From ks at mic.dtu.dk Mon May 5 11:16:11 2003 From: ks at mic.dtu.dk (Kurt Stokbro) Date: Mon, 5 May 2003 11:16:11 +0200 Subject: [Pw_forum] compilation of pwscf with lahey compiler Message-ID: I have tried to compile pwscf with the lahey compiler. I get the error: 2372-S: "constants.F90", line 76: Invalid number of actual arguments for intri nsic subroutine ERROR. It seems that there is a build in function called error. How do I avoid the comfusion between this build in function and the supplied function? My make.sys library looks like: SHOME=/home/ks/pwscf # # System-dependent Make definitions for Linux PCs, pgi compiler # Edit according to your needs # # Precompiler # we assume that the GNU precompiler is used. The option -traditional # prevent string concatenation '//' from being interpreted as a comment # CPP = /lib/cpp # # Add -DADD_BLAS_ONE_UNDERSCORE if your blas/lapack library names contain # two underscores at the end # Define FFTW library names with one underscore less than they have in # the library (one underscore is added by the compiler) # CPPFLAGS = -P -traditional -I$(OSHOME)/include -DPC -DPGI -DFFTW \ -D"FFTWND_F77_ONE=fftwnd_f77_one" \ -D"FFTW3D_F77_CREATE_PLAN=fftw3d_f77_create_plan" # # Fortran compilers: # Please note: -r8 is necessary for numerical stability .. F90 =lf95 #F90FLAGS = -fast -r8 # # This is needed to tell the compiler where modules are # MODULEFLAG= -I$(OSHOME)/Modules -I$(OSHOME)/PW -I$(OSHOME)/PH # # Loader: # this below uses precompiled pgi libraries ( not very efficient but stable ) # we assume that fftw is placed in /usr/local/lib # LD=$(F90) #LIBS = -L/usr/local/pgi/linux86/lib/ -llapack -lblas -latlas -L/usr/local/lib -lfftw LIBS = -llapack -lblas -latlas -L/usr/local/lib -lfftw # LFLAGS = $(LIBS) $(MODULEFLAG) # # ar: # AR = ar ARFLAGS = ruv Best Regards, Kurt Stokbro From degironc at sissa.it Mon May 5 11:20:30 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Mon, 05 May 2003 11:20:30 +0200 Subject: [Pw_forum] compilation of pwscf with lahey compiler References: Message-ID: <3EB62CDE.2020200@sissa.it> a simple fix that we have adopted in the developping version is to change the name of the pwscf "error" routine to a name that does not cause conflicts, for instance "errore". There are about 1000 calls to error in the code .... Stefano de Gironcoli Kurt Stokbro wrote: >I have tried to compile pwscf with the lahey compiler. I get the >error: >2372-S: "constants.F90", line 76: Invalid number of actual arguments for intri >nsic subroutine ERROR. > >It seems that there is a build in function called error. How do I avoid the comfusion between this build in function and the supplied function? > >My make.sys library looks like: > > > From giannozz at nest.sns.it Mon May 5 11:33:08 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 5 May 2003 11:33:08 +0200 Subject: [Pw_forum] compilation of pwscf with lahey compiler In-Reply-To: References: Message-ID: <200305051133.08603.giannozz@nest.sns.it> Hi Kurt > I have tried to compile pwscf with the lahey compiler [...] > It seems that there is a build in function called error. How do I avoid the > confusion between this build in function and the supplied function? the current version supports the lahey compiler (./configure pc_lahey), thanks to help from Marcos Verissimo Alves.You can download the current version using anonymous cvs (see attached instruction). If you have problems, please let us know 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 -------------- next part -------------- The current version is available using anonymous CVS. Define environmental variables: setenv CVS_RSH ssh setenv CVSROOT :pserver:cvsanon at democritos.sissa.it:/home/cvs (tcsh/csh) or export CVS_RSH=ssh export CVSROOT=:pserver:cvsanon at democritos.sissa.it:/home/cvs (sh/bash). Then: cvs login (password: cvsanon). For the first code download: cvs co O-sesame for the entire repository (the code appears in directory O-sesame/). Alternatively, "cvs co pwscf", or "cp", or "fpmd" will download only PWscf, CP, FPMD, respectively, in directories with the same name. For updating the code to the current version: cvs update -d in the directory containing the distribution. PLEASE NOTE: re-run "./configure" if files have been moved/added/removed since the last checkout, otherwise "make" may not work properly due to obsolete or missing dependencies. Do not blindly re-use a "make.sys" file from a preceding version: it may no longer work. PLEASE ALSO NOTE: the development version may not work properly, and sometimes not even compile properly. Use at your own risk. From ks at mic.dtu.dk Mon May 5 13:33:55 2003 From: ks at mic.dtu.dk (Kurt Stokbro) Date: Mon, 5 May 2003 13:33:55 +0200 Subject: [Pw_forum] compilation of pwscf with lahey compiler Message-ID: Hi Paolo, Thanks for the help. I can now compile the program, but I have problems with linking. I have all the blas, lapack and atlas libraries compiled in /usr/lib and I have also installed fftw-3.0 . I have tried to fix the problems in make.sys but there are still errors. I include make.sys and the error log . Thanks for the help. By the way, I like your webpage and the logo. Who made this? I looks very professional. How many users do you have now? Best Regards -----Original Message----- From: Paolo Giannozzi [mailto:giannozz at nest.sns.it] Sent: Mon 05-05-2003 11:33 To: pw_forum at pwscf.org Cc: Subject: Re: [Pw_forum] compilation of pwscf with lahey compiler Hi Kurt > I have tried to compile pwscf with the lahey compiler [...] > It seems that there is a build in function called error. How do I avoid the > confusion between this build in function and the supplied function? the current version supports the lahey compiler (./configure pc_lahey), thanks to help from Marcos Verissimo Alves.You can download the current version using anonymous cvs (see attached instruction). If you have problems, please let us know 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 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 18471 bytes Desc: not available Url : /pipermail/attachments/20030505/6df7b71e/attachment.bin From giannozz at nest.sns.it Mon May 5 14:04:54 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 5 May 2003 14:04:54 +0200 Subject: [Pw_forum] compilation of pwscf with lahey compiler In-Reply-To: References: Message-ID: <200305051404.54923.giannozz@nest.sns.it> Hi Kurt > Thanks for the help. I can now compile the program, but I have > problems with linking. I have all the blas, lapack and atlas libraries > compiled in /usr/lib and I have also installed fftw-3.0 don't use fftw-3.0 ! the code is written for fftw-2 . They are NOT compatible. This explains many missing symbols. Other are missing because the f90-to-C linking for Lahey is not implemented (it's a recent change). Remaining missing symbols are in the lapack library (likely compiled with gcc). Try to load -lg2c or -lm 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 ks at mic.dtu.dk Mon May 5 17:38:28 2003 From: ks at mic.dtu.dk (Kurt Stokbro) Date: Mon, 5 May 2003 17:38:28 +0200 Subject: [Pw_forum] upf for siesta pseudopotentials Message-ID: I would like to read siesta pseudopotentials with pwscf. Has anybody made a converter to upf? The siesta pseudopotentials do not have info on the pseudowavefunctions. I tried to remove pseudo wave functions in the upf format and got an cdiaghg error. Can pwscf run without information about the pseudowavefunctions? Kurt From giannozz at nest.sns.it Mon May 5 19:18:50 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 5 May 2003 19:18:50 +0200 Subject: [Pw_forum] upf for siesta pseudopotentials In-Reply-To: References: Message-ID: <200305051918.50162.giannozz@nest.sns.it> Hi Kurt > I would like to read siesta pseudopotentials with pwscf. > Has anybody made a converter to upf? not that I know. It shouldn't be that difficult, but remember: PWscf uses separable PP only, and only projectors are stored in the upf format, not pseudopotentials in semilocal form. In order to build the former from the latter, you need in any case the pseudowavefunctions. > Can pwscf run without information about the pseudowavefunctions? it should, unless you want to calculate projections on atomic wavefunctions and similar stuff 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 eyvaz_isaev at yahoo.com Mon May 5 19:30:48 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 5 May 2003 10:30:48 -0700 (PDT) Subject: [Pw_forum] upf for siesta pseudopotentials In-Reply-To: Message-ID: <20030505173048.12500.qmail@web80504.mail.yahoo.com> Dear Kurt, I assume, not. Because right after generating of k-points initial potential generation is being started and it uses atomic wfc (for initial charge density generation). Sorry, if I am wrong. Regards, Eyvaz Isaev Theoretical Physics Department Moscow Steel and Alloys Institute --- Kurt Stokbro wrote: > I would like to read siesta pseudopotentials with > pwscf. > Has anybody made a converter to upf? > The siesta pseudopotentials do not have info on the > pseudowavefunctions. I tried to remove pseudo wave > functions > in the upf format and got an cdiaghg error. Can > pwscf run without information about the > pseudowavefunctions? > Kurt > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com From francesco.antoniella at aquila.infn.it Tue May 6 14:33:56 2003 From: francesco.antoniella at aquila.infn.it (antonief) Date: Tue, 6 May 2003 14:33:56 +0200 (CEST) Subject: [Pw_forum] upf for siesta pseudopotentials In-Reply-To: <20030505173048.12500.qmail@web80504.mail.yahoo.com> Message-ID: On Mon, 5 May 2003, Eyvaz Isaev wrote: > Dear Kurt, > > I assume, not. Because right after generating of > k-points initial potential generation is being started > and it uses atomic wfc (for initial charge density > generation). > > Sorry, if I am wrong. This appears like a non-problem, try to use starting_wfc='random'. Hi all , Francesco > > Regards, > Eyvaz Isaev > Theoretical Physics Department > Moscow Steel and Alloys Institute > > --- Kurt Stokbro wrote: > > I would like to read siesta pseudopotentials with > > pwscf. > > Has anybody made a converter to upf? > > The siesta pseudopotentials do not have info on the > > pseudowavefunctions. I tried to remove pseudo wave > > functions > > in the upf format and got an cdiaghg error. Can > > pwscf run without information about the > > pseudowavefunctions? > > Kurt > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > __________________________________ > Do you Yahoo!? > The New Yahoo! Search - Faster. Easier. Bingo. > http://search.yahoo.com > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From astroppa at ts.infn.it Tue May 6 15:15:03 2003 From: astroppa at ts.infn.it (Alessandro Stroppa) Date: Tue, 6 May 2003 15:15:03 +0200 (MET DST) Subject: [Pw_forum] Virtual Crystal Message-ID: Is there someone who have a program for Virtual Crystal calculation?I mean, a program which starts with 2 pseudopotential and give a Virtual Potential. Thanks,A. From yanming_ma at hotmail.com Tue May 6 15:56:33 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Tue, 06 May 2003 21:56:33 +0800 Subject: [Pw_forum] upf for siesta pseudopotentials Message-ID: Hi, I am pretty sure the PPs used by Siesta and PWSCF can be converted directly each other. Siesta and PWSCF can both use Troullier Martins PPs, although the atomic orbital is used to expanse the bloch function in Siesta, while in PWSCF the plane waves is used. I use both code, but I have not tried to convert