From pushpa at jncasr.ac.in Fri Aug 1 07:44:42 2003 From: pushpa at jncasr.ac.in (Raghani Pushpa) Date: Fri, 1 Aug 2003 11:14:42 +0530 (IST) Subject: [Pw_forum] (no subject) Message-ID: Dear users, I am calculating the phonon spectrum of a bct structure. When I calculate the force constant file with q2r.x program, it comes out to be the same with zasr=.true. and zasr=.false. It should actually be different in the two cases. Where do you think the problem is arising? I calculated the force constants for gaas which is given as the examples in pwscf and with the same program q2r.x, the force constants are different with zasr=.true. and zasr=.false. i.e. there is no problem with the program as such. Pushpa -- From degironc at sissa.it Fri Aug 1 09:32:20 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Fri, 01 Aug 2003 09:32:20 +0200 Subject: [Pw_forum] (no subject) References: Message-ID: <3F2A1784.30303@sissa.it> Raghani Pushpa wrote: >Dear users, >I am calculating the phonon spectrum of a bct structure. When I calculate >the force constant file with q2r.x program, it comes out to be the same >with zasr=.true. and zasr=.false. It should actually be different in the >two cases. Where do you think the problem is arising? > >I calculated the force constants for gaas which is given as the examples >in pwscf and with the same program q2r.x, the force constants are >different with zasr=.true. and zasr=.false. i.e. there is no problem with >the program as such. > >Pushpa > Accurate caculation of Zstars and dielectric constants requires a more accurate BZ sampling than other properties. zstar=.true. option is provided because experience showed that good phonon frequencies can be obtained from Zstars computed with non converged BZ sampling (that therefore do not satisfy the acustic sume rule accurately) by imposing this rule by hand. Maybe in your case the Acustic Sum Rule for the effective charges is already well satisfied by the Zstar computed directly by ph.x (as it should be when one makes calculation with large enough number of k-points) and the zasr=.true. option does not change significantly the computed values of Zstar. stefano From sprokuda at dnk.ru Mon Aug 4 14:02:46 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Mon, 4 Aug 2003 16:02:46 +0400 Subject: [Pw_forum] For B pseudopotential In-Reply-To: References: Message-ID: <663556068.20030804160246@dnk.ru> Dear PWscf users! I wanna ask you a question, which has no direct concern to PWscf. Generally my problem in to computer the eigenvector for bulk Germanium. I'mv short in time with this problem, and causeof that I've tried to use the code from O.H. Nielsen, W. Weber, Computer Physics Communications 18 (1979) 101-108. It is based on bond-charged model and gives good results for frequencies, which are the result of the solution eigenvalue problem for dynamical matrix (6*6 size for Ge). As for eigenvectors the results seems strange and wrong. First - the results of eigenvectors is strongly compilier dependent - that is my firsr question - how it could be and what could be reason for that? Eigenvector here is a complex vector of 6*1 size. First three digits describes eigendisplasement of first atom in unit cell, second three digits describes eigendisplasement of second atom. For each three e1 and e2 the condition must be satisfied: e1,2 = c1,2*exp(-i*phi1,2), where e1,2 is the eigendisplacement, c1,2 is a real vectors of dimention 3*1, i - imaginary unit.This condition is not satisfied . More in detail, in some cases this condition is satisfied, but there only in some points along [q q 0] direction. The second question what could be the reason for such a strange behavior of comutation of eigenvectors (while the results for frequencies is rather good)?. The numerical routine for the solution of eigenvector problem is well-tested, namely I use for that IMSL numerical library. I would be greatly thankfull for any information of suggestions. Best wishes, Sergey Prokudaylo. From degironc at sissa.it Mon Aug 4 14:16:35 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Mon, 04 Aug 2003 14:16:35 +0200 Subject: [Pw_forum] For B pseudopotential References: <663556068.20030804160246@dnk.ru> Message-ID: <3F2E4EA3.1040906@sissa.it> A few comments a) About machine dependency (excluding presence of bugs in the code): 1) eigenvectors of non-degenerate eigenvalues are defined up to a phase factor 2) eigenvectors of degenerate eigenvalues are defined up to a unitary rotation in the degenerate subspace 3) different machines/diagonalization routines may choose a different phase factors or unitary transformation among the infinit number of legittimate choises. b) About bond charge models: 1) Bond charge models have been fitted to experimental frequencies, not to eigenvectors. These happen to be poorly described by the model if compared with ab-initio results. This is certainly true for GaAs but I bet it is so also for Ge. 2) Paolo Giannozzi worked several years ago about re-fitting bond charge parameters to a ab-initio eigenvalues AND eigenvectors, maybe he still has some data about that. Stefano de Gironcoli sprokuda at dnk.ru wrote: >Dear PWscf users! >I wanna ask you a question, which has no direct concern to PWscf. >Generally my problem in to computer the eigenvector for bulk >Germanium. I'mv short in time with this problem, and causeof that I've >tried to use the code from >O.H. Nielsen, W. Weber, Computer Physics Communications 18 (1979) 101-108. > >It is based on bond-charged model and gives good results for frequencies, which are the result of the solution eigenvalue >problem for dynamical matrix (6*6 size for Ge). >As for eigenvectors the results seems strange and wrong. >First - the results of eigenvectors is strongly compilier dependent - >that is my firsr question - how it could be and what could be reason >for that? >Eigenvector here is a complex vector of 6*1 size. First three digits describes eigendisplasement of first atom in unit >cell, second three digits describes eigendisplasement of second atom. For each three e1 and e2 the condition must be >satisfied: e1,2 = c1,2*exp(-i*phi1,2), where e1,2 is the eigendisplacement, c1,2 is a real vectors of dimention 3*1, i - >imaginary unit.This condition is not satisfied . More in detail, in some cases this condition is satisfied, but there only >in some points along [q q 0] direction. >The second question what could be the reason for such a strange behavior of comutation of eigenvectors (while the results for >frequencies is rather good)?. >The numerical routine for the solution of eigenvector problem is >well-tested, namely I use for that IMSL numerical library. > >I would be greatly thankfull for any information of suggestions. >Best wishes, Sergey Prokudaylo. > > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > From taner at nist.gov Mon Aug 4 23:19:26 2003 From: taner at nist.gov (Taner Yildirim) Date: Mon, 4 Aug 2003 17:19:26 -0400 Subject: [Pw_forum] LDA+U in PWSCF Message-ID: <019b01c35ace$17ef2fe0$94790681@tanlap> Dear Users, I would like to play with LDA+U in pwscf. I was just wondering if anybody has done such a calculations using pwscf. I couldn't find anything in the user manual on this subject. Thanks in advance, Sincerely Taner ******************************************************************** Dr. Taner Yildirim, Ph.D. Physicist, Condensed Matter Science Group, taner at nist.gov NIST Center for Neutron Research, PHONE: 301-975-6228 National Institute of Standards & Technology FAX : 301-921-9847 Gaithersburg, MD 20899-8562 Web : http://webster.ncnr.nist.gov/staff/taner ******************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030804/57f67129/attachment.htm From proffess at yandex.ru Tue Aug 5 07:04:35 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Tue, 5 Aug 2003 09:04:35 +0400 (MSD) Subject: [Pw_forum] Exchange-correlation = ? In-Reply-To: <200307281030.35554.giannozz@nest.sns.it> References: <3EE86CDE.000001.22552@pantene.yandex.ru> <200306121547.02106.giannozz@nest.sns.it> <3F24B08A.000005.23897@ariel.yandex.ru> <200307281030.35554.giannozz@nest.sns.it> Message-ID: <3F2F3AE3.000001.20551@camay.yandex.ru> Dear PW users, I have question. I use PW with old type of pseudopotential. I found the strange line in my output file (section "Exchange-correlation"): ..... bravais-lattice index = 0 lattice parameter (a_0) = 41.4843 a.u. unit-cell volume = 17848.0255 (a.u.)^3 number of atoms/cell = 98 number of atomic types = 1 kinetic-energy cutoff = 35.0000 Ry charge density cutoff = 144.0000 Ry convergence threshold = 1.0E-01 beta = 0.1000 number of iterations used = 8 TF mixing Exchange-correlation = ? (1100) iswitch = 3 nstep = 100 I check it with new type of pseudo. It looks as: Exchange-correlation = SLA PZ NOGX NOGC (1100) I really can use only old format of pseudo. Is it error:Exchange-correlation = ? (1100) Thanks, Best wishes, Sergey From degironc at sissa.it Tue Aug 5 09:36:57 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Tue, 05 Aug 2003 09:36:57 +0200 Subject: [Pw_forum] Exchange-correlation = ? References: <3EE86CDE.000001.22552@pantene.yandex.ru> <200306121547.02106.giannozz@nest.sns.it> <3F24B08A.000005.23897@ariel.yandex.ru> <200307281030.35554.giannozz@nest.sns.it> <3F2F3AE3.000001.20551@camay.yandex.ru> Message-ID: <3F2F5E99.2050302@sissa.it> > > >I check it with new type of pseudo. It looks as: > Exchange-correlation = SLA PZ NOGX NOGC (1100) > >I really can use only old format of pseudo. Is it error:Exchange-correlation = ? (1100) > > > There is no error. (1100) is the internal encoding of SLA PZ NOGX NOGC and the code uses only this information. In the new pseudopotential format exchange-correlation is written as a human meaningful string, (sort of, like SLA PZ NOGX NOGC ) and is converted to the four digit encoding inside the code. In the old pseudopotential version exchange-correlation was given directly in the internal encoding. The question mark you get on output from PWSCF when using the old format simply says that the human meaningful string is missing. There are a number of converters from the old to the new format in the subdirectory upftools Stefano de Gironcoli From sprokuda at dnk.ru Thu Aug 7 10:37:25 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Thu, 7 Aug 2003 12:37:25 +0400 Subject: [Pw_forum] MKL question In-Reply-To: References: Message-ID: <71644877.20030807123725@dnk.ru> Dear Matteo! Thanks a lot for file make.sys for PWscf you sent me last time. But neverethelss I've failed to compile PWscf, and the compilation log is attched to this message. It seems compiler can't find the files in Modules directory, but why - I can understand. Could you please make a look into the file attached, perhaps you will see what is the problem. As for compiler - I use ifc, and MKL6.0 library. Thank you in anvance, Prokudaylo Sergey. -------------- next part -------------- A non-text attachment was scrubbed... Name: matteo.log Type: application/octet-stream Size: 356 bytes Desc: not available Url : /pipermail/attachments/20030807/2d1b49d0/attachment.obj From proffess at yandex.ru Thu Aug 7 16:05:00 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Thu, 7 Aug 2003 18:05:00 +0400 (MSD) Subject: [Pw_forum] some processors have no planes In-Reply-To: <71644877.20030807123725@dnk.ru> References: <71644877.20030807123725@dnk.ru> Message-ID: <3F325C8C.000008.29097@soapbox.yandex.ru> Dear PW users, I have some error (or comment), but PW does not stop: Program PWSCF 1.2.0 starts ... Today is 7Aug2003 at 11:36:40 Parallel version (MPI) Number of processors in use: 128 R & G space division: nprocp = 128 Ultrasoft (Vanderbilt) Pseudopotentials Complex Hamiltonian current dimensions of program pwscf are: ntypx= 6 npsx = 6 lmax = 3 npk =40000 nbrx = 6 lqmax = 7 nqfm = 8 Reading pseudopotential file in UPF format Reading pseudopotential file in UPF format Reading pseudopotential file in UPF format Reading pseudopotential file in UPF format Reading pseudopotential file in UPF format Reading pseudopotential file in UPF format ...... %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from data_structure : error # -1 some processors have no planes %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% ...... --- Executing new GGEN Loop --- bravais-lattice index = 1 lattice parameter (a_0) = 25.0000 a.u. unit-cell volume = 15625.0000 (a.u.)^3 number of atoms/cell = 69 number of atomic types = 1 kinetic-energy cutoff = 35.0000 Ry What does it mean? Thanks. Sergey From eyvaz_isaev at yahoo.com Thu Aug 7 17:59:44 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 7 Aug 2003 08:59:44 -0700 (PDT) Subject: [Pw_forum] some processors have no planes In-Reply-To: <3F325C8C.000008.29097@soapbox.yandex.ru> Message-ID: <20030807155944.34466.qmail@web80503.mail.yahoo.com> Dear Sergei, It means that you asked more CPUs than they are required for your task. So, some of processors out of work. Do you have any successful test on Compaq /Alpha/Linux? Regards, Eyvaz. --- Sergei Lisenkov wrote: > > Dear PW users, > > I have some error (or comment), but PW does not > stop: > > > Program PWSCF 1.2.0 starts ... > Today is 7Aug2003 at 11:36:40 > > Parallel version (MPI) > Number of processors in use: 128 > R & G space division: nprocp = 128 > > > Ultrasoft (Vanderbilt) Pseudopotentials > Complex Hamiltonian > > current dimensions of program pwscf are: > ntypx= 6 npsx = 6 lmax = 3 npk =40000 > nbrx = 6 lqmax = 7 nqfm = 8 > Reading pseudopotential file in UPF format > Reading pseudopotential file in UPF format > Reading pseudopotential file in UPF format > Reading pseudopotential file in UPF format > Reading pseudopotential file in UPF format > Reading pseudopotential file in UPF format > > ...... > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from data_structure : error # -1 > some processors have no planes > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > ...... > --- Executing new GGEN Loop --- > > > > > bravais-lattice index = 1 > lattice parameter (a_0) = 25.0000 a.u. > unit-cell volume = 15625.0000 > (a.u.)