they use >> pw2casino as an intermediate step and I'm not sure whether the code is >> released. Hope this helps. >> >> Luke Shulenburger >> Geophysical Laboratory >> Carnegie Institute of Washington > > > > From lshulenburger at ciw.edu Fri Jan 23 07:31:57 2009 From: lshulenburger at ciw.edu (lshulenburger at ciw.edu) Date: Fri, 23 Jan 2009 01:31:57 -0500 (EST) Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: <56212.71.246.243.74.1232692317.squirrel@mail.ciw.edu> I'll defer to the experts on the list for a direct answer to your question. I've typically used the code in serial fashion whereby it collects the wavefunctions from all of the processors into a single largish pwfn.data file. I've found very little documentation for the utility and am not sure if it is designed to work in a parallel fashion. It also seems possible that pw2casino could be padding the wavefunction for some reason, you might look for masses of coefficients that are exactly 0, but this is pure speculation on my part. Luke Shulenburger Geophysical Laboratory Carnegie Institution of Washington > > Hi, > > I have tried out to generate the list of wavefunction coefficients by > running pw2casino.x, on the same number of processors (Nproc), the pw.x > calculation was running with. It generated Nproc pieces of files of the > Cu111.pwfn.data* kind, each containing a list of wavefunction coefficients > for a certain fraction of the G vectors. Each of these files indicates the > number of G vectors contained, under "Number of G-vectors". Surprisingly, > these numbers are not the same as what are listed on the top of the output > of the pw.x calculation (under "columns G" and "(wavefct grid)"), for each > processor. What is the reason for this difference? I have used version > 4.0.4 of QE. > Thanks: > > Karoly > > >>> Message: 2 >>> Date: Thu, 22 Jan 2009 12:57:32 -0500 (EST) >>> From: lshulenburger at ciw.edu >>> Subject: Re: [Pw_forum] wavefunction coefficients >>> To: "PWSCF Forum" >>> Message-ID: <59239.10.1.20.105.1232647052.squirrel at mail.ciw.edu> >>> Content-Type: text/plain;charset=iso-8859-1 >>> >>> The pw2casino utility may be of use for this case. Run as a >>> postprocessing step after a scf or nscf calculation, it writes a >>> plaintext >>> file which first lists the G vectors in order, then for each k-point >>> the >>> coefficients of those G vectors for each band. It's output is rather >>> verbose, but it is reasonably self documenting so it might be worth a >>> look. As for hdf output, I know the group developing the qmcpack code >>> at >>> the University of Illinois does exactly this, but for the moment they >>> use >>> pw2casino as an intermediate step and I'm not sure whether the code is >>> released. Hope this helps. >>> >>> Luke Shulenburger >>> Geophysical Laboratory >>> Carnegie Institute of Washington >> >> >> >> > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From paulatto at sissa.it Fri Jan 23 09:45:39 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Fri, 23 Jan 2009 09:45:39 +0100 Subject: [Pw_forum] Trying to find reasonable parallelization parameters In-Reply-To: References: <448661.52977.qm@web50904.mail.re2.yahoo.com> Message-ID: On Thu, 22 Jan 2009 19:37:00 +0100, Axel Kohlmeyer wrote: > very simple. -npools can be maximally the actual number of k-points that > you are using, and that should be exploited to the maximum. True, but the maximum may not be the number of kpoints; using the parallel davidson algorithm (default) the hamiltonian matrix will be distributed and diagonalized on the larger perfect square number of processor *inside* each pool. If you are low on memory you may need to spread the hamiltonian over several processors in order to fit in. I.e. for a 3000 electrons job I'm trying these days, I have to use at least 100 procs per pool (and set diago_david_ndim=2) to stay (well) under ~2GB. cheers -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From sezginaydin at gazi.edu.tr Fri Jan 23 12:52:22 2009 From: sezginaydin at gazi.edu.tr (sezginaydin at gazi.edu.tr) Date: Fri, 23 Jan 2009 11:52:22 -0000 Subject: [Pw_forum] (no subject) Message-ID: Dear PWScf users, I work on the PtN Rocksalt structure. My input file for pw.x as foolowing: ------------------------------- PtN 2 atoms, NaCl bulk &control calculation='scf' restart_mode='from_scratch', prefix='PtN' pseudo_dir = '/home/sezgin/espresso-4.0.1/pseudo', outdir='./' / &system ibrav = 2, celldm(1) = 7.7 nat= 2, ntyp= 2, ecutwfc = 20 / &electrons conv_thr = 1.0e-8 mixing_beta = 0.7 mixing_mode = 'plain' / ATOMIC_SPECIES N 14.0 N.pw91-van_ak.UPF Pt 190.0 Pt.pw91-n-van.UPF ATOMIC_POSITIONS N 0.00 0.00 0.00 Pt 0.5 0.5 0.5 K_POINTS {automatic} 3 3 3 0 0 0 ----------------------------- when i work pw.x, i get the error: WARNING: integrated charge= 14.30000000, expected= 15.00000000 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from electrons : error # 1 charge is wrong %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... what is the problem? thanks for your interest From sezginaydin at gazi.edu.tr Fri Jan 23 13:06:28 2009 From: sezginaydin at gazi.edu.tr (sezginaydin at gazi.edu.tr) Date: Fri, 23 Jan 2009 12:06:28 -0000 Subject: [Pw_forum] charge is wrong Message-ID: Dear PWScf users, I work on the PtN Rocksalt structure. My input file for pw.x as foolowing: ------------------------------- PtN 2 atoms, NaCl bulk &control calculation='scf' restart_mode='from_scratch', prefix='PtN' pseudo_dir = '/home/sezgin/espresso-4.0.1/pseudo', outdir='./' / &system ibrav = 2, celldm(1) = 4.4 nat= 2, ntyp= 2, ecutwfc = 20 / &electrons conv_thr = 1.0e-8 mixing_beta = 0.7 mixing_mode = 'plain' / ATOMIC_SPECIES N 14.0 N.pw91-van_ak.UPF Pt 190.0 Pt.pw91-n-van.UPF ATOMIC_POSITIONS N 0.00 0.00 0.00 Pt 0.5 0.5 0.5 K_POINTS {automatic} 3 3 3 0 0 0 ----------------------------- when i work pw.x, i get the error: WARNING: integrated charge= 14.30000000, expected= 15.00000000 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from electrons : error # 1 charge is wrong %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... what is the problem? thanks for your interest From lataprem29 at gmail.com Fri Jan 23 13:18:24 2009 From: lataprem29 at gmail.com (premlata pandit) Date: Fri, 23 Jan 2009 17:48:24 +0530 Subject: [Pw_forum] generating pseudopotential problem In-Reply-To: <49773D9F.9080501@sissa.it> References: <6f289440901210055m7d3444c0xc55950bd7370c174@mail.gmail.com> <49773D9F.9080501@sissa.it> Message-ID: <6f289440901230418m18b30c2do65d4f5005a3ab5f0@mail.gmail.com> First thanks for your suggestions I did calculation as per your suggestion but still I couldn't get success. there is an error (error in DGETRF) has come in output. you suggest to include p and d projector, so my question is which orbital (inner, outer or both), we should include in calculation. kindly give me useful suggestion regarding my problem. Thank You sir On Wed, Jan 21, 2009 at 8:52 PM, Gabriele Sclauzero wrote: > I've not much experience in producing PP, but I recognize some problems in > your input > file. I can point them out, while you wait for a more definitive reply... > > premlata pandit wrote: > > Dear Pwscf User > > We have generated ultrasoft pseudopotential for Pm by using following > > configuration > > 4F 4 3 5.00 0.00 2.10 2.10 2.50 > > 4F 4 3 0.00 -0.10 2.10 2.10 3.50 > > These two rows look strange to me. If you want to generate US PP, you need > at least 2 > energies for each channels and specify the 2nd cutoff radius bigger than > the first. They > are the same for the channels you want pseudize within the norm-conserving > recipe. > I don't understand if your input above is for US with rel=1 or for NC with > rel=2. > In the relativistic case (rel=2), for each l, the number of projectors > doubles (you have > one with j=l+1/2 and another with j=l-1/2, where j is the number you > specify in the last > column). > So, your input for relativistic USPP should look more like: > > 4F 4 3 5.00 0.00 2.10 2.40 2.50 > 4F 4 3 0.00 -0.10 2.10 2.40 2.50 > 4F 4 3 0.00 0.00 2.10 2.40 3.50 > 4F 4 3 0.00 -0.10 2.10 2.40 3.50 > > Of course this is not sufficient to get a relativistic US-PP, you need to > play a lot with > cut-off radius and energies as well as maybe adding p or d projectors... it > is usually a > difficult task to obtain a working PP for rare earth elements. > > > 6S 1 0 2.00 0.00 2.40 2.40 0.50 > > > > By using above configuration, we have got output. > > This does not mean that the PP will work correctly elsewhere than the atom > in the > configuration used for generating the PP... > > My suggestion is to look at the input files for PP generation in > espresso/atomic_doc/pseudo_library/LDA/REL/ and understand what you need to > specify. > > Regards > > GS > > > > >But when we used > > split 4F configuration ( i.e. more than 2 ), we found some error. > > The input file Pm2.in, which is used to generated pp, is also attached > > with this mail. > > There are some question - > > 1. is input file and generated pseudopotential is correct? > > 2. Why we did not get results, when we used split 4F configuration ? > > anyone can help me , i would be greatful. > > > > Premlata Pandit > > Ph.d. student, > > Barkatullah university, > > Bhopal 462026, > > MP, India > > > > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > -- > > > o ------------------------------------------------ o > | Gabriele Sclauzero, PhD Student | > | c/o: SISSA & CNR-INFM Democritos, | > | via Beirut 2-4, 34014 Trieste (Italy) | > | email: sclauzer at sissa.it | > | phone: +39 040 3787 511 | > | skype: gurlonotturno | > o ------------------------------------------------ o > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- With Regards 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/20090123/95c79155/attachment.htm From eyvaz_isaev at yahoo.com Fri Jan 23 13:58:18 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 23 Jan 2009 04:58:18 -0800 (PST) Subject: [Pw_forum] charge is wrong In-Reply-To: Message-ID: <11701.85748.qm@web65713.mail.ac4.yahoo.com> Dear user, Congratulations, you have "rediscovered" this problem, but it was explained many times in the forum. Let me try explain in somewhat different way. What is the charge density? It is 2\sum_{n,k}|\psi_{n,k}|^2, right? (n is bands number, and k is number of k-points). Imagine, that you did not include a valence band to this sum, then will you have total number of electrons you expected? The same question if you include more conduction bands to the sum. If you can find the right answers then you can fix the problem easily (if you have a look at INPUT_PW file as well). I can suggest that what you do is a teaching purpose only as there is no B1 PtN exists. There was a paper in PRB stating that B1 structure suggested for PtN was wrong, finally it turned out to be FeS2, not CaF2. Please also keep in mind that submitting a question to this forum requires specifying your name and affiliation. ------------------------------------------------------------------- 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 Fri, 1/23/09, sezginaydin at gazi.edu.tr wrote: > From: sezginaydin at gazi.edu.tr > Subject: [Pw_forum] charge is wrong > To: pw_forum at pwscf.org > Date: Friday, January 23, 2009, 3:06 PM > Dear PWScf users, > I work on the PtN Rocksalt structure. My input file for > pw.x as foolowing: > ------------------------------- > PtN > 2 atoms, NaCl bulk > &control > calculation='scf' > restart_mode='from_scratch', > prefix='PtN' > pseudo_dir = > '/home/sezgin/espresso-4.0.1/pseudo', > outdir='./' > / > &system > ibrav = 2, > celldm(1) = 4.4 > nat= 2, > ntyp= 2, > ecutwfc = 20 > / > &electrons > conv_thr = 1.0e-8 > mixing_beta = 0.7 > mixing_mode = 'plain' > / > ATOMIC_SPECIES > N 14.0 N.pw91-van_ak.UPF > Pt 190.0 Pt.pw91-n-van.UPF > ATOMIC_POSITIONS > N 0.00 0.00 0.00 > Pt 0.5 0.5 0.5 > K_POINTS {automatic} > 3 3 3 0 0 0 > ----------------------------- > > when i work pw.x, i get the error: > > WARNING: integrated charge= 14.30000000, expected= > 15.00000000 > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from electrons : error # 1 > charge is wrong > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > what is the problem? > thanks for your interest > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From eyvaz_isaev at yahoo.com Fri Jan 23 14:15:42 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 23 Jan 2009 05:15:42 -0800 (PST) Subject: [Pw_forum] charge is wrong In-Reply-To: Message-ID: <458189.23490.qm@web65708.mail.ac4.yahoo.com> In addition to my previous mail: 1. Hopefully, you can answer a question why "expected= 15.00000000". 2. > ecutwfc = 20 Very bad choice, at least 25Ry (30Ry is very safe choice) for US PsP. 3. > 3 3 3 0 0 0 Very small number of k-points. Please study how your result (say, total energy) is changed if you increase k-points number. The last two comments are not related to the question you posted, they are connected to the quality of calculations. 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 Fri, 1/23/09, sezginaydin at gazi.edu.tr wrote: > From: sezginaydin at gazi.edu.tr > Subject: [Pw_forum] charge is wrong > To: pw_forum at pwscf.org > Date: Friday, January 23, 2009, 3:06 PM > Dear PWScf users, > I work on the PtN Rocksalt structure. My input file for > pw.x as foolowing: > ------------------------------- > PtN > 2 atoms, NaCl bulk > &control > calculation='scf' > restart_mode='from_scratch', > prefix='PtN' > pseudo_dir = > '/home/sezgin/espresso-4.0.1/pseudo', > outdir='./' > / > &system > ibrav = 2, > celldm(1) = 4.4 > nat= 2, > ntyp= 2, > ecutwfc = 20 > / > &electrons > conv_thr = 1.0e-8 > mixing_beta = 0.7 > mixing_mode = 'plain' > / > ATOMIC_SPECIES > N 14.0 N.pw91-van_ak.UPF > Pt 190.0 Pt.pw91-n-van.UPF > ATOMIC_POSITIONS > N 0.00 0.00 0.00 > Pt 0.5 0.5 0.5 > K_POINTS {automatic} > 3 3 3 0 0 0 > ----------------------------- > > when i work pw.x, i get the error: > > WARNING: integrated charge= 14.30000000, expected= > 15.00000000 > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from electrons : error # 1 > charge is wrong > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > what is the problem? > thanks for your interest > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From jjr19 at uakron.edu Fri Jan 23 14:50:02 2009 From: jjr19 at uakron.edu (J. J. Ramsey) Date: Fri, 23 Jan 2009 05:50:02 -0800 (PST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters References: <448661.52977.qm@web50904.mail.re2.yahoo.com> Message-ID: <496033.89475.qm@web50901.mail.re2.yahoo.com> ----- Original Message ---- > From: Lorenzo Paulatto > To: Axel Kohlmeyer ; PWSCF Forum > Sent: Friday, January 23, 2009 3:45:39 AM > Subject: Re: [Pw_forum] Trying to find reasonable parallelization parameters > > On Thu, 22 Jan 2009 19:37:00 +0100, Axel Kohlmeyer > wrote: > > very simple. -npools can be maximally the actual number of k-points that > > you are using, and that should be exploited to the maximum. > > True, but the maximum may not be the number of kpoints; using the parallel > davidson algorithm (default) the hamiltonian matrix will be distributed and > diagonalized on the larger perfect square number of processor *inside* each > pool. *Each* pool? I had gotten the impression from the documentation () that if, say, I had 36 CPU cores and 4 pools, the diagonalization would by default take place over only 9 cores, not 9 cores for each pool. (FWIW, I also understand that the diagonalization takes place on a subset of the processors performing the FFTs, so it's not like I have some cores doing diagonalization and a completely different set of cores handling the plane wave parallelization. Not sure if that's right, though.) BTW, thanks for the tip on diago_david_ndim. I may give that a try. From paulatto at sissa.it Fri Jan 23 15:01:08 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Fri, 23 Jan 2009 15:01:08 +0100 Subject: [Pw_forum] Trying to find reasonable parallelization parameters In-Reply-To: <496033.89475.qm@web50901.mail.re2.yahoo.com> References: <448661.52977.qm@web50904.mail.re2.yahoo.com> <496033.89475.qm@web50901.mail.re2.yahoo.com> Message-ID: On Fri, 23 Jan 2009 14:50:02 +0100, J. J. Ramsey wrote: > *Each* pool? I had gotten the impression from the documentation > () > that if, say, I had 36 CPU cores and 4 pools, the diagonalization would > by default take place over only 9 cores, not 9 cores for each pool. No, the documentation is a bit misleading, but if you read it carefully you'll see it agrees with me (I was mislead as well in the beginning); furthermore you can check using -ndiag X that you cannot specify X > number of processor i a pool. > (FWIW, I also understand that the diagonalization takes place on a > subset of the processors performing the FFTs, so it's not like I have > some cores doing diagonalization and a completely different set of cores > handling the plane wave parallelization. Not sure if that's right, > though.) Not really, they are two different and independet levels of parallelizations: at certain points in the code processors will be divided in task groups to do the FFTs, at other points they are divided in diag groups (ortho groups in CP code) that do the diagonalization. I'm not 100% sure on this part, though. > BTW, thanks for the tip on diago_david_ndim. I may give that a try. It reduces memory usage quite a lot, but requires more iteration of the diagonalization algorithm; I haven't done any serious benchmark on the performance. cheers -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From giannozz at democritos.it Fri Jan 23 15:02:11 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Fri, 23 Jan 2009 15:02:11 +0100 Subject: [Pw_forum] wavefunction coefficients Message-ID: <20090123150211.hcqns7d23oosk4sc@mail.democritos.it> Quoting Karoly Nemeth : > I have tried out to generate the list of wavefunction coefficients by > running pw2casino.x, on the same number of processors (Nproc), the pw.x > calculation was running with. It generated Nproc pieces of files of the > Cu111.pwfn.data* kind, each containing a list of wavefunction coefficients > for a certain fraction of the G vectors. Each of these files indicates the > number of G vectors contained, under "Number of G-vectors". Surprisingly, > these numbers are not the same as what are listed on the top of the output > of the pw.x calculation (under "columns G" and "(wavefct grid)"), for each > processor. What is the reason for this difference? I think the reason is that pw.x uses k=G vectors for each k-point, while the casino code for which this small utility [*] was written requires the same set of G vectors for all k-points, or something like this. Not sure that it works in parallel, though (it should but it will produce distributed output files). If you want to be on the safe side, follow Luke's advice: use wf_collect and run pw2casino on a single processor [*]I believed that pw2casino was a "inutility" and I am pleased to learn that it is actually useful to somebody. Paolo Giannozzi --- Democritos and Udine University ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From akohlmey at cmm.chem.upenn.edu Fri Jan 23 15:57:53 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Fri, 23 Jan 2009 09:57:53 -0500 (EST) Subject: [Pw_forum] Trying to find reasonable parallelization parameters In-Reply-To: References: <448661.52977.qm@web50904.mail.re2.yahoo.com> Message-ID: On Fri, 23 Jan 2009, Lorenzo Paulatto wrote: LP> On Thu, 22 Jan 2009 19:37:00 +0100, Axel Kohlmeyer LP> wrote: LP> >very simple. -npools can be maximally the actual number of k-points that LP> >you are using, and that should be exploited to the maximum. LP> True, but the maximum may not be the number of kpoints; using the parallel LP> davidson algorithm (default) the hamiltonian matrix will be distributed and LP> diagonalized on the larger perfect square number of processor *inside* each lorenzo, please note, that my first and most important statement was, that the only reliable source of what is the best solution, is to make benchmarks and find it out. as i was mentioning before when using multi-core processors theses days, choices can be quite confusing, i.e. you can run much faster by not using a significant amount of your compute power. there are many parameters that affect the performance of a job and they are different from the theoretical performance of an algorithm. LP> pool. If you are low on memory you may need to spread the hamiltonian over LP> several processors in order to fit in. I.e. for a 3000 electrons job I'm LP> trying these days, I have to use at least 100 procs per pool (and set LP> diago_david_ndim=2) to stay (well) under ~2GB. well, one could also see this as a manifestation that you have to beg to get stefano to buy you a larger machine. ;-) not an easy task these days in italy, though. :-( with all that in mind, i still maintain that most people should pay more attention to maximizing use of -npools, as this is not chosen automatically, and thus quite often people