them. Because, for TM PPs, one can generate PPs in Siesta using ATOM code and one can generate PPs in PWSCF using FHI code by choosing the similar core radius cutoff for s, p, d, and f orbital in a common sense. Best Regards 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 lgzhou at i12.com Thu May 8 18:04:52 2003 From: lgzhou at i12.com (Longguang Zhou) Date: Thu, 8 May 2003 11:04:52 -0500 Subject: [Pw_forum] constrained relaxation In-Reply-To: <20030507053506.5994.13840.Mailman@democritos.sissa.it> Message-ID: <000001c3157b$9084aa40$c0127180@lgzhounotebook> Would anybody please help me how could I keep the size of simulation cell constant in one dimension and allow the other two change during "vc-relax" minimization? Thanks. From proffess at yandex.ru Sat May 10 09:09:04 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Sat, 10 May 2003 11:09:04 +0400 (MSD) Subject: [Pw_forum] Symmetry Message-ID: <3EBCA590.00000C.07963@tide.yandex.ru> Dear PWSCF users, It is very strange, but PWscf does not find the any symmetry in fullerene C60. Why? Thanks, Sergey From baroni at sissa.it Sat May 10 10:47:41 2003 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 10 May 2003 10:47:41 +0200 Subject: [Pw_forum] Symmetry In-Reply-To: <3EBCA590.00000C.07963@tide.yandex.ru> References: <3EBCA590.00000C.07963@tide.yandex.ru> Message-ID: <3EBCBCAD.2040401@sissa.it> Do you mean the symmetry of the isolated molecule, or that of the crystal? The symmetry group of the isolated molecule is the icosahedral group (Ih, 120 operations) and, as such, it is not the subgroup of any space group (the largest point group compatible with translations -- "crystallographic point group" -- is Oh, 48 operations). That's why QUASI-crystals are not crystals, and this is why PWscf (or any other code dealing with periodic systems, for what it matters) cannot identify all the symmetry operations of an icosahedrally symmetric system). See, e.g. http://mathworld.wolfram.com/CrystallographyRestriction.html Hope this solve your doubts. Yours, Stefano Baroni Sergei Lisenkov wrote: > Dear PWSCF users, > > It is very strange, but PWscf does not find the any symmetry in fullerene C60. Why? > > Thanks, > Sergey > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From giannozz at nest.sns.it Mon May 12 10:28:39 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 12 May 2003 10:28:39 +0200 Subject: [Pw_forum] Symmetry In-Reply-To: <3EBCA590.00000C.07963@tide.yandex.ru> References: <3EBCA590.00000C.07963@tide.yandex.ru> Message-ID: <200305121028.39063.giannozz@nest.sns.it> On Saturday 10 May 2003 09:09, Sergei Lisenkov wrote: > It is very strange, but PWscf does not find the any symmetry in fullerene > C60. Why? Hi, part of the answer is already contained in Stefano's message. The following will appear in the next version of the users' guide. I hope this will help. 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 ------------------------------------------------- \item {\em pw.x does not find all the symmetries you expected.} {\tt pw.x} determines first the symmetry operations (rotations) of the Bravais lattice; then checks which of these are symmetry operations of the system (including if needed fractional translations). This is done by rotating (and translating if needed) the atoms in the unit cell and verifying if the rotated unit cell coincides with the original one. You may miss symmetry operations because: 1) The number of significant figures in the atomic positions is not big enough. In {\tt PP/checksym.f90}, the variable {\tt accep} is used to decide whether a rotation is a symmetry operation. Its current value ($10^{-5}$) is quite strict: a rotated atom must coincide with another atom to 5 significant digits. You may change the value of {\tt accep} and recompile. 2) They are not acceptable symmetry operations of the Bravais lattice. This is the case for C$_{60}$, for instance: the $I_h$ icosahedral group of C$_{60}$ contains 5-fold rotations that are incompatible with translation symmetry 3) The system is rotated with respect to symmetry axis. For instance: a C$_{60}$ molecule in the fcc lattice will have 24 symmetry operations ($T_h$ group) only if the double bond is aligned along one of the crystal axis; if C$_{60}$ is rotated in some arbitrary way, pw.x may not find any symmetry, apart from inversion. 4) They contain a fractional translation that is incompatible with the FFT grid. Typical fractional translations are 1/2 or 1/3 of a lattice vector. If the FFT grid is not divisible respectively by 2 or by 3, the symmetry operation will not transform the FFT grid into itself. Such symmetry operations are disabled to prevent problems with symmetrization. Note that if you change cutoff or unit cell volume, the automatically computed FFT grid changes. and this may explain changes in symmetry (and number of k-points as a consequence) for no apparent good reason (only if you have fractional transations in the system, though). 5) A fractional translation, without rotation, is a symmetry operation of the system (meaning that the cell is actually a supercell). In this case, all symmetry operations containing fractional translations are disabled. The reason is that in this rather exotic case there is no (simple) way to select those symmetry operations forming a group (in the strict mathematical sense). ---------------------------------------------------- From mtoso at ts.infn.it Mon May 12 16:24:14 2003 From: mtoso at ts.infn.it (Michele Tosolini) Date: Mon, 12 May 2003 16:24:14 +0200 (MET DST) Subject: [Pw_forum] recover from crash Message-ID: Dear PWSCF users, there were some problems on cineca (8 cpu's used) during a nscf calculation (pw1.0.3 F77 version). I need the 8 punchfile's for a band structure calculation but in the scratch directory i just have the 8 fileigk's, the 8 punchfile.wfc's , not all the 8 restart files and none of the 8 punchfile's. May i restart from somewhere (also with the F90 code) sparing cpu time or have i to restart from scratch ? Thanks. From zyli at 263.sina.com Fri May 16 14:29:30 2003 From: zyli at 263.sina.com (zyli) Date: Fri, 16 May 2003 20:29:30 +0800 Subject: [Pw_forum] A little question about total energy Message-ID: <20030516122930.5090.qmail@sina.com> Dear PWSCF users In the output of PWSCF, the energy partitions are list following the total energy as I pasted below. I found that the total energy is not equal to the sum of those listed partitions. Then what are the unlisted contributions to the total energy? Seems not only the energy of ion-ion interaction. ------------------------------------------------------------ ! total energy = -62.07569469 ryd estimated scf accuracy < 0.00000799 ryd band energy sum = -14.82580512 ryd one-electron contribution = -98.79712762 ryd hartree contribution = 51.10053218 ryd xc contribution = -14.02297161 ryd ewald contribution = -0.34484385 ryd scf in/out correction = 0.02901784 ryd correction for metals = -0.01128379 ryd -------------------------------------------------------------- Best Wishes zyli =================================================================== From zyli at 263.sina.com Fri May 16 14:52:58 2003 From: zyli at 263.sina.com (zyli) Date: Fri, 16 May 2003 20:52:58 +0800 Subject: [Pw_forum] A little question about total energy Message-ID: <20030516125258.18661.qmail@sina.com> Dear PWSCF users In the output of PWSCF, the energy partitions are list following the total energy as I pasted below. I found that the total energy is not equal to the sum of those listed partitions. Then what are the unlisted contributions to the total energy? Seems not only the energy of ion-ion interaction. ------------------------------------------------------------ ! total energy = -62.07569469 ryd estimated scf accuracy < 0.00000799 ryd band energy sum = -14.82580512 ryd one-electron contribution = -98.79712762 ryd hartree contribution = 51.10053218 ryd xc contribution = -14.02297161 ryd ewald contribution = -0.34484385 ryd scf in/out correction = 0.02901784 ryd correction for metals = -0.01128379 ryd -------------------------------------------------------------- Best Wishes zyli =================================================================== From giannozz at nest.sns.it Fri May 16 15:28:49 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 16 May 2003 15:28:49 +0200 Subject: [Pw_forum] A little question about total energy In-Reply-To: <20030516122930.5090.qmail@sina.com> References: <20030516122930.5090.qmail@sina.com> Message-ID: <200305161528.49022.giannozz@nest.sns.it> Hi > In the output of PWSCF, the energy partitions are list > following the total energy as I pasted below. I found > that the total energy is not equal to the sum of those listed > partitions. Total Energy = sum of terms marked with "+" below = one-electron + hartree + xc + ewald + correction for metals ! total energy = -62.07569469 ryd estimated scf accuracy < 0.00000799 ryd band energy sum = -14.82580512 ryd + one-electron contribution = -98.79712762 ryd + hartree contribution = 51.10053218 ryd + xc contribution = -14.02297161 ryd + ewald contribution = -0.34484385 ryd scf in/out correction = 0.02901784 ryd + correction for metals = -0.01128379 ryd The "one-electron contribution" is the kinetic + ionic term: \sum The "band energy sum" is exactly what it claims to be: \sum epsilon = \sum The "correction for metals" makes forces and energies consistent if a broadening scheme is used. The "scf in/out correction" is no longer used (and no longer useful) The "estimated scf accuracy" gives an estimate of the error on the energy due to imperfect self-consistency. Admittedly, it is slightly confusing. 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 breezejd at sbu.ac.uk Fri May 16 15:44:22 2003 From: breezejd at sbu.ac.uk (Jonathan Breeze) Date: Fri, 16 May 2003 14:44:22 +0100 Subject: [Pw_forum] third order anharmonic coefficients Message-ID: <3EC4EB36.6030402@sbu.ac.uk> Dear PWScf users, Does anyone have experience of calculating the third order anharmonic coefficients? I have been trying in vain to calculate these for MgO using the instructions in the user guide, but have not been able to generate the files required (drho etc.). I posted two messages a while ago, but have received no response. If anyone could be of assistance it would be greatly appreciated. Even just to hear that success in this endeavour is within reach! Kind regards, Jonathan Breeze Centre for Physical Electronics and Materials Faculty of Engineering, Science and Technology South Bank University 103 Borough Road London SE1 0AA From marzari at MIT.EDU Sat May 17 19:25:49 2003 From: marzari at MIT.EDU (Nicola Marzari) Date: Sat, 17 May 2003 13:25:49 -0400 Subject: [Pw_forum] A little question about total energy References: <20030516122930.5090.qmail@sina.com> <200305161528.49022.giannozz@nest.sns.it> Message-ID: <3EC6709D.90907@mit.edu> To add to this - when you have a broadening the PWSCF total energy is what others call the total free energy F=E-TS. E.g., if you choose a Fermi-Dirac broadening, the PWSCF total energy is really the Mermin-Kohn-Sham free energy. Last, if you are studying isolated atoms, but still use a broadening, you have an entropy S different from zero if there are degeneracies giving rise to fractional occupation numbers. So, you might want to remove the -TS (the correction for metals), and also make sure your results are converged in the limit of small T (i.e. small broadening temperatures). Best, nicola marzari Paolo Giannozzi wrote: > Hi > > >> In the output of PWSCF, the energy partitions are list >>following the total energy as I pasted below. I found >>that the total energy is not equal to the sum of those listed >>partitions. > > > Total Energy = sum of terms marked with "+" below = > one-electron + hartree + xc + ewald + correction for metals > > ! total energy = -62.07569469 ryd > estimated scf accuracy < 0.00000799 ryd > > band energy sum = -14.82580512 ryd > + one-electron contribution = -98.79712762 ryd > + hartree contribution = 51.10053218 ryd > + xc contribution = -14.02297161 ryd > + ewald contribution = -0.34484385 ryd > scf in/out correction = 0.02901784 ryd > + correction for metals = -0.01128379 ryd > > The "one-electron contribution" is the kinetic + ionic term: > \sum > The "band energy sum" is exactly what it claims to be: > \sum epsilon = \sum > The "correction for metals" makes forces and energies > consistent if a broadening scheme is used. > The "scf in/out correction" is no longer used > (and no longer useful) > The "estimated scf accuracy" gives an estimate of the > error on the energy due to imperfect self-consistency. > > Admittedly, it is slightly confusing. > > Paolo > -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 617.2586534 marzari at mit.edu http://nnn.mit.edu From proffess at yandex.ru Sun May 18 14:42:39 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Sun, 18 May 2003 16:42:39 +0400 (MSD) Subject: [Pw_forum] Warning or error? Message-ID: <3EC77FBF.000003.31915@ariel.yandex.ru> Dear PWscf users, I use the Alpha machine clusters with Linux. I got it in my output file: Reading pseudopotential file in UPF format from data_structure : error # -1 some processors have no planes from data_structure : error # -1 some processors have no smooth planes ^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@ ^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@ ^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@ ^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@ current restart_mode = from_scratch current disk_io mode = default RECOVER from restart file has been switched off on input Writing file C120.save file written %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from data_structure : error # -1 some processors have no planes %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from data_structure : error # -1 some processors have no smooth planes %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Planes per process (thick) : nr3 = 64 npp = 1 ncplane = 6144 --- Executing new GGEN Loop --- bravais-lattice index = 8 lattice parameter (a_0) = 23.8742 a.u. unit-cell volume = 6590.7667 (a.u.)^3 number of atoms/cell = 120 number of atomic types = 1 kinetic-energy cutoff = 35.0000 Ry charge density cutoff = 144.0000 Ry convergence threshold = 1.0E-08 beta = 0.1000 number of iterations used = 8 plain mixing Exchange-correlation = SLA PZ NOGX NOGC (1100) iswitch = 3 nstep = 100 celldm(1)= 23.87420 celldm(2)= 0.69167 celldm(3)= 0.70025 celldm(4)= 0.00000 celldm(5)= 0.00000 celldm(6)= 0.00000 ....... The code is running without stop. What is it? Some ideas? Thanks, Sergey From eyvaz_isaev at yahoo.com Mon May 19 16:45:34 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 19 May 2003 07:45:34 -0700 (PDT) Subject: [Pw_forum] Occupation numbers in CP and FPMD In-Reply-To: <3EC6709D.90907@mit.edu> Message-ID: <20030519144534.12035.qmail@web80505.mail.yahoo.com> Dear all, Now I am trying to do some calculations with CP and FPMD. But I was surprised very much due to the "occupations" option. It turned out that they are fixed, i.e. valance and conduction bands are separated as in semiconductors and and insulators. At least, smearing is not used in CP and FPMD. Does it mean that this kind of calculations could not be carried out for metallic systems? Regards, Eyvaz. __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com From giannozz at nest.sns.it Mon May 19 19:13:29 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 19 May 2003 19:13:29 +0200 Subject: [Pw_forum] Warning or error? In-Reply-To: <3EC77FBF.000003.31915@ariel.yandex.ru> References: <3EC77FBF.000003.31915@ariel.yandex.ru> Message-ID: <200305191913.29256.giannozz@nest.sns.it> On Sunday 18 May 2003 14:42, Sergei Lisenkov wrote: > from data_structure : error # -1 > some processors have no planes this usually means that you are using too many processors for a given system (in particolar, more processors than the fft dimension in the z axis, nr3 or nr3s) 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 May 19 19:41:46 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 19 May 2003 19:41:46 +0200 Subject: [Pw_forum] Occupation numbers in CP and FPMD In-Reply-To: <20030519144534.12035.qmail@web80505.mail.yahoo.com> References: <20030519144534.12035.qmail@web80505.mail.yahoo.com> Message-ID: <200305191941.46416.giannozz@nest.sns.it> On Monday 19 May 2003 16:45, Eyvaz Isaev wrote: > Dear all, > > Now I am trying to do some calculations with CP and > FPMD. But I was surprised very much due to the > "occupations" option. It turned out that they are > fixed, i.e. valance and conduction bands are separated > as in semiconductors and and insulators. At least, > smearing is not used in CP and FPMD. Does it mean that > this kind of calculations could not be carried out for > metallic systems? Car-Parrinello dynamics and metallic systems don't go along very well: see for instance G. Pastore, E. Smargiassi, F. Buda, Phys. Rev. A 44, 6334 (1991). In order to perform CP simulations in systems with very small or no gap, one has to use Nose' thermostats on electronic degrees of freedom, or similar techniques. Such techniques may require some serious experience. By the way: there is a bug in the CP distribution on the web that may cause a crash or a memory leak (invariably leading to a crash): lines 2039-2040 of CPV/cprsub.f90: deallocate(gradr) end if should be reversed: end if deallocate(gradr) 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 subhra at physics.rutgers.edu Mon May 19 22:25:10 2003 From: subhra at physics.rutgers.edu (Subhradip Ghosh) Date: Mon, 19 May 2003 16:25:10 -0400 (EDT) Subject: [Pw_forum] Phonon DOS Message-ID: Hi, I would like to know how to calculate the vibrational densities of states(total as well as component-projected) using PWSCF 1.0.1. It seems to me that there is a code(dos.f) that does it for electronic case but it's not clear how to deal with the phonon case. Thanks in advance, Subhradip ############################################################################ Dr Subhradip Ghosh Post Doctoral Fellow Condensed matter theory group Department of Physics and Astronomy ############################################################################ Office: | Residence: --------------------------------------------|------------------------------- Serin E255 |132, South Adelaide Avenue Department of Physics and Astronomy |Apartment# 2A Rutgers, The State University of New Jersey |Highland Park 136 Frelinghuysen Road |NJ 08904 Piscataway, NJ 08854-8019 |USA USA | Email address: subhra at physics.rutgers.edu, |Telephone # :(732)4489266 subhradip at yahoo.com | Telephone # : (732)445-4604 | ############################################################################## From eyvaz_isaev at yahoo.com Tue May 20 01:27:54 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 19 May 2003 16:27:54 -0700 (PDT) Subject: [Pw_forum] Phonon DOS In-Reply-To: Message-ID: <20030519232754.28953.qmail@web80510.mail.yahoo.com> Hi, You are quite right about calculatioins of the total phonon DOS using dos.f (or dos.f90 depending on PWSCF version). Yes, it is able calculate only the total PDOS, but not porjected DOS. Nevertheless, it is very easy to calculate PHDOS. Sure you shoud compile it typing "make dos" (if you do not like compile other programs in the /tools directory). Then you can copy it or make a link to the current directory where your frequency file is placed. So, just type "dos.x" and answer questions which are clear. But I suppose it is much better to calculate PHDOS using the v.1.2. In this version integration over the BZ is done using the tetrahedra method instead of gaussian smearing in dos.f (dos.f90). An example file is attached. I also guess that it is a good idea to have a look matdyn.f90 file. Regards, Eyvaz Isaev Theoretical Physics Department Moscow Steel and Alloys Institute --- Subhradip Ghosh wrote: > Hi, > > I would like to know how to calculate the > vibrational densities of > states(total as well as component-projected) using > PWSCF 1.0.1. It seems > to me that there is a code(dos.f) that does it for > electronic case but > it's not clear how to deal with the phonon case. > > Thanks in advance, > > Subhradip > > ############################################################################ > Dr Subhradip Ghosh > Post Doctoral Fellow > Condensed matter theory group > Department of Physics and Astronomy > ############################################################################ > Office: | > Residence: > --------------------------------------------|------------------------------- > Serin E255 |132, > South Adelaide Avenue > Department of Physics and Astronomy > |Apartment# 2A > Rutgers, The State University of New Jersey > |Highland Park > 136 Frelinghuysen Road |NJ > 08904 > Piscataway, NJ 08854-8019 |USA > USA | > Email address: subhra at physics.rutgers.edu, > |Telephone # :(732)4489266 > subhradip at yahoo.com | > Telephone # : (732)445-4604 | > ############################################################################## > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com -------------- next part -------------- A non-text attachment was scrubbed... Name: run_example_Ph_DOS Type: application/octet-stream Size: 448 bytes Desc: run_example_Ph_DOS Url : /pipermail/attachments/20030519/a81471aa/attachment.obj From lemiesz at mitr.p.lodz.pl Tue May 20 20:11:15 2003 From: lemiesz at mitr.p.lodz.pl (Michal Lemieszewski) Date: Tue, 20 May 2003 20:11:15 +0200 Subject: [Pw_forum] Bonding Energy Message-ID: <004801c31efb$3dd4d9d0$0cfa4dd5@ml> Dear PWscf users, I have tried to calculate bonding energy between Ni-Ni and Pd-Pd. Then I would try to calculate energy of a layer. The file to calculate Ni-Ni I have used from example1 file and modify it to my case, but I got "strange" results (to both calculations). Therefore I have two questions involving my calculations: 1. Can PWscf make so type of calculations? If so can You tell my what am I doing wrong? 