^3 > number of atoms/cell = 69 > number of atomic types = 1 > kinetic-energy cutoff = 35.0000 Ry > > What does it mean? > > Thanks. > Sergey > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From eyvaz_isaev at yahoo.com Thu Aug 7 18:08:28 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 7 Aug 2003 09:08:28 -0700 (PDT) Subject: [Pw_forum] LDA+U in PWSCF In-Reply-To: <019b01c35ace$17ef2fe0$94790681@tanlap> Message-ID: <20030807160828.63709.qmail@web80504.mail.yahoo.com> Dear Taner, I did some LDA+U calculations last year. You should just add the next lines in the SYSTEM section: lda_plus_u =.true. Hubbard_U= -6 (as example) Hubbard_alpha=-1 (as example) Regards, Eyvaz Isaev Theoretical Physics Department Moscow Steel and Alloys Institute --- Taner Yildirim wrote: > Dear Users, > > I would like to play with LDA+U in pwscf. I was just > wondering > if anybody has done such a calculations using pwscf. > I couldn't > find anything in the user manual on this subject. > > Thanks in advance, > > Sincerely > Taner > > ******************************************************************** > Dr. Taner Yildirim, Ph.D. Physicist, Condensed > Matter Science Group, > taner at nist.gov NIST Center for > Neutron Research, > PHONE: 301-975-6228 National Institute of > Standards & Technology > FAX : 301-921-9847 > Gaithersburg, MD 20899-8562 > Web : http://webster.ncnr.nist.gov/staff/taner > ******************************************************************** > > __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From yanming_ma at hotmail.com Thu Aug 7 23:59:44 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Fri, 08 Aug 2003 05:59:44 +0800 Subject: [Pw_forum] setup error Message-ID: Dear Users, I tried to run scf calculation in my Boron case (space group R-3m; a=b=c, alpha=58.02) The following is my input: cat > B.scf.in<< EOF B Rhom &CONTROL calculation='scf', tstress=.t.,tprnfor=.t., PSEUDO_DIR = '$HOME/yanming/pwwork/pseudo/', outdir = '/scratch/B/B_Rho/', / &SYSTEM ibrav= 5, celldm(1) =9.6, celldm(4)=0.529031, nat= 1, ntyp= 1, occupations='fixed', ecutwfc =30.0, / &ELECTRONS mixing_beta = 0.7, conv_thr = 1.0d-8, / ATOMIC_SPECIES B 10.811 B.pz-vbc.UPF ATOMIC_POSITIONS {crystal} B 0.0 0.0 0.0 K_POINTS{automatic} 8 8 8 0 0 0 EOF The error message is the following: Writing file pwscf.save from setup : error # 1 the crystal is a metal stopping ... file written from setup : error # 1 the crystal is a metal stopping ... %%%%%%%%%%%%%%%%%%%%%% from setup : error # 1 the crystal is a metal %%%%%%%%%%%%%%%%%%%%%% stopping ... Any comments on this error? Thanks in advance. Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 _________________________________________________________________ ?????????????? MSN Messenger: http://messenger.msn.com/cn From degironc at sissa.it Fri Aug 8 09:22:33 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Fri, 08 Aug 2003 09:22:33 +0200 Subject: [Pw_forum] setup error References: Message-ID: <3F334FB9.6030508@sissa.it> As the error message states, the system is a metal since nelec=3 is odd and it is not possible to use fixed occupations (that is to fully occupy the lowest nelec/2 bands). You should provide information on how you want to treat Fermi energy crossings. occupations='smearing' or 'tethrahedra' are possible choices, occupations='fixed' can only be applied to insulators with a gap. If you choose the smearing option you need to provide also (in the &SYSTEM namelist) 1) the smearing type: smearing='gaussian','methfessel-paxton','marzari-vanderbilt', of 'fermi-dirac' 2) the smearing width: degauss= xx in Rydberg energy units. If you choose the tetrahedra option you need to use automatically generated uniform k-point grids. Please read information in file pwdocs/INPUT_PW regarding these variables. Stefano de Gironcoli ma Yanming wrote: > > Dear Users, > > I tried to run scf calculation in my Boron case (space group R-3m; > a=b=c, alpha=58.02) > The following is my input: > > cat > B.scf.in<< EOF > B > Rhom > &CONTROL > calculation='scf', tstress=.t.,tprnfor=.t., > PSEUDO_DIR = '$HOME/yanming/pwwork/pseudo/', > outdir = '/scratch/B/B_Rho/', > / > &SYSTEM > ibrav= 5, celldm(1) =9.6, celldm(4)=0.529031, > nat= 1, ntyp= 1, > occupations='fixed', > ecutwfc =30.0, > / > &ELECTRONS mixing_beta = 0.7, > conv_thr = 1.0d-8, > / > ATOMIC_SPECIES > B 10.811 B.pz-vbc.UPF > ATOMIC_POSITIONS {crystal} > B 0.0 0.0 0.0 > K_POINTS{automatic} > 8 8 8 0 0 0 > EOF > > The error message is the following: > > Writing file pwscf.save from setup : error # 1 the crystal is a metal > stopping ... > file written > from setup : error # 1 > the crystal is a metal > stopping ... > %%%%%%%%%%%%%%%%%%%%%% > from setup : error # 1 the crystal is a metal > %%%%%%%%%%%%%%%%%%%%%% > stopping ... > > > Any comments on this error? > Thanks in advance. > > Yanming Ma PhD > Steacie Institute for Molecular Sciences, > National Research Councils of Canada. > 100 Sussex > K1A 0R6 > > _________________________________________________________________ > ?????????????? MSN Messenger: http://messenger.msn.com/cn > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From yanming_ma at hotmail.com Fri Aug 8 15:22:46 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Fri, 08 Aug 2003 21:22:46 +0800 Subject: [Pw_forum] setup error Message-ID: Dear Stefano, Thanks for your reply. Actually, I know exactly how to treat the metals in the calculation. I also aware that Boron is not a real metal so I try to choose occupation as "fixed" to avoid that. I never try to run example with Rhom. symmetry. So I don't know the correct input. Now I know the problem is from the odd valence electrons. It seems that the correct input should include even number of Borons. Best Wishes. Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 >From: Stefano de Gironcoli >Reply-To: pw_forum at pwscf.org >To: pw_forum at pwscf.org >Subject: Re: [Pw_forum] setup error >Date: Fri, 08 Aug 2003 09:22:33 +0200 > >As the error message states, the system is a metal since nelec=3 is odd >and it is not possible to use fixed occupations (that is to fully occupy >the lowest nelec/2 bands). >You should provide information on how you want to treat Fermi energy >crossings. > >occupations='smearing' or 'tethrahedra' are possible choices, >occupations='fixed' can only be applied to insulators with a gap. > >If you choose the smearing option you need to provide also (in the >&SYSTEM namelist) >1) the smearing type: >smearing='gaussian','methfessel-paxton','marzari-vanderbilt', of >'fermi-dirac' >2) the smearing width: degauss= xx in Rydberg energy units. >If you choose the tetrahedra option you need to use automatically >generated uniform k-point grids. > >Please read information in file pwdocs/INPUT_PW regarding these variables. > >Stefano de Gironcoli > >ma Yanming wrote: > > > > > Dear Users, > > > > I tried to run scf calculation in my Boron case (space group R-3m; > > a=b=c, alpha=58.02) > > The following is my input: > > > > cat > B.scf.in<< EOF > > B > > Rhom > > &CONTROL > > calculation='scf', tstress=.t.,tprnfor=.t., > > PSEUDO_DIR = '$HOME/yanming/pwwork/pseudo/', > > outdir = '/scratch/B/B_Rho/', > > / > > &SYSTEM > > ibrav= 5, celldm(1) =9.6, celldm(4)=0.529031, > > nat= 1, ntyp= 1, > > occupations='fixed', > > ecutwfc =30.0, > > / > > &ELECTRONS mixing_beta = 0.7, > > conv_thr = 1.0d-8, > > / > > ATOMIC_SPECIES > > B 10.811 B.pz-vbc.UPF > > ATOMIC_POSITIONS {crystal} > > B 0.0 0.0 0.0 > > K_POINTS{automatic} > > 8 8 8 0 0 0 > > EOF > > > > The error message is the following: > > > > Writing file pwscf.save from setup : error # 1 the crystal is a metal > > stopping ... > > file written > > from setup : error # 1 > > the crystal is a metal > > stopping ... > > %%%%%%%%%%%%%%%%%%%%%% > > from setup : error # 1 the crystal is a metal > > %%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > > > > > Any comments on this error? > > Thanks in advance. > > > > Yanming Ma PhD > > Steacie Institute for Molecular Sciences, > > National Research Councils of Canada. > > 100 Sussex > > K1A 0R6 > > > > _________________________________________________________________ > > ?????????????? MSN Messenger: http://messenger.msn.com/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 _________________________________________________________________ ?????????????? MSN Messenger: http://messenger.msn.com/cn From proffess at yandex.ru Mon Aug 11 07:50:56 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Mon, 11 Aug 2003 09:50:56 +0400 (MSD) Subject: [Pw_forum] error - "not implemented" In-Reply-To: <200306121547.02106.giannozz@nest.sns.it> References: <3EE86CDE.000001.22552@pantene.yandex.ru> <200306121547.02106.giannozz@nest.sns.it> Message-ID: <3F372EC0.000004.05643@camay.yandex.ru> Dear PW users, Recently I found the next errors: ... Parallel version (MPI) Number of processors in use: 50 R & G space division: nprocp = 50 Ultrasoft (Vanderbilt) Pseudopotentials Complex Hamiltonian current dimensions of program pwscf are: ntypx= 6 npsx = 6 lmax = 3 npk =40000 nbrx = 6 lqmax = 7 nqfm = 8 current restart_mode = from_scratch current disk_io mode = default RECOVER from restart file has been switched off on input Writing file pwscf.save file written Planes per process (thick) : nr3 = 96 npp = 2 ncplane = 144 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from data_structure : error # 1 some processors have no pencils, not yet implemented %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... Another error appears on my PC machine. I calculated the structure with only Gamma point. I used the pwg.x version. But I got: .... %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from c_bands : error # 1 not implemented %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% What is "not implemented"? I used also pw.x version - it is works. Many thanks , Sergey From goranka.bilalbegovic at zg.hinet.hr Mon Aug 11 09:09:32 2003 From: goranka.bilalbegovic at zg.hinet.hr (Goranka Bilalbegovic) Date: Mon, 11 Aug 2003 09:09:32 +0200 Subject: [Pw_forum] postproc on linux cluster Message-ID: <000c01c35fd7$baab4620$09821dc3@gost> Hello. I am not able to use postprocessing routines on the linux cluster (mpich, PBS). I have learnt how to use PPs using mpi on SGI O2K, but on the linux cluster it does not work. The error is: from postproc : error # wrong no. of arguments I found this place in the code, but whatever I put in the PBS script on the run the MPI executable line it does not work. On the page 12 of the manual there is a sentence: "The final result is collected into a single file, whose format is processor-number-independent, if the code is compiled with -D__NEW_PUNCH in CPPFLAGS." Postprocessing routines do not work on the linux cluster I use regardless of this -D_NEW_PUNCH option. I noticed that with this option the file *.pun does not form, but the number of *.wfc* files is again equal to the number of processors. I was expecting only one *.wfc to form. Thanks for help. Best regards, Goranka -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030811/88f44d99/attachment.htm From degironc at sissa.it Mon Aug 11 09:31:10 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Mon, 11 Aug 2003 09:31:10 +0200 Subject: [Pw_forum] error - "not implemented" References: <3EE86CDE.000001.22552@pantene.yandex.ru> <200306121547.02106.giannozz@nest.sns.it> <3F372EC0.000004.05643@camay.yandex.ru> Message-ID: <3F37463E.7070700@sissa.it> Sergei Lisenkov wrote: >... >Parallel version (MPI) > Number of processors in use: 50 > R & G space division: nprocp = 50 > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from data_structure : error # 1 > some processors have no pencils, not yet implemented > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > > > In the parallel version of PWSCF, plane wave are distributed among processors according several criteria. In paricular the G-space grid (G=n1*b1+n2*b2+n3*b3) is distributed assigning a number of colums (or pencils) of G vectors (fixed n1&n2 and n3=1,...,nr3) to different processors. Only non empty pencils (containing G vectors inside the cutoff sphere) are taken into consideration. The code stops if the number of processors is larger than the number of non empty pencils. Use less processors or if you are performing a calculation with more than one k-point consider using pools of p;rocessors to deal with different k-points. > >.... > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from c_bands : error # 1 > not implemented > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > >What is "not implemented"? I used also pw.x version - it is works. > > > in the version at Gamma only Davidson diagonalization is currently implemented. best regards, Stefano de Gironcoli From giannozz at nest.sns.it Mon Aug 11 18:07:17 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 11 Aug 2003 18:07:17 +0200 Subject: [Pw_forum] postproc on linux cluster In-Reply-To: <000c01c35fd7$baab4620$09821dc3@gost> References: <000c01c35fd7$baab4620$09821dc3@gost> Message-ID: <200308111807.17096.giannozz@nest.sns.it> On Monday 11 August 2003 09:09, Goranka Bilalbegovic wrote: > I am not able to use postprocessing routines on the linux cluster (mpich, > PBS). I have learnt how to use PPs using mpi on SGI O2K, but on the linux > cluster it does not work. The error is: from postproc : error # > wrong no. of arguments try to remove everything from PP/start_postproc.f90 but the following: version = 'POSTPROC-121' (or whatever you like) nodenumber = ' ' call startup (nodenumber, version) #ifdef __PARA call init_pool #endif With these changes, pp.x will properly read arguments, provided pw.x does. Note that this may work only on recent CVS version. Originally the postprocessing code used to read data from a file, whose file name was given as an argument, in parallel execution. This is no longer useful nor desirable, so all the crap contained in start_postproc.f90 can (should) (will!) be removed. > On the page 12 of the manual there is a sentence: "The final result is > collected into a single file, whose format is processor-number-independent, > if the code is compiled with -D__NEW_PUNCH in CPPFLAGS." Postprocessing > routines do not work on the linux cluster I use regardless of this > -D_NEW_PUNCH option. I noticed that with this option the file *.pun does > not form, but the number of *.wfc* files is again equal to the number of > processors. I was expecting only one *.wfc to form. I am afraid that the -D__NEW_PUNCH option and postprocessing are not yet compatible. Remember that without -D__NEW_PUNCH you need to run the postprocessing code on the same number of processors and of pools that were used to generate data files. Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From proffess at yandex.ru Wed Aug 13 07:44:57 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Wed, 13 Aug 2003 09:44:57 +0400 (MSD) Subject: [Pw_forum] vc-relax In-Reply-To: <3F37463E.7070700@sissa.it> References: <3EE86CDE.000001.22552@pantene.yandex.ru> <200306121547.02106.giannozz@nest.sns.it> <3F372EC0.000004.05643@camay.yandex.ru> <3F37463E.7070700@sissa.it> Message-ID: <3F39D059.000001.32197@soapbox.yandex.ru> Dear PW authors and users, I do my calculations using "vc-relax" with Parrinello-Rahman method with diagonalization="CG". I would like to know, when my calculations will finished. I use nstep=1000. Is the code will be wait the convergence of calculations or will do all steps? Another question. With 'vc-relax' code start calculations whith T=0 K. Is it possible to start calculations with T=300 K ? Thanks, Best wishes, Sergey From goranka.bilalbegovic at zg.hinet.hr Wed Aug 13 12:34:23 2003 From: goranka.bilalbegovic at zg.hinet.hr (Goranka Bilalbegovic) Date: Wed, 13 Aug 2003 12:34:23 +0200 Subject: [Pw_forum] io_base (O-sesame) Message-ID: <002f01c36186$acd10fc0$ca841dc3@gost> Hello. In the compilation of the O-sesame code (version 11.8. 2003.), using the ifc 7.1 compiler, on the PC cluster (Red Hat 7.2), I have the following error in a new io_base.f90: ----- ... module subroutine WRITE_RESTART_ELECTRONS1 module subroutine WRITE_RESTART_ELECTRONS2 module subroutine READ_RESTART_ELECTRONS1 module subroutine READ_RESTART_ELECTRONS2 module subroutine WRITE_RESTART_WFC1 module subroutine WRITE_RESTART_WFC2 module subroutine READ_RESTART_WFC1 *Compiler Internal Error* : Please report to Intel(R) Corporation compilation aborted for io_base.f90 (code 1) ----- The pwscf 1.2.0 version of io_base.f90 on the same system does not show this error. Since this is a working version of pwscf, and ifc is used by many users, I am reporting here and not to Intel. Best regards, Goranka -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030813/745a258a/attachment.htm From giannozz at nest.sns.it Wed Aug 13 13:06:07 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 13 Aug 2003 13:06:07 +0200 Subject: [Pw_forum] io_base (O-sesame) In-Reply-To: <002f01c36186$acd10fc0$ca841dc3@gost> References: <002f01c36186$acd10fc0$ca841dc3@gost> Message-ID: <200308131306.07991.giannozz@nest.sns.it> On Wednesday 13 August 2003 12:34, Goranka Bilalbegovic wrote: > In the compilation of the O-sesame code (version 11.8. 