will run with much reduced efficiency. this is particularly true for the majority of calculations that are smaller than yours and run on smaller machines, quite often only with GigE interconnect, which is saturated very fast with only g-space parallelization. with best regards, axel. LP> cheers 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 nemeth at anl.gov Fri Jan 23 22:20:07 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Fri, 23 Jan 2009 15:20:07 -0600 (CST) Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: Thanks Luke and Paolo for the comments. The difference between the number of plane waves indicated at the top of the pw.x output file (for my case 7403) and the ones generated by pw2casino.x (5843) is huge. This latter number turns out to be independent from whether running pw.x with the wf_collect=.TRUE. option, or not. I tried out both, in the latter case only one Cu111.pwfn.data file was produced (as it should; in this case I ran pw2casino.x on a single processor). This was a simple calculation on a Cu(111) surface slab. Did the authors of pw2casino.x have left any documentation for this difference somewhere? Perhaps I should email to them as well. Thanks: Karoly *********************************************************************** Karoly Nemeth, Ph.D. Argonne National Laboratory 9700 S. Cass Ave. Argonne, IL 60439. Advanced Photon Source Accelerator Physics and Operations Group Building 401, Room B2200 Tel. 630-252-5813 Fax: 630-252-4732 *********************************************************************** > I'll defer to the experts on the list for a direct answer to your > question. I've > typically used the code in serial fashion whereby it collects the > wavefunctions from all of > the processors into a single largish pwfn.data file. I've found very > little documentation > for the utility and am not sure if it is designed to work in a parallel > fashion. It also > seems possible that pw2casino could be padding the wavefunction for some > reason, > you might look for masses of coefficients that are exactly 0, but this is > pure > speculation on my part. > > Luke Shulenburger > Geophysical Laboratory > Carnegie Institution of Washington > From nemeth at anl.gov Fri Jan 23 23:51:01 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Fri, 23 Jan 2009 16:51:01 -0600 (CST) Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: Another comment on the use of the pw2casino.x code. The extracted wavefunction coefficients do not norm to one. Even if I use the single file collection and run it on one processor only, the norm of the coefficients vector is not one, but, e.g. 0.87449. It appears that pw2casino.x may not collect all the coefficients for every G vector. Karoly From djaithania at hotmail.com Sat Jan 24 11:48:18 2009 From: djaithania at hotmail.com (hania djani-ait aissa) Date: Sat, 24 Jan 2009 10:48:18 +0000 Subject: [Pw_forum] vc-relax Message-ID: Dear all, i am doing calculation on orthorombic system containing 36 atoms (Bi2WO6). My computer is a workstation HP xw6400 quad-core Xeon, 2GB. Vc-relax on this system spent more than 01 mounth and still not finish. the energy is converging but forces fluctuate. is there something wrong in my input in terms of umass or dt? or have i to capitulate to this very slow convergence? here is the input file. Any suggestions will welcome. &control calculation = 'vc-relax' restart_mode = 'from_scratch' pseudo_dir = '/root/pwscf/pseudo/' outdir = '/root/tmp/' tstress = .TRUE. tprnfor = .TRUE. nstep = 55 etot_conv_thr= 1e-5, forc_conv_thr= 1e-4, dt = 30, / &system ibrav=8 celldm(1)=10.37, celldm(2)=0.996, celldm(3)=3.00, nat=36 ntyp=3 ecutwfc=35.0 ecutrho=200.0, / &electrons conv_thr = 1e-8, mixing_beta=0.7, / &ions /&cell cell_dynamics = 'damp-w', press = 0.00, wmass = 0.004,/ ATOMIC_SPECIES W 183.84 W.pz-bhs.UPF Bi 208.98 083-Bi-ca-d-vgrp_ji.uspp.UPF O 15.9994 O.pz-van_ak.UPF ATOMIC_POSITIONS alatW 0.000000000 0.005016973 0.747761262 0 1 1W 0.500000000 0.990983027 2.252238738 0 1 1W 0.500000000 0.492983027 0.747761262 0 1 1W 0.000000000 0.503016973 2.252238738 0 1 1Bi 0.964931181 0.523336245 1.268793038Bi 0.464931181 0.472663755 1.731206962Bi 0.464931181 0.970663755 1.268793038Bi 0.964931181 0.025336245 1.731206962Bi 0.981726782 0.473760364 0.230569958Bi 0.481726782 0.522239636 2.769430042Bi 0.481726782 0.024239636 0.230569958Bi 0.981726782 0.971760364 2.769430042O 0.100459245 0.081203101 0.435231801O 0.600459245 0.914796899 2.564768199O 0.600459245 0.416796899 0.435231801O 0.100459245 0.579203101 2.564768199O 0.264260826 0.262143525 3.000064967O 0.764260826 0.733856475 -0.000064967O 0.764260826 0.235856475 3.000064967O 0.264260826 0.760143525 -0.000064967O 0.247153777 0.235070329 1.501190932O 0.747153777 0.760929671 1.498809068O 0.747153777 0.262929671 1.501190932O 0.247153777 0.733070329 1.498809068O 0.283152650 0.701520345 0.680189839O 0.783152650 0.294479655 2.319810161O 0.783152650 0.792479655 0.680189839O 0.283152650 0.203520345 2.319810161O 0.335776234 0.239505837 0.816308565O 0.835776234 0.756494163 2.183691435O 0.835776234 0.258494163 0.816308565O 0.335776234 0.737505837 2.183691435O 0.586065178 0.587540848 1.058853940O 0.086065178 0.408459152 1.941146060O 0.086065178 0.906459152 1.058853940O 0.586065178 0.089540848 1.941146060 K_POINTS {automatic} 5 5 5 1 1 1 thanks a lot Hania Djani-ait aissa PHD student Algeria _________________________________________________________________ Sur Windows Live Ideas, d?couvrez en exclusivit? de nouveaux services en ligne... si nouveaux qu'ils ne sont pas encore sortis officiellement sur le march? ! http://ideas.live.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090124/434aadee/attachment.htm From davidfarrell2008 at u.northwestern.edu Sat Jan 24 15:46:08 2009 From: davidfarrell2008 at u.northwestern.edu (David Farrell) Date: Sat, 24 Jan 2009 08:46:08 -0600 Subject: [Pw_forum] vc-relax In-Reply-To: References: Message-ID: <05A835BB-2A8A-4175-B69A-0AEAF2397B56@u.northwestern.edu> I have run into convergence issues (fluctuations about a minimum) with larger cells and too tight of a force and energy tolerance (yours are quite tight). You may want to back off the tolerances a bit, then restart it with tighter tolerances, incrementing until you get the desired accuracy. That or only set 1 of the tolerances tight. Since both criteria need to be met, you may be over-constraining the system and so you get fluctuations. One approach may be to back off the energy tolerance, so that the forces are tightly converged (which often occurs after the energy has converged well) Or I could be completely wrong. Dave On Jan 24, 2009, at 4:48 AM, hania djani-ait aissa wrote: > Dear all, > i am doing calculation on orthorombic system containing 36 atoms > (Bi2WO6). My computer is a workstation HP xw6400 quad-core Xeon, > 2GB. Vc-relax on this system spent more than 01 mounth and still not > finish. the energy is converging but forces fluctuate. is there > something wrong in my input in terms of umass or dt? or have i to > capitulate to this very slow convergence? > here is the input file. Any suggestions will welcome. > &control > calculation = 'vc-relax' > restart_mode = 'from_scratch' > pseudo_dir = '/root/pwscf/pseudo/' > outdir = '/root/tmp/' > tstress = .TRUE. > tprnfor = .TRUE. > nstep = 55 > etot_conv_thr= 1e-5, > forc_conv_thr= 1e-4, > dt = 30, > / > &system > ibrav=8 > celldm(1)=10.37, > celldm(2)=0.996, > celldm(3)=3.00, > nat=36 > ntyp=3 > ecutwfc=35.0 > ecutrho=200.0, > > / > &electrons > conv_thr = 1e-8, > mixing_beta=0.7, > / > &ions > / > &cell > cell_dynamics = 'damp-w', > press = 0.00, > wmass = 0.004, > / > ATOMIC_SPECIES > > W 183.84 W.pz-bhs.UPF > Bi 208.98 083-Bi-ca-d-vgrp_ji.uspp.UPF > O 15.9994 O.pz-van_ak.UPF > > ATOMIC_POSITIONS alat > W 0.000000000 0.005016973 0.747761262 0 1 1 > W 0.500000000 0.990983027 2.252238738 0 1 1 > W 0.500000000 0.492983027 0.747761262 0 1 1 > W 0.000000000 0.503016973 2.252238738 0 1 1 > Bi 0.964931181 0.523336245 1.268793038 > Bi 0.464931181 0.472663755 1.731206962 > Bi 0.464931181 0.970663755 1.268793038 > Bi 0.964931181 0.025336245 1.731206962 > Bi 0.981726782 0.473760364 0.230569958 > Bi 0.481726782 0.522239636 2.769430042 > Bi 0.481726782 0.024239636 0.230569958 > Bi 0.981726782 0.971760364 2.769430042 > O 0.100459245 0.081203101 0.435231801 > O 0.600459245 0.914796899 2.564768199 > O 0.600459245 0.416796899 0.435231801 > O 0.100459245 0.579203101 2.564768199 > O 0.264260826 0.262143525 3.000064967 > O 0.764260826 0.733856475 -0.000064967 > O 0.764260826 0.235856475 3.000064967 > O 0.264260826 0.760143525 -0.000064967 > O 0.247153777 0.235070329 1.501190932 > O 0.747153777 0.760929671 1.498809068 > O 0.747153777 0.262929671 1.501190932 > O 0.247153777 0.733070329 1.498809068 > O 0.283152650 0.701520345 0.680189839 > O 0.783152650 0.294479655 2.319810161 > O 0.783152650 0.792479655 0.680189839 > O 0.283152650 0.203520345 2.319810161 > O 0.335776234 0.239505837 0.816308565 > O 0.835776234 0.756494163 2.183691435 > O 0.835776234 0.258494163 0.816308565 > O 0.335776234 0.737505837 2.183691435 > O 0.586065178 0.587540848 1.058853940 > O 0.086065178 0.408459152 1.941146060 > O 0.086065178 0.906459152 1.058853940 > O 0.586065178 0.089540848 1.941146060 > > > K_POINTS {automatic} > 5 5 5 1 1 1 > > thanks a lot > > Hania Djani-ait aissa > PHD student > Algeria > > > > T?l?chargez le nouveau Windows Live Messenger ! T?l?chargez > Messenger, c'est gratuit ! > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum David E. Farrell Post-Doctoral Fellow Department of Materials Science and Engineering Northwestern University email: d-farrell2 at northwestern.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090124/3da98891/attachment-0001.htm From akohlmey at cmm.chem.upenn.edu Sat Jan 24 18:05:49 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 24 Jan 2009 12:05:49 -0500 (EST) Subject: [Pw_forum] vc-relax In-Reply-To: References: Message-ID: On Sat, 24 Jan 2009, hania djani-ait aissa wrote: HDA> HDA> Dear all, HDA> i am doing calculation on orthorombic system containing 36 atoms HDA> (Bi2WO6). My computer is a workstation HP xw6400 quad-core Xeon, HDA> 2GB. Vc-relax on this system spent more than 01 mounth and still congratulations for your patience. i would have thought about alternate ways bring this to a result long ago. HDA> not finish. the energy is converging but forces fluctuate. is there HDA> something wrong in my input in terms of umass or dt? or have i to HDA> capitulate to this very slow convergence? one thing to consider is the fact, that your oxygen atoms are over ten times lighter than your other atoms. that means that the (much) heavier atoms can kick them around good, even with little movement. for structural relaxation, mass doen't matter, so i would set all atoms to the same atomic mass. HDA> here is the input file. Any suggestions will welcome. since you have constrained computational resources, have you tried running a smaller version of this system first and finding out the optimal parameters for the cutoffs and convergence thresholds? also, it looks like you have imbalanced k-point sampling. cheers, axel. HDA> &control calculation = 'vc-relax' restart_mode = 'from_scratch' HDA>pseudo_dir = '/root/pwscf/pseudo/' outdir = '/root/tmp/' tstress = are you running this as root user???? if yes, be glad you are not a phd student of mine... [...] HDA> thanks a lot HDA> HDA> Hania Djani-ait aissa HDA> PHD student HDA> Algeria HDA> HDA> HDA> _________________________________________________________________ HDA> Sur Windows Live Ideas, d?couvrez en exclusivit? de nouveaux services en ligne... si nouveaux qu'ils ne sont pas encore sortis officiellement sur le march? ! HDA> http://ideas.live.com -- ======================================================================= 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 thmason24 at yahoo.com Sat Jan 24 18:47:02 2009 From: thmason24 at yahoo.com (Timothy Mason) Date: Sat, 24 Jan 2009 09:47:02 -0800 (PST) Subject: [Pw_forum] vc-relax Message-ID: <557638.65079.qm@web81807.mail.mud.yahoo.com> Dear Hania, as someone else struggling with VC-relax, I'd like to see a grep of your cell volumes. (just grep on volume). I find that in these cases where the forces oscillate but the engergies decrease (all cases for me) the volumes continue to expand about linearly with time. I'm not sure what to make of it. It even happens starting with experimental structures often increasing them by 50 percent before convergence thresholds are met. would you please provide them? Tim Mason University of Missouri St. Louis From akohlmey at cmm.chem.upenn.edu Sat Jan 24 19:12:13 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 24 Jan 2009 13:12:13 -0500 (EST) Subject: [Pw_forum] vc-relax In-Reply-To: <557638.65079.qm@web81807.mail.mud.yahoo.com> References: <557638.65079.qm@web81807.mail.mud.yahoo.com> Message-ID: On Sat, 24 Jan 2009, Timothy Mason wrote: TM> Dear Hania, TM> TM> as someone else struggling with VC-relax, I'd like to see a grep of TM> your cell volumes. (just grep on volume). I find that in these TM> cases where the forces oscillate but the engergies decrease (all TM> cases for me) the volumes continue to expand about linearly with TM> time. I'm not sure what to make of it. It even happens starting good point. that could very likely mean, that your stress tensor is not sufficiently converged wrt. to the wavefunction or density cutoff. one of the most problematic issues with vc-relax calculations is that the stress tensor need a much higher wavefunction cutoff (and very much highe density cutoff in case of ultra-soft pseudos) to be well converged. to make matters worse, in a vc-relax calculations, you do not keep those cutoffs constant, but the number of plane waves. so for an expanding cell you effectively decrease the cutoffs, and if you stress tensor is not well converged, it will become quickly not converged. there are some tricks to make this effect less of a problem, but the immediate solution is to start at a lattice constant that is definitely too large, i.e. the cell will never expand beyond it. you still have consider pulay stress, but that is relatively minor concern. TM> with experimental structures often increasing them by 50 percent TM> before convergence thresholds are met. a 50% increase sounds unphysical. have you tried restarting from the final configuration (resets the "effective" basis set size) and see if the results are consistent? cheers, axel. TM> TM> would you please provide them? TM> TM> Tim Mason TM> University of Missouri St. Louis TM> _______________________________________________ TM> Pw_forum mailing list TM> Pw_forum at pwscf.org TM> http://www.democritos.it/mailman/listinfo/pw_forum TM> -- ======================================================================= 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 thmason24 at yahoo.com Sat Jan 24 19:36:57 2009 From: thmason24 at yahoo.com (Timothy Mason) Date: Sat, 24 Jan 2009 10:36:57 -0800 (PST) Subject: [Pw_forum] vc-relax Message-ID: <843763.90245.qm@web81802.mail.mud.yahoo.com> Axel, thanks for the insight. That is outstanding information. I expect many people are running into this problem (I've seen the force oscillation energy decrease phenomenom posted many times). I am using what I thought were conservative ecuts (40 ryd) but perhaps not big enough for the stress tensor when using USPP. This characteristic would be a great thing to warn people about in the wiki vc-relax section. I'd offer to do it myself but I'd feel better leaving it to an expert. I'll report back if this improves my calculation but with the higher ecuts it won't be for a while. Is the issue confined to USPP? would you recommend NCPP for VC-relax as a work around? Tim Mason University of Missouri St. Louis On Sat, Jan 24, 2009 at 12:12 PM, Axel Kohlmeyer wrote: On Sat, 24 Jan 2009, Timothy Mason wrote: TM> Dear Hania, TM> TM> as someone else struggling with VC-relax, I'd like to see a grep of TM> your cell volumes. (just grep on volume). I find that in these TM> cases where the forces oscillate but the engergies decrease (all TM> cases for me) the volumes continue to expand about linearly with TM> time. I'm not sure what to make of it. It even happens starting good point. that could very likely mean, that your stress tensor is not sufficiently converged wrt. to the wavefunction or density cutoff. one of the most problematic issues with vc-relax calculations is that the stress tensor need a much higher wavefunction cutoff (and very much highe density cutoff in case of ultra-soft pseudos) to be well converged. to make matters worse, in a vc-relax calculations, you do not keep those cutoffs constant, but the number of plane waves. so for an expanding cell you effectively decrease the cutoffs, and if you stress tensor is not well converged, it will become quickly not converged. there are some tricks to make this effect less of a problem, but the immediate solution is to start at a lattice constant that is definitely too large, i.e. the cell will never expand beyond it. you still have consider pulay stress, but that is relatively minor concern. TM> with experimental structures often increasing them by 50 percent TM> before convergence thresholds are met. a 50% increase sounds unphysical. have you tried restarting from the final configuration (resets the "effective" basis set size) and see if the results are consistent? cheers, axel. From djaithania at hotmail.com Sat Jan 24 20:14:58 2009 From: djaithania at hotmail.com (hania djani-ait aissa) Date: Sat, 24 Jan 2009 19:14:58 +0000 Subject: [Pw_forum] vc-relax In-Reply-To: References: Message-ID: Dear Axel, Maybe the ignorance mixed with generosity can lead to great patience... What do you mean by imbalanced k-point sampling? Running as root user is the bad habit of our Lab. I will try to impose sessions. If i was a Phd student of you, this woul be a different story :) thanks for your response, Hania Djani Ait Aissa PHd student Algeria> Date: Sat, 24 Jan 2009 12:05:49 -0500> From: akohlmey at cmm.chem.upenn.edu> To: djaithania at hotmail.com> CC: pw_forum at pwscf.org> Subject: Re: [Pw_forum] vc-relax> > On Sat, 24 Jan 2009, hania djani-ait aissa wrote:> > HDA> > HDA> Dear all,> > HDA> i am doing calculation on orthorombic system containing 36 atoms > HDA> (Bi2WO6). My computer is a workstation HP xw6400 quad-core Xeon, > HDA> 2GB. Vc-relax on this system spent more than 01 mounth and still > > congratulations for your patience. i would have thought about > alternate ways bring this to a result long ago.> > HDA> not finish. the energy is converging but forces fluctuate. is there > HDA> something wrong in my input in terms of umass or dt? or have i to > HDA> capitulate to this very slow convergence?> > one thing to consider is the fact, that your oxygen atoms> are over ten times lighter than your other atoms. that means> that the (much) heavier atoms can kick them around good, even> with little movement. for structural relaxation, mass doen't > matter, so i would set all atoms to the same atomic mass.> > HDA> here is the input file. Any suggestions will welcome. > > since you have constrained computational resources, have you> tried running a smaller version of this system first and finding> out the optimal parameters for the cutoffs and convergence thresholds?> > also, it looks like you have imbalanced k-point sampling.> > cheers,> axel.> > HDA> &control calculation = 'vc-relax' restart_mode = 'from_scratch' > HDA>pseudo_dir = '/root/pwscf/pseudo/' outdir = '/root/tmp/' tstress = > > are you running this as root user???? > if yes, be glad you are not a phd student of mine...> > > [...]> > HDA> thanks a lot> HDA> > HDA> Hania Djani-ait aissa> HDA> PHD student> HDA> Algeria> HDA> > HDA> > HDA> _________________________________________________________________> HDA> Sur Windows Live Ideas, d?couvrez en exclusivit? de nouveaux services en ligne... si nouveaux qu'ils ne sont pas encore sortis officiellement sur le march? !