2. How can I calculate celldm(1) in general (for example for Pd), I know the cell parameters but I don't know how to translate them to celldm(1). I think it should be exact value (or could it be estimated?). Please forgive me if those are trivial questions, but I am new in this field and I am learning ( I am undergraduate student). I also want to ask if You could recommend me some articles or ("good") web pages about Ni, Pd (e.g. experimental values) k-points and things which You think I should read to understand more about program and my problem (catalysis). I enclose example files for Ni. Thanks, Michal # # self-consistent calculation # cat > ni.scf0.in << EOF &control calculation='scf' restart_mode='from_scratch', pseudo_dir = '$PSEUDO_DIR/', outdir='$TMP_DIR/' prefix='ni' tprnfor = .true., tstress = .true. / &system ibrav=2, celldm(1) =6.48, nat=1, ntyp=1, nspin = 2, starting_magnetization(1)=0.7, ecutwfc = 24.0, ecutrho = 288.0, occupations='smearing', smearing='methfessel-paxton', degauss=0.02 / &electrons conv_thr = 1.0e-8 mixing_beta = 0.7 / ATOMIC_SPECIES Ni 58.69 NiUS.RRKJ3.UPF ATOMIC_POSITIONS{angstrom} Ni 0.0 0.0 0.0 K_POINTS{gamma} EOF # # self-consistent calculation # for i in `seq 1.0 0.1 5.0`;do cat > ni.scf$i.in << EOF &control calculation='scf' restart_mode='from_scratch', pseudo_dir = '$PSEUDO_DIR/', outdir='$TMP_DIR/' prefix='ni' tprnfor = .true., tstress = .true. / &system ibrav=2, celldm(1) =6.48, nat=2, ntyp=1, nspin = 2, starting_magnetization(1)=0.7, ecutwfc = 24.0, ecutrho = 288.0, occupations='smearing', smearing='methfessel-paxton', degauss=0.02 / &electrons conv_thr = 1.0e-8 mixing_beta = 0.7 / ATOMIC_SPECIES Ni 58.69 NiUS.RRKJ3.UPF ATOMIC_POSITIONS{angstrom} Ni 0.0 0.0 0.0 Ni $i 0.0 0.0 K_POINTS{gamma} EOF From eyvaz_isaev at yahoo.com Tue May 20 20:56:10 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 20 May 2003 11:56:10 -0700 (PDT) Subject: [Pw_forum] Bonding Energy In-Reply-To: <004801c31efb$3dd4d9d0$0cfa4dd5@ml> Message-ID: <20030520185610.19084.qmail@web80508.mail.yahoo.com> Dear Michal, I have just three comments: 1. You should not use only \Gamma point for small systems (where number of atoms <10). Rather you have to use integration via Monkhorst-Pack k-points which are generated automatically if you enter: K_POINTS{automatic} 10 10 10 0 0 0 which means that k-points will be generated according to Monkhorst-Pack schem, each direction of reciprocal basis vectors is divided by 10 and there is no shift of origin of k-vectors. 2. celldim(i) are given in atomic units, so, if you have lattice parameter of Pd which is 3.89Angstrom, you can convert it to atomic unit using 1Ang=1./0.529177, i.e. 3.89Ang=3.89/0.529177 (a.u.) For cubic structures only celldim(1) is required. 3. Probably, in the second script something is wrong. You chose fcc lattice within 2 atoms. Sure, this happens, but I suggest you should consider 2 atoms in the simple cubic cell (like isolated atoms). If so, you should vary (increase) cell parameter (celldim(1)) in order get isolated atoms. Besides your > ATOMIC_POSITIONS{angstrom} > Ni 0.0 0.0 0.0 > Ni $i 0.0 0.0 looks not correct. You mention {angstrom}, but I suggest in your case it should be {alat}. Though it depends on what do you want. What concerns good (really good) papers I advice M.C.Payne, M.P.Teter et. al., Reviews of Modern Physics, 64 (1992), p.1045. It should be as just starting paper. Regards, Eyvaz. Theoretical Physics Department Moscow Steel and Alloys Institute --- Michal Lemieszewski wrote: > Dear PWscf users, > > I have tried to calculate bonding energy between > Ni-Ni and Pd-Pd. Then I > would try to calculate energy of a layer. > The file to calculate Ni-Ni I have used from > example1 file and modify it to > my case, but I got "strange" results (to both > calculations). > Therefore I have two questions involving my > calculations: > 1. Can PWscf make so type of calculations? If so can > You tell my what am I > doing wrong? > 2. How can I calculate celldm(1) in general (for > example for Pd), I know the > cell parameters but I > don't know how to translate them to celldm(1). > I think it should be exact value (or could it be > estimated?). > > Please forgive me if those are trivial questions, > but I am new in this field > and I am learning ( I am undergraduate student). > I also want to ask if You could recommend me some > articles or ("good") web > pages about Ni, Pd (e.g. experimental values) > k-points and > things which You think I should read to understand > more about program and my > problem (catalysis). > > I enclose example files for Ni. > > Thanks, > Michal > > # > # self-consistent calculation > # > cat > ni.scf0.in << EOF > &control > calculation='scf' > restart_mode='from_scratch', > pseudo_dir = '$PSEUDO_DIR/', > outdir='$TMP_DIR/' > prefix='ni' > tprnfor = .true., > tstress = .true. > / > &system > ibrav=2, celldm(1) =6.48, nat=1, ntyp=1, > nspin = 2, starting_magnetization(1)=0.7, > ecutwfc = 24.0, ecutrho = 288.0, > occupations='smearing', > smearing='methfessel-paxton', degauss=0.02 > / > &electrons > conv_thr = 1.0e-8 > mixing_beta = 0.7 > / > ATOMIC_SPECIES > Ni 58.69 NiUS.RRKJ3.UPF > ATOMIC_POSITIONS{angstrom} > Ni 0.0 0.0 0.0 > K_POINTS{gamma} > EOF > > # > # self-consistent calculation > # > for i in `seq 1.0 0.1 5.0`;do > cat > ni.scf$i.in << EOF > &control > calculation='scf' > restart_mode='from_scratch', > pseudo_dir = '$PSEUDO_DIR/', > outdir='$TMP_DIR/' > prefix='ni' > tprnfor = .true., > tstress = .true. > / > &system > ibrav=2, celldm(1) =6.48, nat=2, ntyp=1, > nspin = 2, starting_magnetization(1)=0.7, > ecutwfc = 24.0, ecutrho = 288.0, > occupations='smearing', > smearing='methfessel-paxton', degauss=0.02 > / > &electrons > conv_thr = 1.0e-8 > mixing_beta = 0.7 > / > ATOMIC_SPECIES > Ni 58.69 NiUS.RRKJ3.UPF > ATOMIC_POSITIONS{angstrom} > Ni 0.0 0.0 0.0 > Ni $i 0.0 0.0 > K_POINTS{gamma} > EOF > > > > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com From felipe at titan.ipicyt.edu.mx Tue May 20 23:34:50 2003 From: felipe at titan.ipicyt.edu.mx (FELIPE VALENCIA) Date: Tue, 20 May 2003 16:34:50 -0500 (CDT) Subject: [Pw_forum] Bonding Energy In-Reply-To: <004801c31efb$3dd4d9d0$0cfa4dd5@ml> Message-ID: Dear MIchael, > Dear PWscf users, > > I have tried to calculate bonding energy between Ni-Ni and Pd-Pd. Then I > would try to calculate energy of a layer. > The file to calculate Ni-Ni I have used from example1 file and modify it to > my case, but I got "strange" results (to both calculations). Well, the files from example1 are intended for crystall calculations, so if you are modeling molecular systems, you must change a few things. First, as stated by Eyvaz, you should better use a simple cubic lattice (ibrav=1) or even better a free lattice (ibrav=0, and take a look to example3) Next, remember that atomic coordinatives are written scaled to the lattice parameter ( celldm(1) ), and that the lattice parameter is written in atomic units. As pwscf makes use of plane wave expantions, whenever you calculate a molecule, you are also considering its "images" wich have a peridicity given by alat (celldm(1)), In order to minimize the ghost interacitions, you should increase alat until the results are no longer dependent on it. That is why it is better to use the free lattice: you can choose celldm(1) = 1.0, so that the atomic coordinates are written in atomic units, and then change the lattice vectores (again see example3) until the image interactions get negligible. Hope it will be usefull, Felipe > Therefore I have two questions involving my calculations: > 1. Can PWscf make so type of calculations? If so can You tell my what am I > doing wrong? > 2. How can I calculate celldm(1) in general (for example for Pd), I know the > cell parameters but I > don't know how to translate them to celldm(1). > I think it should be exact value (or could it be estimated?). > > Please forgive me if those are trivial questions, but I am new in this field > and I am learning ( I am undergraduate student). > I also want to ask if You could recommend me some articles or ("good") web > pages about Ni, Pd (e.g. experimental values) k-points and > things which You think I should read to understand more about program and my > problem (catalysis). > > I enclose example files for Ni. > > Thanks, > Michal > > # > # self-consistent calculation > # > cat > ni.scf0.in << EOF > &control > calculation='scf' > restart_mode='from_scratch', > pseudo_dir = '$PSEUDO_DIR/', > outdir='$TMP_DIR/' > prefix='ni' > tprnfor = .true., > tstress = .true. > / > &system > ibrav=2, celldm(1) =6.48, nat=1, ntyp=1, > nspin = 2, starting_magnetization(1)=0.7, > ecutwfc = 24.0, ecutrho = 288.0, > occupations='smearing', smearing='methfessel-paxton', degauss=0.02 > / > &electrons > conv_thr = 1.0e-8 > mixing_beta = 0.7 > / > ATOMIC_SPECIES > Ni 58.69 NiUS.RRKJ3.UPF > ATOMIC_POSITIONS{angstrom} > Ni 0.0 0.0 0.0 > K_POINTS{gamma} > EOF > > # > # self-consistent calculation > # > for i in `seq 1.0 0.1 5.0`;do > cat > ni.scf$i.in << EOF > &control > calculation='scf' > restart_mode='from_scratch', > pseudo_dir = '$PSEUDO_DIR/', > outdir='$TMP_DIR/' > prefix='ni' > tprnfor = .true., > tstress = .true. > / > &system > ibrav=2, celldm(1) =6.48, nat=2, ntyp=1, > nspin = 2, starting_magnetization(1)=0.7, > ecutwfc = 24.0, ecutrho = 288.0, > occupations='smearing', smearing='methfessel-paxton', degauss=0.02 > / > &electrons > conv_thr = 1.0e-8 > mixing_beta = 0.7 > / > ATOMIC_SPECIES > Ni 58.69 NiUS.RRKJ3.UPF > ATOMIC_POSITIONS{angstrom} > Ni 0.0 0.0 0.0 > Ni $i 0.0 0.0 > K_POINTS{gamma} > EOF > > > > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From felipe at titan.ipicyt.edu.mx Tue May 20 23:41:44 2003 From: felipe at titan.ipicyt.edu.mx (FELIPE VALENCIA) Date: Tue, 20 May 2003 16:41:44 -0500 (CDT) Subject: [Pw_forum] Bonding Energy In-Reply-To: <20030520185610.19084.qmail@web80508.mail.yahoo.com> Message-ID: > 1. You should not use only \Gamma point for small > systems (where number of atoms <10). Rather you have > to use integration via Monkhorst-Pack k-points which > are generated automatically if you enter: > > K_POINTS{automatic} > 10 10 10 > 0 0 0 > > which means that k-points will be generated according > to Monkhorst-Pack schem, each direction of reciprocal > basis vectors is divided by 10 and there is no shift > of origin of k-vectors. > Concernig this point I disagree with Eyvaz, It is true that in your actual input files you are dealing with small systes, but if your final goal is to study molecular systems, you are going to increase the lattice parameter... the real system has no peridicity (i.e. an infinite lattice parameter) and the reciprocal space reduces then to the gamma point, so that Gamma point seems to me the logical for molecules... Felipe From eyvaz_isaev at yahoo.com Wed May 21 00:05:57 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 20 May 2003 15:05:57 -0700 (PDT) Subject: [Pw_forum] Bonding Energy In-Reply-To: Message-ID: <20030520220557.4210.qmail@web80509.mail.yahoo.com> Dear Felipe, There is no disagree, sure more k-points for small systems as I quoted in my previous mail, but \gamma point for a big cell (supercell) with large cell parameters even there is just one atom. Regards, Eyvaz. --- FELIPE VALENCIA wrote: > > 1. You should not use only \Gamma point for small > > systems (where number of atoms <10). Rather you > have > > to use integration via Monkhorst-Pack k-points > which > > are generated automatically if you enter: > > > > K_POINTS{automatic} > > 10 10 10 > > 0 0 0 > > > > which means that k-points will be generated > according > > to Monkhorst-Pack schem, each direction of > reciprocal > > basis vectors is divided by 10 and there is no > shift > > of origin of k-vectors. > > > > Concernig this point I disagree with Eyvaz, It is > true that in your > actual input files you are dealing with small > systes, but if your final > goal is to study molecular systems, you are going > to increase the lattice > parameter... the real system has no peridicity (i.e. > an infinite lattice > parameter) and the reciprocal space reduces then to > the gamma point, so > that Gamma point seems to me the logical for > molecules... > > Felipe > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com From xhongjun at mail.ustc.edu.cn Wed May 21 08:26:34 2003 From: xhongjun at mail.ustc.edu.cn (xhongjun) Date: Wed, 21 May 2003 14:26:34 +0800 Subject: [Pw_forum] about symmetrize the self-consistent potential of the perturbations