2003.), using > ifc 7.1 compiler, on the PC cluster (Red Hat 7.2), I have the following > error in a new io_base.f90: [...] > *Compiler Internal Error* : Please report to Intel(R) Corporation In some cases, this may be a consequence of an error in the code, but since everything works fine with ifc 6.0, it must be a ifc 7.1 issue. This is what the manual says: -------------------------------------------------- If you get ``internal compiler error'' or similar messages, lower the optimization level, or remove optimization, just for the routine that has problems. Some earlier versions of fortran-90 compilers were not especially reliable. If you have weird problems, try to use the most recent version of the compiler you can find, or to install patches for the compiler (most vendors release at least a few patches for free). -------------------------------------------------- The most recent version of ifc v.7.1 (7.1.028) was released yesterday. You might try it and let us know. Unfortunately you may need to download the entire package: 80 Mb or so !!! Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From sprokuda at dnk.ru Wed Aug 13 17:47:48 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Wed, 13 Aug 2003 19:47:48 +0400 Subject: [Pw_forum] FFTW problem In-Reply-To: <002f01c36186$acd10fc0$ca841dc3@gost> References: <002f01c36186$acd10fc0$ca841dc3@gost> Message-ID: <77478331.20030813194748@dnk.ru> Dear PWscf users! Probably osmebody could help me with my problem: I've compiled PWscf by ifc6.0 with MKL6.0 and tried to run examples. That is what I get in *.out file. Program PWSCF 1.2.0 starts ... Today is 13Aug2003 at 15:47:27 Ultrasoft (Vanderbilt) Pseudopotentials Complex Hamiltonian current dimensions of program pwscf are: ntypx= 6 npsx = 6 lmax = 3 npk =40000 nbrx = 6 lqmax = 7 nqfm = 8 Reading pseudopotential file in UPF format current restart_mode = from_scratch current disk_io mode = default RECOVER from restart file has been switched off on input Writing file silicon.save file written --- Executing new GGEN Loop --- bravais-lattice index = 2 lattice parameter (a_0) = 10.2000 a.u. unit-cell volume = 265.3020 (a.u.)^3 number of atoms/cell = 2 number of atomic types = 1 kinetic-energy cutoff = 18.0000 Ry charge density cutoff = 72.0000 Ry convergence threshold = 1.0E-08 beta = 0.7000 number of iterations used = 8 plain mixing Exchange-correlation = SLA PZ NOGX NOGC (1100) iswitch = 0 celldm(1)= 10.20000 celldm(2)= 0.00000 celldm(3)= 0.00000 celldm(4)= 0.00000 celldm(5)= 0.00000 celldm(6)= 0.00000 crystal axes: (cart. coord. in units of a_0) a(1) = ( -0.5000 0.0000 0.5000 ) a(2) = ( 0.0000 0.5000 0.5000 ) a(3) = ( -0.5000 0.5000 0.0000 ) reciprocal axes: (cart. coord. in units 2 pi/a_0) b(1) = ( -1.0000 -1.0000 1.0000 ) b(2) = ( 1.0000 1.0000 1.0000 ) b(3) = ( -1.0000 1.0000 -1.0000 ) PSEUDO 1 is Si zval = 4.0 lmax= 1 lloc= 0 (in numerical form: 431 grid points, xmin = 0.00, dx = 0.0000) atomic species valence mass pseudopotential Si 4.00 28.08600 Si( 1.00) 48 Sym.Ops. (with inversion) Cartesian axes site n. atom positions (a_0 units) 1 Si tau( 1) = ( 0.0000000 0.0000000 0.0000000 ) 2 Si tau( 2) = ( 0.2500000 0.2500000 0.2500000 ) number of k points= 10 cart. coord. in units 2pi/a_0 k( 1) = ( 0.1250000 0.1250000 0.1250000), wk = 0.0625000 k( 2) = ( 0.1250000 0.1250000 0.3750000), wk = 0.1875000 k( 3) = ( 0.1250000 0.1250000 0.6250000), wk = 0.1875000 k( 4) = ( 0.1250000 0.1250000 0.8750000), wk = 0.1875000 k( 5) = ( 0.1250000 0.3750000 0.3750000), wk = 0.1875000 k( 6) = ( 0.1250000 0.3750000 0.6250000), wk = 0.3750000 k( 7) = ( 0.1250000 0.3750000 0.8750000), wk = 0.3750000 k( 8) = ( 0.1250000 0.6250000 0.6250000), wk = 0.1875000 k( 9) = ( 0.3750000 0.3750000 0.3750000), wk = 0.0625000 k( 10) = ( 0.3750000 0.3750000 0.6250000), wk = 0.1875000 G cutoff = 189.7462 ( 2733 G-vectors) FFT grid: ( 20, 20, 20) nbndx = 16 nbnd = 4 natomwfc = 8 npwx = 350 nelec = 8.00 nkb = 8 ngl = 65 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from cft_3 : error # 1 machine-specific routine not available %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... I've made a look into cft_3.f90 file and it seems that the symbol FFTW is not defined in this case ( #ifdef FFTW ). what should I do to fix this problem ? Best wishes, Prokudaylo Sergey. From sprokuda at dnk.ru Wed Aug 13 17:52:30 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Wed, 13 Aug 2003 19:52:30 +0400 Subject: [Pw_forum] FFTW problem In-Reply-To: <77478331.20030813194748@dnk.ru> References: <002f01c36186$acd10fc0$ca841dc3@gost> <77478331.20030813194748@dnk.ru> Message-ID: <147760577.20030813195230@dnk.ru> Dear PWscf users! Can anybody say could the pseudopotential for Ge cited on PWscf.org be used for latticedynamics simulation of Germanium ? Best wishes, Prokudaylo Sergey. From gjohnson at srrc.ars.usda.gov Wed Aug 13 18:50:20 2003 From: gjohnson at srrc.ars.usda.gov (Glenn Johnson) Date: Wed, 13 Aug 2003 11:50:20 -0500 Subject: [Pw_forum] problem with input redirection with plotrho.x Message-ID: <20030813165020.GA20811@node1.cluster.srrc.usda.gov> I am a newbie to pwscf so have patience. I just compiled pwscf-1.2 with Linux IFC-7.1, Build 20030307Z. I downloaded the pseudopotentials from the pwscf.org Web site for the elements needed for the examples and then I ran all of the examples. I did have to change the names in the run_example scripts to match the names of the downloaded pseudopotentials. Everything seems to work fine except for two things: - In example2, I get a "FORTRAN STOP 1" at the end of the "phonon calculation at X for a single mode". The output seems okay though. - In example5, I get an error with plotrho.x (see below). The error is due to redirection of the input file to plotrho.x; if I execute plotrho.x without input redirection and enter the appropriate values when prompted, the program works fine. running ./run_example: This example shows how to use pw.x and pp.x to calculate the the scf charge density and plot it in the [110] plane. PW_ROOT is defined as /home/glenn/pwscf PSEUDO_DIR is defined as /home/glenn/pwscf/pseudopotentials TMP_DIR is defined as /usr/scratch/pwscf/ checking that needed directories and files exist; done cleaning /usr/scratch/pwscf/ ; done running the scf calculation; done running the post-processing for charge density; done running chdens.x to make 2-d plot with plotrho.x ; done running plotrho.x to generate rho.ps ; Filename missing or blank - please enter name UNIT 1? Input/Output Error 153: Input file ended In Procedure: main program At Line: 89 Statement: List-Directed READ Unit: 5 Connected To: Stdin Form: Formatted Access: Sequential Records Read : 2 Records Written: 0 ** Address Error ** Here is the output into si.plotrho.out: usage: plotrho [input file] r0 : 0.0000 0.0000 0.0000 tau1 : 2.0000 2.0000 0.0000 tau2 : 0.0000 0.0000 2.0000 2 atomic positions read output file > nx, ny (output) > I looked at the Fortran code in plotrho.f90 but nothing jumped out at me as being wrong. Has anyone seen this before? Thanks. -- Glenn Johnson USDA, ARS, SRRC Phone: (504) 286-4252 New Orleans, LA 70124 e-mail: gjohnson at srrc.ars.usda.gov From giannozz at nest.sns.it Wed Aug 13 18:52:34 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 13 Aug 2003 18:52:34 +0200 Subject: [Pw_forum] FFTW problem In-Reply-To: <77478331.20030813194748@dnk.ru> References: <002f01c36186$acd10fc0$ca841dc3@gost> <77478331.20030813194748@dnk.ru> Message-ID: <200308131852.34204.giannozz@nest.sns.it> On Wednesday 13 August 2003 17:47, sprokuda at dnk.ru wrote: > from cft_3 : error # 1 > machine-specific routine not available you need to recompile adding -DFFTW to CPPFLAGS (in "make.sys") Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From goranka.bilalbegovic at zg.hinet.hr Thu Aug 14 08:55:59 2003 From: goranka.bilalbegovic at zg.hinet.hr (Goranka Bilalbegovic) Date: Thu, 14 Aug 2003 08:55:59 +0200 Subject: [Pw_forum] io_base (O-sesame) Message-ID: <002301c36231$60be2720$b8331dc3@gost> >The most recent version of ifc v.7.1 (7.1.028) was released >yesterday. You might try it and let us know. I did not find this new ifc 7.1.028. Only 7.1.008, which is exactly the version I already installed. I even did a new registration, but they only offer to me 7.1.008. Perhaps, that was news from someone who has a commercial version of ifc. I only have a free unsupported version. I have also searched the Intel fortran forum and web, and did not find anything connected to a new version of ifc, or some patch. Best regards, Goranka -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030814/a6f7c3a5/attachment.htm From cvasse at hotmail.com Thu Aug 14 10:47:30 2003 From: cvasse at hotmail.com (Vasse chis) Date: Thu, 14 Aug 2003 10:47:30 +0200 Subject: [Pw_forum] q2r.x Message-ID: Hi. Could somebody tell me if one have made phonon calculations on a not-generated Monkhorst-Pack grid, how the points of the grid could be inserted in the q2r.x input file. thus, instead of nr1=#,nr2=#,nr3=#, if one could insert a list of k-points? Thanks. Best regards Vasse _________________________________________________________________ The new MSN 8: smart spam protection and 2 months FREE* http://join.msn.com/?page=features/junkmail From proffess at yandex.ru Thu Aug 14 11:32:18 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Thu, 14 Aug 2003 13:32:18 +0400 (MSD) Subject: [Pw_forum] io_base (O-sesame) In-Reply-To: <002301c36231$60be2720$b8331dc3@gost> References: <002301c36231$60be2720$b8331dc3@gost> Message-ID: <3F3B5722.00000B.18860@soapbox.yandex.ru> Dear Goranka, I got two days ago the new version of IFC Paolo wrote. But I am not sure that it helps. The problem is with new version of IFC (>7.0). If you have ifc v.6.0 - you will not have problems (it is very stable version). Good luck, Sergey From eyvaz_isaev at yahoo.com Thu Aug 14 11:36:13 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 14 Aug 2003 02:36:13 -0700 (PDT) Subject: [Pw_forum] q2r.x In-Reply-To: Message-ID: <20030814093613.57601.qmail@web80513.mail.yahoo.com> Hi Vasse, I am not sure that it is possible directly. Probalby you have to edit q2r.F90 in order be able to read your input file. But I am afraid in this case you can meet problem with the FFT transormation. Regards, Eyvaz. --- Vasse chis wrote: > Hi. > > Could somebody tell me if one have made phonon > calculations on a > not-generated Monkhorst-Pack grid, how the points of > the grid could be > inserted in the q2r.x input file. > thus, instead of nr1=#,nr2=#,nr3=#, if one could > insert a list of k-points? > > Thanks. > Best regards Vasse > > _________________________________________________________________ > The new MSN 8: smart spam protection and 2 months > FREE* > http://join.msn.com/?page=features/junkmail > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From sprokuda at dnk.ru Thu Aug 14 12:50:57 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Thu, 14 Aug 2003 14:50:57 +0400 Subject: [Pw_forum] FFTW problem In-Reply-To: <200308131852.34204.giannozz@nest.sns.it> References: <002f01c36186$acd10fc0$ca841dc3@gost> <77478331.20030813194748@dnk.ru> <200308131852.34204.giannozz@nest.sns.it> Message-ID: <61500073.20030814145057@dnk.ru> Dear PWscf users! Probably anybody could help me with my problem ?. I had problem running exaples ( PWscf compiled by ifc6.0 with MKL6.0) > from cft_3 : error # 1 > machine-specific routine not available After direction from Paolo Giannozz I've added -DFFTW to CPPFLAGS (in "make.sys"). Now I've got a new probelm in compilation: ifc -Vaxlib -O2 -tpp6 -fpp -I/home/sprokuda/PWscf/include/ -D__LINUX -D__INTEL - D__FFTW -DFFTW -D"FFTWND_F77_ONE=fftwnd_f77_one" -D"FFTW3D_F77_CREATE_PLAN=fftw3 d_f77_create_plan" -D"FFTW_F77=fftw_f77" -D"FFTW_F77_CREATE_PLAN=fftw_f77_create _plan" -nomodule -I /home/sprokuda/PWscf/Modules -I/home/sprokuda/PWscf/PW -I/ho me/sprokuda/PWscf/PH -c cft_3.f90 external subroutine CFT_3 C_POINTER, save :: plan(2,2) ^ Error 24 at (43:cft_3.f90) : syntax error C_POINTER, save :: plan(2,2) ^ Error 24 at (43:cft_3.f90) : syntax error data plan/0,0,0,0/ ^ Comment 17 at (44:cft_3.f90) : A data statement amongst executable statements is obsolescent in Fortran 95 Error 355 : In program unit CFT_3 variable C_POINTER has not been given a type Error 355 : In program unit CFT_3 variable PLAN has not been given a type 4 Errors compilation aborted for cft_3.f90 (code 1) make[1]: *** [cft_3.o] Error 1 make[1]: Leaving directory `/home/sprokuda/PWscf/PW' make: *** [pw] Error 2 probably it would be well to cite also the log of my " ./configure pc_ifc " command. [root at sprokuda PWscf]# ./configure pc_ifc creating catalog file for Intel compiler ./configure: line 30: /home/sprokuda/PWscf/flib/intel.pcl: No such file or directory ./configure: line 31: /home/sprokuda/PWscf/flib/intel.pcl: No such file or directory cc -o moduledep.x moduledep.c cc -o sizeof.x sizeof.c SIZEOF ( void * ) . . . . = 4 SIZEOF ( double ) . . . . = 8 SIZEOF ( float ). . . . . = 4 SIZEOF ( int ). . . . . . = 4 SIZEOF ( char ) . . . . . = 1 ./shdep: line 47: cd: ../FPMD: No such file or directory ./shdep: line 54: cd: ../flib: No such file or directory ls: ../FPMD/*.f90: No such file or directory ./shdep: line 61: cd: ../CPV: No such file or directory it seems that everything is not so fine here. Best wishes, Sergey Prokudaylo. From matteoc at MIT.EDU Thu Aug 14 16:48:35 2003 From: matteoc at MIT.EDU (Matteo Cococcioni) Date: Thu, 14 Aug 2003 10:48:35 -0400 (EDT) Subject: [Pw_forum] LDA+U in PWSCF In-Reply-To: <019b01c35ace$17ef2fe0$94790681@tanlap> Message-ID: Dear Taner, the LDA+U in pwscf was implemented by Stefano de Gironcoli and me. To help you I can provide my PhD thesis where the theory of this part is illustrated, and also some examples of input for pwscf. Anyway to run a LDA+U calculation is quite simple: you just need to give a non zero value to the input variable Hubbard_U and set the logical variable lda_plus_u to .true. . You can find some explanation about this in the pwdocs/INPUT_PW file. Let me know if you want my thesis or have some problems with lda+U calculations. best, Matteo On Mon, 4 Aug 2003, Taner Yildirim wrote: > Dear Users, > > I would like to play with LDA+U in pwscf. I was just wondering > if anybody has done such a calculations using pwscf. I couldn't > find anything in the user manual on this subject. > > Thanks in advance, > > Sincerely > Taner > > ******************************************************************** > Dr. Taner Yildirim, Ph.D. Physicist, Condensed Matter Science Group, > taner at nist.gov NIST Center for Neutron Research, > PHONE: 301-975-6228 National Institute of Standards & Technology > FAX : 301-921-9847 Gaithersburg, MD 20899-8562 > Web : http://webster.ncnr.nist.gov/staff/taner > ******************************************************************** > > From eyvaz_isaev at yahoo.com Thu Aug 14 17:05:04 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 14 Aug 2003 08:05:04 -0700 (PDT) Subject: [Pw_forum] LDA+U in PWSCF In-Reply-To: Message-ID: <20030814150504.34569.qmail@web80513.mail.yahoo.com> Dear Matteo, I am also interested getting your Thesis. Regards, Eyvaz Isaev --- Matteo Cococcioni wrote: > > Dear Taner, > > the LDA+U in pwscf was implemented by Stefano de > Gironcoli and me. To help > you I can provide my PhD thesis where the theory of > this part is > illustrated, and also some examples of input for > pwscf. Anyway to run a > LDA+U calculation is quite simple: you just need to > give a non zero value > to the input variable Hubbard_U and set the logical > variable lda_plus_u > to .true. . You can find some explanation about this > in the > pwdocs/INPUT_PW file. > Let me know if you want my thesis or have some > problems with lda+U > calculations. > > best, > > Matteo > > On Mon, 4 Aug 2003, Taner Yildirim wrote: > > > Dear Users, > > > > I would like to play with LDA+U in pwscf. I was > just wondering > > if anybody has done such a calculations using > pwscf. I couldn't > > find anything in the user manual on this subject. > > > > Thanks in advance, > > > > Sincerely > > Taner > > > > > ******************************************************************** > > Dr. Taner Yildirim, Ph.D. Physicist, Condensed > Matter Science Group, > > taner at nist.gov NIST Center for > Neutron Research, > > PHONE: 301-975-6228 National Institute of > Standards & Technology > > FAX : 301-921-9847 > Gaithersburg, MD 20899-8562 > > Web : http://webster.ncnr.nist.gov/staff/taner > > > ******************************************************************** > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From sprokuda at dnk.ru Thu Aug 14 17:33:45 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Thu, 14 Aug 2003 19:33:45 +0400 Subject: [Pw_forum] LDA+U in PWSCF In-Reply-To: References: Message-ID: <183574825.20030814193345@dnk.ru> Dear Matteo! I would like , if possible, to get you thesis. best wishes, Sergey Prokudaylo. From goranka.bilalbegovic at zg.hinet.hr Thu Aug 14 17:51:43 2003 From: goranka.bilalbegovic at zg.hinet.hr (Goranka Bilalbegovic) Date: Thu, 14 Aug 2003 17:51:43 +0200 Subject: [Pw_forum] LDA+U in PWSCF and Germanium References: Message-ID: <001e01c3627c$104797e0$5a861dc3@gost> > Let me know if you want my thesis or have some problems with lda+U > calculations. > Dear Matteo, I think that you should put your thesis on your web page (a simple link does not take much time). I am not working on the lda+U problem, but often I would like to learn something new and this is good for all users of PWscf. Actually, I already did search for your thesis when Taner asked this question and did not find it. I know for your PRB paper, but this is not the same. I was postdoc at SISSA (looong time ago, working on the glue MD with Furio Ercolessi and Erio Tosatti) and this is why I know that a written Ph.D thesis at SISSA is a very good source for learning. I do not know why recent theses are not on the SISSA web, but for example, >Generally my problem in to computer the eigenvector for bulk Germanium... >Can anybody say could the pseudopotential for Ge cited on PWscf.org be >used for latticedynamics simulation of Germanium ? >Prokudaylo Sergey. I think that Sergey should look at the thesis of Pasquale Pavone: http://www.sissa.it/cm/thesis/1991/pavone.ps.gz Best regards, Goranka From ps at ned.sims.nrc.ca Thu Aug 14 13:40:15 2003 From: ps at ned.sims.nrc.ca (Serguei Patchkovskii) Date: Thu, 14 Aug 2003 07:40:15 -0400 (EDT) Subject: [Pw_forum] io_base (O-sesame) In-Reply-To: <002301c36231$60be2720$b8331dc3@gost> Message-ID: On Thu, 14 Aug 2003, Goranka Bilalbegovic wrote: > >The most recent version of ifc v.7.1 (7.1.028) was released > >yesterday. You might try it and let us know. > > > I did not find this new ifc 7.1.028. Only 7.1.008, which is exactly the > version I already installed. I even did a new registration, but they > only offer to me 7.1.008. Perhaps, You can get ifc updates from Intel Premier support site. See: http://developer.intel.com/software/products/support/#premier for the information. Premier site is not public; it requires registration and approval from Intel. Serguei --- Dr. Serguei Patchkovskii Tel: +1-(613)-990-0945 Fax: +1-(613)-947-2838 E-mail: Serguei.Patchkovskii at nrc.ca Coordinator of Modelling Software Theory and Computation Group Steacie Institute for Molecular Sciences National Research Council Canada Room 2011, 100 Sussex Drive Ottawa, Ontario K1A 0R6 Canada From giannozz at nest.sns.it Thu Aug 14 22:38:20 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 14 Aug 2003 22:38:20 +0200 Subject: [Pw_forum] FFTW problem In-Reply-To: <61500073.20030814145057@dnk.ru> References: <002f01c36186$acd10fc0$ca841dc3@gost> <200308131852.34204.giannozz@nest.sns.it> <61500073.20030814145057@dnk.ru> Message-ID: <200308142238.20237.giannozz@nest.sns.it> On Thursday 14 August 2003 12:50, sprokuda at dnk.ru wrote: > ifc -Vaxlib -O2 -tpp6 -fpp -I/home/sprokuda/PWscf/include/ -D__LINUX > -D__INTEL - D__FFTW -DFFTW -D"FFTWND_F77_ONE=fftwnd_f77_one" > -D"FFTW3D_F77_CREATE_PLAN=fftw3 d_f77_create_plan" -D"FFTW_F77=fftw_f77" > -D"FFTW_F77_CREATE_PLAN=fftw_f77_create _plan" -nomodule -I > /home/sprokuda/PWscf/Modules -I/home/sprokuda/PWscf/PW -I/ho > me/sprokuda/PWscf/PH -c cft_3.f90 > external subroutine CFT_3 > > C_POINTER, save :: plan(2,2) > ^ > Error 24 at (43:cft_3.f90) : syntax error which version of the code are you using ? the one on the web, the one or cvs,...? C_POINTER is defined in include/machine.h > probably it would be well to cite also the log of my " ./configure pc_ifc " > command. it is irrelevant. Paolo From sprokuda at dnk.ru Fri Aug 15 07:57:49 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Fri, 15 Aug 2003 09:57:49 +0400 Subject: [Pw_forum] FFTW problem In-Reply-To: <200308142238.20237.giannozz@nest.sns.it> References: <002f01c36186$acd10fc0$ca841dc3@gost> <200308131852.34204.giannozz@nest.sns.it> <61500073.20030814145057@dnk.ru> <200308142238.20237.giannozz@nest.sns.it> Message-ID: <154637978.20030815095749@dnk.ru> Dear Paolo! following my question: > ifc -Vaxlib -O2 -tpp6 -fpp -I/home/sprokuda/PWscf/include/ -D__LINUX > -D__INTEL - D__FFTW -DFFTW -D"FFTWND_F77_ONE=fftwnd_f77_one" > -D"FFTW3D_F77_CREATE_PLAN=fftw3 d_f77_create_plan" -D"FFTW_F77=fftw_f77" > -D"FFTW_F77_CREATE_PLAN=fftw_f77_create _plan" -nomodule -I > /home/sprokuda/PWscf/Modules -I/home/sprokuda/PWscf/PW -I/ho > me/sprokuda/PWscf/PH -c cft_3.f90 > external subroutine CFT_3 > > C_POINTER, save :: plan(2,2) > ^ > Error 24 at (43:cft_3.f90) : syntax error which version of the code are you using ? the one on the web, the one or cvs,...? C_POINTER is defined in include/machine.h I'm using the version which is on the web. Hope you will help me with this problem. > probably it would be well to cite also the log of my " ./configure pc_ifc " > command. it is irrelevant. Paolo From sprokuda at dnk.ru Fri Aug 15 09:23:38 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Fri, 15 Aug 2003 11:23:38 +0400 Subject: [Pw_forum] FFTW problem In-Reply-To: <200308142238.20237.giannozz@nest.sns.it> References: <002f01c36186$acd10fc0$ca841dc3@gost> <200308131852.34204.giannozz@nest.sns.it> <61500073.20030814145057@dnk.ru> <200308142238.20237.giannozz@nest.sns.it> Message-ID: <1661359247.20030815112338@dnk.ru> Dear Paolo! I've reported you aobut my problem with #define FFTW, I had following compilation error: > ifc -Vaxlib -O2 -tpp6 -fpp -I/home/sprokuda/PWscf/include/ -D__LINUX > -D__INTEL - D__FFTW -DFFTW -D"FFTWND_F77_ONE=fftwnd_f77_one" > -D"FFTW3D_F77_CREATE_PLAN=fftw3 d_f77_create_plan" -D"FFTW_F77=fftw_f77" > -D"FFTW_F77_CREATE_PLAN=fftw_f77_create _plan" -nomodule -I > /home/sprokuda/PWscf/Modules -I/home/sprokuda/PWscf/PW -I/ho > me/sprokuda/PWscf/PH -c cft_3.f90 > external subroutine CFT_3 > > C_POINTER, save :: plan(2,2) > ^ > Error 24 at (43:cft_3.f90) : syntax error you replied : which version of the code are you using ? the one on the web, the one or cvs,...? C_POINTER is defined in include/machine.h I've modified machine.h as following : !#if defined(CRAYY) || defined(SGI) || defined(PC) || defined(AIX) || !defined(T3D) //||defined(HITACHI) || defined(SUN) #define C_POINTER integer !#endif it seems my Linux doesn't understoond "PC", and now I have following after issuing "make pw" : cft_3.o: In function `cft_3_': cft_3.o(.text+0xea): undefined reference to `fftw3d_f77_create_plan_' cft_3.o(.text+0x118): undefined reference to `fftwnd_f77_one_' cft_fftw.o: In function `cft_1_': cft_fftw.o(.text+0x5d): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x80): undefined reference to `fftw_f77_' cft_fftw.o: In function `cft_1s_': cft_fftw.o(.text+0x159): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x178): undefined reference to `fftw_f77_' cft_fftw.o: In function `cft_2_': cft_fftw.o(.text+0x2db): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x30e): undefined reference to `fftw_f77_' cft_fftw.o(.text+0x33b): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x392): undefined reference to `fftw_f77_' cft_fftw.o(.text+0x433): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x48a): undefined reference to `fftw_f77_' cft_fftw.o(.text+0x4c9): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x4fc): undefined reference to `fftw_f77_' cft_fftw.o: In function `cft_2s_': cft_fftw.o(.text+0x68d): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x6c2): undefined reference to `fftw_f77_' cft_fftw.o(.text+0x6f2): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x750): undefined reference to `fftw_f77_' cft_fftw.o(.text+0x7f3): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x851): undefined reference to `fftw_f77_' cft_fftw.o(.text+0x889): undefined reference to `fftw_f77_create_plan_' cft_fftw.o(.text+0x8be): undefined reference to `fftw_f77_' make[1]: *** [all] Error 1 make[1]: Leaving directory `/home/sprokuda/PWscf/PW' make: *** [pw] Error 2 I was looking for routines and files with name fftw_f77_*, but I've found nothing (except fftw.f90 in Modules, but there is nothing relevant). what could be the reason for those errors ? Best wishes, Sergey Prokudaylo. From taner at nist.gov Sun Aug 17 00:27:21 2003 From: taner at nist.gov (Taner Yildirim) Date: Sat, 16 Aug 2003 18:27:21 -0400 Subject: [Pw_forum] Re: LDA+U References: Message-ID: <002f01c36445$9178cd00$94790681@tanlap> Dear Matteo, Thank you very much for your thesis, which will be very usefull. I have one quick question: Don't we need to input Racah parameters (which can be taken from atomic limits) for U? In PWSCF, what does U mean, an average Coulomb interaction?? What about the J??? Thanks again for your help, Best Taner ******************************************************************** Dr. Taner Yildirim, Ph.D. Physicist, Condensed Matter Science Group, taner at nist.gov NIST Center for Neutron Research, PHONE: 301-975-6228 National Institute of Standards & Technology FAX : 301-921-9847 Gaithersburg, MD 20899-8562 Web : http://webster.ncnr.nist.gov/staff/taner ******************************************************************** ----- Original Message ----- From: "Matteo Cococcioni" To: "Taner Yildirim" Cc: "Eyvaz Isaev" ; ; "Goranka Bilalbegovic" ; "Ponniah Ravindran" Sent: Thursday, August 14, 2003 6:07 PM > > Dear all > > I'm sending you a copy of my PhD thesis as attachment to this message. > The original part of it is the method to compute the Hubbard U which we > used to study some real systems (compounds of iron). The results presented > in my thesis are not all correct (in particular the structural > properties of FeO) because we realized there were some > mistakes after my defense and corrected them afterwards. Unfortunately the > paper with the correct results is not ready yet. Anyway the method > described in the thesis is the one we actually used. > Let me know if you have questions or problems with LDA+U. > > best, > > Matteo > From giannozz at nest.sns.it Mon Aug 18 11:39:42 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 18 Aug 2003 11:39:42 +0200 Subject: [Pw_forum] FFTW problem In-Reply-To: <1661359247.20030815112338@dnk.ru> References: <002f01c36186$acd10fc0$ca841dc3@gost> <200308142238.20237.giannozz@nest.sns.it> <1661359247.20030815112338@dnk.ru> Message-ID: <200308181139.42640.giannozz@nest.sns.it> On Friday 15 August 2003 09:23, sprokuda at dnk.ru wrote: > it seems my Linux doesn't understood "PC" You have to supply "PC" as a precompilation option. Please follow the instructions THAT REFER TO THE VERSION OF THE CODE THAT YOU ARE USING, in particular those contained in the sample makefiles (install/Make.pc_ifc) and in the manual (pwdocs/manual.tex) > cft_3.o(.text+0xea): undefined reference to > `fftw3d_f77_create_plan_' [...] this problem has been discussed several times in the mailing list. You have to - link to fftw library, and - compile with the appropriate number of underscores, as explained in the manual Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From giannozz at nest.sns.it Mon Aug 18 12:13:55 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 18 Aug 2003 12:13:55 +0200 Subject: [Pw_forum] q2r.x In-Reply-To: References: Message-ID: <200308181213.55379.giannozz@nest.sns.it> On Thursday 14 August 2003 10:47, Vasse chis wrote: > Could somebody tell me if one have made phonon calculations on a > not-generated Monkhorst-Pack grid what is a "not-generated" Monkhorst-Pack grid ?!? > how the points of the grid could be inserted in the q2r.x input file. > thus, instead of nr1=#,nr2=#,nr3=#, if one could insert a list of k-points? if your "not-generated" grid corresponds to a "generated" (?) Monkhorst-Pack grid of given nr1, nr2, nr3, you can specify nr1, nr2, nr3, and give a list of dynamical matrices, the one for k=0 first, those for other k-vectors - only those in the irreducible Brillouin Zone - following in any order. Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From matteoc at MIT.EDU Mon Aug 18 17:25:56 2003 From: matteoc at MIT.EDU (Matteo Cococcioni) Date: Mon, 18 Aug 2003 11:25:56 -0400 (EDT) Subject: [Pw_forum] LDA+U in PWSCF In-Reply-To: Message-ID: Dear Jiang-long Wang the full procedure to compute the Hubbard parameters (i.e. to process the data coming from the linear response study with pwscf) is explained in my thesis in full detail. To obtain the necesary data using pwscf do the following: 1) lda (Hubbard_U = Hubbard_alpha = 0) calculation up to scf 2) put Hubbard_alpha =/= 0 on ONE of the d ions (it needs to be inequivalent to others, so you have to describe it as of different kind thus breaking some possible symmetries) and, starting from the wavefunction and the scf potential obtained in 1) (startingwfc = 'file', startingpot = 'file') you make a run up to scf. Remember to lower the threshold for the diagonalizaion of the Hamiltonian (ethr) to the value used in the last iteration of the run in 1). Probably you will have to add ethr among the input variables and make the code read it from the input namelist. This is because you require high precision also at the first iteration of the perturbed run. 3) repeat 2) for several values of Hubbard_alpha around 0 (typical |values| are from 0.01 to 0.1 eV) checking a linear behaviour of the d state occupations with respect to Hubbard_alpha expecially around 0. 