> HDA> http://ideas.live.com> > -- > =======================================================================> 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. _________________________________________________________________ D?couvrez toutes les possibilit?s de communication avec vos proches http://www.microsoft.com/windows/windowslive/default.aspx -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090124/9da265f5/attachment.htm From baroni at sissa.it Sun Jan 25 09:25:30 2009 From: baroni at sissa.it (Stefano Baroni) Date: Sun, 25 Jan 2009 09:25:30 +0100 Subject: [Pw_forum] TD-DFT In-Reply-To: References: <11338.79.6.168.51.1232294679.squirrel@mail.unito.it> <79CA99D6-A6B0-4CA2-A7E2-2A358FE68A1F@sissa.it> <19B3EBBD-0A06-48DA-88BF-1FDE38DFE187@sissa.it> <738DCC90-F2F1-4EF5-83FC-D9E2003A6BC7@sissa.it> Message-ID: <5D99BBB3-7B7A-4824-8513-9F2CE89022C9@sissa.it> On Jan 24, 2009, at 4:47 PM, lan haiping wrote: > Thanks for the instruction! > Does it mean that i should try to use BSE method to examine > excitons' properties? yes (TDDFT with exact exchange would also do, though) > I do wonder whether this failure is due to my solid systems . My > basic knowledge is that tddft cannot predict proper > optical spectra for solids system but wonderful results for isolated > systems due to no standard functionals for this scheme. yes - this is the populace's (mis-) belief. the reason why TDDFT completely fails for excitons is exactly the same why it behaves (sometimes) poorly in finite systems. eg, it cannot predict Rydberg series. After all, what is the physical difference between an infinite system and a (large) finite one? shouldn't the results for the latter converge to those of the former when the size grows large enough? Stefano > > > Best, > Hai-Ping > > On Sat, Jan 24, 2009 at 10:27 PM, Stefano Baroni > wrote: > ian haipin: no problem with the size of the system. > bad news about excitons - no standard (lda/gga) implementation of > tddft cane ever hope to get them right (to "get them" altogether, I > would say) > Stafano > > On Jan 24, 2009, at 10:06 AM, lan haiping wrote: > >> Dear Stefano & Baris: >> >> I am sorry for delayed reply due to our Chinese New year holidays. >> I feel interested in exictonic properties of tinanites' derivates, >> and want to examine their optical properties with TDDFT framework. >> Though i have read some related papers include your works on this >> issue, >> i wonder about the computational load in this scheme since their >> relative large systems i should handle with (about 50 atoms 300 >> electrons). >> >> Thanks for your help and offers. >> >> Best Wishes, >> >> Hai-Ping >> >> >> On Tue, Jan 20, 2009 at 6:27 AM, Stefano Baroni >> wrote: >> Dear Haiping, >> >> please have a read at our paper first, and then revert to us with a >> proposal on what to do with the code. We will be glad to help with >> any suitable project you may have. No problem at all in writing to >> each other personally. >> >> Best regards >> Stefano >> >> On Jan 19, 2009, at 9:27 PM, lan haiping wrote: >> >>> Dear Stefano, >>> >>> >>> Is it possible for me to have a pre-public version for tddft >>> module of QE ? >>> I am reading the LNP book on TDDFT, and curious about >>> your work on this issue . >>> >>> Sorry for mailing you directly. >>> >>> Best wishes, >>> >>> Hai-Ping >>> >>> On Mon, Jan 19, 2009 at 4:25 AM, Stefano Baroni >>> wrote: >>> Anna: we do have an experimental code implementing TDDFT in QE, >>> see: J. Chem. Phys. 128, 154105 (2008). It is not yet public, >>> though. If you want to give it a try, please contact me or Baris >>> Malcioglu who is reading me in cc. Cheers - Stefano >>> >>> >>> On Jan 18, 2009, at 5:04 PM, anna.ferrari at unito.it wrote: >>> >>>> dear all, >>>> TD-DFT is implemented in PWSCF? >>>> In which codes is TDDFT implemented for solids? >>>> >>>> thanks a lot >>>> >>>> anna ferrari >>>> dip. chimica IFM >>>> univ. Torino >>>> _______________________________________________ >>>> 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 >>> >>> >>> >>> >>> -- >>> Hai-Ping Lan >>> Department of Electronics , >>> Peking University , Bejing, 100871 >>> lanhaiping at gmail.com, hplan at pku.edu.cn >> >> --- >> 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 >> >> >> >> >> >> >> >> >> >> -- >> Hai-Ping Lan >> Department of Electronics , >> Peking University , Bejing, 100871 >> lanhaiping at gmail.com, hplan at pku.edu.cn > > --- > 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 > > > > > > > > > > -- > Hai-Ping Lan > Department of Electronics , > Peking University , Bejing, 100871 > lanhaiping at gmail.com, hplan at pku.edu.cn --- 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/20090125/8b6deeeb/attachment-0001.htm From sezginaydin at gazi.edu.tr Sun Jan 25 14:06:11 2009 From: sezginaydin at gazi.edu.tr (Sezgin AYDIN) Date: Sun, 25 Jan 2009 15:06:11 +0200 Subject: [Pw_forum] spin-orbit coupling Message-ID: <000001c97eed$b1fff3a0$15ffdae0$@edu.tr> Dear PWScf users, How can I calculate the spin-orbit coupling, Thanks for your interest Sezgin -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090125/69310cad/attachment.htm From eyvaz_isaev at yahoo.com Sun Jan 25 14:27:11 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sun, 25 Jan 2009 05:27:11 -0800 (PST) Subject: [Pw_forum] spin-orbit coupling In-Reply-To: <000001c97eed$b1fff3a0$15ffdae0$@edu.tr> Message-ID: <106779.58463.qm@web65712.mail.ac4.yahoo.com> Dear Sezgin, Most likely, you should see /examples/example22. Besides, you will need a fully-relativistic pseudopotential for this purpose. 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 Sun, 1/25/09, Sezgin AYDIN wrote: > From: Sezgin AYDIN > Subject: [Pw_forum] spin-orbit coupling > To: "PWScf" > Date: Sunday, January 25, 2009, 4:06 PM > Dear PWScf users, > > How can I calculate the spin-orbit coupling, > > > > Thanks for your interest > > Sezgin > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From djaithania at hotmail.com Sun Jan 25 15:20:26 2009 From: djaithania at hotmail.com (hania djani-ait aissa) Date: Sun, 25 Jan 2009 14:20:26 +0000 Subject: [Pw_forum] vc-relax In-Reply-To: <843763.90245.qm@web81802.mail.mud.yahoo.com> References: <843763.90245.qm@web81802.mail.mud.yahoo.com> Message-ID: Dear Tim,here attached my grep on volume. cheers Hania Djani ait aissa Algeria> Date: Sat, 24 Jan 2009 10:36:57 -0800> From: thmason24 at yahoo.com> To: pw_forum at pwscf.org> CC: akohlmey at cmm.chem.upenn.edu> Subject: Re: [Pw_forum] vc-relax> > Axel,> > thanks for the insight. That is outstanding information. I expect many people are running into this problem (I've seen the force oscillation energy decrease phenomenom posted many times). I am using what I thought were conservative ecuts (40 ryd) but perhaps not big enough for the stress tensor when using USPP. This characteristic would be a great thing to warn people about in the wiki vc-relax section. I'd offer to do it myself but I'd feel better leaving it to an expert. I'll report back if this improves my calculation but with the higher ecuts it won't be for a while. Is the issue confined to USPP? would you recommend NCPP for VC-relax as a work around?> > > > > Tim Mason> University of Missouri St. Louis> > > > > > > On Sat, Jan 24, 2009 at 12:12 PM, Axel Kohlmeyer wrote:> > On Sat, 24 Jan 2009, Timothy Mason wrote:> > TM> Dear Hania,> TM>> > TM> as someone else struggling with VC-relax, I'd like to see a grep of> TM> your cell volumes. (just grep on volume). I find that in these> TM> cases where the forces oscillate but the engergies decrease (all> TM> cases for me) the volumes continue to expand about linearly with> TM> time. I'm not sure what to make of it. It even happens starting> > good point. that could very likely mean, that your stress> tensor is not sufficiently converged wrt. to the wavefunction> or density cutoff. one of the most problematic issues with> vc-relax calculations is that the stress tensor need a much> higher wavefunction cutoff (and very much highe density cutoff> in case of ultra-soft pseudos) to be well converged. to make> matters worse, in a vc-relax calculations, you do not keep> those cutoffs constant, but the number of plane waves. so for> an expanding cell you effectively decrease the cutoffs, and> if you stress tensor is not well converged, it will become> quickly not converged. there are some tricks to make this> effect less of a problem, but the immediate solution is to> start at a lattice constant that is definitely too large,> i.e. the cell will never expand beyond it. you still have> consider pulay stress, but that is relatively minor concern.> > TM> with experimental structures often increasing them by 50 percent> TM> before convergence thresholds are met.> > a 50% increase sounds unphysical. have you tried restarting> from the final configuration (resets the "effective" basis> set size) and see if the results are consistent?> > cheers,> axel.> > > _______________________________________________> Pw_forum mailing list> Pw_forum at pwscf.org> http://www.democritos.it/mailman/listinfo/pw_forum _________________________________________________________________ D?couvrez toutes les possibilit?s de communication avec vos proches http://www.microsoft.com/windows/windowslive/default.aspx -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090125/9b417a9c/attachment-0001.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: grep vol.pdf Type: application/pdf Size: 14930 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20090125/9b417a9c/attachment-0001.pdf From sezginaydin at gazi.edu.tr Mon Jan 26 09:12:59 2009 From: sezginaydin at gazi.edu.tr (sezginaydin at gazi.edu.tr) Date: Mon, 26 Jan 2009 08:12:59 -0000 Subject: [Pw_forum] elastic constant and bulk modulus Message-ID: Dear PWScf users, how can I calculate elastic constant and bulk modulus by using PWScf package thanks for your help -sezgin From eyvaz_isaev at yahoo.com Mon Jan 26 10:49:12 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 26 Jan 2009 01:49:12 -0800 (PST) Subject: [Pw_forum] elastic constant and bulk modulus In-Reply-To: Message-ID: <312108.51848.qm@web65710.mail.ac4.yahoo.com> Dear Sezgin, You should find the same question on the forum. Nevertheless, you can calculate elastic constants and bulk modulus applying strains according to original paper of Mike Mehl et al., Phys.Rev.B 41, 10311 for cubic and tetragonal lattices. For hexagonal lattices see a paper of R.Cohen on elastic constants of iron published also in PRB. Anyway, search PRB for straines for different lattices. Of course, you could calculate elastic constants as a slope of phonon dispersion relations in the long-wave approximation i.e. for q-points near Gamma point, but this way seems to be more expensive, as the derivative is quite sensitive to details of phonon calculations and one needs converged phonon calculations for this purpose. And I am not sure whether elastic constants calculations in DFPT according to Hamann et al. is implemented. At least, I never heard about this opportunity. For bulk modulus only calculations just calculate E(V) or (E(a) for cubic lattices) and then use ev.x from /tools directory to find B. Hope this hels. 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, 1/26/09, sezginaydin at gazi.edu.tr wrote: > From: sezginaydin at gazi.edu.tr > Subject: [Pw_forum] elastic constant and bulk modulus > To: pw_forum at pwscf.org > Date: Monday, January 26, 2009, 11:12 AM > Dear PWScf users, > how can I calculate elastic constant and bulk modulus by > using PWScf package > > thanks for your help > -sezgin > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From baroni at sissa.it Mon Jan 26 10:56:18 2009 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 26 Jan 2009 10:56:18 +0100 Subject: [Pw_forum] elastic constant and bulk modulus In-Reply-To: <312108.51848.qm@web65710.mail.ac4.yahoo.com> References: <312108.51848.qm@web65710.mail.ac4.yahoo.com> Message-ID: <894EDF7B-863B-4930-931A-15564A3DBB55@sissa.it> On Jan 26, 2009, at 10:49 AM, Eyvaz Isaev wrote: > > And I am not sure whether elastic constants calculations in DFPT > according to Hamann et al. is implemented. At least, I never heard > about this opportunity. Hamann et al's algorithm used to be implemented in a prehistoric version of QE ages before they "invented" it, see: PRL, 59, 2662 (1987). Unfortunately, it is no longer there. Anybody volunteering to reimplement it? Stefano --- 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/20090126/c51359f9/attachment.htm From eyvaz_isaev at yahoo.com Mon Jan 26 11:34:00 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 26 Jan 2009 02:34:00 -0800 (PST) Subject: [Pw_forum] elastic constant and bulk modulus In-Reply-To: <894EDF7B-863B-4930-931A-15564A3DBB55@sissa.it> Message-ID: <275616.99407.qm@web65704.mail.ac4.yahoo.com> Hi Stefano, Thanks for the Reference. Bests, Eyvaz. --- On Mon, 1/26/09, Stefano Baroni wrote: > From: Stefano Baroni > Subject: Re: [Pw_forum] elastic constant and bulk modulus > To: eyvaz_isaev at yahoo.com, "PWSCF Forum" > Date: Monday, January 26, 2009, 12:56 PM > On Jan 26, 2009, at 10:49 AM, Eyvaz Isaev wrote: > > > > And I am not sure whether elastic constants > calculations in DFPT according to Hamann et al. is > implemented. At least, I never heard about this opportunity. > > Hamann et al's algorithm used to be implemented in a > prehistoric version of QE ages before they > "invented" it, see: PRL, 59, 2662 (1987). > Unfortunately, it is no longer there. Anybody volunteering > to reimplement it? > > Stefano > > --- > 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 From adams_mejl at hotmail.com Mon Jan 26 16:44:51 2009 From: adams_mejl at hotmail.com (Adam Jacobsson) Date: Mon, 26 Jan 2009 16:44:51 +0100 Subject: [Pw_forum] Phonon calculations and LDA+U In-Reply-To: <275616.99407.qm@web65704.mail.ac4.yahoo.com> References: <894EDF7B-863B-4930-931A-15564A3DBB55@sissa.it> <275616.99407.qm@web65704.mail.ac4.yahoo.com> Message-ID: Dear all, I wonder about the status of the implementation of LDA+U into the phonon code. Has it been implemented in the newest version or is it still an unsolved issue? Thanks in advance, Adam Jacobsson Master Thesis student Uppsala University _________________________________________________________________ 25 GB gratis lagring p? n?tet - Skaffa SkyDrive nu! http://skydrive.live.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090126/931922c8/attachment.htm From eyvaz_isaev at yahoo.com Mon Jan 26 17:52:43 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 26 Jan 2009 08:52:43 -0800 (PST) Subject: [Pw_forum] Phonon calculations and LDA+U In-Reply-To: Message-ID: <810273.31794.qm@web65715.mail.ac4.yahoo.com> Dear Adam, As far as know it is still under testing. I remember, Matteo told me that I am the first in the queue list for this one. 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, 1/26/09, Adam Jacobsson wrote: > From: Adam Jacobsson > Subject: [Pw_forum] Phonon calculations and LDA+U > To: pw_forum at pwscf.org > Date: Monday, January 26, 2009, 6:44 PM > Dear all, > > I wonder about the status of the implementation of LDA+U > into the phonon code. Has it been implemented in the newest > version or is it still an unsolved issue? > > Thanks in advance, > Adam Jacobsson > Master Thesis student > Uppsala University > > _________________________________________________________________ > 25 GB gratis lagring p? n?tet - Skaffa SkyDrive nu! > http://skydrive.live.com > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From matteo at umn.edu Mon Jan 26 18:27:22 2009 From: matteo at umn.edu (Matteo Cococcioni) Date: Mon, 26 Jan 2009 11:27:22 -0600 Subject: [Pw_forum] Phonon calculations and LDA+U In-Reply-To: <810273.31794.qm@web65715.mail.ac4.yahoo.com> References: <810273.31794.qm@web65715.mail.ac4.yahoo.com> Message-ID: <497DF27A.2060700@umn.edu> The implementation of the LDA+U into the phonon code is still under construction. Andrea Floris, Stefano de Gironcoli and I have been working at this project quite much in the last months and I think we are now much closer to a working implementation. There are still few (hopefully small) things that need to be clarified and then it will be ready. We will announce when the implementation will be ported into the CVS version of the code. Regards, Matteo Eyvaz Isaev wrote: > Dear Adam, > > As far as know it is still under testing. I remember, Matteo told me that I am the first in the queue list for this one. > > 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, 1/26/09, Adam Jacobsson wrote: > > >> From: Adam Jacobsson >> Subject: [Pw_forum] Phonon calculations and LDA+U >> To: pw_forum at pwscf.org >> Date: Monday, January 26, 2009, 6:44 PM >> Dear all, >> >> I wonder about the status of the implementation of LDA+U >> into the phonon code. Has it been implemented in the newest >> version or is it still an unsolved issue? >> >> Thanks in advance, >> Adam Jacobsson >> Master Thesis student >> Uppsala University >> >> _________________________________________________________________ >> 25 GB gratis lagring p? n?tet - Skaffa SkyDrive nu! >> http://skydrive.live.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 > -------------- 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/20090126/28a3cefe/attachment.vcf From baris.malcioglu at gmail.com Mon Jan 26 18:22:52 2009 From: baris.malcioglu at gmail.com (O. Baris Malcioglu) Date: Mon, 26 Jan 2009 19:22:52 +0200 Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: Dear Karoly, If you are willing to write an extension (including post processing codes) to PWSCF, I would really recommend you to use already present subroutines. For example in your case, you may have a look at 'davcio'. It depends on a number of preset constants. An example on how to use it might be like this (I am omitting initialization which, I am afraid, you will have to tailor according to your needs) : --------------------------------------------------------------------------------------------------------------------------- use klist, only : nks, xk use gvect, only : ngm, g, ecutwfc use io_files, only : nwordwfc, iunwfc, prefix use wvfct, only : npw, igk, nbnd, g2kin, npwx use wavefunctions_module, only : evc .....(whatever initialization necessary, have a look at PW/read_file.f90 for reading already computed ones)... nwordwfc = 2 * nbnd * npwx !The 'size' of the each wavefunction segment written in file ! call diropn ( iunwfc, 'wfc', nwordwfc, exst) ! This opens the .wfc files in a machine dependent way. You have to set 'prefix' first. ! if (.not.exst) call errore('lr_read_wfc', TRIM( prefix )//'.wfc'//' not found',1) !Everyone here had its nice experiences with "errore" right? The error reporting mechanism of PWSCF ! do ik=1,nks ! k-point loop for this pool ! !CALL gk_sort( xk(1,ik), ngm, g, ( ecutwfc / tpiba2 ), npw, igk, g2kin ) !sorts the g vectors for identification purposes, this assures that each processor has the same correspondence list (probably you need this for your purposes ) ! call davcio(evc(1,1,ik),nwordwfc,iunwfc,ik,-1) !Reads the wfc. Parameters: The variable where the wavefunction is stored, the block size, the unit of the file, the k point you are interested in, and an integer <0 reads >0 writes ! enddo ! CLOSE( UNIT = iunwfc) ! And we are done with the wfc files !Note The wafefunctions are stored as : evc(plane_wave_index,band_index,k_point_index) --------------------------------------------------------------------------------------------------------------------------- For finding the number of plane waves (for each k point) you may either investigate a pw.x screen output where they are (were?) reported in the beginning, or use "PW/n_plane_waves.f90" Hope this helps, O. Baris Malcioglu SISSA CM Sector. On Thu, Jan 22, 2009 at 7:33 PM, Karoly Nemeth wrote: > > Thank you Gabriele! Since most of the time the *.wfc* are split among a > few files, a utility that uses as input only the number of processors, > nspin, and perhaps the number of basis functions could be useful. If the > wfc files would be in hdf5 format it would be probably easier to figure > out all their structure, but since this is not the case, a utility program > could be quite useful for this purpose. > > Karoly > > *********************************************************************** > Karoly Nemeth, Ph.D. > Argonne National Laboratory > 9700 S. Cass Ave. Argonne, IL 60439. > Advanced Photon Source > Accelerator Physics and Operations Group > Building 401, Room B2200 > Tel. 630-252-5813 > Fax: 630-252-4732 > *********************************************************************** > > > >> Message: 3 >> Date: Thu, 22 Jan 2009 09:54:17 +0100 >> From: Gabriele Sclauzero >> Subject: Re: [Pw_forum] wavefunction coefficients >> To: PWSCF Forum >> Message-ID: <49783439.40104 at sissa.it> >> Content-Type: text/plain; charset=us-ascii; format=flowed >> >> Hi, >> >> of course Paolo and Axel could help you much more than me on this issue, but inthe >> meanwhile: >> >> Karoly Nemeth wrote: >>> Hi, >>> >>> I would like to obtain the wavefunction coefficients u(i,k,G), where "i" >>> is the band-index, "k" is the k-space vector used in the discretization of >>> the Brillouin-zone in a PW calculation and "G" is the plane-wave basis >>> function's wave-vector (assuming a complex form as exp[i(k+G)r] ). >> >> I think that the quantities you are interested in are store in the variable "evc" of the >> pw program (but only in certains parts of the program and only for one k-point at a time; >> the wavefunctions for all the k-points are stored on the .wfc files). >> >>> >>> Is there any utility program that is independent from the rest of PWSCF >>> and would provide some sort of tabulation on u(i,k,G) as a function >>> of i,k and G , for a parallel calculation? Perhaps a quick quide to the >>> format of this information in the *.wfc* files written by a parallel run? >> >> Be careful that in parallel runs the G vectors (i.e. plane waves) are