In-Reply-To: <004801c31efb$3dd4d9d0$0cfa4dd5@ml> References: <004801c31efb$3dd4d9d0$0cfa4dd5@ml> Message-ID: <3ECB1C1A.1010905@mail.ustc.edu.cn> Dear PWSCF users: I don't understand the principles that symdvscf uses to symmetrize the self-consistent potential of the perturbations when I am reading the pwscf code. Thanks for any suggestion. Best regards. xianghjun From giannozz at nest.sns.it Wed May 21 11:04:07 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 21 May 2003 11:04:07 +0200 Subject: [Pw_forum] Bonding Energy In-Reply-To: <20030520220557.4210.qmail@web80509.mail.yahoo.com> References: <20030520220557.4210.qmail@web80509.mail.yahoo.com> Message-ID: <200305211104.07480.giannozz@nest.sns.it> Hi > There is no disagree, sure more k-points for small > systems as I quoted in my previous mail, but \gamma > point for a big cell (supercell) with large cell > parameters even there is just one atom. for molecular systems, the choice of Gamma is justified even if the supercell is not especially big. If Michal wants to calculate the energy of the Ni_2 dimer (is this correct?) Gamma is fine, and a large supercell is not required. > you should better use a simple cubic lattice (ibrav=1) > or even better a free lattice (ibrav=0, and take a look to example3) in principle, however, one can use whatever form of the supercell, provided that it is big enough to keep periodic replicas at a sufficient distance. > remember that atomic coordinatives are written scaled > to the lattice parameter ( celldm(1) ) this is the default, but it is overridden by the option in the ATOMIC_POSITIONS card: > ATOMIC_POSITIONS{angstrom} means that input atomic positions are in Angstrom (they are reprinted in a0=celldm(1) units on output) Warning: transition metals are nasty beasts. You may have trouble with both pseudopotentials and DFT. LDA badly overestimates cohesive 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 Damien.Connetable at lpmcn.univ-lyon1.fr Wed May 21 17:46:19 2003 From: Damien.Connetable at lpmcn.univ-lyon1.fr (Damien Connetable) Date: Wed, 21 May 2003 17:46:19 +0200 Subject: [Pw_forum] projected dos on local basis Message-ID: <3ECB9F4B.90803@lpmcn.univ-lyon1.fr> Dear PWSCF users, I would like to plot the density of state on atom orbitals (projected on the s, p and d shells). I suppose that the programme named " projected_dos.f90" can probably do it. But there is no documentation, and I don't know how to use it. Is there somebody who arrives to use it, and arrives to explain me how to use it? Or maybe is there another way to plot this quantity with PWSCF? Thanks Damien From francescoantoniella at yahoo.it Wed May 21 23:02:29 2003 From: francescoantoniella at yahoo.it (Francesco Antoniella) Date: 21 May 2003 23:02:29 +0200 Subject: [Pw_forum] projected dos on local basis In-Reply-To: <3ECB9F4B.90803@lpmcn.univ-lyon1.fr> References: <3ECB9F4B.90803@lpmcn.univ-lyon1.fr> Message-ID: <1053550982.2274.16.camel@localhost.localdomain> Il mer, 2003-05-21 alle 17:46, Damien Connetable ha scritto: > Dear PWSCF users, > > I would like to plot the density of state on atom orbitals (projected > on the s, p and d shells). > I suppose that the programme named " projected_dos.f90" can probably do it. > But there is no documentation, and I don't know how to use it. > Is there somebody who arrives to use it, and arrives to explain me how > to use it? Ok guy, I'm the writer of the routine. You need before to create the projections on the atomic orbitals via the projefc.x program. I'assume you are aware of how to do this step. After you must check if your projwfc is the version modified by me that give TWO output files : the one you named in the input of projwfc.x and the same with ".raw" appended. Is the latter the one needed for the program projected_dos You must give a namelist like &input filename="site/where/.rawfile/is" alpha=0.05 !the smearing of the DOS ei=10 ef=5 ! to have the dos plot form -10:5 ev , efermi beeing the 0 nstep=501 (default) number of interpolation steps. / If you don't have the modified projwfc.x I will give you the (few) lines to be put in the source. BestRegards Francesco > Or maybe is there another way to plot this quantity with PWSCF? > > Thanks > > Damien > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From Damien.Connetable at lpmcn.univ-lyon1.fr Thu May 22 13:51:16 2003 From: Damien.Connetable at lpmcn.univ-lyon1.fr (Damien Connetable) Date: Thu, 22 May 2003 13:51:16 +0200 Subject: [Pw_forum] projected dos on local basis In-Reply-To: <1053550982.2274.16.camel@localhost.localdomain> References: <3ECB9F4B.90803@lpmcn.univ-lyon1.fr> <1053550982.2274.16.camel@localhost.localdomain> Message-ID: <3ECCB9B4.70605@lpmcn.univ-lyon1.fr> Hi, I arrived to use the ``projefc.x'' program, however I did not arrive to use the ``projected_dos'' program, because the output format is not good. Then, I have tried to find in the different PWSCF versions the good ``projefc.x'' which can generate the good output file, whithout results. Could you specify to me the modifications ? Thank you in advance for your kindness. Regards, Damien >Ok guy, I'm the writer of the routine. >You need before to create the projections on the atomic orbitals via the >projefc.x program. >I'assume you are aware of how to do this step. >After you must check if your projwfc is the version modified by me that >give TWO output files : the one you named in the input of projwfc.x and >the same with ".raw" appended. >Is the latter the one needed for the program projected_dos >You must give a namelist like >&input >filename="site/where/.rawfile/is" >alpha=0.05 !the smearing of the DOS >ei=10 >ef=5 ! to have the dos plot form -10:5 ev , efermi beeing the 0 >nstep=501 (default) number of interpolation steps. >/ > >If you don't have the modified projwfc.x I will give you the (few) lines >to be put in the source. >BestRegards >Francesco > > From francesco.antoniella at aquila.infn.it Thu May 22 14:10:45 2003 From: francesco.antoniella at aquila.infn.it (Francesco Antoniella) Date: 22 May 2003 14:10:45 +0200 Subject: [Pw_forum] projected dos on local basis In-Reply-To: <3ECCB9B4.70605@lpmcn.univ-lyon1.fr> References: <3ECB9F4B.90803@lpmcn.univ-lyon1.fr> <1053550982.2274.16.camel@localhost.localdomain> <3ECCB9B4.70605@lpmcn.univ-lyon1.fr> Message-ID: <1053605455.2083.7.camel@localhost.localdomain> Il gio, 2003-05-22 alle 13:51, Damien Connetable ha scritto: > Hi, > > I arrived to use the ``projefc.x'' program, however > I did not arrive to use the ``projected_dos'' program, because the > output format is not good. Then, I have tried to find in the different PWSCF versions the > good ``projefc.x'' which can generate the good output file, whithout results. > Could you specify to me the modifications ? > > Thank you in advance for your kindness. > Regards, I will send you the whole projwave.f90 because there are modifications at least in two places. This file refers to the PW_112 version but I think it's rather simple to attach to other versions. Try before to make a diff to see wich places are different!! Ciao Francesco > > Damien > > >Ok guy, I'm the writer of the routine. > >You need before to create the projections on the atomic orbitals via the > >projefc.x program. > >I'assume you are aware of how to do this step. > >After you must check if your projwfc is the version modified by me that > >give TWO output files : the one you named in the input of projwfc.x and > >the same with ".raw" appended. > >Is the latter the one needed for the program projected_dos > >You must give a namelist like > >&input > >filename="site/where/.rawfile/is" > >alpha=0.05 !the smearing of the DOS > >ei=10 > >ef=5 ! to have the dos plot form -10:5 ev , efermi beeing the 0 > >nstep=501 (default) number of interpolation steps. > >/ > > > >If you don't have the modified projwfc.x I will give you the (few) lines > >to be put in the source. > >BestRegards > >Francesco > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- ! ! Copyright (C) 2001 PWSCF group ! This file is distributed under the terms of the ! GNU General Public License. See the file `License' ! in the root directory of the present distribution, ! or http://www.gnu.org/copyleft/gpl.txt . ! !----------------------------------------------------------------------- subroutine do_projwfc (nodenumber) !----------------------------------------------------------------------- ! ! projects wavefunctions onto orthogonalized atomic wavefunctions ! calculates Lowdin charges, spilling parameter ! input: namelist "&inputpp", with variables ! filpun input file saved by program pwscf ! tmp_dir temporary directory where file filpun resides ! filproj output file containing the results ! use pwcom use io implicit none character (len=3) :: nodenumber character (len=14) :: filproj integer :: ios namelist / inputpp / tmp_dir, filpun, filproj ! nd_nmbr = nodenumber ! ! set default values for variables in namelist ! filpun = ' ' tmp_dir = './' filproj = ' ' ! read (5, inputpp, err = 200, iostat = ios) 200 call error ('projwave', 'reading inputpp namelist', abs (ios) ) ! ! Now allocate space for pwscf variables, read and check them. ! call read_file call openfil ! call projwave (filproj) ! return end subroutine do_projwfc !----------------------------------------------------------------------- subroutine projwave (filproj) !----------------------------------------------------------------------- ! #include "machine.h" use allocate use pwcom use io #ifdef PARA use para #endif implicit none character (len=14) :: filproj character (len=18) :: filprojraw ! type wfc_label integer na, n, l, m end type wfc_label type(wfc_label), allocatable :: nlmchi(:) ! integer :: ik, ibnd, i, j, k, na, nb, nt, isym, n, m, m1, l, lm, nwfc,& nwfc1, lmax_wfc, np, npseu logical :: exst real(kind=DP) :: psum, totcharge_up, totcharge_down real(kind=DP), allocatable :: e (:), proj (:,:,:), charges_up(:,:), charges_down(:,:) complex(kind=DP), allocatable :: wfcatom (:,:), overlap (:,:), & work (:,:), work1(:), proj0(:,:) integer, allocatable :: index(:) ! ! if (filproj.eq.' ') return filprojraw=trim(filproj)//'.raw' write (6, '(/5x,''Calling projwave .... '', & & /5x,''Projections are written on file '',a,'' and '',a)') & filproj, filprojraw if (lsda) then write(6,'(/5x,'' Spin polarized calculation '')') if (.not.