4) repeat 2) and 3) for every non equivalent d-ion in your unit cell. 5) built the matrices \chi and \chi_0 and compute U from them. 6) repeat everything in larger and larger supercells up to convergence of the U This procedure guarantees plenty consideration of the screening of the Hubbard parameter so that U should not be overstimated anymore. Good luck, Matteo On Mon, 18 Aug 2003 jlwang at theory.issp.ac.cn wrote: > > Dear Dr. Matteo, > > Thank you very much for your thesis ! > > Could you give me a example or explanation to calculate > Hubbardparameter with PWSCF code. For example, the vale of > Hubbard_alpha, how to calculate the local dos. > > I used to calculate the Hubbard parameter with TB-LMTO method by > constrained LDA calculation. The U parameter is usually overestimated for > metal e.g 3d transition metal series, it is hard to take the screening > effect into account correctly. > > I want do some calculation with the LDA+U utility in PWSCF code. > -- > Yours sincerely, > > Jiang-long Wang > > ................................. > > Institute of Solid State Physics, > Academia Sinica, P. O. Box 1129, > 230031 Hefei, Anhui, P. R. China > > Tel: 86-551-5591464 > Fax: 86-551-5591434 > > E-mail: jlwang at theory.issp.ac.cn > > > From sprokuda at dnk.ru Mon Aug 18 19:10:09 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Mon, 18 Aug 2003 21:10:09 +0400 Subject: [Pw_forum] FFTW problem In-Reply-To: <200308181139.42640.giannozz@nest.sns.it> References: <002f01c36186$acd10fc0$ca841dc3@gost> <200308142238.20237.giannozz@nest.sns.it> <1661359247.20030815112338@dnk.ru> <200308181139.42640.giannozz@nest.sns.it> Message-ID: <741828921.20030818211009@dnk.ru> Dear Paolo ! I'm really sorry that I'm asking the trivial question that has been asked several times. The thing is that I can't find nothing with names like fftw_f77_...so...on neither in MKL6.0 no in ATLAS. you see - when I issue : [root at sprokuda PWscf]# nm /home/sprokuda/ATLAS/lib/Linux_PII/*.a | grep T | grep -i fftw_f77 [root at sprokuda PWscf]# nm /opt/intel/mkl60/lib/32/*.a | grep T | grep -i fftw_f77 I get on both nothing. ( /home/sprokuda/ATLAS/lib/Linux_PII/ - ATLAS libs, /opt/intel/mkl60/lib/32/ - MKL libs ), but when issue : [root at sprokuda PWscf]# nm /opt/intel/mkl60/lib/32/*.a | grep T | grep -i dcopy 00000000 T __MKL_BLAS_def_dcopy 00000000 T __MKL_BLAS_dcopy 00000000 T __MKL_BLAS_p3_dcopy 00000000 T __MKL_BLAS_p4_dcopy 00000000 T __MKL_BLAS_p4_dcopycr0 00000000 T __MKL_BLAS_p4_dcopycr1 00000000 T __MKL_BLAS_p4_dcopycrb 00000000 T __MKL_BLAS_F77_DCOPY 00000000 T DCOPY 00000000 T dcopy 00000000 T dcopy_ 00000000 T cblas_dcopy or : [root at sprokuda PWscf]# nm /home/sprokuda/ATLAS/lib/Linux_PII/*.a | grep T | grep -i dcopy U ATL_dcopy U ATL_dcopy ATL_dcopy.o: 00000000 T ATL_dcopy U ATL_dcopy_xp0yp0aXbX U ATL_dcopy_xp1yp1aXbX ATL_dcopy_xp1yp1aXbX.o: 00000000 T ATL_dcopy_xp1yp1aXbX ATL_dcopy_xp0yp0aXbX.o: 00000000 T ATL_dcopy_xp0yp0aXbX U ATL_dcopy U ATL_dcopy U ATL_dcopy U ATL_dcopy U ATL_dcopy U ATL_dcopy U ATL_dcopy U ATL_dcopy U ATL_dcopy U ATL_dcopy 00000000 T cblas_dcopy 00000000 T dcopy_ ATL_F77wrap_dcopy.o: U ATL_dcopy 00000000 T atl_f77wrap_dcopy__ - that is clear even for such a horrible Linux looser like me, that there are no function with names I need neither in MKL , no in ATLAS. Probably the thing is that fftw_f77 routines should be found in some other place ? if yes, where ? You have to supply "PC" as a precompilation option. I have it, and this problem is solved - thank you. Please follow the instructions THAT REFER TO THE VERSION OF THE CODE THAT YOU ARE USING, in particular those contained in the sample makefiles (install/Make.pc_ifc) and in the manual (pwdocs/manual.tex) > cft_3.o(.text+0xea): undefined reference to > `fftw3d_f77_create_plan_' [...] this problem has been discussed several times in the mailing list. I'm really apologizing , but I guess I'm subscribed for not so long time - I've found nothig ... You have to - link to fftw library, and yes, that's the point ... - compile with the appropriate number of underscores, as explained in the manual I've tried with one, two and three - no result... Sorry for thre distorsion... Sergey Prokudaylo. From baroni at sissa.it Mon Aug 18 19:27:06 2003 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 18 Aug 2003 19:27:06 +0200 Subject: [Pw_forum] FFTW problem In-Reply-To: <741828921.20030818211009@dnk.ru> References: <002f01c36186$acd10fc0$ca841dc3@gost> <200308142238.20237.giannozz@nest.sns.it> <1661359247.20030815112338@dnk.ru> <200308181139.42640.giannozz@nest.sns.it> <741828921.20030818211009@dnk.ru> Message-ID: <3F410C6A.40401@sissa.it> sprokuda at dnk.ru wrote: > Dear Paolo ! > > - that is clear even for such a horrible Linux looser like me, that > there are no function with names I need neither in MKL , no in > ATLAS. Probably the thing is that fftw_f77 routines should be found > in some other place ? if yes, where ? http://www.fftw.org Stefano Baroni > > You have to supply "PC" as a precompilation option. > > I have it, and this problem is solved - thank you. > > Please follow the instructions THAT REFER TO THE > VERSION OF THE CODE THAT YOU ARE USING, > in particular those contained in the sample makefiles > (install/Make.pc_ifc) and in the manual (pwdocs/manual.tex) > > >>cft_3.o(.text+0xea): undefined reference to >>`fftw3d_f77_create_plan_' [...] > > > this problem has been discussed several times in the > mailing list. > > I'm really apologizing , but I guess I'm subscribed for not so long > time - I've found nothig ... > > You have to > - link to fftw library, and > > yes, that's the point ... > > - compile with the appropriate number of underscores, > as explained in the manual > > I've tried with one, two and three - no result... > > Sorry for thre distorsion... > > Sergey Prokudaylo. > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From aaron at nemo.physics.ncsu.edu Fri Aug 22 14:07:05 2003 From: aaron at nemo.physics.ncsu.edu (aaron at nemo.physics.ncsu.edu) Date: Fri, 22 Aug 2003 08:07:05 -0400 (EDT) Subject: [Pw_forum] Intel compile pw.1.1.2 problem Message-ID: I'm having a compile problem, an Intel machine under Linux, with pw.1.2.0. The error is similar to that of posting by Mahmoud Payami on pw_forum ( http://www.democritos.it/pipermail/pw_forum/2002-March/subject.html ) for march 2002 (unfortunately I have not seen a reply in this thread). Thanks in advance for any help you can provide. Aaron From giannozz at nest.sns.it Fri Aug 22 16:12:20 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 22 Aug 2003 16:12:20 +0200 Subject: [Pw_forum] Intel compile pw.1.1.2 problem In-Reply-To: References: Message-ID: <200308221612.20418.giannozz@nest.sns.it> On Friday 22 August 2003 14:07, aaron at nemo.physics.ncsu.edu wrote: > I'm having a compile problem, an Intel machine under Linux, with > pw.1.2.0. The error is similar to that of posting by Mahmoud Payami > on pw_forum [...] for march 2002 so the reason should also be similar: > ifc: Command line warning: ignoring unknown option '-nomodule' option "-nomodule" is for ifc v. 7 and later. For ifc version 6 and earlier, use MODULEFLAG= -cl,./intel.pcl instead of MODULEFLAG= -nomodule -I $(OSHOME)/Modules -I$(OSHOME)/PW -I$(OSHOME)/PH in make.sys -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From yanming_ma at hotmail.com Mon Aug 25 19:43:54 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Tue, 26 Aug 2003 01:43:54 +0800 Subject: [Pw_forum] Input atomic positions for alpha_Ga structure Message-ID: Dear PWSCF users I have a problem when I tried to input the atomic positions for alpha_Ga structure. The space group is Cmca in Orthorhombic Cell. The Wyckoff positions for Ga atoms are in 8f. The 8f positions are the following from the international Crystallographic Table in conventional settings. (0, 0, 0) + (1/2, 1/2, 0) (0, y, z) (0, - y + 1/2 , z + 1/2 ) (0, y + 1/2 , - z + 1/2 ) (0, - y, - z) y=0.158 z=0.081 I tried to transform the 8f positions in conventional setting with 8 positions to a primitive cell with 4 positions using the following transformation matix. 0.5 0.5 0.0 -0.5 0.5 0.0 0.0 0.0 1.0 ,instead of the bravias lattice for Base centered Orthorhombic structure in PWSCF, a/2 b/c 0.0 -a/2 b/2 0.0 0.0 0.0 c The resulting positions are the following ( - y, y, z) (y + 1/2 , - y + 1/2 , z + 1/2 ) ( - y + 1/2 , y + 1/2 , - z + 1/2 ) (y, - y, - z) I input the above four positions into PWSCF. I do find 8 system operations. But the calculated bandstructure is different with the one calculated with Full potential Code, WIEN2K. My question is if my input atomic positions are wrong? if so, what is the correct input for Orthorhombic cases. If someone has the successful experiences in Orthorhombic structure calculation using PWSCF, please share with me. I will highly appreciate your help. Best Regards Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 _________________________________________________________________ ??????????????? MSN Hotmail? http://www.hotmail.com From g930102 at ccit.edu.tw Wed Aug 27 04:21:33 2003 From: g930102 at ccit.edu.tw (g930102 at ccit.edu.tw) Date: Wed, 27 Aug 2003 10:21:33 +0800 (CST) Subject: [Pw_forum] PWscf(1.2) work well in Windows platform Message-ID: <200308270221.h7R2LXmj039518@smtp.ccit.edu.tw> Dear PWSCF users I have compiled an Intel machine under Windows 2000, with pw.1.2.0. Here is my configurations of Hardware and Software. PC Machine: Intel P-3 Operation System: Microsoft Windows 2000 sp4 Unixlike Environment in Windows: cygwin Compiler: Intel(R) Fortran Compiler, Ver 7.1 Build 20030801z Compiler: Intel(R) C++ Compiler, Ver 7.1 Build 20030801z MKL Library: Intel? Math Kernel Library, Ver6.1 FFT Library: FFTW package(http://www.fftw.org/), FFTW 2.1.5.(http://www.fftw.org/fftw-2.1.5.tar.gz) The content of make.sys as follows, ------------------------------------- # OSHOME=/home/LuFF/PWscf # # System-dependent Make definitions for cygwin in Windows platform and Intel compiler (v.7) # Here just for trial in win32 porting # # add -DMKL if using the Intel Mathematical Kernel Library, Ver 6.1 # add -DFFTW if using the FFTW library, Ver. 2.1.5 or above # CPP = /bin/cpp CPPFLAGS = -I$(OSHOME)/include/ -DPC -DMKL -DFFTW \ -D"FFTWND_F77_ONE=fftwnd_f77_one" \ -D"FFTW3D_F77_CREATE_PLAN=fftw3d_f77_create_plan" \ -D"FFTW_F77=fftw_f77" \ -D"FFTW_F77_CREATE_PLAN=fftw_f77_create_plan" # # Fortran compiler: # F90=/cygdrive/c/Progra~1/Intel/Compiler70/IA32/Bin/ifl.exe # # fortran options: # /4Yportlib enables the use of portability libraries (getarg # and maybe other library calls) # /w disable warnings # /O3 more aggressive optimization # /G6 Pentium Pro, 2, 3 optimization # /G7 Pentium 4 optimization # FFLAGS =/4Yportlib /G6 /w /O3 /I../include/ F90FLAGS= $(FFLAGS) /fpp $(CPPFLAGS) # # This is needed to tell the compiler where modules are # use this for version = 7 MODULEFLAG= -nomodule -I $(OSHOME)/Modules -I$(OSHOME)/PW -I$(OSHOME)/PH # # Loader: # # The following is for Intel MKL libraries v. 6.1 on Pentium-3 or above # # Here we use the fftw-2.1.5, which compiled by Intel compiler(v.7) # LIBS= ../libfftw.lib ../mkl_c.lib # # Linker: # LD=/cygdrive/c/Progra~1/Intel/Compiler70/IA32/Bin/icl.exe LFLAGS = /4Yportlib $(LIBS) # # Library Tools: # ar: # AR = ar ARFLAGS = rv ------------------------------------- I have run all examples, It's seems ok and work well. It may be help for someone who need to try it in windows paltform. Best Regards ======================================================= Lu Fu-Fa, Doctoral Student, E-mail: g930102 at ccit.edu.tw Semiconductors Lab., Dept. of Applied Physics Chung Cheng Institute of Technology (CCIT), Taiwan ======================================================= ___________________________________ NOCC, http://nocc.sourceforge.net From zyli at sina.com Wed Aug 27 16:53:51 2003 From: zyli at sina.com (Zhenyu Li) Date: Wed, 27 Aug 2003 22:53:51 +0800 Subject: [Pw_forum] error information Message-ID: <200308271454.WAA22310@mx1.ustc.edu.cn> Dear PWSCF users When using PWSCF 1.2.0 do some geometry optimization, I got the following error message: ------------------------------------------------------------------------------------- iteration # 1 ecut= 24.00 ryd beta=0.30 Davidson diagonalization with overlap %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from cdiaghg : error # 730 info =/= 0 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... -------------------------------------------------------------------------------------- What's the problem of my calculations? Best wishes Zhenyu LI ================================================= Zhenyu LI Ph.D. Candidate Open Laboratory of Bond Selective Chemistry, University of Science and Technology of China, Hefei, Anhui, 230026, People's Republic of China Tel.: 86-551-3603748,3603418 Fax.: 86-551-3602969 http://www.bsc.ustc.edu.cn/~zyli/whoami/ ===================================================?????????????? From eyvaz_isaev at yahoo.com Wed Aug 27 17:15:20 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 27 Aug 2003 08:15:20 -0700 (PDT) Subject: [Pw_forum] error information In-Reply-To: <200308271454.WAA22310@mx1.ustc.edu.cn> Message-ID: <20030827151520.72140.qmail@web80508.mail.yahoo.com> Dear Zhenyu, The error was discussed early in the forum. It means that the diagonalization procedure was not converged within machine-specific precision. You can slightly change your input file (Ecut and Ecutrho) and the error should disappear. Regards, Eyvaz Isaev, Theoretical Physics Department Moscow State Intitute of Steel and Alloys --- Zhenyu Li wrote: > Dear PWSCF users > > When using PWSCF 1.2.0 do some geometry > optimization, > I got the following error message: > ------------------------------------------------------------------------------------- > iteration # 1 ecut= 24.00 ryd > beta=0.30 > Davidson diagonalization with overlap > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from cdiaghg : error # 730 > info =/= 0 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > -------------------------------------------------------------------------------------- > What's the problem of my calculations? > > Best wishes > Zhenyu LI > > > > ================================================= > Zhenyu LI > Ph.D. Candidate > Open Laboratory of Bond Selective Chemistry, > University of Science and Technology of China, > Hefei, Anhui, 230026, > People's Republic of China > Tel.: 86-551-3603748,3603418 > Fax.: 86-551-3602969 > http://www.bsc.ustc.edu.cn/~zyli/whoami/ > ===================================================???????????????????????????? > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From giannozz at nest.sns.it Wed Aug 27 18:55:32 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 27 Aug 2003 18:55:32 +0200 Subject: [Pw_forum] error information In-Reply-To: <20030827151520.72140.qmail@web80508.mail.yahoo.com> References: <20030827151520.72140.qmail@web80508.mail.yahoo.com> Message-ID: <200308271855.32529.giannozz@nest.sns.it> On Wednesday 27 August 2003 17:15, Eyvaz Isaev wrote: > Dear Zhenyu, > > The error was discussed early in the forum [...] > You can slightly change your input file (Ecut and > Ecutrho) and the error should disappear. unfortunately it is not that simple. This is an error that should never happen. If it does, something is seriously wrong. This is what the manual says: - serious error in data, such as bad atomic positions or bad crystal structure/supercell; - a bad PP (for instance, with a ghost); - IBM SP3: under some circumstances (typically a large number of k-points) we get an error in cdiaghg that is reproducible but disappears if we change anything in the calculation. We don't know what happens and why. Try to use conjugate-gradient diagonalization (diagonalization='cg'). - HP-Compaq alphas with cxml libraries: try to use compiled Blas/Lapack (or better, Atlas) instead of those contained in cxml (just load them before cxml). It might also be a consequence of the following case: - Structural optimization goes wild after first or second step. The algorithm used in structural optimization is not very robust. If you start too far away from minimum, it may lead to badly wrong atomic positions. Restart from a better starting point. Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From yanming_ma at hotmail.com Wed Aug 27 20:30:02 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Thu, 28 Aug 2003 02:30:02 +0800 Subject: [Pw_forum] error information Message-ID: Dear Zhenyu, I also met the same problem when I did the optimization with variable cell using PWSCF. You can read the PWSCF manual, this is what Paolo replied you. For my experiences, one main factor for this problem is that your starting structure is far from the final results. In this case, you may not choose the option "potential_extrapolation='wfc2'". Just use the default value. For my case, I tried to change the ecut and ecutrho. But it did not help if I start with a crazy structure and with potential_extrapolation='wfc2'on. So if you can not guess the starting structure, it will help if you do not choose the option "potential_extrapolation='wfc2'". Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 >From: "Zhenyu Li" >Reply-To: pw_forum at pwscf.org >To: "pw_forum at pwscf.org" >Subject: [Pw_forum] error information >Date: Wed, 27 Aug 2003 22:53:51 +0800 > >Dear PWSCF users > > When using PWSCF 1.2.0 do some geometry optimization, >I got the following error message: >------------------------------------------------------------------------------------- > iteration # 1 ecut= 24.00 ryd beta=0.30 > Davidson diagonalization with overlap > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from cdiaghg : error # 730 > info =/= 0 > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... >-------------------------------------------------------------------------------------- >What's the problem of my calculations? > >Best wishes >Zhenyu LI > > > >================================================= > Zhenyu LI > Ph.D. Candidate > Open Laboratory of Bond Selective Chemistry, > University of Science and Technology of China, > Hefei, Anhui, 230026, > People's Republic of China > Tel.: 86-551-3603748,3603418 > Fax.: 86-551-3602969 > http://www.bsc.ustc.edu.cn/~zyli/whoami/ >===================================================???????????? ?? > > > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum _________________________________________________________________ ???? MSN Explorer: http://explorer.msn.com/lccn From zyli at sina.com Thu Aug 28 03:37:42 2003 From: zyli at sina.com (Zhenyu Li) Date: Thu, 28 Aug 2003 9:37:42 +0800 Subject: [Pw_forum] error information Message-ID: <200308280138.JAA20815@mx1.ustc.edu.cn> Dear All, Thank all of you for your kindly help. I will try again according to your advices. Best wishes Zhenyu LI ======= 2003-08-27 18:55:00 ????????======= >On Wednesday 27 August 2003 17:15, Eyvaz Isaev wrote: >> Dear Zhenyu, >> >> The error was discussed early in the forum [...] >> You can slightly change your input file (Ecut and >> Ecutrho) and the error should disappear. > >unfortunately it is not that simple. This is an error >that should never happen. If it does, something >is seriously wrong. This is what the manual says: > - serious error in data, such as bad atomic positions or bad > crystal structure/supercell; > - a bad PP (for instance, with a ghost); > - IBM SP3: under some circumstances (typically a large number > of k-points) we get an error in cdiaghg that is reproducible > but disappears if we change anything in the calculation. > We don't know what happens and why. Try to use > conjugate-gradient diagonalization (diagonalization='cg'). > - HP-Compaq alphas with cxml libraries: try to use > compiled Blas/Lapack (or better, Atlas) instead of those > contained in cxml (just load them before cxml). >It might also be a consequence of the following case: > - Structural optimization goes wild after first or second step. > The algorithm used in structural optimization is not very robust. > If you start too far away from minimum, it may lead to badly > wrong atomic positions. Restart from a better starting point. >Paolo >-- >Paolo Giannozzi e-mail: giannozz at nest.sns.it >Scuola Normale Superiore Phone: +39/050509412 >Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 >I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > >. = = = = = = = = = = = = = = = = = = = = ????????? ?? ????????Zhenyu Li ????????zyli at sina.com ??????????2003-08-28 ================================================= Zhenyu LI Ph.D. Candidate Open Laboratory of Bond Selective Chemistry, University of Science and Technology of China, Hefei, Anhui, 230026, People's Republic of China Tel.: 86-551-3603748,3603418 Fax.: 86-551-3602969 http://www.bsc.ustc.edu.cn/~zyli/whoami/ ===================================================?????????????? From zyli at sina.com Thu Aug 28 14:32:34 2003 From: zyli at sina.com (Zhenyu Li) Date: Thu, 28 Aug 2003 20:32:34 +0800 Subject: [Pw_forum] error information Message-ID: <200308281233.UAA01229@mx1.ustc.edu.cn> Dear All, By checking my input files, I found the most probably problem may be the ecutrho problem. But after increasing the values of ecut and ecutrho, I get another error: --------------------------------------------------------------------------------- Total force = 46.997979 Total SCF correction = 0.000555 searching for next position (pslinmin)... Eold = -1031.97922186 Etot = -1035.35596607 DEold = -13.55214176 DEtot = -11.39808412 linmin: 3rd order interpolation %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from linmin : error # 2 unexpected error %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... ----------------------------------------------------------------------------------- How to solve this problem now? Zhenyu Li ======= 2003-08-28 09:37:00 ????????======= >Dear All, > > Thank all of you for your kindly help. I will try >again according to your advices. > >Best wishes >Zhenyu LI > > > >======= 2003-08-27 18:55:00 ????????======= > >>On Wednesday 27 August 2003 17:15, Eyvaz Isaev wrote: >>> Dear Zhenyu, >>> >>> The error was discussed early in the forum [...] >>> You can slightly change your input file (Ecut and >>> Ecutrho) and the error should disappear. >> >>unfortunately it is not that simple. This is an error >>that should never happen. If it does, something >>is seriously wrong. This is what the manual says: >> - serious error in data, such as bad atomic positions or bad >> crystal structure/supercell; >> - a bad PP (for instance, with a ghost); >> - IBM SP3: under some circumstances (typically a large number >> of k-points) we get an error in cdiaghg that is reproducible >> but disappears if we change anything in the calculation. >> We don't know what happens and why. Try to use >> conjugate-gradient diagonalization (diagonalization='cg'). >> - HP-Compaq alphas with cxml libraries: try to use >> compiled Blas/Lapack (or better, Atlas) instead of those >> contained in cxml (just load them before cxml). >>It might also be a consequence of the following case: >> - Structural optimization goes wild after first or second step. >> The algorithm used in structural optimization is not very robust. >> If you start too far away from minimum, it may lead to badly >> wrong atomic positions. Restart from a better starting point. >>Paolo >>-- >>Paolo Giannozzi e-mail: giannozz at nest.sns.it >>Scuola Normale Superiore Phone: +39/050509412 >>Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 >>I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 >> >>_______________________________________________ >>Pw_forum mailing list >>Pw_forum at pwscf.org >>http://www.democritos.it/mailman/listinfo/pw_forum >> >> >>. > >= = = = = = = = = = = = = = = = = = = = > > >????????? >?? > > >????????Zhenyu Li >????????zyli at sina.com >??????????2003-08-28 > >================================================= > Zhenyu LI > Ph.D. Candidate > Open Laboratory of Bond Selective Chemistry, > University of Science and Technology of China, > Hefei, Anhui, 230026, > People's Republic of China > Tel.: 86-551-3603748,3603418 > Fax.: 86-551-3602969 > http://www.bsc.ustc.edu.cn/~zyli/whoami/ >===================================================?????????????? > > > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum = = = = = = = = = = = = = = = = = = = = ????????? ?? ????????Zhenyu Li ????????zyli at sina.com ??????????2003-08-28 ================================================= Zhenyu LI Ph.D. Candidate Open Laboratory of Bond Selective Chemistry, University of Science and Technology of China, Hefei, Anhui, 230026, People's Republic of China Tel.: 86-551-3603748,3603418 Fax.: 86-551-3602969 http://www.bsc.ustc.edu.cn/~zyli/whoami/ ===================================================?????????????? From giannozz at nest.sns.it Thu Aug 28 15:54:23 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 28 Aug 2003 15:54:23 +0200 Subject: [Pw_forum] error information In-Reply-To: <200308281233.UAA01229@mx1.ustc.edu.cn> References: <200308281233.UAA01229@mx1.ustc.edu.cn> Message-ID: <200308281554.23440.giannozz@nest.sns.it> On Thursday 28 August 2003 14:32, Zhenyu Li wrote: > [...] after increasing the values of ecut and ecutrho, I get another error: > [...] from linmin : error # 2 > unexpected error >How to solve this problem now? see my preceding message, towards the end: - Structural optimization goes wild after first or second step. The algorithm used in structural optimization is not very robust. If you start too far away from minimum, it may lead to badly wrong atomic positions. Restart from a better starting point. and also (from the next version of the manual): - Structural optimization stops in linmin with "unexpected error". The line minimization algorithm (linmin) looks for a minimum -- using atomic positions and forces at the current and preceding step, and 3rd- or 2nd-degree polynomial interpolation -- along a given direction (determined by the BFGS algorithm). If this minimum turns out to be where it shouldn't be (for instance, on the wrong side with respect to forces), the code stops with "unexpected error". In practice, this means that there is an inconsistency between forces and energies Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From yanming_ma at hotmail.com Thu Aug 28 17:00:38 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Thu, 28 Aug 2003 23:00:38 +0800 Subject: [Pw_forum] for q2r.x errors Message-ID: Dear All, When I tried to calculate the force constant in Real space using q2r.x, I met the following error message: reading dyn.mat. from file Case_3.dyn nqs= 2 q= 0.00000000 0.00000000 0.82109896 q= 0.00000000 0.00000000 -0.82109896 1 1 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from init : error # 1 nc already filled %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... I know if one chooses the wrong q points set to calculate the force constant, then PWSCF will find the duplications.The error message will occur as above. But I am confident that mine is not this case. My question is If my input atomic positions are wrong,will this error message occur? I still doubt about the input atomic coordinate for C-based Orthorhombic unit cell as my previous email. But for that email, nobody reply me. Does anyone have the successful experiences in calculating the phonon in C_based orthorhombic unit cell? Any Comment? Thanks in advance. Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 _________________________________________________________________ ???? MSN Explorer: http://explorer.msn.com/lccn/ From bungaro at physics.rutgers.edu Thu Aug 28 17:24:31 2003 From: bungaro at physics.rutgers.edu (CLAUDIA BUNGARO) Date: Thu, 28 Aug 2003 11:24:31 -0400 (EDT) Subject: [Pw_forum] for q2r.x errors In-Reply-To: References: Message-ID: Dear Yanming, How did you define your unit cell in the input for the pw.x runs? If you used ibrav=0 and then specified the lattice vectors using: CELL_PARAMETERS { cubic | hexagonal } a(1,1) a(2,1) a(3,1) a(1,2) a(2,2) a(3,2) a(1,3) a(2,3) a(3,3) I am afraid this is the source of your problem running q2r.x The q2r program, as it is now written, reads in input dynamical matrices generated using ibrav=/=0 and uses the corresponding "automatic" choice for the lattice vector (generated by latgen). best regards, Claudia __________________________________________________ Dr. Claudia Bungaro Dept. of Physics and Astronomy, Rutgers University 136 Frelinghuysen Road, Piscataway, NJ 08854-8019 tel: +1 732 445 4197 fax: +1 732 445 4343 email: bungaro at physics.rutgers.edu __________________________________________________ On Thu, 28 Aug 2003, ma Yanming wrote: > Dear All, > > When I tried to calculate the force constant in Real space using q2r.x, I > met the following error message: > > reading dyn.mat. from file Case_3.dyn > nqs= 2 > q= 0.00000000 0.00000000 0.82109896 > q= 0.00000000 0.00000000 -0.82109896 > 1 1 2 > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from init : error # 1 > nc already filled > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > > I know if one chooses the wrong q points set to calculate the force > constant, then PWSCF will find the duplications.The error message will > occur as above. But I am confident that mine is not this case. > > My question is If my input atomic positions are wrong,will this error > message occur? > I still doubt about the input atomic coordinate for C-based Orthorhombic > unit cell as my previous email. But for that email, nobody reply me. > Does anyone have the successful experiences in calculating the phonon in > C_based orthorhombic unit cell? > > Any Comment? > > Thanks in advance. > > Yanming Ma PhD > Steacie Institute for Molecular Sciences, > National Research Councils of Canada. > 100 Sussex > K1A 0R6 > > _________________________________________________________________ > ???? MSN Explorer: http://explorer.msn.com/lccn/ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From yanming_ma at hotmail.com Thu Aug 28 17:52:15 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Thu, 28 Aug 2003 23:52:15 +0800 Subject: [Pw_forum] error information Message-ID: Dear Zhenyu, > By checking my input files, I found the most probably problem >may be the ecutrho problem. But after increasing the values of ecut >and ecutrho, I get another error: I do not think your problem is from ecut and ecutrho. The smaller ecut will result in inaccurate results, but not as serious as it stops running... >--------------------------------------------------------------------------------- >Total force = 46.997979 Total SCF correction = 0.000555 Your total force is getting crazy. Are you sure your input is correct? Send your input file. Yanming Ma > > >======= 2003-08-28 09:37:00 ????????