split across >> processor, so that each processor may be working only with a slab of the full grid. >> The option wfc_collect=.TRUE. collects all wavefunctions on a single file after the >> parallel run (maybe it can help you understand how to gather all the coefficients). >> >> GS >> >> PS: The forum community would be grateful if you specify your affiliation... thanks. >> >>> >>> I would appreciate any hint that could save me some time from going >>> through the source code for this info. >>> Thanks a lot in advance: >>> >>> Karoly >>> > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From baris.malcioglu at gmail.com Mon Jan 26 18:48:55 2009 From: baris.malcioglu at gmail.com (O. Baris Malcioglu) Date: Mon, 26 Jan 2009 19:48:55 +0200 Subject: [Pw_forum] (no subject) In-Reply-To: References: Message-ID: Dear Sezgin, The problem is the exact same thing reported to you, integrated charge is not what is expected. You should agree that since this is a DFT code, this error is quite dramatic. There are a couple of things you may try, first of all, you may try increasing the ecutrho. Unlike norm conserving PPs, it is better (almost always mandatory) if you set ecutrho yourself when using ultrasoft pseudo potentials. You may give "10 times the ecutwfc" a try first, and adjust according to your needs. This, in fact, is a very commonly asked question. May I suggest you to look at the archives for further reference? Also, as a rule of etiquette, it would be better if you add your affiliation and contact information to your mail. Baris Malcioglu SISSA, CM Sector On Fri, Jan 23, 2009 at 1:52 PM, wrote: > Dear PWScf users, > I work on the PtN Rocksalt structure. My input file for pw.x as foolowing: > ------------------------------- > PtN > 2 atoms, NaCl bulk > &control > calculation='scf' > restart_mode='from_scratch', > prefix='PtN' > pseudo_dir = '/home/sezgin/espresso-4.0.1/pseudo', > outdir='./' > / > &system > ibrav = 2, > celldm(1) = 7.7 > nat= 2, > ntyp= 2, > ecutwfc = 20 > / > &electrons > conv_thr = 1.0e-8 > mixing_beta = 0.7 > mixing_mode = 'plain' > / > ATOMIC_SPECIES > N 14.0 N.pw91-van_ak.UPF > Pt 190.0 Pt.pw91-n-van.UPF > ATOMIC_POSITIONS > N 0.00 0.00 0.00 > Pt 0.5 0.5 0.5 > K_POINTS {automatic} > 3 3 3 0 0 0 > ----------------------------- > > when i work pw.x, i get the error: > > WARNING: integrated charge= 14.30000000, expected= 15.00000000 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from electrons : error # 1 > charge is wrong > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > what is the problem? > thanks for your interest > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From eyvaz_isaev at yahoo.com Mon Jan 26 20:21:32 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 26 Jan 2009 11:21:32 -0800 (PST) Subject: [Pw_forum] Phonon calculations and LDA+U In-Reply-To: <497DF27A.2060700@umn.edu> Message-ID: <926704.34045.qm@web65704.mail.ac4.yahoo.com> Hi Matteo, Thanks for the clarification. Bests, Eyvaz. --- On Mon, 1/26/09, Matteo Cococcioni wrote: > From: Matteo Cococcioni > Subject: Re: [Pw_forum] Phonon calculations and LDA+U > To: eyvaz_isaev at yahoo.com, "PWSCF Forum" > Date: Monday, January 26, 2009, 8:27 PM > The implementation of the LDA+U into the phonon code is > still under construction. Andrea Floris, Stefano de > Gironcoli and I have been > working at this project quite much in the last months and I > think we are now much closer to a working implementation. > There are still > few (hopefully small) things that need to be clarified and > then it will be ready. > We will announce when the implementation will be ported > into the CVS version of the code. > > Regards, > > Matteo > > > > Eyvaz Isaev wrote: > > Dear Adam, > > > > As far as know it is still under testing. I remember, > Matteo told me that I am the first in the queue list for > this one. > > > > 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, 1/26/09, Adam Jacobsson > wrote: > > > > > >> From: Adam Jacobsson > > >> Subject: [Pw_forum] Phonon calculations and LDA+U > >> To: pw_forum at pwscf.org > >> Date: Monday, January 26, 2009, 6:44 PM > >> Dear all, > >> > >> I wonder about the status of the implementation of > LDA+U > >> into the phonon code. Has it been implemented in > the newest > >> version or is it still an unsolved issue? > >> > >> Thanks in advance, > >> Adam Jacobsson > >> Master Thesis student > >> Uppsala University > >> > >> > _________________________________________________________________ > >> 25 GB gratis lagring p? n?tet - Skaffa SkyDrive > nu! > >> http://skydrive.live.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 > > From lpmc_med at yahoo.com Mon Jan 26 22:56:41 2009 From: lpmc_med at yahoo.com (Lpmc Med) Date: Mon, 26 Jan 2009 13:56:41 -0800 (PST) Subject: [Pw_forum] How to calculate the Partial Phonon Density Of States ? Message-ID: <730870.66977.qm@web112217.mail.gq1.yahoo.com> Dear PWscf users, ? I performed a phonon calculation but I can't calculate the?Partial?Phonon Density Of States. Could?you tell me how it can be calculated ? please thank you very much ? Hadj Med University of Oran (Algeria) Condensed Matter Laboratory ? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090126/8400dc98/attachment.htm From eyvaz_isaev at yahoo.com Mon Jan 26 23:24:15 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 26 Jan 2009 14:24:15 -0800 (PST) Subject: [Pw_forum] How to calculate the Partial Phonon Density Of States ? In-Reply-To: <730870.66977.qm@web112217.mail.gq1.yahoo.com> Message-ID: <222609.18789.qm@web65715.mail.ac4.yahoo.com> Dear Hadj, Well, I will send you the code. 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, 1/27/09, Lpmc Med wrote: > From: Lpmc Med > Subject: [Pw_forum] How to calculate the Partial Phonon Density Of States ? > To: pw_forum at pwscf.org > Date: Tuesday, January 27, 2009, 12:56 AM > Dear PWscf users, > ? > I performed a phonon calculation but I can't calculate > the?Partial?Phonon Density Of States. > Could?you tell me how it can be calculated ? please > thank you very much > ? > Hadj Med > University of Oran (Algeria) > Condensed Matter Laboratory > ? > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From jasius_1 at yahoo.com Tue Jan 27 17:41:35 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Tue, 27 Jan 2009 08:41:35 -0800 (PST) Subject: [Pw_forum] convergence issues In-Reply-To: <1102475624-1232321057-cardhu_decombobulator_blackberry.rim.net-536706383-@bxe249.bisx.prod.on.blackberry> Message-ID: <622399.17326.qm@web54407.mail.yahoo.com> Hi, the simplest polymer system wouldn'd converge the first SCF cycle after 100 iterations. I would appreciate any comments Jonas University of Iowa iteration #100 ecut= 25.00 Ry beta=0.50 Davidson diagonalization with overlap ethr = 4.79E-07, avg # of iterations = 8.9 negative rho (up, down): 0.466E-01 0.000E+00 total cpu time spent up to now is 35393.00 secs total energy = -323.97794967 Ry Harris-Foulkes estimate = -323.97794161 Ry estimated scf accuracy < 0.00009414 Ry End of self-consistent calculation convergence NOT achieved after 100 iterations: stopping INPUT &control calculation='relax', prefix='44tpcb_PBE_optimization', pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', outdir='/home/jbaltrus/atom-lsda/', nstep=200 / &system ibrav=12, nat=38, ntyp=4, a=11.367, b=9.749, c=14.273, cosab=-0.36922777, ecutwfc=25.0 / &electrons mixing_beta=0.5D0, conv_thr = 1.D-8 / &ions ion_dynamics = ?none? / ATOMIC_SPECIES O 15.9994 O.pbe-van_ak.UPF H 1.00794 H.pbe-van_ak.UPF N 14.00674 N.pbe-van_ak.UPF C 12.0107 C.pbe-van_ak.UPF ATOMIC_POSITIONS O 0.5906d0 0.1506d0 -0.0226d0 H 0.5262d0 0.1780d0 -0.0127d0 O 0.4362d0 0.4661d0 -0.3017d0 H 0.3777d0 0.4763d0 -0.2797d0 N 0.3866d0 0.2227d0 0.0225d0 N -0.2359d0 0.4717d0 0.2407d0 C 0.5970d0 0.2136d0 -0.1054d0 C 0.5078d0 0.3087d0 -0.1618d0 H 0.4379d0 0.3310d0 -0.1433d0 C 0.5203d0 0.3707d0 -0.2445d0 C 0.6226d0 0.3399d0 -0.2715d0 H 0.6319d0 0.3834d0 -0.3279d0 C 0.7110d0 0.2453d0 -0.2152d0 H 0.7808d0 0.2233d0 -0.2338d0 C 0.6996d0 0.1818d0 -0.1322d0 H 0.7612d0 0.1172d0 -0.0939d0 C 0.3805d0 0.3404d0 0.0679d0 H 0.4571d0 0.3888d0 0.1017d0 C 0.2693d0 0.3958d0 0.0688d0 H 0.2707d0 0.4799d0 0.1028d0 C 0.1556d0 0.3288d0 0.0201d0 C 0.1626d0 0.2053d0 -0.0245d0 H 0.0878d0 0.1537d0 -0.0576d0 C 0.2765d0 0.1569d0 -0.0215d0 H 0.2778d0 0.0710d0 -0.0525d0 C 0.0295d0 0.3910d0 0.0066d0 H -0.0351d0 0.3174d0 -0.0023d0 C 0.0213d0 0.4989d0 0.0827d0 H 0.1082d0 0.5246d0 0.1298d0 C -0.0637d0 0.4764d0 0.1406d0 C -0.0454d0 0.5536d0 0.2264d0 H 0.0275d0 0.6101d0 0.2536d0 C -0.1322d0 0.5488d0 0.2720d0 H -0.1172d0 0.6043d0 0.3299d0 C -0.2497d0 0.3918d0 0.1611d0 H -0.3209d0 0.3323d0 0.1380d0 C -0.1675d0 0.3903d0 0.1106d0 H -0.1821d0 0.3300d0 0.0551d0 K_POINTS {automatic} 4 4 4 0 0 0 From sclauzer at sissa.it Tue Jan 27 18:02:50 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Tue, 27 Jan 2009 18:02:50 +0100 Subject: [Pw_forum] convergence issues In-Reply-To: <622399.17326.qm@web54407.mail.yahoo.com> References: <622399.17326.qm@web54407.mail.yahoo.com> Message-ID: <497F3E3A.1080105@sissa.it> Hi Jonas Jonas Baltrusaitis wrote: > > INPUT > > &control > calculation='relax', > prefix='44tpcb_PBE_optimization', > pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', > outdir='/home/jbaltrus/atom-lsda/', > nstep=200 > / > &system > ibrav=12, > nat=38, > ntyp=4, > a=11.367, > b=9.749, > c=14.273, > cosab=-0.36922777, > ecutwfc=25.0 As a first hint I would suggest to specify ecutrho, since you're using Ultrasoft PPs, and test convergence of energy (or other quantities you're interested in) w.r.t. to this parameter (I usually choose ecutrho ~ 10*ecutwfc). If you want to do spin polarized calculations, you need to specify nspin=2 and a non-zero starting magnetization for one species at least. > / > &electrons > mixing_beta=0.5D0, Then maybe decrease the mixing parameter, or use another mixing (for instance, mixing_mode='local-TF' works well for highly inhomogeneous systems). > conv_thr = 1.D-8 > / > &ions > ion_dynamics = ?none? Are you doing a relaxation or what? > / > K_POINTS {automatic} > 4 4 4 0 0 0 Do you want to study an isolated molecule? If so you don't need a k-point sampling, simply use > K_POINTS GAMMA Regards GS > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From jasius_1 at yahoo.com Tue Jan 27 18:06:51 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Tue, 27 Jan 2009 09:06:51 -0800 (PST) Subject: [Pw_forum] convergence issues In-Reply-To: <497F3E3A.1080105@sissa.it> Message-ID: <219843.56157.qm@web54408.mail.yahoo.com> Somewhat confused: I am doing a 3-D crystal from the experimental structure optimization. Why are the last two parts of my input wrong with this respect? > > / > > &ions > > ion_dynamics = ?none? > > Are you doing a relaxation or what? > > > / > > K_POINTS {automatic} > > 4 4 4 0 0 0 > > Do you want to study an isolated molecule? If so you > don't need a k-point sampling, simply > use > > > K_POINTS GAMMA --- On Tue, 1/27/09, Gabriele Sclauzero wrote: > From: Gabriele Sclauzero > Subject: Re: [Pw_forum] convergence issues > To: "PWSCF Forum" > Date: Tuesday, January 27, 2009, 9:02 AM > Hi Jonas > > Jonas Baltrusaitis wrote: > > > > INPUT > > > > &control > > calculation='relax', > > prefix='44tpcb_PBE_optimization', > > > pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', > > outdir='/home/jbaltrus/atom-lsda/', > > nstep=200 > > / > > &system > > ibrav=12, > > nat=38, > > ntyp=4, > > a=11.367, > > b=9.749, > > c=14.273, > > cosab=-0.36922777, > > ecutwfc=25.0 > > As a first hint I would suggest to specify ecutrho, since > you're using Ultrasoft PPs, and > test convergence of energy (or other quantities you're > interested in) w.r.t. to this > parameter (I usually choose ecutrho ~ 10*ecutwfc). > > If you want to do spin polarized calculations, you need to > specify nspin=2 and a non-zero > starting magnetization for one species at least. > > > / > > &electrons > > mixing_beta=0.5D0, > > Then maybe decrease the mixing parameter, or use another > mixing (for instance, > mixing_mode='local-TF' works well for highly > inhomogeneous systems). > > > > conv_thr = 1.D-8 > > / > > &ions > > ion_dynamics = ?none? > > Are you doing a relaxation or what? > > > / > > K_POINTS {automatic} > > 4 4 4 0 0 0 > > Do you want to study an isolated molecule? If so you > don't need a k-point sampling, simply > use > > > K_POINTS GAMMA > > Regards > > GS > > > > > > > > > > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > -- > > > o ------------------------------------------------ o > | Gabriele Sclauzero, PhD Student | > | c/o: SISSA & CNR-INFM Democritos, | > | via Beirut 2-4, 34014 Trieste (Italy) | > | email: sclauzer at sissa.it | > | phone: +39 040 3787 511 | > | skype: gurlonotturno | > o ------------------------------------------------ o > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From miguel.martinez at ehu.es Tue Jan 27 17:57:07 2009 From: miguel.martinez at ehu.es (Miguel =?UTF-8?B?TWFydMOtbmV6?=) Date: Tue, 27 Jan 2009 11:57:07 -0500 Subject: [Pw_forum] convergence issues In-Reply-To: <622399.17326.qm@web54407.mail.yahoo.com> References: <1102475624-1232321057-cardhu_decombobulator_blackberry.rim.net-536706383-@bxe249.bisx.prod.on.blackberry> <622399.17326.qm@web54407.mail.yahoo.com> Message-ID: <20090127115707.31ec2ec7@lcpybm> Hello Jonas, After a quick glance, I'd say that you aren't using a high enough cutoff, at least for the charge density. As you are using an ultrasoft pseudopotential, you probably need to set ecutrho to somewhere close to 10 times ecutwfc. The mixing parameter also seems a tad too low. Also, to do a relaxation you may be interested to set all the masses to a similar value, since the light mass of the hydrogen atoms are making them more susceptible to forces. Regards, Miguel -- ---------------------------------------- Miguel Mart?nez Canales Condensed Matter Physics Dpt. UPV/EHU Faculty of Science and Technology Apdo. 644 48080 Bilbao (Spain) Fax: +34 94 601 3500 Tlf: +34 94 601 5326 ---------------------------------------- "If you have an apple and I have an apple and we exchange these apples then you and I will still each have one apple. But if you have an idea and I have an idea and we exchange these ideas, then each of us will have two ideas." George Bernard Shaw From giannozz at democritos.it Tue Jan 27 20:28:24 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Tue, 27 Jan 2009 20:28:24 +0100 Subject: [Pw_forum] convergence issues In-Reply-To: <622399.17326.qm@web54407.mail.yahoo.com> References: <622399.17326.qm@web54407.mail.yahoo.com> Message-ID: <20090127202824.ihs921q3ok0ks844@mail.democritos.it> Quoting Jonas Baltrusaitis : > the simplest polymer system wouldn'd converge the first SCF cycle > after 100 iterations. I would appreciate any comments there is a section dedicated to this kind of problems in the user guide. Paolo ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From jasius_1 at yahoo.com Wed Jan 28 01:50:23 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Tue, 27 Jan 2009 16:50:23 -0800 (PST) Subject: [Pw_forum] convergence issues In-Reply-To: <497F3E3A.1080105@sissa.it> Message-ID: <573132.63151.qm@web54409.mail.yahoo.com> WIth all being said, how do I test pseudopotentials, grid and cutoff for my system? My understanding is this: select pseudopotential, use it on a real system, have grid fixed, calculate total energy at , say, 20 Ry and several ecutrho values, the increase cutoff value, calculate total energy at sevarl ecutrho again etc. The take the lowest cutoff and ecutrho values (when it doesn;t change anymore) and increase grid while having them constant. Am I thinking right? I am a cluster type of guy so it's somewhat out of my understanding thanks JOnas University of Iowa --- On Tue, 1/27/09, Gabriele Sclauzero wrote: > From: Gabriele Sclauzero > Subject: Re: [Pw_forum] convergence issues > To: "PWSCF Forum" > Date: Tuesday, January 27, 2009, 9:02 AM > Hi Jonas > > Jonas Baltrusaitis wrote: > > > > INPUT > > > > &control > > calculation='relax', > > prefix='44tpcb_PBE_optimization', > > > pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', > > outdir='/home/jbaltrus/atom-lsda/', > > nstep=200 > > / > > &system > > ibrav=12, > > nat=38, > > ntyp=4, > > a=11.367, > > b=9.749, > > c=14.273, > > cosab=-0.36922777, > > ecutwfc=25.0 > > As a first hint I would suggest to specify ecutrho, since > you're using Ultrasoft PPs, and > test convergence of energy (or other quantities you're > interested in) w.r.t. to this > parameter (I usually choose ecutrho ~ 10*ecutwfc). > > If you want to do spin polarized calculations, you need to > specify nspin=2 and a non-zero > starting magnetization for one species at least. > > > / > > &electrons > > mixing_beta=0.5D0, > > Then maybe decrease the mixing parameter, or use another > mixing (for instance, > mixing_mode='local-TF' works well for highly > inhomogeneous systems). > > > > conv_thr = 1.D-8 > > / > > &ions > > ion_dynamics = ?none? > > Are you doing a relaxation or what? > > > / > > K_POINTS {automatic} > > 4 4 4 0 0 0 > > Do you want to study an isolated molecule? If so you > don't need a k-point sampling, simply > use > > > K_POINTS GAMMA > > Regards > > GS > > > > > > > > > > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > -- > > > o ------------------------------------------------ o > | Gabriele Sclauzero, PhD Student | > | c/o: SISSA & CNR-INFM Democritos, | > | via Beirut 2-4, 34014 Trieste (Italy) | > | email: sclauzer at sissa.it | > | phone: +39 040 3787 511 | > | skype: gurlonotturno | > o ------------------------------------------------ o > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From akohlmey at cmm.chem.upenn.edu Wed Jan 28 02:12:21 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Tue, 27 Jan 2009 20:12:21 -0500 (EST) Subject: [Pw_forum] convergence issues In-Reply-To: <573132.63151.qm@web54409.mail.yahoo.com> References: <573132.63151.qm@web54409.mail.yahoo.com> Message-ID: On Tue, 27 Jan 2009, Jonas Baltrusaitis wrote: JB> WIth all being said, how do I test pseudopotentials, grid and cutoff JB> for my system? My understanding is this: JB> JB> select pseudopotential, use it on a real system, have grid fixed, using a real system at the beginning is serious waste of computer time. an initial test would be to converge the lattice constant of a corresponding bulk cystal system for each element using E/V curves. nicola marzari posted some examples on carbon a while ago, IIRC. not sure whether they made it to the wiki, but a search through the mailing list archives, should get you there. for ultrasoft psps, you increase the wavefunction cutoff (leaving density cutoff at the default of 4*ecutwfc) until convergence, keep that density cutoff and lower ecutwfc until it starts to diverge (usually at 25-35 ry). JB> calculate total energy at , say, 20 Ry and several ecutrho values, JB> the increase cutoff value, calculate total energy at sevarl ecutrho JB> again etc. The take the lowest cutoff and ecutrho values (when it JB> doesn;t change anymore) and increase grid while having them JB> constant. nope. grids are determined by the cutoffs. JB> Am I thinking right? I am a cluster type of guy so it's somewhat out JB> of my understanding get a good text book then! asking questions on a mailing list is no replacement for learning the basics. sorry. TANSTAFL. cheers, axel. JB> thanks JB> JB> JOnas JB> University of Iowa JB> JB> -- ======================================================================= 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 jasius_1 at yahoo.com Wed Jan 28 05:22:00 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Tue, 27 Jan 2009 20:22:00 -0800 (PST) Subject: [Pw_forum] charge is wrong and system is non-metallic In-Reply-To: Message-ID: <641484.17463.qm@web54409.mail.yahoo.com> My crystal get's recognized as metallic even though it's a pure insulator (urea) and calculations abort with charge is wrong message. I want to use this simple molecule as a compound containing O,N,H,C for threshold and pseudopotential testing &control calculation='scf', prefix='urea_test1', pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', outdir='/home/jbaltrus/urea_test1/', nstep=200 / &system ibrav=6, nat=5, ntyp=4, a=5.5760, c=4.6920, ecutwfc=30.0 / &electrons mixing_beta=0.7D0, conv_thr = 1.D-8 / ATOMIC_SPECIES O 15.9994 O.pbe-van_ak.UPF H 1.00794 H.pbe-van_ak.UPF N 14.00674 N.pbe-van_ak.UPF C 12.0107 C.pbe-van_ak.UPF ATOMIC_POSITIONS C 0.00000 0.50000 0.32900 O 0.00000 0.50000 0.59700 N 0.14550 0.64550 0.17910 H 0.25610 0.75610 0.28550 H 0.14290 0.64290 -0.03700 K_POINTS {automatic} 2 2 2 0 0 0 From akohlmey at cmm.chem.upenn.edu Wed Jan 28 05:28:17 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Tue, 27 Jan 2009 23:28:17 -0500 (EST) Subject: [Pw_forum] charge is wrong and system is non-metallic In-Reply-To: <641484.17463.qm@web54409.mail.yahoo.com> References: <641484.17463.qm@web54409.mail.yahoo.com> Message-ID: On Tue, 27 Jan 2009, Jonas Baltrusaitis wrote: JB> My crystal get's recognized as metallic even though it's a pure how do you deduce that? JB> insulator (urea) and calculations abort with charge is wrong this is the correct answer for your input. you request a close shell calculation with an odd number of electons. JB> message. I want to use this simple molecule as a compound containing JB> O,N,H,C for threshold and pseudopotential testing this is _not_ urea! you are an NH2 group short. know thy chemistry! cheers, axel. JB> JB> &control JB> calculation='scf', JB> prefix='urea_test1', JB> pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', JB> outdir='/home/jbaltrus/urea_test1/', JB> nstep=200 JB> / JB> &system JB> ibrav=6, JB> nat=5, JB> ntyp=4, JB> a=5.5760, c=4.6920, JB> ecutwfc=30.0 JB> / JB> &electrons JB> mixing_beta=0.7D0, JB> conv_thr = 1.D-8 JB> / JB> ATOMIC_SPECIES JB> O 15.9994 O.pbe-van_ak.UPF JB> H 1.00794 H.pbe-van_ak.UPF JB> N 14.00674 N.pbe-van_ak.UPF JB> C 12.0107 C.pbe-van_ak.UPF JB> ATOMIC_POSITIONS JB> C 0.00000 0.50000 0.32900 JB> O 0.00000 0.50000 0.59700 JB> N 0.14550 0.64550 0.17910 JB> H 0.25610 0.75610 0.28550 JB> H 0.14290 0.64290 -0.03700 JB> K_POINTS {automatic} JB> 2 2 2 0 0 0 JB> JB> JB> JB> JB> _______________________________________________ JB> Pw_forum mailing list JB> Pw_forum at pwscf.org JB> http://www.democritos.it/mailman/listinfo/pw_forum JB> -- ======================================================================= 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 lataprem29 at gmail.com Wed Jan 28 08:40:07 2009 From: lataprem29 at gmail.com (premlata pandit) Date: Wed, 28 Jan 2009 13:10:07 +0530 Subject: [Pw_forum] pseudo-potential problem Message-ID: <6f289440901272340t5c2f3cd0x7758b7acf0ba7672@mail.gmail.com> Dear pwscf users, I have been trying to generate the ultrasoft pseudo potential for Pm rare earth compound, but I've some problem Here is my input file &input title='Pm', zed=61.0, rel=2, rlderiv=2.50, eminld=-4.0, emaxld=4.0, deld=0.02, nld=3, config='[Xe]4f5.00 6s2.0 5d0.0 6p0.00', iswitch=3, dft='LDA' / &inputp pseudotype=3, lloc=0, nlcc=.true., rcore=1.5, rcloc=2.5, file_pseudopw='Pm.RRKJ3.UPF', / 11 4F 4 3 5.00 0.00 2.20 2.30 2.50 4F 4 3 0.00 -0.20 2.20 2.30 2.50 4F 4 3 0.00 -0.20 2.20 2.30 3.50 4F 4 3 0.00 -0.20 2.20 2.30 3.50 5D 3 2 0.00 -0.30 1.80 2.40 1.50 5D 3 2 0.00 -0.30 1.80 2.40 1.50 5D 3 2 0.00 -0.30 1.80 2.40 2.50 5D 3 2 0.00 -0.30 1.80 2.40 2.50 6P 2 1 0.00 0.00 2.30 2.30 0.50 6P 2 1 0.00 0.00 2.30 2.30 1.50 6S 1 0 2.00 0.00 2.40 2.40 0.50 At the end ld1.x calculation we found the error-- Wfc 5D rcut= 1.802 Estimated cut-off energy= 69.67 Ry This function has 0 nodes for 0 < r < 1.802 Wfc-us 5D rcutus= 2.373 Estimated cut-off energy= 19.47 Ry Message from routine compute phi: negative determinant ld= 3.166003 f2ae -0.121907 faenor 0.027736 1 1260 1259 100.899617435058 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from compute_chi : error # 1 n is too large %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... could someone suggest me what should i do to generate pseudopotential for Pm compound and what i doing wrong hare. -- 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/20090128/d2ff9736/attachment.htm From sclauzer at sissa.it Wed Jan 28 09:54:50 2009 From: sclauzer at sissa.it (Gabriele Sclauzero) Date: Wed, 28 Jan 2009 09:54:50 +0100 Subject: [Pw_forum] pseudo-potential problem In-Reply-To: <6f289440901272340t5c2f3cd0x7758b7acf0ba7672@mail.gmail.com> References: <6f289440901272340t5c2f3cd0x7758b7acf0ba7672@mail.gmail.com> Message-ID: <49801D5A.9010903@sissa.it> To generate such pseudopotential is not an easy task, especially for those who have not much experience (included me). I think it's unprobable that someone of this community will do it for you in a short time (or maybe even in a long time). If you want to generate by yourself and you want some help, you have at least to show that you understand some basics on USPPs. So, please first read some documentation/paper/tutorial about Vanderbilt USPPs (maybe you'll find some on Vanderbilt's site http://www.physics.rutgers.edu/~dhv/uspp/), then maybe try to generate and test a scalar-relativistic PP and then switch to fully-relativistic. GS premlata pandit wrote: > Dear pwscf users, > I have been trying to generate the ultrasoft pseudo potential for Pm > rare earth compound, but I've some problem > Here is my input file > > &input > title='Pm', > zed=61.0, > rel=2, > rlderiv=2.50, > eminld=-4.0, > emaxld=4.0, > deld=0.02, > nld=3, > config='[Xe]4f5.00 6s2.0 5d0.0 6p0.00', > iswitch=3, > dft='LDA' > / > &inputp > pseudotype=3, > lloc=0, > nlcc=.true., > rcore=1.5, > rcloc=2.5, > file_pseudopw='Pm.RRKJ3.UPF', > / > 11 > 4F 4 3 5.00 0.00 2.20 2.30 2.50 > 4F 4 3 0.00 -0.20 2.20 2.30 2.50 > 4F 4 3 0.00 -0.20 2.20 2.30 3.50 > 4F 4 3 0.00 -0.20 2.20 2.30 3.50 > 5D 3 2 0.00 -0.30 1.80 2.40 1.50 > 5D 3 2 0.00 -0.30 1.80 2.40 1.50 > 5D 3 2 0.00 -0.30 1.80 2.40 2.50 > 5D 3 2 0.00 -0.30 1.80 2.40 2.50 > 6P 2 1 0.00 0.00 2.30 2.30 0.50 > 6P 2 1 0.00 0.00 2.30 2.30 1.50 > 6S 1 0 2.00 0.00 2.40 2.40 0.50 > > At the end ld1.x calculation we found the error-- > > Wfc 5D rcut= 1.802 Estimated cut-off energy= 69.67 Ry > This function has 0 nodes for 0 < r < 1.802 > Wfc-us 5D rcutus= 2.373 Estimated cut-off energy= 19.47 Ry > Message from routine compute phi: > negative determinant > > ld= 3.166003 f2ae -0.121907 faenor 0.027736 > 1 1260 1259 100.899617435058 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from compute_chi : error # 1 > n is too large > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > could someone suggest me what should i do to generate pseudopotential > for Pm > compound and what i doing wrong hare. > -- > Premlata Pandit > Ph.d. student, > Barkatullah university, > Bhopal 462026, > MP, India > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- o ------------------------------------------------ o | Gabriele Sclauzero, PhD Student | | c/o: SISSA & CNR-INFM Democritos, | | via Beirut 2-4, 34014 Trieste (Italy) | | email: sclauzer at sissa.it | | phone: +39 040 3787 511 | | skype: gurlonotturno | o ------------------------------------------------ o From baris.malcioglu at gmail.com Wed Jan 28 12:04:09 2009 From: baris.malcioglu at gmail.com (O. Baris Malcioglu) Date: Wed, 28 Jan 2009 12:04:09 +0100 Subject: [Pw_forum] wavefunction coefficients In-Reply-To: References: Message-ID: Dear Karoly, > the norm of the coefficients vector is not one, but, e.g. 0.87449. I do not know the intrinsic of pw2casino, but I hope you are making sure the changed metric (i.e. S in USPPs) is taken into account in your calculation. For example, In the example I have sent you previously, you will have to calculate it separately using init_us* calbec spsi et al. Normally, pwscf considers the metric within a reduced superoperator like (( H - e S ) * evc = 0 ). Please have a look at PW/regterg.f90 for details. O. Baris Malcioglu. SISSA CM Sector From m.abbasnejad at gmail.com Wed Jan 28 13:33:06 2009 From: m.abbasnejad at gmail.com (mohaddeseh abbasnejad) Date: Wed, 28 Jan 2009 16:03:06 +0330 Subject: [Pw_forum] converting pseudopotential Message-ID: <7699a1950901280433j760f2c2am426645099b021e17@mail.gmail.com> Dear All, I have found that there is a utility in PWscf (versions >=4) called casino2upf.x that will convert the tabulated CASINO pseudopotentials from the CASINO site into upf files that PWscf can use. But I couldn't find it. Would you please tell me how I can use it? Thanks in advance. Best regards, Mohaddeseh Abbasnejad -- -------------------------------------------------------------------------------- Mohaddeseh Abbasnejad, Room No. 323, Department of Physics, University of Tehran, End of North Karegar Ave., Tehran, P.O. Box: 14395-547- IRAN Tel. No.: +98 21 6111 8634 & Fax No.: +98 21 8800 4781 E-Mail: m.abbasnejad at gmail.com Website: http://physics.ut.ac.ir -------------------------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090128/198649db/attachment.htm From paulatto at sissa.it Wed Jan 28 13:54:21 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Wed, 28 Jan 2009 13:54:21 +0100 Subject: [Pw_forum] converting pseudopotential In-Reply-To: <7699a1950901280433j760f2c2am426645099b021e17@mail.gmail.com> References: <7699a1950901280433j760f2c2am426645099b021e17@mail.gmail.com> Message-ID: On Wed, 28 Jan 2009 13:33:06 +0100, mohaddeseh abbasnejad wrote: > I have found that there is a utility in PWscf > [...] > But I couldn't find it. Well, have you found it or not? ;-) In either case it is in upftools/; just eneter the directory and type "make all". > Would you please tell me how I can use it? One you have compiled it you can run it from upftools/ like ./casino2upf.x it'll ask you to provide the name of casino file (for input) and of the UPF file (for output). cheers -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From jasius_1 at yahoo.com Wed Jan 28 15:00:18 2009 From: jasius_1 at yahoo.com (Jonas Baltrusaitis) Date: Wed, 28 Jan 2009 06:00:18 -0800 (PST) Subject: [Pw_forum] charge is wrong and system is non-metallic In-Reply-To: Message-ID: <162301.59979.qm@web54401.mail.yahoo.com> I deduce it from the error output (in fact it explicitly says that) Below is an output from the .cif file, obviously NH2 group is symmetrical to one another. this somewhat brings me to another question I asked earlier about how to use data in .cif files to setup an pwscf input. I looked at the lecture notes on QE website how to use Wyckoff coordinates to setup an input but in this case it wasn't useful _chemical_formula_sum 'C O N2 H4' _cell_length_a 5.576 _cell_length_b 5.576 _cell_length_c 4.692 _cell_angle_alpha 90 _cell_angle_beta 90 _cell_angle_gamma 90 _cell_volume 145.883 _symmetry_space_group_name_H-M 'P -4 21 m' loop_ _symmetry_equiv_pos_as_xyz 'x,y,z' '1/2-y,1/2-x,z' 'y,-x,-z' '1/2-x,1/2+y,-z' '-x,-y,z' '1/2+y,1/2+x,z' '-y,x,-z' '1/2+x,1/2-y,-z' loop_ _atom_site_label _atom_site_fract_x _atom_site_fract_y _atom_site_fract_z C 0.00000 0.50000 0.32900 O 0.00000 0.50000 0.59700 N 0.14550 0.64550 0.17910 H1 0.25610 0.75610 0.28550 H2 0.14290 0.64290 -0.03700 --- On Tue, 1/27/09, Axel Kohlmeyer wrote: > From: Axel Kohlmeyer > Subject: Re: [Pw_forum] charge is wrong and system is non-metallic > To: "Jonas Baltrusaitis" > Cc: pw_forum at pwscf.org > Date: Tuesday, January 27, 2009, 8:28 PM > On Tue, 27 Jan 2009, Jonas Baltrusaitis wrote: > > JB> My crystal get's recognized as metallic even > though it's a pure > > how do you deduce that? > > JB> insulator (urea) and calculations abort with charge > is wrong > > this is the correct answer for your input. > you request a close shell calculation with > an odd number of electons. > > JB> message. I want to use this simple molecule as a > compound containing > JB> O,N,H,C for threshold and pseudopotential testing > > this is _not_ urea! you are an NH2 group short. > > know thy chemistry! > > cheers, > axel. > > > JB> > JB> &control > JB> calculation='scf', > JB> prefix='urea_test1', > JB> > pseudo_dir='/share/apps/espresso-4.0.4/pseudo/', > JB> outdir='/home/jbaltrus/urea_test1/', > JB> nstep=200 > JB> / > JB> &system > JB> ibrav=6, > JB> nat=5, > JB> ntyp=4, > JB> a=5.5760, c=4.6920, > JB> ecutwfc=30.0 > JB> / > JB> &electrons > JB> mixing_beta=0.7D0, > JB> conv_thr = 1.D-8 > JB> / > JB> ATOMIC_SPECIES > JB> O 15.9994 O.pbe-van_ak.UPF > JB> H 1.00794 H.pbe-van_ak.UPF > JB> N 14.00674 N.pbe-van_ak.UPF > JB> C 12.0107 C.pbe-van_ak.UPF > JB> ATOMIC_POSITIONS > JB> C 0.00000 0.50000 0.32900 > JB> O 0.00000 0.50000 0.59700 > JB> N 0.14550 0.64550 0.17910 > JB> H 0.25610 0.75610 0.28550 > JB> H 0.14290 0.64290 -0.03700 > JB> K_POINTS {automatic} > JB> 2 2 2 0 0 0 > JB> > JB> > JB> > JB> > JB> _______________________________________________ > JB> Pw_forum mailing list > JB> Pw_forum at pwscf.org > JB> http://www.democritos.it/mailman/listinfo/pw_forum > JB> > > -- > ======================================================================= > 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 davidfarrell2008 at u.northwestern.edu Wed Jan 28 20:46:12 2009 From: davidfarrell2008 at u.northwestern.edu (David Farrell) Date: Wed, 28 Jan 2009 13:46:12 -0600 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P Message-ID: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> I am trying to run a 1152 atom, 2560 electron pw MD system on a BG/P, and I believe I am running up against memory issues (not a surprise...)- but I am not exactly sure how to debug & solve the issue. I am trying to run on 1024 procs (I've tried in smp, dual and vn mode), though I guess I may have to go higher - but I am not certain yet. I have kept npools =1 , nimage = 1, since the didn't seem applicable to my run. I have tried varying ntg to 1, 2 & 32 and kept ndiag to the default. While varying the number of taskgroups (ntg), I would get output like the following: Parallel version (MPI) Number of processors in use: 1024 R & G space division: proc/pool = 1024 wavefunctions fft division: fft/group = 2 For Norm-Conserving or Ultrasoft (Vanderbilt) Pseudopotentials or PAW Current dimensions of program pwscf are: Max number of different atomic species (ntypx) = 10 Max number of k-points (npk) = 40000 Max angular momentum in pseudopotentials (lmaxx) = 3 gamma-point specific algorithms are used Iterative solution of the eigenvalue problem a parallel distributed memory algorithm will be used, eigenstates matrixes will be distributed block like on ortho sub-group = 32* 32 procs Message from routine data_structure: some processors have no planes Message from routine data_structure: some processors have no smooth planes Planes per process (thick) : nr3 =480 npp = 1 ncplane =***** Proc/ planes cols G planes cols G columns G Pool (dense grid) (smooth grid) (wavefct grid) 1 1 162 50122 1 162 50122 42 6294 2 0 162 50122 0 162 50122 42 6294 3 1 162 50122 1 162 50122 42 6294 4 0 162 50122 0 162 50122 42 6294 5 1 162 50122 1 162 50122 42 6294 (continues similarly for each of the 1024 procs) So the number of FFT planes that need to be distributed is 480. The output below that made it seem like there were processors that which still weren't taking part in the calculation, and presumably weren't helping out in the distribution of the data. My understanding is that the processors of each taskgroup would take part in the FFT calculation for the plane associated with the task group. So my first question is - is the fact that some procs don't have planes in my output actually an issue? the output continues and the run finally dies here: Largest allocated arrays est. size (Mb) dimensions Kohn-Sham Wavefunctions 73.76 Mb ( 3147,1536) NL pseudopotentials 227.42 Mb ( 3147,4736) Each V/rho on FFT grid 3.52 Mb ( 230400) Each G-vector array 0.19 Mb ( 25061) G-vector shells 0.08 Mb ( 10422) Largest temporary arrays est. size (Mb) dimensions Auxiliary wavefunctions 73.76 Mb ( 3147,3072) Each subspace H/S matrix 72.00 Mb ( 3072,3072) Each matrix 55.50 Mb ( 4736,1536) Arrays for rho mixing 28.12 Mb ( 230400, 8) Initial potential from superposition of free atoms Check: negative starting charge= -7.401460 starting charge 2556.45492, renormalised to 2560.00000 negative rho (up, down): 0.741E+01 0.000E+00 Starting wfc are 2944 atomic wfcs total cpu time spent up to now is 704.01 secs per-process dynamical memory: 13.6 Mb Self-consistent Calculation iteration # 1 ecut= 38.22 Ry beta=0.70 Davidson diagonalization with overlap ethr = 1.00E-02, avg # of iterations = 2.0 process group 2362 has completed with an like this in the stderr file: Abort(1) on node 210 (rank 210 in comm 1140850688): Fatal error in MPI_Scatterv: Other MPI error, error sta ck: MPI_Scatterv(360): MPI_Scatterv(sbuf=0x36c02010, scnts=0x7fffa940, displs=0x7fffb940, MPI_DOUBLE_PRECISION, rbuf=0x4b83010, rcount=230400, MPI_DOUBLE_PRECISION, root=0, comm=0x84000002) failed MPI_Scatterv(100): Out of memory So I figure I am running out of memory on a node at some point... but not entirely sure where (seems to be in the first electronic step) or how to get around it. Any help would be appreciated. Dave David E. Farrell Post-Doctoral Fellow Department of Materials Science and Engineering Northwestern University email: d-farrell2 at northwestern.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090128/11246cce/attachment-0001.htm From akohlmey at cmm.chem.upenn.edu Wed Jan 28 21:04:53 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 28 Jan 2009 15:04:53 -0500 (EST) Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> Message-ID: On Wed, 28 Jan 2009, David Farrell wrote: [...] DF> Largest allocated arrays est. size (Mb) dimensions DF> Kohn-Sham Wavefunctions 73.76 Mb ( 3147,1536) DF> NL pseudopotentials 227.42 Mb ( 3147,4736) DF> Each V/rho on FFT grid 3.52 Mb ( 230400) DF> Each G-vector array 0.19 Mb ( 25061) DF> G-vector shells 0.08 Mb ( 10422) DF> Largest temporary arrays est. size (Mb) dimensions DF> Auxiliary wavefunctions 73.76 Mb ( 3147,3072) DF> Each subspace H/S matrix 72.00 Mb ( 3072,3072) DF> Each matrix 55.50 Mb ( 4736,1536) DF> Arrays for rho mixing 28.12 Mb ( 230400, 8) DF> [...] DF> with an like this in the stderr file: DF> DF> Abort(1) on node 210 (rank 210 in comm 1140850688): Fatal error in DF> MPI_Scatterv: Other MPI error, error sta DF> ck: DF> MPI_Scatterv(360): MPI_Scatterv(sbuf=0x36c02010, scnts=0x7fffa940, DF> displs=0x7fffb940, MPI_DOUBLE_PRECISION, DF> rbuf=0x4b83010, rcount=230400, MPI_DOUBLE_PRECISION, root=0, DF> comm=0x84000002) failed DF> MPI_Scatterv(100): Out of memory DF> DF> So I figure I am running out of memory on a node at some point... but not DF> entirely sure where (seems to be in the first electronic step) or how to get DF> around it. it dies on the processor calling MPI_Scatterv, probably the (group)master(s). what is interesting is that the rcount size matches the "arrays for rho mixing", so i would suggest to first have a look there and try to determine how large the combined send buffers are. cheers, axel. DF> DF> Any help would be appreciated. DF> DF> Dave DF> DF> DF> DF> DF> David E. Farrell DF> Post-Doctoral Fellow DF> Department of Materials Science and Engineering DF> Northwestern University DF> email: d-farrell2 at northwestern.edu DF> DF> -- ======================================================================= 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 nemeth at anl.gov Wed Jan 28 21:40:04 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Wed, 28 Jan 2009 14:40:04 -0600 (CST) Subject: [Pw_forum] Pw_forum Digest, Vol 19, Issue 51 In-Reply-To: References: Message-ID: Hi Baris! thank you very much for your help! I have forgotten about the metric, you are right. Somehow I have missed your previous email, could you please resend it to me? Thanks: Karoly > Date: Wed, 28 Jan 2009 12:04:09 +0100 > From: "O. Baris Malcioglu" > Subject: Re: [Pw_forum] wavefunction coefficients > To: PWSCF Forum > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > Dear Karoly, > >> the norm of the coefficients vector is not one, but, e.g. 0.87449. > > I do not know the intrinsic of pw2casino, but I hope you are making > sure the changed metric (i.e. S in USPPs) is taken into account in > your calculation. > > For example, In the example I have sent you previously, you will have > to calculate it separately using init_us* calbec spsi et al. > Normally, pwscf considers the metric within a reduced superoperator > like (( H - e S ) * evc = 0 ). Please have a look at PW/regterg.f90 > for details. > > O. Baris Malcioglu. > SISSA CM Sector > From davidfarrell2008 at u.northwestern.edu Wed Jan 28 22:18:17 2009 From: davidfarrell2008 at u.northwestern.edu (David Farrell) Date: Wed, 28 Jan 2009 15:18:17 -0600 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> Message-ID: Oddly enough, the same input file, run in dual mode with 1 taskgroup (so each process should have access to 1 GB of RAM), doesn't spit out the previous error, but rather this one: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%% from pdpotf : error # 1 problems computing cholesky decomposition %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%% I would have expected this one to fail the same way. The solution to this from the mailing list seems to be to disable the parallel cholesky decomposition, but that doesn't seem a very good option in my case. I am trying to re-run this case to see if this error is reproducible, and trying the smp-mode version with 1 taskgroup to see if I can get a better read on where the MPI_Scatterv is being called from (there was no core file for the master process for some reason.)