(mod(nkstot,2).eq.0)) then write(6,'(/5x,'' Wrong no of k points ... exiting'')') stop end if end if ! allocate(swfcatom (npwx , natomwfc ) ) allocate(wfcatom (npwx, natomwfc) ) allocate(proj (natomwfc, nbnd, nkstot) ) allocate(overlap (natomwfc, natomwfc) ) allocate(e (natomwfc) ) proj = 0.d0 overlap= (0.d0,0.d0) ! ! initialize D_Sl for l=1 and l=2, for l=0 D_S0 is 1 ! call d_matrix (d1, d2) ! ! fill structure nlmchi ! allocate (nlmchi(natomwfc)) nwfc=0 lmax_wfc = 0 do na = 1, nat nt = ityp (na) do n = 1, nchi (nt) if (oc (n, nt) .gt.0.d0.or..not.newpseudo (nt) ) then l = lchi (n, nt) lmax_wfc = max (lmax_wfc, l ) do m = 1, 2 * l + 1 nwfc=nwfc+1 nlmchi(nwfc)%na = na nlmchi(nwfc)%n = n nlmchi(nwfc)%l = l nlmchi(nwfc)%m = m enddo endif enddo enddo if (lmax_wfc.gt.2) call error ('projwave', 'l > 2 not yet implemented', 1) if (nwfc.ne.natomwfc) call error ('projwave', 'wrong # of atomic wfcs?', 1) ! ! loop on k points ! call init_us_1 ! do ik = 1, nks npw = npwx call gk_sort (xk (1, ik), ngm, g, ecut (maxter) / tpiba2, npw, & igk, g2kin) call davcio (evc, nwordwfc, iunwfc, ik, - 1) call atomic_wfc (ik, wfcatom) call init_us_2 (npw, igk, xk (1, ik), vkb) call ccalbec (nkb, npwx, npw, natomwfc, becp, vkb, wfcatom) call s_psi (npwx, npw, natomwfc, wfcatom, swfcatom) ! ! wfcatom = |phi_i> , swfcatom = \hat S |phi_i> ! calculate overlap matrix O_ij = ! call ZGEMM ('c', 'n', natomwfc, natomwfc, npw, (1.d0, 0.d0) , & wfcatom, npwx, swfcatom, npwx, (0.d0, 0.d0) , overlap, natomwfc) #ifdef PARA call reduce (2 * natomwfc * natomwfc, overlap) #endif ! ! calculate O^{-1/2} ! allocate(work (natomwfc, natomwfc) ) call cdiagh (natomwfc, overlap, natomwfc, e, work) do i = 1, natomwfc e (i) = 1.d0 / dsqrt (e (i) ) enddo do i = 1, natomwfc do j = i, natomwfc overlap (i, j) = (0.d0, 0.d0) do k = 1, natomwfc overlap (i, j) = overlap (i, j) + e (k) * work (j, k) * conjg (work (i, k) ) enddo if (j.ne.i) overlap (j, i) = conjg (overlap (i, j)) enddo enddo deallocate (work) ! ! calculate wfcatom = O^{-1/2} \hat S | phi> ! call ZGEMM ('n', 't', npw, natomwfc, natomwfc, (1.d0, 0.d0) , & swfcatom, npwx, overlap, natomwfc, (0.d0, 0.d0), wfcatom, npwx) ! ! make the projection ! allocate(proj0(natomwfc,nbnd) ) call ZGEMM ('c', 'n', natomwfc, nbnd, npw, (1.d0, 0.d0) , & wfcatom, npwx, evc, npwx, (0.d0, 0.d0) , proj0, natomwfc) #ifdef PARA call reduce (2 * natomwfc * nbnd, proj0) #endif ! ! symmetrize the projections ! allocate(work1 (nbnd) ) do nwfc = 1, natomwfc ! ! atomic wavefunction nwfc is on atom na ! na= nlmchi(nwfc)%na n = nlmchi(nwfc)%n l = nlmchi(nwfc)%l m = nlmchi(nwfc)%m ! !write(*,'(3(2x,2f18.12))') proj0((nwfc1+1):(nwfc1+3),2) do isym = 1, nsym nb = irt (isym, na) do nwfc1 =1, natomwfc if (nlmchi(nwfc1)%na.eq. nb .and. & nlmchi(nwfc1)%n .eq. nlmchi(nwfc)%n .and. & nlmchi(nwfc1)%l .eq. nlmchi(nwfc)%l .and. & nlmchi(nwfc1)%m .eq. 1 ) go to 10 end do call error('projwave','cannot symmetrize',1) 10 nwfc1=nwfc1-1 !print*, nwfc, nwfc1, isym, na, nb ! ! nwfc1 is the first rotated atomic wfc corresponding to nwfc ! if (l.eq.0) then work1(:) = proj0 (nwfc1 + 1,:) else if (l.eq.1) then work1(:) = 0.d0 do m1 = 1, 3 work1(:) = work1(:) + d1 (m, m1, isym) * & proj0 (nwfc1 + m1,:) enddo ! print*, isym, work1(2) !print*, proj0((nwfc1 + 1):(nwfc1+3),2) else if (l.eq.2) then work1(:) = 0.d0 do m1 = 1, 5 work1(:) = work1(:) + d2 (m, m1, isym) * & proj0 (nwfc1 + m1,:) enddo endif do ibnd = 1, nbnd proj (nwfc, ibnd, ik) = proj (nwfc, ibnd, ik) + & work1(ibnd) * conjg (work1(ibnd)) / nsym !print*,proj(nwfc,2,ik) enddo write(*,'(f18.12,2x,2f18.12)') proj(nwfc,2,ik), work1(2) enddo enddo deallocate (work1) deallocate (proj0 ) ! on k-points enddo #ifdef PARA ! ! recover the vector proj over the pools ! call poolrecover (et, nbndx, nkstot, nks) call poolrecover (proj, nbnd * natomwfc, nkstot, nks) ! if (me.eq.1.and.mypool.eq.1) then #endif ! ! write on the output file ! call seqopn (4, filproj, 'formatted', exst) call seqopn (29, filprojraw, 'formatted', exst) write (4,'(/"Projection on atomic states:"/)') write (29, * ) natomwfc, nbnd, nkstot, ef write (29, * ) ntyp, nat, lsda do np = 1, ntyp write (29, * ) np, nchi (np), (lchi (n, np), n = 1, nchi (np) ) enddo write (29, * ) (ityp (na), na = 1, nat) do nwfc = 1, natomwfc write(4,'(5x,"state #",i3,": atom ",i3," (",a3,"), wfc ",i2, & & " (l=",i1," m=",i2,")")') & nwfc, nlmchi(nwfc)%na, atm(ityp(nlmchi(nwfc)%na)), & nlmchi(nwfc)%n, nlmchi(nwfc)%l, nlmchi(nwfc)%m end do ! allocate(index (natomwfc) ) do ik = 1, nkstot write (29, '(4f14.10)') (xk (i, ik) , i = 1, 3) , wk (ik) if (lsda) then if (ik.eq.1) write(4,'(/,6x,''----------SPIN UP----------'')') if (ik.eq.(nkstot/2+1)) write(4,'(/,6x,''----------SPIN DOWN----------'')') end if write (4, '(/" k = ",3f14.10)') (xk (i, ik) , i = 1, 3) do ibnd = 1, nbnd write (29, '(f14.10)') et (ibnd, ik) * rytoev write (29, '(8f10.6)') (proj (i, ibnd, ik) , i = 1, natomwfc+1) write (4, '(5x,"e = ",f14.10," eV")') et (ibnd, ik) * rytoev ! ! sort projections by magnitude, in decreasing order ! do nwfc = 1, natomwfc index (nwfc) = 0 e (nwfc) = - proj (nwfc, ibnd, ik) end do call hpsort (natomwfc, e, index) ! ! only projections that are larger than 0.001 are written ! do nwfc = 1, natomwfc e (nwfc) = - e(nwfc) if ( abs (e(nwfc)).lt.0.001 ) go to 20 end do nwfc = natomwfc + 1 20 nwfc = nwfc -1 ! ! fancy (?!?) formatting ! write (4, '(5x,"psi = ",5(f5.3,"*[#",i3,"]+"))') & (e (i), index(i), i = 1, min(5,nwfc)) do j = 1, (nwfc-1)/5 write (4, '(10x,"+",5(f5.3,"*[#",i3,"]+"))') & (e (i), index(i), i = 5*j+1, min(5*(j+1),nwfc)) end do psum = 0.d0 do nwfc = 1, natomwfc psum = psum + proj (nwfc, ibnd, ik) end do write (4, '(4x,"|psi|^2 = ",f5.3)') psum ! enddo enddo deallocate (index) ! ! estimate partial charges (Loewdin) on each atom ! allocate ( charges_up (nat, 0:lmax_wfc ) ) allocate ( charges_down (nat, 0:lmax_wfc ) ) charges_up=0.0 charges_down=0.0 do ik = 1, nkstot do ibnd = 1, nbnd do nwfc = 1, natomwfc na= nlmchi(nwfc)%na l = nlmchi(nwfc)%l if((.not.(lsda)).or.((lsda).and.(ik.le.(nkstot/2)))) then charges_up(na,l) = charges_up(na,l) + wg (ibnd,ik) * & proj (nwfc, ibnd, ik) else charges_down(na,l) = charges_down(na,l) + wg (ibnd,ik) * & proj (nwfc, ibnd, ik) end if enddo end do end do ! write (4, '(/"Lowdin Charges: "/)') ! psum = 0.0 do na = 1, nat totcharge_up = 0.d0 totcharge_down = 0.d0 do l = 0, lmax_wfc totcharge_up = totcharge_up + charges_up(na,l) totcharge_down = totcharge_down + charges_down(na,l) end do psum = psum + totcharge_up+totcharge_down if(.not.(lsda)) then write (4, '(5x,"Atom # ",i3,": total charge = ",f8.4, & & ", s, p, d = ",4f8.4 )') & na, totcharge_up, ( charges_up(na,l), l= 0,lmax_wfc) else write (4, '(5x,"Atom # ",i3,": total charge spin up = ",f8.4, & & ", s, p, d = ",4f8.4 )') & na, totcharge_up, ( charges_up(na,l), l= 0,lmax_wfc) write (4, '(5x," ",i3,": total charge spin down = ",f8.4, & & ", s, p, d = ",4f8.4 )') & na, totcharge_down, ( charges_down(na,l), l= 0,lmax_wfc) write (4, '(5x," ",i3,": total pol. (up-down) = ",f8.4, & & ", s, p, d = ",4f8.4 )') & na, totcharge_up-totcharge_down,& ( charges_up(na,l)-charges_down(na,l), l= 0,lmax_wfc) end if end do psum = psum / nelec write (4, '(5x,"Spilling Parameter: ",f8.4)') 1.0 - psum ! ! Sanchez-Portal et al., Sol. State Commun. 95, 685 (1995). ! The spilling parameter measures the ability of the basis provided by ! the pseudo-atomic wfc to represent the PW eigenstates, ! by measuring how much of the subspace of the Hamiltonian ! eigenstates falls outside the subspace spanned by the atomic basis ! close (unit=4) deallocate (charges_up,charges_down) #ifdef PARA endif #endif deallocate (nlmchi) deallocate (e) deallocate (overlap) deallocate (proj) deallocate (wfcatom) deallocate (swfcatom) return end subroutine projwave From yanming_ma at hotmail.com Fri May 23 16:37:27 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Fri, 23 May 2003 22:37:27 +0800 Subject: [Pw_forum] finding symmetry Message-ID: Hi, all, Concerning a problem about finding symmetry in PWSCF. In my several cases, when I run scf calculation with changing unit cell or ecutwfc. The code can not find symmetry as many as it should find. The only reason is I have changed the unit cell or ecutwfc. As Paolo's previous email about symmetry, this problem is from that the fractional translation is incompatible with the FFT grid. Is this behavior of finding less symmetries harmful to the calculation without concerning the calculating time? How to solve it if I have to fix the unit cell and ecutwfc? Best Wishes! Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 _________________________________________________________________ ??????????????? MSN Hotmail? http://www.hotmail.com From Damien.Connetable at lpmcn.univ-lyon1.fr Fri May 23 16:47:29 2003 From: Damien.Connetable at lpmcn.univ-lyon1.fr (Damien Connetable) Date: Fri, 23 May 2003 16:47:29 +0200 Subject: [Pw_forum] projected dos on local basis In-Reply-To: <1053605455.2083.7.camel@localhost.localdomain> References: <3ECB9F4B.90803@lpmcn.univ-lyon1.fr> <1053550982.2274.16.camel@localhost.localdomain> <3ECCB9B4.70605@lpmcn.univ-lyon1.fr> <1053605455.2083.7.camel@localhost.localdomain> Message-ID: <3ECE3481.9060106@lpmcn.univ-lyon1.fr> Thanks a lot, it works fine now damien >Il gio, 2003-05-22 alle 13:51, Damien Connetable ha scritto: > > >> Hi, >> >>I arrived to use the ``projefc.x'' program, however >>I did not arrive to use the ``projected_dos'' program, because the >>output format is not good. Then, I have tried to find in the different PWSCF versions the >>good ``projefc.x'' which can generate the good output file, whithout results. >>Could you specify to me the modifications ? >> >>Thank you in advance for your kindness. >>Regards, >> >> >I will send you the whole projwave.f90 because there are modifications >at least in two places. >This file refers to the PW_112 version but I think it's rather simple to >attach to other versions. >Try before to make a diff to see wich places are different!! >Ciao >Francesco > > > >>Damien >> >> >> >>>Ok guy, I'm the writer of the routine. >>>You need before to create the projections on the atomic orbitals via the >>>projefc.x program. >>>I'assume you are aware of how to do this step. >>>After you must check if your projwfc is the version modified by me that >>>give TWO output files : the one you named in the input of projwfc.x and >>>the same with ".raw" appended. >>>Is the latter the one needed for the program projected_dos >>>You must give a namelist like >>>&input >>>filename="site/where/.rawfile/is" >>>alpha=0.05 !the smearing of the DOS >>>ei=10 >>>ef=5 ! to have the dos plot form -10:5 ev , efermi beeing the 0 >>>nstep=501 (default) number of interpolation steps. >>>/ >>> >>>If you don't have the modified projwfc.x I will give you the (few) lines >>>to be put in the source. >>>BestRegards >>>Francesco >>> >>> >>> >>> >>_______________________________________________ >>Pw_forum mailing list >>Pw_forum at pwscf.org >>http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> > > > From baknoo at tokyo.rist.or.jp Mon May 26 13:43:07 2003 From: baknoo at tokyo.rist.or.jp (NoahPark) Date: Mon, 26 May 2003 20:43:07 +0900 Subject: [Pw_forum] A little question about total energy References: <20030516122930.5090.qmail@sina.com> <200305161528.49022.giannozz@nest.sns.it> Message-ID: <000e01c3237b$fa1add40$4b00a8c0@computat030u9q> Dear Users; If USPP and NCPP are simultaneously used in the description of a certain heterogeneous structure(with the same exchange-correlation functional), is there any problem to figure out the total energy and binding energy ? Best regards, Noah ----- Original Message ----- From: "Paolo Giannozzi" To: Sent: Friday, May 16, 2003 10:28 PM Subject: Re: [Pw_forum] A little question about total energy > Hi > > > In the output of PWSCF, the energy partitions are list > > following the total energy as I pasted below. I found > > that the total energy is not equal to the sum of those listed > > partitions. > > Total Energy = sum of terms marked with "+" below = > one-electron + hartree + xc + ewald + correction for metals > > ! total energy = -62.07569469 ryd > estimated scf accuracy < 0.00000799 ryd > > band energy sum = -14.82580512 ryd > + one-electron contribution = -98.79712762 ryd > + hartree contribution = 51.10053218 ryd > + xc contribution = -14.02297161 ryd > + ewald contribution = -0.34484385 ryd > scf in/out correction = 0.02901784 ryd > + correction for metals = -0.01128379 ryd > > The "one-electron contribution" is the kinetic + ionic term: > \sum > The "band energy sum" is exactly what it claims to be: > \sum epsilon = \sum > The "correction for metals" makes forces and energies > consistent if a broadening scheme is used. > The "scf in/out correction" is no longer used > (and no longer useful) > The "estimated scf accuracy" gives an estimate of the > error on the energy due to imperfect self-consistency. > > Admittedly, it is slightly confusing. > > 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 From giannozz at nest.sns.it Tue May 27 18:50:00 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 May 2003 18:50:00 +0200 Subject: [Pw_forum] A little question about total energy In-Reply-To: <000e01c3237b$fa1add40$4b00a8c0@computat030u9q> References: <20030516122930.5090.qmail@sina.com> <200305161528.49022.giannozz@nest.sns.it> <000e01c3237b$fa1add40$4b00a8c0@computat030u9q> Message-ID: <200305271850.00456.giannozz@nest.sns.it> On Monday 26 May 2003 13:43, NoahPark wrote: > If USPP and NCPP are simultaneously used in the description of a certain > heterogeneous structure(with the same exchange-correlation functional), > is there any problem to figure out the total energy and binding energy ? I am not aware of any problem 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 Tue May 27 19:00:06 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 27 May 2003 19:00:06 +0200 Subject: [Pw_forum] finding symmetry In-Reply-To: References: Message-ID: <200305271900.06486.giannozz@nest.sns.it> On Friday 23 May 2003 16:37, ma Yanming wrote: > [...] this problem is from that the fractional translation > is incompatible with the FFT grid. > Is this behavior of finding less symmetries harmful to the calculation > without concerning the calculating time? no, it is not. The system must converge to the solution with the good symmetry, whether it is imposed or not. > How to solve it if I have to fix the unit cell and ecutwfc? you can specify in input a slightly enlarged (or reduced, at your risk) FFT grid that is commensurate with the fractional translation 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 Tue May 27 20:56:33 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Wed, 28 May 2003 02:56:33 +0800 Subject: [Pw_forum] finding symmetry Message-ID: Dear paolo Thanks Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 >From: Paolo Giannozzi >Reply-To: pw_forum at pwscf.org >To: pw_forum at pwscf.org >Subject: Re: [Pw_forum] finding symmetry >Date: Tue, 27 May 2003 19:00:06 +0200 > >On Friday 23 May 2003 16:37, ma Yanming wrote: > > [...] this problem is from that the fractional translation > > is incompatible with the FFT grid. > > Is this behavior of finding less symmetries harmful to the calculation > > without concerning the calculating time? > >no, it is not. The system must converge to the solution with the >good symmetry, whether it is imposed or not. > > > How to solve it if I have to fix the unit cell and ecutwfc? > >you can specify in input a slightly enlarged (or reduced, at your risk) >FFT grid that is commensurate with the fractional translation > >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 _________________________________________________________________ ?????????????? MSN Messenger: http://messenger.msn.com/cn From rn_a at rambler.ru Thu May 29 10:10:44 2003 From: rn_a at rambler.ru (Роман Назаров) Date: Thu, 29 May 2003 12:10:44 +0400 (MSD) Subject: [Pw_forum] [Fwd: ] Message-ID: <3ED5C084.AA16536@mb1.rambler.ru> ___________________ ????????? "???????-2" - ?? ???????????, ???? ?????? 40% ?? ?????? ??? ???: http://www.rambler.ru/nettrader/ -------------- next part -------------- An embedded message was scrubbed... From: ????? ??????? Subject: Date: Wed, 28 May 2003 09:18:46 +0400 (MSD) Size: 1350 Url: /pipermail/attachments/20030529/7852f372/attachment.eml From eyvaz_isaev at yahoo.com Thu May 29 12:58:38 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 29 May 2003 03:58:38 -0700 (PDT) Subject: [Pw_forum] [Fwd: ] In-Reply-To: <3ED5C084.AA16536@mb1.rambler.ru> Message-ID: <20030529105838.55930.qmail@web80511.mail.yahoo.com> Dear Roman and Anton, You tried to compile using IBM XLF-compiler which, certainly, does not work on Linux/Intel (AMD) PCs. Besides, sure, it is not installed on your computer. It is commercial product. Please let me know what is your compiler (IFC, PGI, Absoft,...) you have installed on your computer. Then I will be able to help you compiling of the code. Regards, Eyvaz Isaev Theoretical Physics Department Moscow Institute of Steel and Alloys. --- ????? ??????? wrote: > > ___________________ > ????????? "???????-2" - ?? ???????????, ???? ?????? > 40% ?? ?????? ??? ???: > http://www.rambler.ru/nettrader/ > > ATTACHMENT part 2 message/rfc822 > From: ????? ??????? > To: pwscf at pwscf.org > CC: > Subject: > Date: Wed, 28 May 2003 09:18:46 +0400 (MSD) > > Dear sirs. > We are trying to use PWscf programme for the first > time. In addition we are > just begining to learn Linux. At this moment we have > the problem: after we > print "make pw", we receive the following answer: > ( cd Modules; make all ) > make[1]: Entering directory > '/home/roman/pwscf1/pwscf1/Modules' > mpxlf -qddim -qnosave -qfree=f90 -qinit=f90ptr > -qalias=noaryovrlp > -qalias=nointptr -I../include -03 -qstrict > -qarch=auto -qtune=auto > qsuffix=cpp=f90 -I/home/roman/pwscf1/pwscf1/install > -WF,DAIX,DPARA,-D__MPI,-DHAS_ZHEGVX > -I/home/roman/pwscf1/pwscf1/Modules > -I/home/roman/pwscf1/pwscf1/PW/ > -I/home/roman/pwscf1/pwscf1/PH/ -c kind.f90 > make[1]:mpxlf: Command not found > make[1]:***[kind.o] Error 127 > make[1]:Leaving directory > '/home/roman/pwscf1/pwscf1/Modules' > make:***[modules] Error2 > > Please, help to understand what are ours mistakes. > > Sincerely yours > Roman Nazarov, Anton Artamonov > > ___________________ > ????????? "???????-2" - ?? ???????????, ???? ?????? > 40% ?? ?????? ??? ???: > http://www.rambler.ru/nettrader/ > __________________________________ Do you Yahoo!? Yahoo! Calendar - Free online calendar with sync to Outlook(TM). http://calendar.yahoo.com From ilyeshamdi at yahoo.fr Fri May 30 12:56:43 2003 From: ilyeshamdi at yahoo.fr (=?iso-8859-1?q?ilyes=20hamdi?=) Date: Fri, 30 May 2003 12:56:43 +0200 (CEST) Subject: [Pw_forum] special points mesh Message-ID: <20030530105644.1654.qmail@web9908.mail.yahoo.com> Dear Users In PWSCF we use six prameters to generate special points in Monkhorst-Pack method. the 3 last prameters are defined to be 0 or 1. What's the meaning of these and where can i find their definition in the program. Last year i worked on CASTEP, and i have only to define the grid nk1,nk2,nk3. exemple structure ZB 4x4x4 gives 10 points in CASTEP is equivalent to 4x4x4 001 in PWSCF. Thank you all and good luck. ___________________________________________________________ Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en fran?ais ! Yahoo! Mail : http://fr.mail.yahoo.com From subhra at physics.rutgers.edu Fri May 30 14:50:34 2003 From: subhra at physics.rutgers.edu (Subhradip Ghosh) Date: Fri, 30 May 2003 08:50:34 -0400 (EDT) Subject: [Pw_forum] Phonon DOS Message-ID: Hi, I am trying to compute Phonon DOS using the code matdyn.f90. I found that it worked for an insulator but in my case the system is metallic and while I ran the code there is some output generated alongwith an error message %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from matdyn : error # -1 non-analytic term for q=0 missing ! %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% I traced it back to the code and found that this error message pops up in case we have no dielectric tensor(probably) written in the real-space force constant file. There is a logical switch in the force constant file which comes out to be F in my case and therefore there are no entries of the variable epsil. I have no idea what the remedy is? Is it a serious error or should I ignore it? Thanks in advance, Subhradip PS: My system is not a simple metal, rather a metallic alloy. ############################################################################ Dr Subhradip Ghosh Post Doctoral Fellow Condensed matter theory group Department of Physics and Astronomy ############################################################################ Office: | Residence: --------------------------------------------|------------------------------- Serin E255 |132, South Adelaide Avenue Department of Physics and Astronomy |Apartment# 2A Rutgers, The State University of New Jersey |Highland Park 136 Frelinghuysen Road |NJ 08904 Piscataway, NJ 08854-8019 |USA USA | Email address: subhra at physics.rutgers.edu, |Telephone # :(732)4489266 subhradip at yahoo.com | Telephone # : (732)445-4604 | ############################################################################## From eyvaz_isaev at yahoo.com Fri May 30 15:11:45 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 30 May 2003 06:11:45 -0700 (PDT) Subject: [Pw_forum] special points mesh In-Reply-To: <20030530105644.1654.qmail@web9908.mail.yahoo.com> Message-ID: <20030530131145.7179.qmail@web80509.mail.yahoo.com> Hi Ilyes, "nkx nky nkz kx ky kz" means that each basis vectors of a reciprocal lattice are divided by nkx, nky, and nkz, respectively. The three last numbers (usually 0 0 0) means that the origin of basis vectors are not shifted. Sometime a shift is required to avoid divergency problem near the \Gamma point. You can read INPUT_PW file in the /pwdocs directory (or in the core directory for old versions). Regards, Eyvaz Isaev Theoretical Physics Department Moscow Steel and Alloys Institute --- ilyes hamdi wrote: > Dear Users > In PWSCF we use six prameters to generate special > points in Monkhorst-Pack method. the 3 last > prameters > are defined to be 0 or 1. What's the meaning of > these > and where can i find their definition in the > program. > Last year i worked on CASTEP, and i have only to > define the grid nk1,nk2,nk3. exemple > structure ZB > 4x4x4 gives 10 points in CASTEP is equivalent to > 4x4x4 001 in PWSCF. > Thank you all > and good luck. > > > ___________________________________________________________ > Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et > en fran?ais ! > Yahoo! Mail : http://fr.mail.yahoo.com > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? Yahoo! Calendar - Free online calendar with sync to Outlook(TM). http://calendar.yahoo.com From eyvaz_isaev at yahoo.com Fri May 30 23:25:17 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 30 May 2003 14:25:17 -0700 (PDT) Subject: [Pw_forum] Phonon DOS In-Reply-To: Message-ID: <20030530212517.99497.qmail@web80507.mail.yahoo.com> Dear Subhradip, I assume it will not affect for your calculations if you have a metallic system. But it is very important for polar semiconductors where splitting of optical modes at the \Gamma point is not possible without applying electric field via calculation of macroscopic dielectric constant. The error means that electric field is not included. A reason for the message might be you did not mention degauss and ngauss parameters for SCF calculations (for old versions). For the new version there should be three paremeters: occupations, smearing, degauss (in "system" section). If there are not, your system is assumed to be semiconductor type. Besides, I remember my promise. I will do that soon. Regards, Eyvaz. Theoretical Physics Department Moscow Steel and Alloys Institute --- Subhradip Ghosh wrote: > Hi, > > I am trying to compute Phonon DOS using the code > matdyn.f90. I found that > it worked for an insulator but in my case the system > is metallic and while > I ran the code there is some output generated > alongwith an error message > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from