======= > > >Dear All, > > > > Thank all of you for your kindly help. I will try > >again according to your advices. > > > >Best wishes > >Zhenyu LI > > > > > > > >======= 2003-08-27 18:55:00 ????????======= > > > >>On Wednesday 27 August 2003 17:15, Eyvaz Isaev wrote: > >>> Dear Zhenyu, > >>> > >>> The error was discussed early in the forum [...] > >>> You can slightly change your input file (Ecut and > >>> Ecutrho) and the error should disappear. > >> > >>unfortunately it is not that simple. This is an error > >>that should never happen. If it does, something > >>is seriously wrong. This is what the manual says: > >> - serious error in data, such as bad atomic positions or bad > >> crystal structure/supercell; > >> - a bad PP (for instance, with a ghost); > >> - IBM SP3: under some circumstances (typically a large number > >> of k-points) we get an error in cdiaghg that is reproducible > >> but disappears if we change anything in the calculation. > >> We don't know what happens and why. Try to use > >> conjugate-gradient diagonalization (diagonalization='cg'). > >> - HP-Compaq alphas with cxml libraries: try to use > >> compiled Blas/Lapack (or better, Atlas) instead of those > >> contained in cxml (just load them before cxml). > >>It might also be a consequence of the following case: > >> - Structural optimization goes wild after first or second step. > >> The algorithm used in structural optimization is not very robust. > >> If you start too far away from minimum, it may lead to badly > >> wrong atomic positions. Restart from a better starting point. > >>Paolo > >>-- > >>Paolo Giannozzi e-mail: giannozz at nest.sns.it > >>Scuola Normale Superiore Phone: +39/050509412 > >>Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 > >>I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 > >> > >>_______________________________________________ > >>Pw_forum mailing list > >>Pw_forum at pwscf.org > >>http://www.democritos.it/mailman/listinfo/pw_forum > >> > >> > >>. > > > >= = = = = = = = = = = = = = = = = = = = > > > > > >????????? > >?? > > > > > >????????Zhenyu Li > >????????zyli at sina.com > >??????????2003-08-28 > > > >================================================= > > Zhenyu LI > > Ph.D. Candidate > > Open Laboratory of Bond Selective Chemistry, > > University of Science and Technology of China, > > Hefei, Anhui, 230026, > > People's Republic of China > > Tel.: 86-551-3603748,3603418 > > Fax.: 86-551-3602969 > > http://www.bsc.ustc.edu.cn/~zyli/whoami/ > >===================================================??????????? ??? > > > > > > > >_______________________________________________ > >Pw_forum mailing list > >Pw_forum at pwscf.org > >http://www.democritos.it/mailman/listinfo/pw_forum > >= = = = = = = = = = = = = = = = = = = = > > >????????? >?? > > >????????Zhenyu Li >????????zyli at sina.com >??????????2003-08-28 > >================================================= > Zhenyu LI > Ph.D. Candidate > Open Laboratory of Bond Selective Chemistry, > University of Science and Technology of China, > Hefei, Anhui, 230026, > People's Republic of China > Tel.: 86-551-3603748,3603418 > Fax.: 86-551-3602969 > http://www.bsc.ustc.edu.cn/~zyli/whoami/ >===================================================???????????? ?? > > > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum _________________________________________________________________ ??????????????? MSN Hotmail? http://www.hotmail.com From yanming_ma at hotmail.com Thu Aug 28 20:57:58 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Fri, 29 Aug 2003 02:57:58 +0800 Subject: [Pw_forum] for q2r.x errors Message-ID: Dear Claudia, >How did you define your unit cell in the input for the pw.x runs? > >If you used ibrav=0 >and then specified the lattice vectors using: >CELL_PARAMETERS { cubic | hexagonal } > a(1,1) a(2,1) a(3,1) > a(1,2) a(2,2) a(3,2) > a(1,3) a(2,3) a(3,3) > >I am afraid this is the source of your problem running q2r.x > >The q2r program, as it is now written, reads in input dynamical >matrices generated using ibrav=/=0 and uses the corresponding >"automatic" choice for the lattice vector (generated by latgen). I did not input in this way as you stated. I check the reference and find that for Alpha_Ga structure, the spacegroup is Cmca and Gallium atoms are in the 8f Wyckoff position, and y=0.1549 and z=0.081. So I set ibrav=9. The following is my input file for PW.x run. cat >Ga.scf.in <best regards, >Claudia > >__________________________________________________ >Dr. Claudia Bungaro >Dept. of Physics and Astronomy, Rutgers University >136 Frelinghuysen Road, Piscataway, NJ 08854-8019 > >tel: +1 732 445 4197 fax: +1 732 445 4343 >email: bungaro at physics.rutgers.edu >__________________________________________________ > > >On Thu, 28 Aug 2003, ma Yanming wrote: > > > Dear All, > > > > When I tried to calculate the force constant in Real space using q2r.x, I > > met the following error message: > > > > reading dyn.mat. from file Case_3.dyn > > nqs= 2 > > q= 0.00000000 0.00000000 0.82109896 > > q= 0.00000000 0.00000000 -0.82109896 > > 1 1 2 > > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > from init : error # 1 > > nc already filled > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > stopping ... > > > > I know if one chooses the wrong q points set to calculate the force > > constant, then PWSCF will find the duplications.The error message will > > occur as above. But I am confident that mine is not this case. > > > > My question is If my input atomic positions are wrong,will this error > > message occur? > > I still doubt about the input atomic coordinate for C-based Orthorhombic > > unit cell as my previous email. But for that email, nobody reply me. > > Does anyone have the successful experiences in calculating the phonon in > > C_based orthorhombic unit cell? > > > > Any Comment? > > > > Thanks in advance. > > > > Yanming Ma PhD > > Steacie Institute for Molecular Sciences, > > National Research Councils of Canada. > > 100 Sussex > > K1A 0R6 > > > > _________________________________________________________________ > > ???? MSN Explorer: http://explorer.msn.com/lccn/ > > > > _______________________________________________ > > 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 _________________________________________________________________ ?????????????? MSN Messenger: http://messenger.msn.com/cn From giannozz at nest.sns.it Thu Aug 28 22:00:55 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 28 Aug 2003 22:00:55 +0200 Subject: [Pw_forum] for q2r.x errors In-Reply-To: References: Message-ID: <200308282200.56439.giannozz@nest.sns.it> On Thursday 28 August 2003 20:57, ma Yanming wrote: > [...] Unfortunately, PWSCF doesn't perform the checking function checking that the input corresponds to what you have in mind ? THAT would be a major achievement!!! PWscf just checks that the input makes sense. In the forthcoming new version, there is an auxiliary code to calculate distances and angles between atom. This is what I use to verify my input data. Paolo From yanming_ma at hotmail.com Fri Aug 29 01:23:33 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Fri, 29 Aug 2003 07:23:33 +0800 Subject: [Pw_forum] for q2r.x errors Message-ID: Dear Paolo >PWscf just checks that the input makes sense. In the forthcoming >new version, I am just wondering when the new version will come out. >there is an auxiliary code to calculate distances >and angles between atom. This is what I use to verify my input >data. That will be great if we can test what we input. One thing I am interested in is the very useful software "Xcrysden". Xcrysden can monitor the input structure for WIEN2K and CRYSTAL. It seems that Xcrysden also tried to connect with PWSCF. I tried to use this function, but the connection is not good. The needed input sturcture file is the structure file for old version . It will be great if Xcryden can visualize the pwscf structure and provide K path for bandstructure as it does for WIEN2k. Yanming ma _________________________________________________________________ ?????????????? MSN Messenger: http://messenger.msn.com/cn From Tone.Kokalj at ijs.si Fri Aug 29 05:03:02 2003 From: Tone.Kokalj at ijs.si (Tone Kokalj) Date: Fri, 29 Aug 2003 05:03:02 +0200 Subject: [Pw_forum] for q2r.x errors In-Reply-To: References: Message-ID: <20030829030302.GA27896@crysden.ijs.si> On Fri, Aug 29, 2003 at 07:23:33AM +0800, ma Yanming wrote: > Dear Paolo > > >PWscf just checks that the input makes sense. In the forthcoming > >new version, > > I am just wondering when the new version will come out. > > >there is an auxiliary code to calculate distances > >and angles between atom. This is what I use to verify my input > >data. > > That will be great if we can test what we input. > One thing I am interested in is the very useful software "Xcrysden". > Xcrysden can monitor the input structure for WIEN2K and CRYSTAL. It seems > that Xcrysden also tried to connect with PWSCF. I tried to use this > function, but the connection is not good. The needed input sturcture file > is the structure file for old version. Indeed, there are some problems in the PWscf-input --> XSF filter, which were solved and will be incorporated in the incoming version 0.9.5. I am attaching the corrected "pwi2xsf.sh" filter. Use it for the PWscf-input to XSF conversion, then you can load the XSF file as "xcrysden --xsf myfile.xsf". I hope this solves your problem. (I guess the filter of the 0.9.3 version should work if you use a lowercase letters for the namelist (i.e. &system, &end, ibrav, celldm, nat).) > It will be great if Xcryden can visualize the pwscf structure and > provide K path for bandstructure as it does for WIEN2k. You can select the k-path by using the Tools->"k-path Selection". Then you can save this path in kind of general k-path file, where the coordinates of the selected k-path points will be printed. You can use this info and make a script that will convert that to a format suitable for PWSCF. Regards, Tone -- +------------------------------------------------------------------------+ | Anton Kokalj Email: Tone.Kokalj at ijs.si | | Department of Physical and Organic Chemistry Phone: x 386 1 477 3523 | | Jozef Stefan Institute Fax: x 386 1 477 3811 | | Jamova 39, SI-1000 Ljubljana | | SLOVENIA | +------------------------------------------------------------------------+ !!! *** please, do not send me attachments in any Microsoft format *** !!! -------------- next part -------------- A non-text attachment was scrubbed... Name: pwi2xsf.sh Type: text/x-sh Size: 3044 bytes Desc: not available Url : /pipermail/attachments/20030829/614a38d2/attachment.sh From sprokuda at dnk.ru Fri Aug 29 14:39:26 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Fri, 29 Aug 2003 16:39:26 +0400 Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: References: Message-ID: <341736498.20030829163926@dnk.ru> Dear All! First of all I would like to thank Stefano de Gironcoli and Eyvaz Isaev for the great help with fftw library. Only with their help I've succesed to compile PWscf. but then I got some new problems ... Trying to run the examples I got "FORTRAN STOP 2" error message in example2, in single mode caluclations - what could it be and why ? Nevetheless, I've done caculation for the bulk Si ( on the base of the example2 scripts) for the high symmetry directions. Everything was fine. Then I've tried to run the same for Germanium. I've just modified celldm(1) =10.6260 Ge 72.61 Ge.pz-bhs.UPF ATOMIC_POSITIONS Ge 0.00 0.00 0.00 Ge 0.25 0.25 0.25 and corresponding items. For the direction [q00] everything was fine, but for [0qq] I got againg "FORTRAN STOP 2", namely for point q = [0 0.1 0.1]. The scpript file I've used for this run is attached. One reamrk probubly is important here: the runs for [0qq] took much nore time per point then runs for points on [q00]. Thank you in adavance. Prokudaylo Sergey. -------------- next part -------------- A non-text attachment was scrubbed... Name: run_ge_011 Type: application/octet-stream Size: 4451 bytes Desc: not available Url : /pipermail/attachments/20030829/34e535ab/attachment.obj From zyli at sina.com Fri Aug 29 15:36:30 2003 From: zyli at sina.com (Zhenyu Li) Date: Fri, 29 Aug 2003 21:36:30 +0800 Subject: [Pw_forum] error information Message-ID: <200308291337.VAA21072@mx1.ustc.edu.cn> Dear All I have attached the input file with this email. My starting geometry is taken from experimental structure, I don't think it will be far from minimum. One thing now I noticed is that I have combined the norm conserving pseudopotential and ultrasoft pseudopotential, will this causes serious problem? Zhenyu LI ======= 2003-08-28 23:52:00 ????????======= >Dear Zhenyu, > >I do not think your problem is from ecut and ecutrho. The smaller ecut will >result in inaccurate results, but not as serious as it stops running... > >Your total force is getting crazy. Are you sure your input is correct? >Send your input file. > >Yanming Ma > ================================================= Zhenyu LI Ph.D. Candidate Open Laboratory of Bond Selective Chemistry, University of Science and Technology of China, Hefei, Anhui, 230026, People's Republic of China Tel.: 86-551-3603748,3603418 Fax.: 86-551-3602969 http://www.bsc.ustc.edu.cn/~zyli/whoami/ ===================================================?????????????? -------------- next part -------------- A non-text attachment was scrubbed... Name: 75.in Type: application/octet-stream Size: 1594 bytes Desc: not available Url : /pipermail/attachments/20030829/59228416/attachment.obj From breezejd at sbu.ac.uk Fri Aug 29 15:48:07 2003 From: breezejd at sbu.ac.uk (Jonathan Breeze) Date: Fri, 29 Aug 2003 14:48:07 +0100 Subject: [Pw_forum] Third order anharmonic dynamical matrix Message-ID: <000d01c36e34$2d5907b0$95039488@breezy> Dear Pwscf Users, Thanks Paolo for the archive containing examples on how to calculate the third order anharmonic coefficients. Everything worked fine for Beryllium and I have managed to calculate the D3 coefficients for a few other metals like Aluminium and Iron. I have now been trying to calculate for insulators but get very small numbers for the TO phonon which don't agree with the numerical calculation. I was wondering if there was a problem with calculating for insulators with wide band-gaps. Kinds regards, Jonathan Breeze From degironc at sissa.it Fri Aug 29 15:42:42 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Fri, 29 Aug 2003 15:42:42 +0200 Subject: [Pw_forum] error information References: <200308291337.VAA21072@mx1.ustc.edu.cn> Message-ID: <3F4F5852.5080200@sissa.it> Zhenyu Li wrote: >One thing now I noticed is that I have combined the norm conserving >pseudopotential and ultrasoft pseudopotential, will this causes serious problem? > > NO your input file contain the line ibrav= 4, celldm(1) = 5.3668, celldm(3) = 3.831338, this means: hexagonal bravais lattice with a=5.3668 bohr and c/a=3.831338 that is c=20.562024 bohr. is this what you want ? do not confuse Bohr with Angstrom or c with c/a stefano From yanming_ma at hotmail.com Fri Aug 29 16:53:51 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Fri, 29 Aug 2003 22:53:51 +0800 Subject: [Pw_forum] error information Message-ID: Dear Zhenyu, I can not find any explicit error in your input file. But one thing I may state that you should be careful with the lattice vectors used in PWSCF. You claim you get your input atomic positions from experiments. But please be sure the experimental lattice vectors are the same with the ones in PWSCF, otherwise you should convert the atomic positions between the two different settings. Good luck! Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 >From: "Zhenyu Li" >Reply-To: pw_forum at pwscf.org >To: "pw_forum at pwscf.org" >Subject: Re: [Pw_forum] error information >Date: Fri, 29 Aug 2003 21:36:30 +0800 > >Dear All > > I have attached the input file with this email. My starting geometry >is taken from experimental structure, I don't think it will be far from >minimum. One thing now I noticed is that I have combined the norm conserving >pseudopotential and ultrasoft pseudopotential, will this causes serious problem? > >Zhenyu LI > >======= 2003-08-28 23:52:00 ????????======= > > >Dear Zhenyu, > > > >I do not think your problem is from ecut and ecutrho. The smaller ecut will > >result in inaccurate results, but not as serious as it stops running... > > > >Your total force is getting crazy. Are you sure your input is correct? > >Send your input file. > > > >Yanming Ma > > > > >================================================= > Zhenyu LI > Ph.D. Candidate > Open Laboratory of Bond Selective Chemistry, > University of Science and Technology of China, > Hefei, Anhui, 230026, > People's Republic of China > Tel.: 86-551-3603748,3603418 > Fax.: 86-551-3602969 > http://www.bsc.ustc.edu.cn/~zyli/whoami/ >===================================================???????????? ?? ><< 75.in >> _________________________________________________________________ ?????????????? MSN Messenger: http://messenger.msn.com/cn From eyvaz_isaev at yahoo.com Fri Aug 29 16:54:50 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 29 Aug 2003 07:54:50 -0700 (PDT) Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: <341736498.20030829163926@dnk.ru> Message-ID: <20030829145450.457.qmail@web80512.mail.yahoo.com> Dear Sergey, Your problem is not connected to q-vectors (less symmetry -> more q-vectors -> more computer time). I can suggest that the nscf-step is not finished, i.e. you have the same error which was reported recently ("cdiaghg" error). Please have a look to the end of nscf.out file. Regards, Eyvaz. --- sprokuda at dnk.ru wrote: > Dear All! > First of all I would like to thank Stefano de > Gironcoli and Eyvaz > Isaev for the great help with fftw library. Only > with their help I've > succesed to compile PWscf. > > but then I got some new problems ... > Trying to run the examples I got "FORTRAN STOP 2" > error message in > example2, in single mode caluclations - what could > it be and why ? > > Nevetheless, I've done caculation for the bulk Si ( > on the base of the > example2 scripts) for the high symmetry directions. > Everything was > fine. > Then I've tried to run the same for Germanium. I've > just modified > celldm(1) =10.6260 > > Ge 72.61 Ge.pz-bhs.UPF > ATOMIC_POSITIONS > Ge 0.00 0.00 0.00 > Ge 0.25 0.25 0.25 > > and corresponding items. For the direction [q00] > everything was > fine, but for [0qq] I got againg "FORTRAN STOP > 2", namely for point > q = [0 0.1 0.1]. The scpript file I've used for > this run is attached. > One reamrk probubly is important here: the runs for > [0qq] took much > nore time per point then runs for points on [q00]. > > Thank you in adavance. > Prokudaylo Sergey. > ATTACHMENT part 2 application/octet-stream name=run_ge_011 __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From yanming_ma at hotmail.com Fri Aug 29 17:04:48 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Fri, 29 Aug 2003 23:04:48 +0800 Subject: [Pw_forum] for q2r.x errors Message-ID: Dear Tone, Thanks for your information. Best Wishes! Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 >From: Tone Kokalj >Reply-To: pw_forum at pwscf.org >To: pw_forum at pwscf.org >Subject: Re: [Pw_forum] for q2r.x errors >Date: Fri, 29 Aug 2003 05:03:02 +0200 > >On Fri, Aug 29, 2003 at 07:23:33AM +0800, ma Yanming wrote: > > Dear Paolo > > > > >PWscf just checks that the input makes sense. In the forthcoming > > >new version, > > > > I am just wondering when the new version will come out. > > > > >there is an auxiliary code to calculate distances > > >and angles between atom. This is what I use to verify my input > > >data. > > > > That will be great if we can test what we input. > > One thing I am interested in is the very useful software "Xcrysden". > > Xcrysden can monitor the input structure for WIEN2K and CRYSTAL. It seems > > that Xcrysden also tried to connect with PWSCF. I tried to use this > > function, but the connection is not good. The needed input sturcture file > > is the structure file for old version. > >Indeed, there are some problems in the PWscf-input --> XSF filter, >which were solved and will be incorporated in the incoming version >0.9.5. I am attaching the corrected "pwi2xsf.sh" filter. Use it for >the PWscf-input to XSF conversion, then you can load the XSF file as >"xcrysden --xsf myfile.xsf". I hope this solves your problem. > >(I guess the filter of the 0.9.3 version should work if you use a >lowercase letters for the namelist (i.e. &system, &end, ibrav, celldm, >nat).) > > > It will be great if Xcryden can visualize the pwscf structure and > > provide K path for bandstructure as it does for WIEN2k. > >You can select the k-path by using the Tools->"k-path Selection". Then >you can save this path in kind of general k-path file, where the >coordinates of the selected k-path points will be printed. You can use >this info and make a script that will convert that to a format >suitable for PWSCF. > >Regards, Tone > >-- >+------------------------------------------------------------------------+ >| Anton Kokalj Email: Tone.Kokalj at ijs.si | >| Department of Physical and Organic Chemistry Phone: x 386 1 477 3523 | >| Jozef Stefan Institute Fax: x 386 1 477 3811 | >| Jamova 39, SI-1000 Ljubljana | >| SLOVENIA | >+------------------------------------------------------------------------+ >!!! *** please, do not send me attachments in any Microsoft format *** !!! ><< pwi2xsf.sh >> _________________________________________________________________ ???? MSN Explorer: http://explorer.msn.com/lccn/ From zyli at sina.com Fri Aug 29 16:52:12 2003 From: zyli at sina.com (Zhenyu Li) Date: Fri, 29 Aug 2003 22:52:12 +0800 Subject: [Pw_forum] error information Message-ID: <200308291453.WAA05440@mx1.ustc.edu.cn> Dear Paolo It's really the case. The experimental cell parameters are a=2.84 A and c=10.811 A. Sincerely Zhenyu LI ======= 2003-08-29 15:42:00 ????????======= >Zhenyu Li wrote: > >>One thing now I noticed is that I have combined the norm conserving >>pseudopotential and ultrasoft pseudopotential, will this causes serious problem? >> >> >NO > >your input file contain the line > >ibrav= 4, celldm(1) = 5.3668, celldm(3) = 3.831338, > >this means: hexagonal bravais lattice with a=5.3668 bohr and >c/a=3.831338 that is c=20.562024 bohr. > >is this what you want ? > >do not confuse Bohr with Angstrom or c with c/a > >stefano > > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > >. = = = = = = = = = = = = = = = = = = = = ????????? ?? ????????Zhenyu Li ????????zyli at sina.com ??????????2003-08-29 ================================================= Zhenyu LI Ph.D. Candidate Open Laboratory of Bond Selective Chemistry, University of Science and Technology of China, Hefei, Anhui, 230026, People's Republic of China Tel.: 86-551-3603748,3603418 Fax.: 86-551-3602969 http://www.bsc.ustc.edu.cn/~zyli/whoami/ ===================================================?????????????? From giannozz at nest.sns.it Fri Aug 29 18:17:20 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 29 Aug 2003 18:17:20 +0200 Subject: [Pw_forum] error information In-Reply-To: <200308291453.WAA05440@mx1.ustc.edu.cn> References: <200308291453.WAA05440@mx1.ustc.edu.cn> Message-ID: <200308291817.20758.giannozz@nest.sns.it> On Friday 29 August 2003 16:52, Zhenyu Li wrote: > Dear Paolo ??? you mean "Dear Stefano" ... > It's really the case. The experimental cell parameters are > a=2.84 A and c=10.811 A. funny structure. Anyway: my distance checker yields for your job the following nearest-neighbor distances: O-Co 1.29A; Na-Na, O-O, Co-Co : 1.42A; Na-O 1.91 A. It doesn't look right to me: they are way too small. Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From giannozz at nest.sns.it Fri Aug 29 18:54:39 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 29 Aug 2003 18:54:39 +0200 Subject: [Pw_forum] for q2r.x errors In-Reply-To: References: Message-ID: <200308291854.39450.giannozz@nest.sns.it> On Friday 29 August 2003 01:23, ma Yanming wrote: > I am just wondering when the new version will come out. in a few days (really!). Note that the version under development is always accessible via anonymous CVS. Coming back to your original question: > from init : error # 1 > nc already filled > stopping ... > My question is > If my input atomic positions are wrong,will this error message occur? No. This error has nothing to do with atomic positions. Only the lattice symmetry and the q-point set used to calculate phonons are relevant here. Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From yanming_ma at hotmail.com Fri Aug 29 19:34:55 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Sat, 30 Aug 2003 01:34:55 +0800 Subject: [Pw_forum] Germanuim [0qq] problem .... Message-ID: Dear Sergey, Thanks again for your MKL6.0 > One reamrk probubly is important here: the runs for [0qq] took much > nore time per point then runs for points on [q00]. There are six phonon branches in phonon disperson for Ge. Due to the high symmetry along [q00] direction, two TA branches degenerate, two TO branches degenerate, plus LA and LO, so along this direction, PWSCF only calculates four phonon values for individual K points. Due to the low symmetry along [0qq] direction, six phonon branches totall split, in this case PWSCF will calculate all six phonon values. Also it will take longer time to calculate individual phonon value with lower symmetry k point. This is the reason why the runs for [0qq] took much more time per point then runs for points on [q00]. Yanming Ma _________________________________________________________________ ?????????????? MSN Messenger: http://messenger.msn.com/cn From felipe at titan.ipicyt.edu.mx Sat Aug 30 00:07:35 2003 From: felipe at titan.ipicyt.edu.mx (FELIPE VALENCIA) Date: Fri, 29 Aug 2003 17:07:35 -0500 (CDT) Subject: [Pw_forum] error information In-Reply-To: <200308291817.20758.giannozz@nest.sns.it> Message-ID: Dear Zhenyu LI, Are you absolutely sure about the input coordinates? I have taken a look to them and they seem certainly amazing... In any case, I am sending you also the XYZ coordinates corresponding to your input file, may you find them useful. Felipe ..................................... 30 Na 0.709997 0.409905 8.160632 Na -0.000000 1.639619 8.160632 Na 2.129990 0.409905 8.160632 Na 0.000000 0.819810 2.720211 Na -0.709997 2.049524 2.720211 Na 1.419994 0.819810 2.720211 Co 0.000000 0.000000 0.000000 Co 1.419994 0.000000 0.000000 Co -0.709997 1.229714 0.000000 Co 0.709997 1.229714 0.000000 Co 0.000000 0.000000 5.440421 Co 1.419994 0.000000 5.440421 Co -0.709997 1.229714 5.440421 Co 0.709997 1.229714 5.440421 O 0.709997 0.409905 0.993421 O -0.000000 1.639619 0.993421 O 2.129990 0.409905 0.993421 O 1.419994 1.639619 0.993421 O 0.709997 0.409905 4.447000 O -0.000000 1.639619 4.447000 O 2.129990 0.409905 4.447000 O 1.419994 1.639619 4.447000 O 0.000000 0.819810 6.433842 O -0.709997 2.049524 6.433842 O 1.419994 0.819810 6.433842 O 0.709997 2.049524 6.433842 O 0.000000 0.819810 9.887422 O -0.709997 2.049524 9.887422 O 1.419994 0.819810 9.887422 O 0.709997 2.049524 9.887422 > > > It's really the case. The experimental cell parameters are > > a=2.84 A and c=10.811 A. > From zyli at sina.com Sat Aug 30 04:32:38 2003 From: zyli at sina.com (Zhenyu Li) Date: Sat, 30 Aug 2003 10:32:38 +0800 Subject: [Pw_forum] error information Message-ID: <200308300233.KAA02290@mx1.ustc.edu.cn> Dear All Thank you for you help, I find that I have really made a mistake in my coordinates. And, I am sorry to Prof. Stefano de Gironcoli for the mistake. Zhenyu LI ======= 2003-08-29 18:17:00 ????????======= >On Friday 29 August 2003 16:52, Zhenyu Li wrote: >> Dear Paolo > >??? you mean "Dear Stefano" ... > >> It's really the case. The experimental cell parameters are >> a=2.84 A and c=10.811 A. > >funny structure. Anyway: my distance checker yields >for your job the following nearest-neighbor distances: >O-Co 1.29A; Na-Na, O-O, Co-Co : 1.42A; Na-O 1.91 A. >It doesn't look right to me: they are way too small. > >Paolo > >-- >Paolo Giannozzi e-mail: giannozz at nest.sns.it >Scuola Normale Superiore Phone: +39/050509412 >Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 >I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > >. = = = = = = = = = = = = = = = = = = = = ????????? ?? ????????Zhenyu Li ????????zyli at sina.com ??????????2003-08-30 ================================================= Zhenyu LI Ph.D. Candidate Open Laboratory of Bond Selective Chemistry, University of Science and Technology of China, Hefei, Anhui, 230026, People's Republic of China Tel.: 86-551-3603748,3603418 Fax.: 86-551-3602969 http://www.bsc.ustc.edu.cn/~zyli/whoami/ ===================================================?????????????? From bbouhafs at mail.com Sun Aug 31 19:55:59 2003 From: bbouhafs at mail.com (Bachir Bouhafs) Date: Sun, 31 Aug 2003 18:55:59 +0100 Subject: [Pw_forum] problem with compilation Message-ID: <20030831175559.19166.qmail@mail.com> Dear All! I would ask for some help to compile the new version of PWscf using absoft compiler, when I compile I got some problems with fftw library that I have downloaded form www.fftw.org web site. Best regards Bachir Dr. Bachir Bouhafs Physics Department University of Sidi Bel Abbes 22000 Sidi Bel Abbes, Algeria Mobil:+213 (0)71 18 21 47 Fax:+213 48 54 11 52 E-mail: bbouhafs at mail.com -- __________________________________________________________ Sign-up for your own personalized E-mail at Mail.com http://www.mail.com/?sr=signup CareerBuilder.com has over 400,000 jobs. Be smarter about your job search http://corp.mail.com/careers From giannozz at nest.sns.it Sun Aug 31 21:59:27 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 31 Aug 2003 21:59:27 +0200 Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: <341736498.20030829163926@dnk.ru> References: <341736498.20030829163926@dnk.ru> Message-ID: <200308312159.27901.giannozz@nest.sns.it> On Fri, 29 Aug 2003 sprokuda at dnk.ru wrote: > Trying to run the examples I got "FORTRAN STOP 2" error message in > example2, in single mode calculations - what could it be and why ? it could be anything: fortran stop 2 means that fortran execution stopped with error 2 (whatever it means). For me, your scripts stops because of an obsolete variable name in a namelist ("tmp_dir" instead of "outdir"). Paolo