... and I am not really sure how to go about finding out the send buffer size (I guess a debugger may be the only option?) Dave On Jan 28, 2009, at 2:04 PM, Axel Kohlmeyer wrote: > On Wed, 28 Jan 2009, David Farrell wrote: > > > [...] > > DF> Largest allocated arrays est. size (Mb) dimensions > DF> Kohn-Sham Wavefunctions 73.76 Mb ( 3147,1536) > DF> NL pseudopotentials 227.42 Mb ( 3147,4736) > DF> Each V/rho on FFT grid 3.52 Mb ( 230400) > DF> Each G-vector array 0.19 Mb ( 25061) > DF> G-vector shells 0.08 Mb ( 10422) > DF> Largest temporary arrays est. size (Mb) dimensions > DF> Auxiliary wavefunctions 73.76 Mb ( 3147,3072) > DF> Each subspace H/S matrix 72.00 Mb ( 3072,3072) > DF> Each matrix 55.50 Mb ( 4736,1536) > DF> Arrays for rho mixing 28.12 Mb ( 230400, 8) > DF> > [...] > DF> with an like this in the stderr file: > DF> > DF> Abort(1) on node 210 (rank 210 in comm 1140850688): Fatal error in > DF> MPI_Scatterv: Other MPI error, error sta > DF> ck: > DF> MPI_Scatterv(360): MPI_Scatterv(sbuf=0x36c02010, scnts=0x7fffa940, > DF> displs=0x7fffb940, MPI_DOUBLE_PRECISION, > DF> rbuf=0x4b83010, rcount=230400, MPI_DOUBLE_PRECISION, root=0, > DF> comm=0x84000002) failed > DF> MPI_Scatterv(100): Out of memory > DF> > DF> So I figure I am running out of memory on a node at some > point... but not > DF> entirely sure where (seems to be in the first electronic step) > or how to get > DF> around it. > > it dies on the processor calling MPI_Scatterv, probably the > (group)master(s). > what is interesting is that the rcount size matches the "arrays for > rho > mixing", so i would suggest to first have a look there and try to > determine how large the combined send buffers are. > > cheers, > axel. > > > DF> > DF> Any help would be appreciated. > DF> > DF> Dave > DF> > DF> > DF> > DF> > DF> David E. Farrell > DF> Post-Doctoral Fellow > DF> Department of Materials Science and Engineering > DF> Northwestern University > DF> email: d-farrell2 at northwestern.edu > DF> > DF> > > -- > = > ====================================================================== > 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. David E. Farrell Post-Doctoral Fellow Department of Materials Science and Engineering Northwestern University email: d-farrell2 at northwestern.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090128/87c97029/attachment.htm From marzari at MIT.EDU Wed Jan 28 22:41:39 2009 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 28 Jan 2009 16:41:39 -0500 Subject: [Pw_forum] convergence issues In-Reply-To: References: <573132.63151.qm@web54409.mail.yahoo.com> Message-ID: <4980D113.1050300@mit.edu> Axel Kohlmeyer wrote: > nicola marzari posted some examples on carbon a while ago, IIRC. > not sure whether they made it to the wiki, but a search through > the mailing list archives, should get you there. Hi Axel, they are in the pseudopotential table, linked to the carbon psp we use for PBE (C.pbe-rrkjus): http://www.quantum-espresso.org/pseudo/upfdetails.php?upf=C.pbe-rrkjus.UPF nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From naromero at gmail.com Wed Jan 28 22:37:06 2009 From: naromero at gmail.com (Nichols A. Romero) Date: Wed, 28 Jan 2009 15:37:06 -0600 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> Message-ID: <6ac064b60901281337k755debfdtb177e93393ae75ff@mail.gmail.com> David, Seems like PWSCF calls Scalapack for its Cholesky decomposition. If you don't care too much about performance, doing the Cholesky decomposition on your system size would not slow you down terribly. However, you might pay a memory penalty. Depends on how the Cholesky decomposition is done. I imagine they probably collect the entire overlap matrix in one place before sending it to Scalapack. On Wed, Jan 28, 2009 at 3:18 PM, David Farrell < davidfarrell2008 at u.northwestern.edu> wrote: > Oddly enough, the same input file, run in dual mode with 1 taskgroup (so > each process should have access to 1 GB of RAM), doesn't spit out the > previous error, but rather this one: > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from pdpotf : error # 1 > problems computing cholesky decomposition > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > I would have expected this one to fail the same way. The solution to this > from the mailing list seems to be to disable the parallel cholesky > decomposition, but that doesn't seem a very good option in my case. > > I am trying to re-run this case to see if this error is reproducible, and > trying the smp-mode version with 1 taskgroup to see if I can get a better > read on where the MPI_Scatterv is being called from (there was no core file > for the master process for some reason.)... and I am not really sure how to > go about finding out the send buffer size (I guess a debugger may be the > only option?) > > Dave > > > > > On Jan 28, 2009, at 2:04 PM, Axel Kohlmeyer wrote: > > On Wed, 28 Jan 2009, David Farrell wrote: > > > [...] > > DF> Largest allocated arrays est. size (Mb) dimensions > DF> Kohn-Sham Wavefunctions 73.76 Mb ( 3147,1536) > DF> NL pseudopotentials 227.42 Mb ( 3147,4736) > DF> Each V/rho on FFT grid 3.52 Mb ( 230400) > DF> Each G-vector array 0.19 Mb ( 25061) > DF> G-vector shells 0.08 Mb ( 10422) > DF> Largest temporary arrays est. size (Mb) dimensions > DF> Auxiliary wavefunctions 73.76 Mb ( 3147,3072) > DF> Each subspace H/S matrix 72.00 Mb ( 3072,3072) > DF> Each matrix 55.50 Mb ( 4736,1536) > DF> Arrays for rho mixing 28.12 Mb ( 230400, 8) > DF> > [...] > DF> with an like this in the stderr file: > DF> > DF> Abort(1) on node 210 (rank 210 in comm 1140850688): Fatal error in > DF> MPI_Scatterv: Other MPI error, error sta > DF> ck: > DF> MPI_Scatterv(360): MPI_Scatterv(sbuf=0x36c02010, scnts=0x7fffa940, > DF> displs=0x7fffb940, MPI_DOUBLE_PRECISION, > DF> rbuf=0x4b83010, rcount=230400, MPI_DOUBLE_PRECISION, root=0, > DF> comm=0x84000002) failed > DF> MPI_Scatterv(100): Out of memory > DF> > DF> So I figure I am running out of memory on a node at some point... but > not > DF> entirely sure where (seems to be in the first electronic step) or how > to get > DF> around it. > > it dies on the processor calling MPI_Scatterv, probably the > (group)master(s). > what is interesting is that the rcount size matches the "arrays for rho > mixing", so i would suggest to first have a look there and try to > determine how large the combined send buffers are. > > cheers, > axel. > > > DF> > DF> Any help would be appreciated. > DF> > DF> Dave > DF> > DF> > DF> > DF> > DF> David E. Farrell > DF> Post-Doctoral Fellow > DF> Department of Materials Science and Engineering > DF> Northwestern University > DF> email: d-farrell2 at northwestern.edu > DF> > DF> > > -- > ======================================================================= > 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. > > > David E. Farrell > Post-Doctoral Fellow > Department of Materials Science and Engineering > Northwestern University > email: d-farrell2 at northwestern.edu > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- Nichols A. Romero, Ph.D. Argonne Leadership Computing Facility Argonne, IL 60490 (630) 252-3441 (O) (630) 470-0462 (C) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090128/f81f81db/attachment-0001.htm From akohlmey at cmm.chem.upenn.edu Wed Jan 28 22:58:54 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 28 Jan 2009 16:58:54 -0500 (EST) Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> Message-ID: On Wed, 28 Jan 2009, David Farrell wrote: DF> I am trying to re-run this case to see if this error is reproducible, and DF> trying the smp-mode version with 1 taskgroup to see if I can get a better DF> read on where the MPI_Scatterv is being called from (there was no core file DF> for the master process for some reason.)... and I am not really sure how to DF> go about finding out the send buffer size (I guess a debugger may be the DF> only option?) grep does the job. MPI_Scatterv is used only in fft_base.f90 and only in three subroutines. pw.x uses only two of them and one is used in exx.f90, which you most likely don't use, however the other use in psymrho.f90/psymrho_mag.f90 is consistent with the problem happening at the rho-mixing stage. you have a "grid size" allocate/deallocate around the call to grid_scatter in psymrho.f90, so that may lead to exhausting the memory when MPI_SCATTERV needs it internally... HTH, axel. DF> DF> Dave DF> DF> DF> DF> DF> On Jan 28, 2009, at 2:04 PM, Axel Kohlmeyer wrote: DF> DF> >On Wed, 28 Jan 2009, David Farrell wrote: DF> > DF> > DF> >[...] DF> > DF> > DF> Largest allocated arrays est. size (Mb) dimensions DF> > DF> Kohn-Sham Wavefunctions 73.76 Mb ( 3147,1536) DF> > DF> NL pseudopotentials 227.42 Mb ( 3147,4736) DF> > DF> Each V/rho on FFT grid 3.52 Mb ( 230400) DF> > DF> Each G-vector array 0.19 Mb ( 25061) DF> > DF> G-vector shells 0.08 Mb ( 10422) DF> > DF> Largest temporary arrays est. size (Mb) dimensions DF> > DF> Auxiliary wavefunctions 73.76 Mb ( 3147,3072) DF> > DF> Each subspace H/S matrix 72.00 Mb ( 3072,3072) DF> > DF> Each matrix 55.50 Mb ( 4736,1536) DF> > DF> Arrays for rho mixing 28.12 Mb ( 230400, 8) DF> > DF> DF> >[...] DF> > DF> with an like this in the stderr file: DF> > DF> DF> > DF> Abort(1) on node 210 (rank 210 in comm 1140850688): Fatal error in DF> > DF> MPI_Scatterv: Other MPI error, error sta DF> > DF> ck: DF> > DF> MPI_Scatterv(360): MPI_Scatterv(sbuf=0x36c02010, scnts=0x7fffa940, DF> > DF> displs=0x7fffb940, MPI_DOUBLE_PRECISION, DF> > DF> rbuf=0x4b83010, rcount=230400, MPI_DOUBLE_PRECISION, root=0, DF> > DF> comm=0x84000002) failed DF> > DF> MPI_Scatterv(100): Out of memory DF> > DF> DF> > DF> So I figure I am running out of memory on a node at some point... but DF> > DF> not DF> > DF> entirely sure where (seems to be in the first electronic step) or how DF> > DF> to get DF> > DF> around it. DF> > DF> >it dies on the processor calling MPI_Scatterv, probably the DF> >(group)master(s). DF> >what is interesting is that the rcount size matches the "arrays for rho DF> >mixing", so i would suggest to first have a look there and try to DF> >determine how large the combined send buffers are. DF> > DF> >cheers, DF> > axel. DF> > DF> > DF> > DF> DF> > DF> Any help would be appreciated. DF> > DF> DF> > DF> Dave DF> > DF> DF> > DF> DF> > DF> DF> > DF> DF> > DF> David E. Farrell DF> > DF> Post-Doctoral Fellow DF> > DF> Department of Materials Science and Engineering DF> > DF> Northwestern University DF> > DF> email: d-farrell2 at northwestern.edu DF> > DF> DF> > DF> DF> > DF> >-- DF> >======================================================================= DF> >Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu DF> > Center for Molecular Modeling -- University of Pennsylvania DF> >Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 DF> >tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 DF> >======================================================================= DF> >If you make something idiot-proof, the universe creates a better idiot. DF> DF> David E. Farrell DF> Post-Doctoral Fellow DF> Department of Materials Science and Engineering DF> Northwestern University DF> email: d-farrell2 at northwestern.edu DF> DF> -- ======================================================================= 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 Wed Jan 28 23:08:41 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Wed, 28 Jan 2009 23:08:41 +0100 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> Message-ID: <20090128230841.x0rzedjqgo4gowo0@mail.democritos.it> Quoting David Farrell : > I am trying to run a 1152 atom, 2560 electron pw MD system on a BG/P, > and I believe I am running up against memory issues set nbnd, diago_david_ndim, mixing_ndim to the smallest possible values to save memory. Use the CVS version and try to compile scalapack (instructions in the wiki) if you have trouble with subspace diagonalization, or else use a smaller set of processors in the "ortho group": 1024 seems to me a lot for a system with O(1000) states. Paolo ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From giannozz at democritos.it Wed Jan 28 23:15:27 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Wed, 28 Jan 2009 23:15:27 +0100 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> Message-ID: <20090128231527.uq7wa69a04okwssc@mail.democritos.it> Quoting Axel Kohlmeyer : > however the other use in psymrho.f90/psymrho_mag.f90 is consistent > with the problem happening at the rho-mixing stage. the way parallel symmetrization is implemented now guarantees that a large system with symmmetry (a strange object, though) will go out of memory. A workaround should be simple, though Paolo ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From naromero at gmail.com Wed Jan 28 23:20:15 2009 From: naromero at gmail.com (Nichols A. Romero) Date: Wed, 28 Jan 2009 16:20:15 -0600 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: <20090128230841.x0rzedjqgo4gowo0@mail.democritos.it> References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> <20090128230841.x0rzedjqgo4gowo0@mail.democritos.it> Message-ID: <6ac064b60901281420t2b32c8e1mc48c99698d0034bf@mail.gmail.com> David, You have the: ortho sub group set to 32*32 Paolo can correct me if I am wrong. This is the Scalapack blacs grid for the cholesky decomposition. It basically takes the overlap matrix whose dimensions are (number of states) by (number of states) and divides into 32-by-32 pieces according to a 2D block cyclic algorithm. You are using 32*32=1024 processors to do the cholesky decomposition of a 2560-by-2560. I would recommend using something like 8*8. On Wed, Jan 28, 2009 at 4:08 PM, wrote: > Quoting David Farrell : > > > I am trying to run a 1152 atom, 2560 electron pw MD system on a BG/P, > > and I believe I am running up against memory issues > > set nbnd, diago_david_ndim, mixing_ndim to the smallest possible > values to save memory. Use the CVS version and try to compile scalapack > (instructions in the wiki) if you have trouble with subspace > diagonalization, or else use a smaller set of processors in the "ortho > group": 1024 seems to me a lot for a system with O(1000) states. > > Paolo > > ---------------------------------------------------------------- > 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 > -- Nichols A. Romero, Ph.D. Argonne Leadership Computing Facility Argonne, IL 60490 (630) 252-3441 (O) (630) 470-0462 (C) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090128/35905611/attachment.htm From naromero at gmail.com Thu Jan 29 01:07:25 2009 From: naromero at gmail.com (Nichols A. Romero) Date: Wed, 28 Jan 2009 18:07:25 -0600 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: <6ac064b60901281420t2b32c8e1mc48c99698d0034bf@mail.gmail.com> References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> <20090128230841.x0rzedjqgo4gowo0@mail.democritos.it> <6ac064b60901281420t2b32c8e1mc48c99698d0034bf@mail.gmail.com> Message-ID: <6ac064b60901281607w5f17f081jc82ec82da415f475@mail.gmail.com> David, You should really start by making estimates of how much memory your calculation needs. To do that you will really need to understand the algorithm otherwise you will just end up playing with parameters forever. ntg is for band parallelization. You have 2560 electrons. ntg = 32 is probably too large. Maybe somebody on this list can make a suggestion? My experience with a real-space code is the more bands per processor the better. You will probably want at least 250-500 bands per processor. On Wed, Jan 28, 2009 at 4:20 PM, Nichols A. Romero wrote: > David, > > You have the: > ortho sub group set to 32*32 > > Paolo can correct me if I am wrong. This is the Scalapack blacs grid for > the > cholesky decomposition. It basically takes the overlap matrix whose > dimensions > are (number of states) by (number of states) and divides into 32-by-32 > pieces > according to a 2D block cyclic algorithm. You are using 32*32=1024 > processors > to do the cholesky decomposition of a 2560-by-2560. > > I would recommend using something like 8*8. > > On Wed, Jan 28, 2009 at 4:08 PM, wrote: > >> Quoting David Farrell : >> >> > I am trying to run a 1152 atom, 2560 electron pw MD system on a BG/P, >> > and I believe I am running up against memory issues >> >> set nbnd, diago_david_ndim, mixing_ndim to the smallest possible >> values to save memory. Use the CVS version and try to compile scalapack >> (instructions in the wiki) if you have trouble with subspace >> diagonalization, or else use a smaller set of processors in the "ortho >> group": 1024 seems to me a lot for a system with O(1000) states. >> >> Paolo >> >> ---------------------------------------------------------------- >> 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 >> > > > > -- > Nichols A. Romero, Ph.D. > Argonne Leadership Computing Facility > Argonne, IL 60490 > (630) 252-3441 (O) > (630) 470-0462 (C) > > -- Nichols A. Romero, Ph.D. Argonne Leadership Computing Facility Argonne, IL 60490 (630) 252-3441 (O) (630) 470-0462 (C) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090128/0c471c7c/attachment.htm From akohlmey at cmm.chem.upenn.edu Thu Jan 29 07:22:51 2009 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 29 Jan 2009 01:22:51 -0500 (EST) Subject: [Pw_forum] charge is wrong and system is non-metallic In-Reply-To: <162301.59979.qm@web54401.mail.yahoo.com> References: <162301.59979.qm@web54401.mail.yahoo.com> Message-ID: On Wed, 28 Jan 2009, Jonas Baltrusaitis wrote: JB> I deduce it from the error output (in fact it explicitly says that) JB> it says "charge is wrong", but nothing about the system being metallic. ...and of course the GIGO principle applies to Q-E. JB> Below is an output from the .cif file, obviously NH2 group is JB> symmetrical to one another. this somewhat brings me to another JB> question I asked earlier about how to use data in .cif files to JB> setup an pwscf input. I looked at the lecture notes on QE website there are cif parsers/readers/converters around, you just have to compile/install one of them and export to cartesian coordinates. JB> how to use Wyckoff coordinates to setup an input but in this case it JB> wasn't useful perhaps you didn't read it properly? [...] JB> loop_ JB> _symmetry_equiv_pos_as_xyz JB> 'x,y,z' JB> '1/2-y,1/2-x,z' JB> 'y,-x,-z' JB> '1/2-x,1/2+y,-z' JB> '-x,-y,z' JB> '1/2+y,1/2+x,z' JB> '-y,x,-z' JB> '1/2+x,1/2-y,-z' JB> loop_ well, doesn't this block provide all you need? i don't know CIF files very well, but with a little bit of common sense once can deduce that you have to apply the listed transformations to the coordinates below. then just remove the coordinates that map onto the same position (remember to consider PBC) and you should be done. can be done in a small script or with a piece of paper and a pencil in not too much time. [...] JB> _atom_site_label JB> _atom_site_fract_x JB> _atom_site_fract_y JB> _atom_site_fract_z JB> C 0.00000 0.50000 0.32900 JB> O 0.00000 0.50000 0.59700 JB> N 0.14550 0.64550 0.17910 JB> H1 0.25610 0.75610 0.28550 JB> H2 0.14290 0.64290 -0.03700 JB> 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 paulatto at sissa.it Thu Jan 29 10:42:07 2009 From: paulatto at sissa.it (Lorenzo Paulatto) Date: Thu, 29 Jan 2009 10:42:07 +0100 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: <6ac064b60901281337k755debfdtb177e93393ae75ff@mail.gmail.com> References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> <6ac064b60901281337k755debfdtb177e93393ae75ff@mail.gmail.com> Message-ID: On Wed, 28 Jan 2009 22:37:06 +0100, Nichols A. Romero wrote: > David, > > Seems like PWSCF calls Scalapack for its Cholesky decomposition. pwscf does not use scalapack, unless specifically compiled for it, using the -D__SCALAPACK flag. Per default the code uses a home-brew version of parallel davidson algorithm implemented by C. Cavazzoni. -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/ From giannozz at democritos.it Thu Jan 29 12:55:03 2009 From: giannozz at democritos.it (giannozz at democritos.it) Date: Thu, 29 Jan 2009 12:55:03 +0100 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> <6ac064b60901281337k755debfdtb177e93393ae75ff@mail.gmail.com> Message-ID: <20090129125503.oeh95kdegc4g40kg@mail.democritos.it> Quoting Lorenzo Paulatto : > pwscf does not use scalapack, unless specifically compiled for it, > using the -D__SCALAPACK flag. > > Per default the code uses a home-brew version right > of parallel davidson algorithm not right: of a parallel conventional diagonalization, used my Davidson and in iterative orthonormalization (cp) > implemented by C. Cavazzoni. right again Paolo ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From davidfarrell2008 at u.northwestern.edu Thu Jan 29 17:21:47 2009 From: davidfarrell2008 at u.northwestern.edu (David Farrell) Date: Thu, 29 Jan 2009 10:21:47 -0600 Subject: [Pw_forum] PW taskgroups and a large run on a BG/P In-Reply-To: <6ac064b60901281607w5f17f081jc82ec82da415f475@mail.gmail.com> References: <2507FF00-6FE9-481B-A3FD-81C4E26131C8@u.northwestern.edu> <20090128230841.x0rzedjqgo4gowo0@mail.democritos.it> <6ac064b60901281420t2b32c8e1mc48c99698d0034bf@mail.gmail.com> <6ac064b60901281607w5f17f081jc82ec82da415f475@mail.gmail.com> Message-ID: <1E4FEB77-157B-4F80-B195-22E8429CCE74@u.northwestern.edu> Yes, I had been trying to take the easy way out. Now is time to do it 'right'. I had been looking through the 'understanding parallelism' section of the user manual, and was a bit confused about what is meant by some of the parameters. Below I will give my impression based on reading the manual - please correct me if I am wrong. World: This is just the MPI_COMM_WORLD ... not really much to it. Images (-nimage n) : for a given run, the allocated processors are divided into n loosely coupled groups of processors, each operating on a different set of data. For relaxations and MD runs, not really important or useful. Pools (-npool n): the procs dedicated to each image are further subdivided into n loosely coupled groups of processors. When k point sampling is used, the kpoints are divided amongst these pools. Within each pool, the planewaves and real space (i.e. 3D FFT) grid points are distributed amongst the processors. So for my gamma point sampling with 1 pool, the planewaves and real space grids *should* be distributed amongst all the procs (from my output this appears to be the case). But since the division of the FFT grid appears to happen plane-wise (not sure which direction though... this part doesn't seem to be mentioned), you run into trouble if the number of procs in a pool is greater than the number of planes in the FFT grid. Task groups (-ntg n): splits the procs in a given pool into n groups, each of which work independently on the 3D FFT. Orthogonalization groups (-ndiag n): a subgroup of n procs from the pool are used in the orthogonormalization or iterative subspace diagonalization of the Hamiltonian (I presume), a matrix that has dimensions of #states x #states. The taskgroup bit is where I run into a good deal of confusion, because even if I set the number of taskgroups such that all processors should have planes (since I assume that the FFT grid would be distributed to each task group such that each task group was doing the same thing). My output posted at the beginning of the thread has a section header like this: Proc/ planes cols G planes cols G columns G Pool (dense grid) (smooth grid) (wavefct grid) it is the first column and planes columns that now confuse me. In my test of 1 pool of 1024 procs with 32 taskgroups, the output looked like this: Proc/ planes cols G planes cols G columns G Pool (dense grid) (smooth grid) (wavefct grid) 1 15 162 50122 15 162 50122 42 6294 2 0 162 50122 0 162 50122 42 6294 3 0 162 50122 0 162 50122 42 6294 4 0 162 50122 0 162 50122 42 6294 5 0 162 50122 0 162 50122 42 6294 ... 