matdyn : error # -1 > non-analytic term for q=0 missing ! > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > I traced it back to the code and found that this > error message pops up in > case we have no dielectric tensor(probably) written > in the real-space > force constant > file. There is a logical switch in the force > constant file which comes out > to be F in my case and therefore there are no > entries of the variable > epsil. I have no idea what the remedy is? Is it a > serious error or should > I ignore it? > > Thanks in advance, > > Subhradip > > PS: My system is not a simple metal, rather a > metallic alloy. > > ############################################################################ > Dr Subhradip Ghosh > Post Doctoral Fellow > Condensed matter theory group > Department of Physics and Astronomy > ############################################################################ > Office: | > Residence: > --------------------------------------------|------------------------------- > Serin E255 |132, > South Adelaide Avenue > Department of Physics and Astronomy > |Apartment# 2A > Rutgers, The State University of New Jersey > |Highland Park > 136 Frelinghuysen Road |NJ > 08904 > Piscataway, NJ 08854-8019 |USA > USA | > Email address: subhra at physics.rutgers.edu, > |Telephone # :(732)4489266 > subhradip at yahoo.com | > Telephone # : (732)445-4604 | > ############################################################################## > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? Yahoo! Calendar - Free online calendar with sync to Outlook(TM). http://calendar.yahoo.com From subhra at physics.rutgers.edu Fri May 30 23:30:44 2003 From: subhra at physics.rutgers.edu (Subhradip Ghosh) Date: Fri, 30 May 2003 17:30:44 -0400 (EDT) Subject: [Pw_forum] Phonon DOS In-Reply-To: <20030530212517.99497.qmail@web80507.mail.yahoo.com> References: <20030530212517.99497.qmail@web80507.mail.yahoo.com> Message-ID: On Fri, 30 May 2003, Eyvaz Isaev wrote: Thanks a lot. However, in the old version I mentioned ngauss and degauss always keeping in mind that it's a metal. That's why I am puzzled a bit. What I understand from your message is that even though the error message remains there is nothing wrong in the calculations and I can trust my results. Regards, Subhradip > Dear Subhradip, > > I assume it will not affect for your calculations if > you have a metallic system. > But it is very important for polar semiconductors > where > splitting of optical modes at the \Gamma point is not > possible without applying electric field via > calculation of macroscopic dielectric constant. > The error means that electric field is not included. > > A reason for the message might be you did not > mention degauss and ngauss parameters for SCF > calculations (for old versions). For the new version > there should be three paremeters: occupations, > smearing, degauss (in "system" section). If there are > not, your system is assumed to be semiconductor type. > > Besides, I remember my promise. I will do that soon. > > Regards, > Eyvaz. > Theoretical Physics Department > Moscow Steel and Alloys Institute > > --- Subhradip Ghosh > wrote: > > Hi, > > > > I am trying to compute Phonon DOS using the code > > matdyn.f90. I found that > > it worked for an insulator but in my case the system > > is metallic and while > > I ran the code there is some output generated > > alongwith an error message > > > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > from matdyn : error # -1 > > non-analytic term for q=0 missing ! > > > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > I traced it back to the code and found that this > > error message pops up in > > case we have no dielectric tensor(probably) written > > in the real-space > > force constant > > file. There is a logical switch in the force > > constant file which comes out > > to be F in my case and therefore there are no > > entries of the variable > > epsil. I have no idea what the remedy is? Is it a > > serious error or should > > I ignore it? > > > > Thanks in advance, > > > > Subhradip > > > > PS: My system is not a simple metal, rather a > > metallic alloy. > > > > > ############################################################################ > > Dr Subhradip Ghosh > > Post Doctoral Fellow > > Condensed matter theory group > > Department of Physics and Astronomy > > > ############################################################################ > > Office: | > > Residence: > > > --------------------------------------------|------------------------------- > > Serin E255 |132, > > South Adelaide Avenue > > Department of Physics and Astronomy > > |Apartment# 2A > > Rutgers, The State University of New Jersey > > |Highland Park > > 136 Frelinghuysen Road |NJ > > 08904 > > Piscataway, NJ 08854-8019 |USA > > USA | > > Email address: subhra at physics.rutgers.edu, > > |Telephone # :(732)4489266 > > subhradip at yahoo.com | > > Telephone # : (732)445-4604 | > > > ############################################################################## > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > __________________________________ > Do you Yahoo!? > Yahoo! Calendar - Free online calendar with sync to Outlook(TM). > http://calendar.yahoo.com > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ############################################################################ Dr Subhradip Ghosh Post Doctoral Fellow Condensed matter theory group Department of Physics and Astronomy ############################################################################ Office: | Residence: --------------------------------------------|------------------------------- Serin E255 |132, South Adelaide Avenue Department of Physics and Astronomy |Apartment# 2A Rutgers, The State University of New Jersey |Highland Park 136 Frelinghuysen Road |NJ 08904 Piscataway, NJ 08854-8019 |USA USA | Email address: subhra at physics.rutgers.edu, |Telephone # :(732)4489266 subhradip at yahoo.com | Telephone # : (732)445-4604 | ############################################################################## From eyvaz_isaev at yahoo.com Sat May 31 00:58:17 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 30 May 2003 15:58:17 -0700 (PDT) Subject: [Pw_forum] Phonon DOS In-Reply-To: Message-ID: <20030530225817.52663.qmail@web80509.mail.yahoo.com> Yes, I think so. You can try epsil=.true. for the \Gamma point and recalculate phonons at the \Gamma point only. Hopefully nothing should be changed for frequencies (there should be additionall lines in the dynamical matrix file) if you have metallic system. Nevertheless an origin of the message is not clear. May be due to smearing you have a little gap. The best test is the electronic DOS calculations. Regards, Eyvaz. --- Subhradip Ghosh wrote: > On Fri, 30 May 2003, Eyvaz Isaev wrote: > > Thanks a lot. However, in the old version I > mentioned ngauss and degauss > always keeping in mind that it's a metal. That's why > I am puzzled a bit. > What I understand from your message is that even > though the error message > remains there is nothing wrong in the calculations > and I can trust my > results. > > Regards, > > Subhradip > > > Dear Subhradip, > > > > I assume it will not affect for your calculations > if > > you have a metallic system. > > But it is very important for polar semiconductors > > where > > splitting of optical modes at the \Gamma point is > not > > possible without applying electric field via > > calculation of macroscopic dielectric constant. > > The error means that electric field is not > included. > > > > A reason for the message might be you did not > > mention degauss and ngauss parameters for SCF > > calculations (for old versions). For the new > version > > there should be three paremeters: occupations, > > smearing, degauss (in "system" section). If there > are > > not, your system is assumed to be semiconductor > type. > > > > Besides, I remember my promise. I will do that > soon. > > > > Regards, > > Eyvaz. > > Theoretical Physics Department > > Moscow Steel and Alloys Institute > > > > --- Subhradip Ghosh > > wrote: > > > Hi, > > > > > > I am trying to compute Phonon DOS using the code > > > matdyn.f90. I found that > > > it worked for an insulator but in my case the > system > > > is metallic and while > > > I ran the code there is some output generated > > > alongwith an error message > > > > > > > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > > from matdyn : error # -1 > > > non-analytic term for q=0 missing ! > > > > > > > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > > I traced it back to the code and found that this > > > error message pops up in > > > case we have no dielectric tensor(probably) > written > > > in the real-space > > > force constant > > > file. There is a logical switch in the force > > > constant file which comes out > > > to be F in my case and therefore there are no > > > entries of the variable > > > epsil. I have no idea what the remedy is? Is it > a > > > serious error or should > > > I ignore it? > > > > > > Thanks in advance, > > > > > > Subhradip > > > > > > PS: My system is not a simple metal, rather a > > > metallic alloy. > > > > > > > > > ############################################################################ > > > Dr Subhradip Ghosh > > > Post Doctoral Fellow > > > Condensed matter theory group > > > Department of Physics and Astronomy > > > > > > ############################################################################ > > > Office: | > > > Residence: > > > > > > --------------------------------------------|------------------------------- > > > Serin E255 > |132, > > > South Adelaide Avenue > > > Department of Physics and Astronomy > > > |Apartment# 2A > > > Rutgers, The State University of New Jersey > > > |Highland Park > > > 136 Frelinghuysen Road |NJ > > > 08904 > > > Piscataway, NJ 08854-8019 |USA > > > USA | > > > Email address: subhra at physics.rutgers.edu, > > > |Telephone # :(732)4489266 > > > subhradip at yahoo.com | > > > Telephone # : (732)445-4604 | > > > > > > ############################################################################## > > > > > > _______________________________________________ > > > Pw_forum mailing list > > > Pw_forum at pwscf.org > > > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > __________________________________ > > Do you Yahoo!? > > Yahoo! Calendar - Free online calendar with sync > to Outlook(TM). > > http://calendar.yahoo.com > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > ############################################################################ > Dr Subhradip Ghosh > Post Doctoral Fellow > Condensed matter theory group > Department of Physics and Astronomy > ############################################################################ > Office: | > Residence: > --------------------------------------------|------------------------------- > Serin E255 |132, > South Adelaide Avenue > Department of Physics and Astronomy > |Apartment# 2A > Rutgers, The State University of New Jersey > |Highland Park > 136 Frelinghuysen Road |NJ > 08904 > Piscataway, NJ 08854-8019 |USA > USA | > Email address: subhra at physics.rutgers.edu, > |Telephone # :(732)4489266 > subhradip at yahoo.com | > Telephone # : (732)445-4604 | > ############################################################################## > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? Yahoo! Calendar - Free online calendar with sync to Outlook(TM). http://calendar.yahoo.com