32 0 164 50136 0 164 50136 42 6290 33 15 164 50136 15 164 50136 42 6290 34 0 164 50136 0 164 50136 42 6290 Which seems to indicate that there are processors that aren't getting any planes and that the FFT grid is *not* being reproduced within each taskgroup. If that was the case, I would expect the 'planes' columns to be either 1 or 0 for procs #1-32, 33-64, etc. Looking at the code, this output implies that the number of planes per proc (npp) is only nonzero on a small number of processors, actually making the balancing situation worse than when ntg = 1. Dave On Jan 28, 2009, at 6:07 PM, Nichols A. Romero wrote: > David, > > You should really start by making estimates of how much memory your > calculation needs. To > do that you will really need to understand the algorithm otherwise > you will just end up playing with > parameters forever. > > ntg is for band parallelization. You have 2560 electrons. > > ntg = 32 is probably too large. > > Maybe somebody on this list can make a suggestion? My experience > with a real-space code is > the more bands per processor the better. You will probably want at > least 250-500 bands per processor. > > On Wed, Jan 28, 2009 at 4:20 PM, Nichols A. Romero > wrote: > David, > > You have the: > ortho sub group set to 32*32 > > Paolo can correct me if I am wrong. This is the Scalapack blacs grid > for the > cholesky decomposition. It basically takes the overlap matrix whose > dimensions > are (number of states) by (number of states) and divides into 32- > by-32 pieces > according to a 2D block cyclic algorithm. You are using 32*32=1024 > processors > to do the cholesky decomposition of a 2560-by-2560. > > I would recommend using something like 8*8. > > On Wed, Jan 28, 2009 at 4:08 PM, wrote: > Quoting David Farrell : > > > I am trying to run a 1152 atom, 2560 electron pw MD system on a BG/ > P, > > and I believe I am running up against memory issues > > set nbnd, diago_david_ndim, mixing_ndim to the smallest possible > values to save memory. Use the CVS version and try to compile > scalapack > (instructions in the wiki) if you have trouble with subspace > diagonalization, or else use a smaller set of processors in the "ortho > group": 1024 seems to me a lot for a system with O(1000) states. > > Paolo > > ---------------------------------------------------------------- > 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 > > > > -- > Nichols A. Romero, Ph.D. > Argonne Leadership Computing Facility > Argonne, IL 60490 > (630) 252-3441 (O) > (630) 470-0462 (C) > > > > > -- > Nichols A. Romero, Ph.D. > Argonne Leadership Computing Facility > Argonne, IL 60490 > (630) 252-3441 (O) > (630) 470-0462 (C) > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum David E. Farrell Post-Doctoral Fellow Department of Materials Science and Engineering Northwestern University email: d-farrell2 at northwestern.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090129/416eb694/attachment.htm From meheut at impmc.jussieu.fr Fri Jan 30 04:19:15 2009 From: meheut at impmc.jussieu.fr (merlin meheut) Date: Fri, 30 Jan 2009 04:19:15 +0100 Subject: [Pw_forum] phonon calculation crashes Message-ID: <20090130041915.20625vqish5eboab@www-ext.impmc.jussieu.fr> Dear all, I have had a strange crash on a phonon calculation, with version 3.2.3: I have calculated the dynamical matrix of the structure at gamma, at q-point (0;0.5;-0.5), both without problems, but I cannot calculate the (0;0;-1) q-point without having this error: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from set_irr : error # 2 npert > max_irr_dim %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% I don't know what this means. The structure I am calculating is very big (80 atoms), so maybe it is related, I don't know. Follows the input files: --------------------------- scf input file: pyrope Pavese 1995 (30K) cubique centre &control calculation = 'scf', restart_mode = 'from_scratch' , prefix = 'PYROPE', disk_io = 'default' , pseudo_dir = './', outdir = '/tmp/', tprnfor = .true., tstress = .true., /&end &system ibrav =3, celldm(1)=21.9096, nat =80, ntyp = 4, ecutwfc = 80.0 /&end &electrons electron_maxstep = 60, conv_thr = 1.d-7 , mixing_mode = 'plain', startingwfc = 'atomic', mixing_beta = 0.4, diagonalization = 'david_overlap', /&end ATOMIC_SPECIES Al 26.9815 Al.pbe2 Mg 23.9850 Mgpbe1_mt.ncpp Si 27.9769 Si.pbe2 O 15.9949 O.pbe ATOMIC_POSITIONS crystal Si 0.62500000 0.62500000 0.25000000 (...) K_POINTS {crystal} 1 0.25 0.25 0.25 1 ------------------------------------------------------------------- ------------------------------------------nscf input file: pyrope Pavese 1995 (30K) cubique centre &control calculation = 'phonon', restart_mode = 'from_scratch' , prefix = 'PYROPE', disk_io = 'default' , pseudo_dir = './', outdir = '/tmp/', tprnfor = .true., tstress = .true., /&end &system ibrav =3, celldm(1)=21.9096, nat =80, ntyp = 4, ecutwfc = 80.0 /&end &electrons electron_maxstep = 60, conv_thr = 1.d-7 , mixing_mode = 'plain', startingwfc = 'atomic', mixing_beta = 0.4, diagonalization = 'david_overlap', /&end &phonon xqq(1)= 0.0000000, xqq(2)= 0.0000000, xqq(3)= -1.0000000 /&end ATOMIC_SPECIES Al 26.9815 Al.pbe2 Mg 23.9850 Mgpbe1_mt.ncpp Si 27.9769 Si.pbe2 O 15.9949 O.pbe ATOMIC_POSITIONS crystal Si 0.62500000 0.62500000 0.25000000 Si 0.87500000 0.87500000 0.75000000 (....) K_POINTS {crystal} 1 0.25 0.25 0.25 1 ------------------------------------------------------------------- -----------------------ph input file: Pyrope m2 &inputph amass(1)=26.9815, amass(2)=23.9850, amass(3)=27.9769, amass(4)=15.9949, alpha_mix(1) = 0.7, tr2_ph = 1.0D-16, prefix='PYROPE', fildyn='mat.strelD.bald.m3', epsil =.false., trans =.true., zue = .false., outdir='/tmp/' /&end 0.0 0.0 -1.0000000 ----------------------------------------------------- Has anyone an idea on the importance or the origin of this error? Thanks for any hints, Best regards, Merlin Meheut ESS department at UCLA -- Merlin M?heut professional adress: 595 Charles Young Drive East 3806 Geology Building Los Angeles, California 90095-1597 United States of America tel: 310 825 7934 cell phone: 310 893 4253 ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From meheut at impmc.jussieu.fr Fri Jan 30 04:21:27 2009 From: meheut at impmc.jussieu.fr (merlin meheut) Date: Fri, 30 Jan 2009 04:21:27 +0100 Subject: [Pw_forum] supercell calculations with matdyn In-Reply-To: <20090120220453.i5jxez58joc0ocgg@mail.democritos.it> References: <20090118034537.14326o4naulse381@www-ext.impmc.jussieu.fr> <20090119224615.8sbly1h5kw8ossk4@mail.democritos.it> <20090120215828.64486y5kroj6nr3o@www-ext.impmc.jussieu.fr> <20090120220453.i5jxez58joc0ocgg@mail.democritos.it> Message-ID: <20090130042127.75716xf8k6hxl8af@www-ext.impmc.jussieu.fr> Hi Paolo, It was just to tell you that everything seems ok with this correction. Best wishes, Merlin Quoting giannozz at democritos.it: > Quoting merlin meheut : > >> I am trying to look at it myself right now. > > excellent initiative. Something that was wrong for sure is the > following: line 243 (of the cvs version) of matdyn.f90 was > > ALLOCATE ( tau (3, nat), ityp(nat), itau_blk(nat_blk) ) > > while it should be instead > > ALLOCATE ( tau (3, nat), ityp(nat), itau_blk(nat) ) > >> At line 289 of file matdyn.f90 >> Fortran runtime error: End of file > > this is quite self-explanatory > > Paolo > --- > Democritos and University of Udine > > ---------------------------------------------------------------- > 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 > -- Merlin M?heut professional adress: 595 Charles Young Drive East 3806 Geology Building Los Angeles, California 90095-1597 United States of America tel: 310 825 7934 cell phone: 310 893 4253 ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. From eyvaz_isaev at yahoo.com Fri Jan 30 10:45:15 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 30 Jan 2009 01:45:15 -0800 (PST) Subject: [Pw_forum] phonon calculation crashes In-Reply-To: <20090130041915.20625vqish5eboab@www-ext.impmc.jussieu.fr> Message-ID: <740492.33195.qm@web65702.mail.ac4.yahoo.com> Dear Merlin, This one is also a famous message, though I did heard about it since long time. Well, this means you have more dimension for irreducible reperesentations that allowed by default, max_irr_dim=4. See phcom.f90 for sure. So, you can just increase this parameter (up to a reasonable value 6, 8, ... ) in phcom.f90 and recompile ph.x. Then it should work. 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 Fri, 1/30/09, merlin meheut wrote: > From: merlin meheut > Subject: [Pw_forum] phonon calculation crashes > To: "PWSCF Forum" > Date: Friday, January 30, 2009, 6:19 AM > Dear all, > > I have had a strange crash on a phonon calculation, with > version > 3.2.3: I have calculated the dynamical matrix of the > structure at > gamma, at q-point (0;0.5;-0.5), both without problems, but > I cannot > calculate the (0;0;-1) q-point without having this error: > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from set_irr : error # 2 > npert > max_irr_dim > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > I don't know what this means. The structure I am > calculating is very > big (80 atoms), so maybe it is related, I don't know. > > Follows the input files: > > --------------------------- scf input file: > pyrope Pavese 1995 (30K) cubique centre > &control > calculation = 'scf', > restart_mode = 'from_scratch' , > prefix = 'PYROPE', > disk_io = 'default' , > pseudo_dir = './', > outdir = '/tmp/', > tprnfor = .true., > tstress = .true., > /&end > &system > ibrav =3, celldm(1)=21.9096, > nat =80, ntyp = 4, ecutwfc = 80.0 > /&end > &electrons > electron_maxstep = 60, > conv_thr = 1.d-7 , > mixing_mode = 'plain', > startingwfc = 'atomic', > mixing_beta = 0.4, > diagonalization = 'david_overlap', > /&end > ATOMIC_SPECIES > Al 26.9815 Al.pbe2 > Mg 23.9850 Mgpbe1_mt.ncpp > Si 27.9769 Si.pbe2 > O 15.9949 O.pbe > ATOMIC_POSITIONS crystal > Si 0.62500000 0.62500000 0.25000000 > (...) > > K_POINTS {crystal} > 1 > 0.25 0.25 0.25 1 > ------------------------------------------------------------------- > > ------------------------------------------nscf input file: > > pyrope Pavese 1995 (30K) cubique centre > &control > calculation = 'phonon', > restart_mode = 'from_scratch' , > prefix = 'PYROPE', > disk_io = 'default' , > pseudo_dir = './', > outdir = '/tmp/', > tprnfor = .true., > tstress = .true., > /&end > &system > ibrav =3, celldm(1)=21.9096, > nat =80, ntyp = 4, ecutwfc = 80.0 > /&end > &electrons > electron_maxstep = 60, > conv_thr = 1.d-7 , > mixing_mode = 'plain', > startingwfc = 'atomic', > mixing_beta = 0.4, > diagonalization = 'david_overlap', > /&end > &phonon > xqq(1)= 0.0000000, xqq(2)= 0.0000000, xqq(3)= -1.0000000 > /&end > ATOMIC_SPECIES > Al 26.9815 Al.pbe2 > Mg 23.9850 Mgpbe1_mt.ncpp > Si 27.9769 Si.pbe2 > O 15.9949 O.pbe > ATOMIC_POSITIONS crystal > Si 0.62500000 0.62500000 0.25000000 > Si 0.87500000 0.87500000 0.75000000 > (....) > > K_POINTS {crystal} > 1 > 0.25 0.25 0.25 1 > ------------------------------------------------------------------- > > -----------------------ph input file: > > Pyrope m2 > &inputph > amass(1)=26.9815, > amass(2)=23.9850, > amass(3)=27.9769, > amass(4)=15.9949, > alpha_mix(1) = 0.7, > tr2_ph = 1.0D-16, > prefix='PYROPE', > fildyn='mat.strelD.bald.m3', > epsil =.false., > trans =.true., > zue = .false., > outdir='/tmp/' > /&end > 0.0 0.0 -1.0000000 > ----------------------------------------------------- > > > Has anyone an idea on the importance or the origin of this > error? > > Thanks for any hints, > > Best regards, > > Merlin Meheut > > ESS department at UCLA > > > > > > > > > -- > Merlin M?heut > professional adress: > 595 Charles Young Drive East > 3806 Geology Building > Los Angeles, California 90095-1597 > United States of America > tel: 310 825 7934 > cell phone: 310 893 4253 > > ---------------------------------------------------------------- > 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 From baroni at sissa.it Fri Jan 30 11:16:17 2009 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 30 Jan 2009 11:16:17 +0100 Subject: [Pw_forum] phonon calculation crashes In-Reply-To: <740492.33195.qm@web65702.mail.ac4.yahoo.com> References: <740492.33195.qm@web65702.mail.ac4.yahoo.com> Message-ID: <68B0D4C9-AB3E-48F1-8585-D22C354B2CAD@sissa.it> I am a bit confused. I thought that the maximum dimension of any irrep of any point group compatible with lattice translations was 3. The number 4 may arise form complications with non-symmorphic groups (such as at the X point of the diamond structure), which I used to me kind of familiar with, but that I do not remember any longer. I am surprised that the need of any irrep larger that 4 may ever arise, but this surprise may well be due to the rustiness of my knowledge in group theory - Cheers - Stefano B On Jan 30, 2009, at 10:45 AM, Eyvaz Isaev wrote: > Dear Merlin, > > This one is also a famous message, though I did heard about it since > long time. > > Well, this means you have more dimension for irreducible > reperesentations that allowed by default, max_irr_dim=4. See > phcom.f90 for sure. > > So, you can just increase this parameter (up to a reasonable value > 6, 8, ... ) in phcom.f90 and recompile ph.x. > Then it should work. > > 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 Fri, 1/30/09, merlin meheut wrote: > >> From: merlin meheut >> Subject: [Pw_forum] phonon calculation crashes >> To: "PWSCF Forum" >> Date: Friday, January 30, 2009, 6:19 AM >> Dear all, >> >> I have had a strange crash on a phonon calculation, with >> version >> 3.2.3: I have calculated the dynamical matrix of the >> structure at >> gamma, at q-point (0;0.5;-0.5), both without problems, but >> I cannot >> calculate the (0;0;-1) q-point without having this error: >> >> >> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% >> %%%%%%%%%% >> from set_irr : error # 2 >> npert > max_irr_dim >> >> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% >> %%%%%%%%%% >> >> I don't know what this means. The structure I am >> calculating is very >> big (80 atoms), so maybe it is related, I don't know. >> >> Follows the input files: >> >> --------------------------- scf input file: >> pyrope Pavese 1995 (30K) cubique centre >> &control >> calculation = 'scf', >> restart_mode = 'from_scratch' , >> prefix = 'PYROPE', >> disk_io = 'default' , >> pseudo_dir = './', >> outdir = '/tmp/', >> tprnfor = .true., >> tstress = .true., >> /&end >> &system >> ibrav =3, celldm(1)=21.9096, >> nat =80, ntyp = 4, ecutwfc = 80.0 >> /&end >> &electrons >> electron_maxstep = 60, >> conv_thr = 1.d-7 , >> mixing_mode = 'plain', >> startingwfc = 'atomic', >> mixing_beta = 0.4, >> diagonalization = 'david_overlap', >> /&end >> ATOMIC_SPECIES >> Al 26.9815 Al.pbe2 >> Mg 23.9850 Mgpbe1_mt.ncpp >> Si 27.9769 Si.pbe2 >> O 15.9949 O.pbe >> ATOMIC_POSITIONS crystal >> Si 0.62500000 0.62500000 0.25000000 >> (...) >> >> K_POINTS {crystal} >> 1 >> 0.25 0.25 0.25 1 >> ------------------------------------------------------------------- >> >> ------------------------------------------nscf input file: >> >> pyrope Pavese 1995 (30K) cubique centre >> &control >> calculation = 'phonon', >> restart_mode = 'from_scratch' , >> prefix = 'PYROPE', >> disk_io = 'default' , >> pseudo_dir = './', >> outdir = '/tmp/', >> tprnfor = .true., >> tstress = .true., >> /&end >> &system >> ibrav =3, celldm(1)=21.9096, >> nat =80, ntyp = 4, ecutwfc = 80.0 >> /&end >> &electrons >> electron_maxstep = 60, >> conv_thr = 1.d-7 , >> mixing_mode = 'plain', >> startingwfc = 'atomic', >> mixing_beta = 0.4, >> diagonalization = 'david_overlap', >> /&end >> &phonon >> xqq(1)= 0.0000000, xqq(2)= 0.0000000, xqq(3)= -1.0000000 >> /&end >> ATOMIC_SPECIES >> Al 26.9815 Al.pbe2 >> Mg 23.9850 Mgpbe1_mt.ncpp >> Si 27.9769 Si.pbe2 >> O 15.9949 O.pbe >> ATOMIC_POSITIONS crystal >> Si 0.62500000 0.62500000 0.25000000 >> Si 0.87500000 0.87500000 0.75000000 >> (....) >> >> K_POINTS {crystal} >> 1 >> 0.25 0.25 0.25 1 >> ------------------------------------------------------------------- >> >> -----------------------ph input file: >> >> Pyrope m2 >> &inputph >> amass(1)=26.9815, >> amass(2)=23.9850, >> amass(3)=27.9769, >> amass(4)=15.9949, >> alpha_mix(1) = 0.7, >> tr2_ph = 1.0D-16, >> prefix='PYROPE', >> fildyn='mat.strelD.bald.m3', >> epsil =.false., >> trans =.true., >> zue = .false., >> outdir='/tmp/' >> /&end >> 0.0 0.0 -1.0000000 >> ----------------------------------------------------- >> >> >> Has anyone an idea on the importance or the origin of this >> error? >> >> Thanks for any hints, >> >> Best regards, >> >> Merlin Meheut >> >> ESS department at UCLA >> >> >> >> >> >> >> >> >> -- >> Merlin M?heut >> professional adress: >> 595 Charles Young Drive East >> 3806 Geology Building >> Los Angeles, California 90095-1597 >> United States of America >> tel: 310 825 7934 >> cell phone: 310 893 4253 >> >> ---------------------------------------------------------------- >> 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 > > > > _______________________________________________ > 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/20090130/1bdf3c3b/attachment-0001.htm From quantumdft at gmail.com Fri Jan 30 17:14:47 2009 From: quantumdft at gmail.com (vega lew) Date: Sat, 31 Jan 2009 00:14:47 +0800 Subject: [Pw_forum] something about Pt.pw91-n-van.UPF Message-ID: <412f6c680901300814u719ea01nc8dc3b52a11d7684@mail.gmail.com> Dear all i have a question about the pseudo potential of Pt.pw91-n-van.UPF. When I want to calculate the total energy of Pt atom in a large box, the calculation could coverge. the box is consistent with the surface calculation. the calculated value of total energy in output file is -88.64000965 Ry which is a little different from the value, -88.64879046471 Ry, in the Pt.pw91-n-van.UPF file. do you think the deviation is acceptable? And when I add more empty band, two or three more, in the system using 'nbnd' command , the calculation can't coverge with 100 scf cycles. How do you think about it? My input files as follows, &CONTROL title = 'Pt' , calculation = 'scf' , restart_mode = 'from_scratch' , outdir = '/tmp/' , wfcdir = '/tmp/' , pseudo_dir = '/home/vega/espresso-4.0/pseudo/' , prefix = 'Pt' , disk_io = 'none' , nstep = 1000 , / &SYSTEM ibrav = 8, celldm(1) =19.6144, celldm(2) = 1.0955, celldm(3) = 1.7218, nat = 1, ntyp = 1, nosym = .ture. , ecutwfc = 30, ecutrho = 300, nbnd = 8, # or nbnd =7, the calculation can't converge for this two value / &ELECTRONS / &IONS / ATOMIC_SPECIES Pt 195.09000 Pt.pw91-n-van.UPF ATOMIC_POSITIONS crystal Pt 0.68061 0.25139 0.40394 1 1 1 K_POINTS gamma 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 ****************************************************************************************************************** Email: vegalew at gmail.com Office: Room A705, Technical Innovation Building, Xinmofan Road 5#, Nanjing, Jiangsu, China ****************************************************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090131/a87cd464/attachment.htm From giannozz at democritos.it Fri Jan 30 20:30:57 2009 From: giannozz at democritos.it (Paolo Giannozzi) Date: Fri, 30 Jan 2009 20:30:57 +0100 Subject: [Pw_forum] phonon calculation crashes In-Reply-To: <20090130041915.20625vqish5eboab@www-ext.impmc.jussieu.fr> References: <20090130041915.20625vqish5eboab@www-ext.impmc.jussieu.fr> Message-ID: <32A81432-50C5-4CAE-B8A1-2F07F64E2469@democritos.it> On Jan 30, 2009, at 4:19 , merlin meheut wrote: > I have had a strange crash on a phonon calculation it is not a crash: the code issues an error message and stops > with version 3.2.3 quite old > from set_irr : error # 2 > npert > max_irr_dim increase max_irr_dim and verify if the maximum irrep dimension is compatible with the symmetry of your system Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From asafis at yahoo.com.br Fri Jan 30 22:10:57 2009 From: asafis at yahoo.com.br (=?iso-8859-1?Q?=C1lvaro_Alves?=) Date: Fri, 30 Jan 2009 13:10:57 -0800 (PST) Subject: [Pw_forum] ph.x Message-ID: <219443.6453.qm@web52311.mail.re2.yahoo.com> Hi all. I?m trying to run ph.x for AlAs (to get phonon frequencies at gamma point only). The mode #? 1 is done, but the mode #? 2 don?t converge. What may be happening ? A. S. Santos Veja quais s?o os assuntos do momento no Yahoo! +Buscados http://br.maisbuscados.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090130/c58e4183/attachment.htm From yay451 at mail.usask.ca Fri Jan 30 23:42:34 2009 From: yay451 at mail.usask.ca (Yansun Yao) Date: Fri, 30 Jan 2009 16:42:34 -0600 Subject: [Pw_forum] Can the cp code do MD calculations in NPT ensemble for metals? Message-ID: <1233355354.4983825a27877@webmail.usask.ca> Dear Quantum-EPSRESSO developers, I guess what I would like to ask is in the title of my post. I found a hint that there were some work been done, from the information that Prof. Marzari left before, http://www.democritos.it/pipermail/pw_forum/2007-September/007296.html Would anybody be so nice letting me know whether this implementaion is finished ? Thank you in advance ! With my best regards, Yansun Yansun Yao Steacie Insititue of Molecular Science National Research Council of Canada From eyvaz_isaev at yahoo.com Sat Jan 31 00:24:44 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 30 Jan 2009 15:24:44 -0800 (PST) Subject: [Pw_forum] ph.x In-Reply-To: <219443.6453.qm@web52311.mail.re2.yahoo.com> Message-ID: <11283.2678.qm@web65701.mail.ac4.yahoo.com> Dear Alvaro, Have you changed any parameter in the example file? Is your calculation parallel or serial? Based on quite scarcy information you posted I think no one can help you. Posting your script or giving more information will be helpful. 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, 1/31/09, ?lvaro Alves wrote: > From: ?lvaro Alves > Subject: [Pw_forum] ph.x > To: "pw" > Date: Saturday, January 31, 2009, 12:10 AM > Hi all. I?m trying to run ph.x for AlAs (to get phonon > frequencies at gamma point only). The mode #? 1 is done, > but the mode #? 2 don?t converge. What may be happening ? > > A. S. Santos > > > Veja quais s?o os assuntos do momento no Yahoo! > +Buscados > http://br.maisbuscados.yahoo.com_______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From eyvaz_isaev at yahoo.com Sat Jan 31 00:40:10 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 30 Jan 2009 15:40:10 -0800 (PST) Subject: [Pw_forum] ph.x In-Reply-To: <11283.2678.qm@web65701.mail.ac4.yahoo.com> Message-ID: <289244.39465.qm@web65710.mail.ac4.yahoo.com> Just to remind you that providing your affiliation in your e-mails is a sign of your respect to the Netiquette accepted in the PWSCF community. Bests, Eyvaz. --- On Sat, 1/31/09, Eyvaz Isaev wrote: > From: Eyvaz Isaev > Subject: Re: [Pw_forum] ph.x > To: asafis at yahoo.com.br, pw_forum at pwscf.org > Date: Saturday, January 31, 2009, 2:24 AM > Dear Alvaro, > > Have you changed any parameter in the example file? Is > your calculation parallel or serial? > > Based on quite scarcy information you posted I think no one > can help you. Posting your script or giving more information > will be helpful. > > 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, 1/31/09, ?lvaro Alves > wrote: > > > From: ?lvaro Alves > > Subject: [Pw_forum] ph.x > > To: "pw" > > Date: Saturday, January 31, 2009, 12:10 AM > > Hi all. I?m trying to run ph.x for AlAs (to get > phonon > > frequencies at gamma point only). The mode #? 1 is > done, > > but the mode #? 2 don?t converge. What may be > happening ? > > > > A. S. Santos > > > > > > Veja quais s?o os assuntos do momento no Yahoo! > > +Buscados > > > http://br.maisbuscados.yahoo.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 From gmacharaga at gmail.com Sat Jan 31 01:02:43 2009 From: gmacharaga at gmail.com (Gratian macharaga) Date: Sat, 31 Jan 2009 08:02:43 +0800 Subject: [Pw_forum] header intact Message-ID: <50c6febb0901301602x14cd27dew90f85152a70683f0@mail.gmail.com> header intact -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090131/92a6abb9/attachment.htm From eyvaz_isaev at yahoo.com Sat Jan 31 01:07:28 2009 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 30 Jan 2009 16:07:28 -0800 (PST) Subject: [Pw_forum] something about Pt.pw91-n-van.UPF In-Reply-To: <412f6c680901300814u719ea01nc8dc3b52a11d7684@mail.gmail.com> Message-ID: <964023.62763.qm@web65713.mail.ac4.yahoo.com> Dear Vega, How about occupations (smearing or tetrahedra) keyword for Pt which is metallic? As there is no "occupations", obviously, your system is considerd as a semiconducting, and there is no way to improve the convergence adding more bands, as all the bands higher than e(n_valence_electrons/2) are considered as empty bands. And you were quite lucky to avoid the message like "metallic system, specify occupations", because the PsP used 10 valence electrons. In this respect your results seem to be incorrect. Bests, Eyvaz. P.S. As you used Pt atom in a large box, of course, you should apply only a smearing technique (your favorite). ------------------------------------------------------------------- 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 Fri, 1/30/09, vega lew wrote: > From: vega lew > Subject: [Pw_forum] something about Pt.pw91-n-van.UPF > To: "PWSCF Forum" > Date: Friday, January 30, 2009, 7:14 PM > Dear all > > i have a question about the pseudo potential of > Pt.pw91-n-van.UPF. > > When I want to calculate the total energy of Pt atom in a > large box, the > calculation could coverge. the box is consistent with the > surface > calculation. the calculated value of total energy in output > file is > -88.64000965 Ry which is a little different from the value, > -88.64879046471 > Ry, in the Pt.pw91-n-van.UPF file. do you think the > deviation is acceptable? > > And when I add more empty band, two or three more, in the > system using > 'nbnd' command , the calculation can't coverge > with 100 scf cycles. How do > you think about it? > > My input files as follows, > > > &CONTROL > > title = 'Pt' , > > calculation = 'scf' > , > restart_mode = 'from_scratch' > , > outdir = '/tmp/' , > wfcdir = '/tmp/' > , > pseudo_dir = > '/home/vega/espresso-4.0/pseudo/' , > prefix = 'Pt' , > > disk_io = 'none' > , > nstep = 1000 > , > / > > &SYSTEM > > ibrav = > 8, > celldm(1) =19.6144, > celldm(2) = 1.0955, > celldm(3) = 1.7218, > nat = > 1, > ntyp = > 1, > nosym = .ture. > , > ecutwfc = 30, > ecutrho = 300, > nbnd = 8, # or > nbnd =7, the > calculation can't converge for this two value > / > > &ELECTRONS > > > > / > > &IONS > > > / > > ATOMIC_SPECIES > > Pt 195.09000 > Pt.pw91-n-van.UPF > ATOMIC_POSITIONS crystal > Pt 0.68061 0.25139 0.40394 1 1 1 > K_POINTS gamma > > 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 > ****************************************************************************************************************** > Email: vegalew at gmail.com > Office: Room A705, Technical Innovation Building, Xinmofan > Road 5#, Nanjing, > Jiangsu, China > ****************************************************************************************************************** > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From nemeth at anl.gov Sat Jan 31 06:45:14 2009 From: nemeth at anl.gov (Karoly Nemeth) Date: Fri, 30 Jan 2009 23:45:14 -0600 (CST) Subject: [Pw_forum] PP by atomic.x In-Reply-To: References: Message-ID: Hi, I am trying to reproduce Ru.pbe-n-van.UPF from the PP library of PWSCF, by atomic.x. I am constantly getting an error message "phi has nodes before r_c" while I am using the same parameters as what is given in the library file provided (my input file for atomic.x is attached). Can someone please help me with this issue? I have used version 4.0.4 of QE. Thanks: Karoly *********************************************************************** Karoly Nemeth, Ph.D. Argonne National Laboratory 9700 S. Cass Ave. Argonne, IL 60439. Advanced Photon Source Accelerator Physics and Operations Group Building 401, Room B2200 Tel. 630-252-5813 Fax: 630-252-4732 *********************************************************************** -------------- next part -------------- &input title='Ru', zed=44.0, rel=1, iswitch=3, rlderiv=2.50, eminld=-4.0, emaxld=4.0, nld=5, deld=0.02, isic=0, config='[Kr] 4d7 5s1 5p0', dft='PBE' / &inputp pseudotype=3, lloc=1, file_pseudopw='RuScRel.UPF', nlcc=.true., / 3 4D 4 2 7.00 0.0 2.00 2.00 5S 5 0 1.00 0.0 2.40 2.40 5P 5 1 0.00 0.0 2.40 2.40 From lataprem29 at gmail.com Sat Jan 31 09:48:44 2009 From: lataprem29 at gmail.com (premlata pandit) Date: Sat, 31 Jan 2009 14:18:44 +0530 Subject: [Pw_forum] pseudo-potential problem In-Reply-To: <49801D5A.9010903@sissa.it> References: <6f289440901272340t5c2f3cd0x7758b7acf0ba7672@mail.gmail.com> <49801D5A.9010903@sissa.it> Message-ID: <6f289440901310048w166d98bbs226c6958852f44f3@mail.gmail.com> Dear PWSCF users I need fully relativistic ultrasoft pseudo-potential for Pm. I tried to generate it , but did not get success. If anyone can help me, I would be grateful. its very important for my Ph.d. work. Thank you On Wed, Jan 28, 2009 at 2:24 PM, Gabriele Sclauzero wrote: > To generate such pseudopotential is not an easy task, especially for those > who have not > much experience (included me). I think it's unprobable that someone of this > community will > do it for you in a short time (or maybe even in a long time). > > If you want to generate by yourself and you want some help, you have at > least to show that > you understand some basics on USPPs. > > So, please first read some documentation/paper/tutorial about Vanderbilt > USPPs (maybe > you'll find some on Vanderbilt's site > http://www.physics.rutgers.edu/~dhv/uspp/), > then > maybe try to generate and test a scalar-relativistic PP and then switch to > fully-relativistic. > > GS > > premlata pandit wrote: > > Dear pwscf users, > > I have been trying to generate the ultrasoft pseudo potential for Pm > > rare earth compound, but I've some problem > > Here is my input file > > > > &input > > title='Pm', > > zed=61.0, > > rel=2, > > rlderiv=2.50, > > eminld=-4.0, > > emaxld=4.0, > > deld=0.02, > > nld=3, > > config='[Xe]4f5.00 6s2.0 5d0.0 6p0.00', > > iswitch=3, > > dft='LDA' > > / > > &inputp > > pseudotype=3, > > lloc=0, > > nlcc=.true., > > rcore=1.5, > > rcloc=2.5, > > file_pseudopw='Pm.RRKJ3.UPF', > > / > > 11 > > 4F 4 3 5.00 0.00 2.20 2.30 2.50 > > 4F 4 3 0.00 -0.20 2.20 2.30 2.50 > > 4F 4 3 0.00 -0.20 2.20 2.30 3.50 > > 4F 4 3 0.00 -0.20 2.20 2.30 3.50 > > 5D 3 2 0.00 -0.30 1.80 2.40 1.50 > > 5D 3 2 0.00 -0.30 1.80 2.40 1.50 > > 5D 3 2 0.00 -0.30 1.80 2.40 2.50 > > 5D 3 2 0.00 -0.30 1.80 2.40 2.50 > > 6P 2 1 0.00 0.00 2.30 2.30 0.50 > > 6P 2 1 0.00 0.00 2.30 2.30 1.50 > > 6S 1 0 2.00 0.00 2.40 2.40 0.50 > > > > At the end ld1.x calculation we found the error-- > > > > Wfc 5D rcut= 1.802 Estimated cut-off energy= 69.67 Ry > > This function has 0 nodes for 0 < r < 1.802 > > Wfc-us 5D rcutus= 2.373 Estimated cut-off energy= 19.47 Ry > > Message from routine compute phi: > > negative determinant > > > > ld= 3.166003 f2ae -0.121907 faenor 0.027736 > > 1 1260 1259 100.899617435058 > > > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > from compute_chi : error # 1 > > n is too large > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > > > stopping ... > > > > could someone suggest me what should i do to generate pseudopotential > > for Pm > > compound and what i doing wrong here. > > -- > > Premlata Pandit > > Ph.d. student, > > Barkatullah university, > > Bhopal 462026, > > MP, India > > > > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > -- > > > o ------------------------------------------------ o > | Gabriele Sclauzero, PhD Student | > | c/o: SISSA & CNR-INFM Democritos, | > | via Beirut 2-4, 34014 Trieste (Italy) | > | email: sclauzer at sissa.it | > | phone: +39 040 3787 511 | > | skype: gurlonotturno | > o ------------------------------------------------ o > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- 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/20090131/ad88bfcd/attachment.htm From bayleyegn_aq at yahoo.com Sat Jan 31 13:36:29 2009 From: bayleyegn_aq at yahoo.com (bayleyegn akane) Date: Sat, 31 Jan 2009 04:36:29 -0800 (PST) Subject: [Pw_forum] How/where to get DISPLACEMENT PATTERN Message-ID: <165831.28261.qm@web45510.mail.sp1.yahoo.com> Hello Dear all I am graduate student at Addis Ababa Univerity-Ethiopia. Currently I am working on my thesis using the Quantum-Espresso package. ? I am trying to do phonon calculation with my system relaxed.?There are?three points that ?I want to be clear about for now: 1. when can I say that a system is fully relaxed? 2. should?I do simple 'relax' prior to a 'vc-relax'? 3. to do a 'vc-relax', what displacement pattern I better follow? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090131/ae63927c/attachment.htm From baroni at sissa.it Sat Jan 31 13:53:38 2009 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 31 Jan 2009 13:53:38 +0100 Subject: [Pw_forum] How/where to get DISPLACEMENT PATTERN In-Reply-To: <165831.28261.qm@web45510.mail.sp1.yahoo.com> References: <165831.28261.qm@web45510.mail.sp1.yahoo.com> Message-ID: Dear Bayleyegn Akane: > I am graduate student at Addis Ababa Univerity-Ethiopia. Currently I > am working on my thesis using the Quantum-Espresso package. welcom on board! > I am trying to do phonon calculation with my system relaxed. There > are three points that I want to be clear about for now: > 1. when can I say that a system is fully relaxed? when the forces acting on atoms and the acroscopic stress are "small enough". How small is "small"? As small as getting it even smaller would not change the quantity you want to calculate (phonon frequencies, in you case) to within your target accuracy. What the target accuracy has to be, only YOU can say ... > 2. should I do simple 'relax' prior to a 'vc-relax'? it depends how far your volume/ cell shape is from their equilibrium state. all in all, it is probably safer do some experimentation by yourself and gain some confidence on what you are doing, than relying on the advice of other people ... > > 3. to do a 'vc-relax', what displacement pattern I better follow? sorry: I do not quite undererstand the question ... take care - Stefano B --- 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/20090131/e717f059/attachment.htm From baroni at sissa.it Sat Jan 31 14:34:22 2009 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 31 Jan 2009 14:34:22 +0100 Subject: [Pw_forum] How/where to get DISPLACEMENT PATTERN In-Reply-To: Message-ID: Dear Prof. Stefano I have taken the your suggestions for the first two of my questions. Regarding the third one, ' 3. to do a 'vc-relax', what displacement pattern I better follow?' What I mean here is that in my in put file for the 'vc-relax' in what way I better displace(disturb) the atoms? I have already done a phonon calculation for my system unrelaxed(the atoms in their ideal positions) and found negative phonon frequencies at gamma, M, and R. --- On Sat, 1/31/09, Stefano Baroni wrote: From: Stefano Baroni Subject: Re: [Pw_forum] How/where to get DISPLACEMENT PATTERN To: bayleyegn_aq at yahoo.com, "PWSCF Forum" Date: Saturday, January 31, 2009, 4:53 AM Dear Bayleyegn Akane: > I am graduate student at Addis Ababa Univerity-Ethiopia. Currently I > am working on my thesis using the Quantum-Espresso package. welcom on board! > I am trying to do phonon calculation with my system relaxed.. > There are three points that I want to be clear about for now: > 1. when can I say that a system is fully relaxed? when the forces acting on atoms and the acroscopic stress are "small enough". How small is "small"? As small as getting it even smaller would not change the quantity you want to calculate (phonon frequencies, in you case) to within your target accuracy. What the target accuracy has to be, only YOU can say ... > 2. should I do simple 'relax' prior to a 'vc-relax'? it depends how far your volume/ cell shape is from their equilibrium state. all in all, it is probably safer do some experimentation by yourself and gain some confidence on what you are doing, than relying on the advice of other people ... > 3. to do a 'vc-relax', what displacement pattern I better follow? sorry: I do not quite undererstand the question ... take care - Stefano B --- 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 --- 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/20090131/bba575c7/attachment-0001.htm From chenhanghuipwscf at gmail.com Sat Jan 31 17:23:56 2009 From: chenhanghuipwscf at gmail.com (hanghui chen) Date: Sat, 31 Jan 2009 11:23:56 -0500 Subject: [Pw_forum] Hubbard U approach in PWSCF Message-ID: <22ae3ca40901310823k698ee8d4g5dca8102850b3530@mail.gmail.com> Dear PWSCF users, Maybe this question is too technical but I would like to give a try. There are two LSDA+U approaches in VASP. One is the so-called 'fully localized limit' developed by Anisimov, et al. In this approach you need to specify both U and J (two adjusting parameters). The other approach is a simplifed one developed by Dudarev et al (PRB 57 1505, 1998) in which there is only one parameter Ueff = U - J. In PWSCF we only have one adjusting parameter Hubbard_U (Hubbard_alpha is something else). Does that mean PWSCF implements the second approach and Hubbard_U = Ueff? However from the menu INPUT_PW.txt, the references you give are the papers from Anisimov et al. Which approach do you implement? If I want to reproduce some results from VASP calculation, shall I turn on Hubbard_U = U-J where U and J are the input parameters in the VASP? Thank you very much. Hanghui Chen Department of Physics Yale University -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090131/2db7ccc3/attachment.htm From matteo at umn.edu Sat Jan 31 20:12:22 2009 From: matteo at umn.edu (Matteo Cococcioni) Date: Sat, 31 Jan 2009 13:12:22 -0600 Subject: [Pw_forum] Hubbard U approach in PWSCF In-Reply-To: <22ae3ca40901310823k698ee8d4g5dca8102850b3530@mail.gmail.com> References: <22ae3ca40901310823k698ee8d4g5dca8102850b3530@mail.gmail.com> Message-ID: <4984A296.9020004@umn.edu> Dear Hanghui, curiously this topic was discussed just few days ago in this forum. presently in pwscf only the simpler scheme is implemented (U_eff = U - J). At the beginning we thought that having J was not so important and so we opted for the most readable and simplest implementation. Some papers appeared recently, however, seem to suggest that J has some role. A prehistoric version of the code exists that has U and J exists but that needs to be re-implemented and put to work in the latest version. regards, Matteo hanghui chen wrote: > Dear PWSCF users, > Maybe this question is too technical but I would like to give a try. > There are two LSDA+U approaches in VASP. One is the so-called > 'fully localized limit' developed by Anisimov, et al. In this approach > you need to specify both U and J (two adjusting parameters). The other > approach is a simplifed one developed by Dudarev et al (PRB 57 1505, > 1998) in which there is only one parameter Ueff = U - J. > In PWSCF we only have one adjusting parameter Hubbard_U > (Hubbard_alpha is something else). Does that mean PWSCF implements the > second approach and Hubbard_U = Ueff? However from the menu > INPUT_PW.txt, the references you give are the papers from Anisimov > et al. Which approach do you implement? > If I want to reproduce some results from VASP calculation, > shall I turn on Hubbard_U = U-J where U and J are the input parameters > in the VASP? > Thank you very much. > > Hanghui Chen > Department of Physics > Yale University > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum >