From Zahraa.Ibrahim at uoit.ca Sat Sep 1 04:52:23 2007 From: Zahraa.Ibrahim at uoit.ca (Zahraa Ibrahim) Date: Fri, 31 Aug 2007 22:52:23 -0400 Subject: [Pw_forum] Temperature extraction from PW References: Message-ID: Hello Antonio Thanks for your reply. The center of mass moves slightly along a given direction. So it is most probably not fixed, does the 3 maybe represent the three rotational degrees of freedom, which due to the translational symmetry of the cell can not be excited? Zahraa Ibrahim UOIT -----Original Message----- From: pw_forum-bounces at pwscf.org on behalf of Antonio Tilocca Sent: Fri 8/31/2007 7:10 AM To: pw_forum at pwscf.org Cc: Subject: Re: [Pw_forum] Temperature extraction from PW I think the extra 3 takes into account the translational (center of mass) degrees of freedom: if the center of mass is fixed, the total number of degrees of freedom decreases by 3. At the same time, the c.o.m. kinetic energy will be zero, so there is probably no need to subtract it from the total kinetic energy in the numerator. Antonio Tilocca UCL Message: 5 Date: Thu, 30 Aug 2007 13:21:11 -0400 From: "Zahraa Ibrahim" Subject: [Pw_forum] Temperature extraction from PW I noticed that the temperature in the espresso version is calculated from T = 2 * (kinetic energy - center of mass energy) / (3N-3) [I am not setting any constraints] while in the democritus version it was calculated by dividing the numerator by (3N). (N being the number of particles). I was wondering what exactly this extra 3 stands for. Another, probably related, question is about the effect of periodic boundary conditions on the temperature. I am simulating thermal motion in a solid with a relatively small number of atoms. Do the periodic boundary conditions change the degrees of freedom of the solid in any ways, that need to be taken into account in the temperature estimation. Greetings Zahraa Ibrahim _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From brsahu at physics.utexas.edu Sat Sep 1 06:23:47 2007 From: brsahu at physics.utexas.edu (brsahu at physics.utexas.edu) Date: Fri, 31 Aug 2007 23:23:47 -0500 Subject: [Pw_forum] compilation on Blue Gene/L Message-ID: <20070831232347.dx2yr27o6nfkg8w0@mail.ph.utexas.edu> Dear PWSCF users, I am trying to compile the latest version of pwscf code on Blue Gene/L. I used the Make.bgl file provided with the distribution and checked the "correctness" of all the entries in it w.r.t Blue Gene here. There is a compilation error at the preprocessing stage: 1501-510 Compilation successful for file iotk_write.f90. /opt/ibmcmp/xlf/9.1/bin/blrts_xlf90 -qalias=noaryovrlp -I../include -O3 -qstrict -qarch=440d -qtune=440 -qsuffix=cpp=f90 -qdpc -qalias=nointptr -Q -I. -I../include -I../Modules -I../PW -I../PH -I/bgl/BlueLight/ppcfloor/bglsys/include -I../iotk/src -qfree=f90 -I../include -WF,-D__AIX,-D__XLF,-D__MPI,-D__PARA,-D__MASS,-D__FFTW,-D__BGL,-D__TRUE_BGL -I. -I../include -I../Modules -I../PW -I../PH -I/bgl/BlueLight/ppcfloor/bglsys/include -I../iotk/src,-D__AIX,-D__XLF,-D__MPI,-D__PARA,-D__MASS,-D__FFTW,-D__BGL,-D__TRUE_BGL -I../Modules -I../PW -I../PH -c iotk_xtox.f90 ** iotk_atol_x === End of Compilation 1 === ** iotk_atoi1 === End of Compilation 2 === ** iotk_itoa1 === End of Compilation 3 === ** iotk_ltoa1 === End of Compilation 4 === 1501-510 Compilation successful for file iotk_xtox.f90. ar libiotk.a iotk_attr+CHARACTER1_0.o iotk_attr+COMPLEX1_0.o iotk_attr+COMPLEX1_3.o iotk_attr+COMPLEX2_0.o iotk_attr+COMPLEX2_3.o iotk_attr+INTEGER1_0.o iotk_attr+INTEGER1_3.o iotk_attr+INTEGER2_0.o iotk_attr+INTEGER2_3.o iotk_attr+LOGICAL1_0.o iotk_attr+LOGICAL1_3.o iotk_attr+LOGICAL2_0.o iotk_attr+LOGICAL2_3.o iotk_attr+REAL1_0.o iotk_attr+REAL1_3.o iotk_attr+REAL2_0.o iotk_attr+REAL2_3.o iotk_attr.o iotk_attr_interf.o iotk_base.o iotk_dat+CHARACTER1_0.o iotk_dat+CHARACTER1_3.o iotk_dat+COMPLEX1_0.o iotk_dat+COMPLEX1_3.o iotk_dat+COMPLEX2_0.o iotk_dat+COMPLEX2_3.o iotk_dat+INTEGER1_0.o iotk_dat+INTEGER1_3.o iotk_dat+INTEGER2_0.o iotk_dat+INTEGER2_3.o iotk_dat+LOGICAL1_0.o iotk_dat+LOGICAL1_3.o iotk_dat+LOGICAL2_0.o iotk_dat+LOGICAL2_3.o iotk_dat+REAL1_0.o iotk_dat+REAL1_3.o iotk_dat+REAL2_0.o iotk_dat+REAL2_3.o iotk_dat.o iotk_dat_interf.o iotk_error.o iotk_error_interf.o iotk_files.o iotk_files_interf.o iotk_fmt.o iotk_fmt_interf.o iotk_misc.o iotk_misc_interf.o iotk_module.o iotk_scan.o iotk_scan_interf.o iotk_str.o iotk_str_interf.o iotk_unit.o iotk_unit_interf.o iotk_write.o iotk_write_interf.o iotk_xtox.o iotk_xtox_interf.o ar: illegal option -- k Usage: ar [emulation options] [-]{dmpqrstx}[abcfilNoPsSuvV] [member-name] [count] archive-file file... ar -M [ Message-ID: On Fri, 31 Aug 2007 brsahu at physics.utexas.edu wrote: this is an error in iotk, the 'k' error is due to the fact, that ARFLAGS is not properly imported. you could try adding them to AR. good luck (you'll be needing it), axel. BS> BS> Dear PWSCF users, BS> BS> I am trying to compile the latest version of pwscf code on Blue BS> Gene/L. I used the BS> Make.bgl file provided with the distribution and checked the BS> "correctness" of all the entries in it w.r.t Blue Gene here. There is BS> a compilation error at the preprocessing stage: BS> BS> 1501-510 Compilation successful for file iotk_write.f90. BS> /opt/ibmcmp/xlf/9.1/bin/blrts_xlf90 -qalias=noaryovrlp -I../include BS> -O3 -qstrict BS> -qarch=440d -qtune=440 -qsuffix=cpp=f90 -qdpc -qalias=nointptr -Q -I. BS> -I../include BS> -I../Modules -I../PW -I../PH -I/bgl/BlueLight/ppcfloor/bglsys/include BS> -I../iotk/src BS> -qfree=f90 -I../include BS> -WF,-D__AIX,-D__XLF,-D__MPI,-D__PARA,-D__MASS,-D__FFTW,-D__BGL,-D__TRUE_BGL BS> -I. BS> -I../include -I../Modules -I../PW -I../PH BS> -I/bgl/BlueLight/ppcfloor/bglsys/include BS> -I../iotk/src,-D__AIX,-D__XLF,-D__MPI,-D__PARA,-D__MASS,-D__FFTW,-D__BGL,-D__TRUE_BGL BS> -I../Modules -I../PW -I../PH -c iotk_xtox.f90 BS> ** iotk_atol_x === End of Compilation 1 === BS> ** iotk_atoi1 === End of Compilation 2 === BS> ** iotk_itoa1 === End of Compilation 3 === BS> ** iotk_ltoa1 === End of Compilation 4 === BS> 1501-510 Compilation successful for file iotk_xtox.f90. BS> ar libiotk.a iotk_attr+CHARACTER1_0.o iotk_attr+COMPLEX1_0.o BS> iotk_attr+COMPLEX1_3.o BS> iotk_attr+COMPLEX2_0.o iotk_attr+COMPLEX2_3.o iotk_attr+INTEGER1_0.o BS> iotk_attr+INTEGER1_3.o iotk_attr+INTEGER2_0.o iotk_attr+INTEGER2_3.o BS> iotk_attr+LOGICAL1_0.o iotk_attr+LOGICAL1_3.o iotk_attr+LOGICAL2_0.o BS> iotk_attr+LOGICAL2_3.o iotk_attr+REAL1_0.o iotk_attr+REAL1_3.o BS> iotk_attr+REAL2_0.o BS> iotk_attr+REAL2_3.o iotk_attr.o iotk_attr_interf.o iotk_base.o BS> iotk_dat+CHARACTER1_0.o BS> iotk_dat+CHARACTER1_3.o iotk_dat+COMPLEX1_0.o iotk_dat+COMPLEX1_3.o BS> iotk_dat+COMPLEX2_0.o BS> iotk_dat+COMPLEX2_3.o iotk_dat+INTEGER1_0.o iotk_dat+INTEGER1_3.o BS> iotk_dat+INTEGER2_0.o BS> iotk_dat+INTEGER2_3.o iotk_dat+LOGICAL1_0.o iotk_dat+LOGICAL1_3.o BS> iotk_dat+LOGICAL2_0.o BS> iotk_dat+LOGICAL2_3.o iotk_dat+REAL1_0.o iotk_dat+REAL1_3.o iotk_dat+REAL2_0.o BS> iotk_dat+REAL2_3.o iotk_dat.o iotk_dat_interf.o iotk_error.o BS> iotk_error_interf.o BS> iotk_files.o iotk_files_interf.o iotk_fmt.o iotk_fmt_interf.o iotk_misc.o BS> iotk_misc_interf.o iotk_module.o iotk_scan.o iotk_scan_interf.o iotk_str.o BS> iotk_str_interf.o iotk_unit.o iotk_unit_interf.o iotk_write.o BS> iotk_write_interf.o BS> iotk_xtox.o iotk_xtox_interf.o BS> ar: illegal option -- k BS> Usage: ar [emulation options] [-]{dmpqrstx}[abcfilNoPsSuvV] BS> [member-name] [count] BS> archive-file file... BS> ar -M [ commands: BS> d - delete file(s) from the archive BS> m[ab] - move file(s) in the archive BS> p - print file(s) found in the archive BS> q[f] - quick append file(s) to the archive BS> r[ab][f][u] - replace existing or insert new file(s) into the archive BS> t - display contents of archive BS> x[o] - extract file(s) from the archive BS> command specific modifiers: BS> [a] - put file(s) after [member-name] BS> [b] - put file(s) before [member-name] (same as [i]) BS> [N] - use instance [count] of name BS> [f] - truncate inserted file names BS> [P] - use full path names when matching BS> [o] - preserve original dates BS> [u] - only replace files that are newer than current BS> archive contents BS> generic modifiers: BS> [c] - do not warn if the library had to be created BS> [s] - create an archive index (cf. ranlib) BS> [S] - do not build a symbol table BS> [v] - be verbose BS> [V] - display the version number BS> emulation options: BS> No emulation specific options BS> ar: supported targets: elf32-powerpc aixcoff-rs6000 elf32-powerpcle BS> ppcboot elf32-little BS> elf32-big elf64-powerpc elf64-powerpcle aixcoff64-rs6000 elf64-little BS> elf64-big srec BS> symbolsrec tekhex binary ihex BS> make[2]: *** [libiotk.a] Error 1 BS> make[2]: Leaving directory `/home/bsahu/pwscf/iotk/src' BS> make[1]: *** [libiotk.a] Error 2 BS> make[1]: Leaving directory `/home/bsahu/pwscf/iotk' BS> make: *** [libiotk] Error 2 BS> BS> It says -- k is an illegal option. But the BS> The portion of make.sys from where the error seems to come is: BS> BS> LIBS = -L$(BGLSYS)/lib -lmpich.rts -lmsglayer.rts -lrts.rts BS> -ldevices.rts -L/home/bekas/LIBS/BGL -lmass -lmassvbgl -ldcopy -lb-d BS> -lnewdczed.rts -llapack_440n -lnewdczed.rts -ldcopy -lb-d -lfftwn BS> -allow-multiple-definition BS> BS> # BS> LD = $(F77) BS> LIBOBJS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a BS> ../iotk/src/libiotk.a BS> LDFLAGS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a BS> $(LIBOBJS) $(LIBS) -qarch=440d BS> BS> AR = ar BS> ARFLAGS = ruv BS> RANLIB = ranlib BS> BS> BS> where I do not see any -- or k after "ar". BS> BS> What could be wrong? BS> BS> BS> BS> Bhagawan Sahu BS> Microelectronics Research Center BS> University of Texas at Austin BS> 10100 Burnet Road, Bldg. 160 BS> MER 1.606H, Mail code R9900, TX 78758-4445 BS> e-mail: brsahu at physics.utexas.edu BS> Phone: (512)-471-1344 FAX: (512)-471-8575 BS> _______________________________________________ BS> Pw_forum mailing list BS> Pw_forum at pwscf.org BS> http://www.democritos.it/mailman/listinfo/pw_forum BS> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From murda at sissa.it Sat Sep 1 09:25:39 2007 From: murda at sissa.it (Garold Murdachaew) Date: Sat, 1 Sep 2007 09:25:39 +0200 (CEST) Subject: [Pw_forum] polarizability perpendicular to metal surface In-Reply-To: References: Message-ID: Dear Pwscf forum members, Can someone please tell me if it is possible to calculate the polarizability (tensor) per unit cell perpendicular to a metal surface (using finite field or other approach)? Ideally, I would prefer to obtain the dynamic polarizability at imaginary frequencies to allow me to combine this with the corresponding property of an adsorbate in the Casimir-Polder formula for the C_6 dispersion coefficient. Then, summing/integrating over the lateral unit cells of the surface I would be able to obtain the C_3 coefficient [as in, e.g., : I. Baraille, M. R?rat, and P. Mora, "Ab initio calculations of the dispersion contribution to the physisorption potential: Application to the N_2-BN system", Phys. Rev. B 73, 075410 (2006)]. Thank you. Garold Murdachaew Postdoctoral Fellow - Condensed Matter Sector SISSA-ISAS (International School for Advanced Studies) email : murda at sissa.it New Building, Room 208 Via Beirut 2-4, 34014 Trieste, Italy tel : +39 040 3787 477 fax : +39 040 3787 528 From benassi.a at gmail.com Sat Sep 1 10:53:56 2007 From: benassi.a at gmail.com (Andrea Benassi) Date: Sat, 1 Sep 2007 10:53:56 +0200 Subject: [Pw_forum] polarizability perpendicular to metal surface In-Reply-To: References: Message-ID: <97c430b70709010153pdcbc96buf8d9002a6291f146@mail.gmail.com> Dear Garold, enclosed in the lates version of Espresso you can find a post processing named epsilon.x and its manual eps_man.tex in the Doc folder. With epsilon.x you can calculate dielectric tensor within the LDA approximation, eels spectra and London transform of the dielectric tensor (epsilon on imm axe). It should be extremely easy to modify the code in order to give the polarizability instead of the dielectric function. The code works as an ordinary pw post processing, it works both for metal and semiconductors and it supports pool paralelization and collinear spin calculation. I used the same code many times to calculate dielectric functions in order to estimate the Casimir force between plates. I recommend you to use the CVS version of epsilon.x Good luck -- Andrea Benassi National Research Center S3, CNR-INFM ( http://www.s3.infm.it ) Dipartimento di Fisica, Universita' di Modena e Reggio Emilia Via Campi 213/A I-41100 Modena, Italy Tel: +39 059 2055314 Fax: +39 059 374794 E-mail: benassi.andrea at unimore.it URL: http://www.nanoscience.unimo.it From zdw2000 at gmail.com Sat Sep 1 13:48:20 2007 From: zdw2000 at gmail.com (Wei Zhou) Date: Sat, 1 Sep 2007 19:48:20 +0800 Subject: [Pw_forum] maximally localized wannier function and k points generated In-Reply-To: <46D5EA9C.1000203@anl.gov> References: <464953.68630.qm@web60316.mail.yahoo.com> <46D5EA9C.1000203@anl.gov> Message-ID: <3e36a8fc0709010448k6c29efci9080f2690485be43@mail.gmail.com> thank both of you for your helpful discuss above, I just to learn to calcuation *maximally localized wannier function, and meet some new question.i have done scf and nscf calculation without problem, but i donnot know how to choose the win_max and froz_max ,is it accrossling to band calculation results? when I do disentangle calculation or wannier calculation ,it always appear error result as below *** from PE : 0 *** in routine : card_wannier_centers *** error msg. : reading line I *** error code : 1 *** aborting *** * it seems that wannier center is wrong ,for my system is sodium , I take the s p orbits into accout, then I set the center as follow atomic 1 0 1 atomic 1 -1 1 atomic 1 -1 2 Is it right ? for wannier calculation ,I am not sure how to control the input date for "wannier_Na.in",would you help me .give some important suggestion? another question is when I calcuate MLWF of the structure with 54 atoms calculated from molecular dyanmics , the symmetry is P1 ,then how I should choose which atom as the wannier center so that the internal atomphere is represent WLWF is a new method for me , so the question maybe is stupid ,excuse me for my unkown. thank you in adviance -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070901/cd34319a/attachment.htm From brsahu at physics.utexas.edu Sat Sep 1 19:13:18 2007 From: brsahu at physics.utexas.edu (brsahu at physics.utexas.edu) Date: Sat, 1 Sep 2007 12:13:18 -0500 Subject: [Pw_forum] compilation on Blue Gene/L In-Reply-To: References: Message-ID: <20070901121318.ntygzbhn0ncwcooc@mail.ph.utexas.edu> Axel, I added to AR and ARFLAGS the -k option as AR = ar -k ARFLAGS = ruv -k but it still gives the same "illegal option" error. Bhagawan Quoting Axel Kohlmeyer : > On Fri, 31 Aug 2007 brsahu at physics.utexas.edu wrote: > > this is an error in iotk, the 'k' error is due to the > fact, that ARFLAGS is not properly imported. you could > try adding them to AR. > > good luck (you'll be needing it), > axel. > > > BS> > BS> Dear PWSCF users, > BS> > BS> I am trying to compile the latest version of pwscf code on Blue > BS> Gene/L. I used the > BS> Make.bgl file provided with the distribution and checked the > BS> "correctness" of all the entries in it w.r.t Blue Gene here. There is > BS> a compilation error at the preprocessing stage: > BS> > BS> 1501-510 Compilation successful for file iotk_write.f90. > BS> /opt/ibmcmp/xlf/9.1/bin/blrts_xlf90 -qalias=noaryovrlp -I../include > BS> -O3 -qstrict > BS> -qarch=440d -qtune=440 -qsuffix=cpp=f90 -qdpc -qalias=nointptr -Q -I. > BS> -I../include > BS> -I../Modules -I../PW -I../PH -I/bgl/BlueLight/ppcfloor/bglsys/include > BS> -I../iotk/src > BS> -qfree=f90 -I../include > BS> > -WF,-D__AIX,-D__XLF,-D__MPI,-D__PARA,-D__MASS,-D__FFTW,-D__BGL,-D__TRUE_BGL > BS> -I. > BS> -I../include -I../Modules -I../PW -I../PH > BS> -I/bgl/BlueLight/ppcfloor/bglsys/include > BS> > -I../iotk/src,-D__AIX,-D__XLF,-D__MPI,-D__PARA,-D__MASS,-D__FFTW,-D__BGL,-D__TRUE_BGL > BS> -I../Modules -I../PW -I../PH -c iotk_xtox.f90 > BS> ** iotk_atol_x === End of Compilation 1 === > BS> ** iotk_atoi1 === End of Compilation 2 === > BS> ** iotk_itoa1 === End of Compilation 3 === > BS> ** iotk_ltoa1 === End of Compilation 4 === > BS> 1501-510 Compilation successful for file iotk_xtox.f90. > BS> ar libiotk.a iotk_attr+CHARACTER1_0.o iotk_attr+COMPLEX1_0.o > BS> iotk_attr+COMPLEX1_3.o > BS> iotk_attr+COMPLEX2_0.o iotk_attr+COMPLEX2_3.o iotk_attr+INTEGER1_0.o > BS> iotk_attr+INTEGER1_3.o iotk_attr+INTEGER2_0.o iotk_attr+INTEGER2_3.o > BS> iotk_attr+LOGICAL1_0.o iotk_attr+LOGICAL1_3.o iotk_attr+LOGICAL2_0.o > BS> iotk_attr+LOGICAL2_3.o iotk_attr+REAL1_0.o iotk_attr+REAL1_3.o > BS> iotk_attr+REAL2_0.o > BS> iotk_attr+REAL2_3.o iotk_attr.o iotk_attr_interf.o iotk_base.o > BS> iotk_dat+CHARACTER1_0.o > BS> iotk_dat+CHARACTER1_3.o iotk_dat+COMPLEX1_0.o iotk_dat+COMPLEX1_3.o > BS> iotk_dat+COMPLEX2_0.o > BS> iotk_dat+COMPLEX2_3.o iotk_dat+INTEGER1_0.o iotk_dat+INTEGER1_3.o > BS> iotk_dat+INTEGER2_0.o > BS> iotk_dat+INTEGER2_3.o iotk_dat+LOGICAL1_0.o iotk_dat+LOGICAL1_3.o > BS> iotk_dat+LOGICAL2_0.o > BS> iotk_dat+LOGICAL2_3.o iotk_dat+REAL1_0.o iotk_dat+REAL1_3.o > iotk_dat+REAL2_0.o > BS> iotk_dat+REAL2_3.o iotk_dat.o iotk_dat_interf.o iotk_error.o > BS> iotk_error_interf.o > BS> iotk_files.o iotk_files_interf.o iotk_fmt.o iotk_fmt_interf.o iotk_misc.o > BS> iotk_misc_interf.o iotk_module.o iotk_scan.o iotk_scan_interf.o > iotk_str.o > BS> iotk_str_interf.o iotk_unit.o iotk_unit_interf.o iotk_write.o > BS> iotk_write_interf.o > BS> iotk_xtox.o iotk_xtox_interf.o > BS> ar: illegal option -- k > BS> Usage: ar [emulation options] [-]{dmpqrstx}[abcfilNoPsSuvV] > BS> [member-name] [count] > BS> archive-file file... > BS> ar -M [ BS> commands: > BS> d - delete file(s) from the archive > BS> m[ab] - move file(s) in the archive > BS> p - print file(s) found in the archive > BS> q[f] - quick append file(s) to the archive > BS> r[ab][f][u] - replace existing or insert new file(s) into the archive > BS> t - display contents of archive > BS> x[o] - extract file(s) from the archive > BS> command specific modifiers: > BS> [a] - put file(s) after [member-name] > BS> [b] - put file(s) before [member-name] (same as [i]) > BS> [N] - use instance [count] of name > BS> [f] - truncate inserted file names > BS> [P] - use full path names when matching > BS> [o] - preserve original dates > BS> [u] - only replace files that are newer than current > BS> archive contents > BS> generic modifiers: > BS> [c] - do not warn if the library had to be created > BS> [s] - create an archive index (cf. ranlib) > BS> [S] - do not build a symbol table > BS> [v] - be verbose > BS> [V] - display the version number > BS> emulation options: > BS> No emulation specific options > BS> ar: supported targets: elf32-powerpc aixcoff-rs6000 elf32-powerpcle > BS> ppcboot elf32-little > BS> elf32-big elf64-powerpc elf64-powerpcle aixcoff64-rs6000 elf64-little > BS> elf64-big srec > BS> symbolsrec tekhex binary ihex > BS> make[2]: *** [libiotk.a] Error 1 > BS> make[2]: Leaving directory `/home/bsahu/pwscf/iotk/src' > BS> make[1]: *** [libiotk.a] Error 2 > BS> make[1]: Leaving directory `/home/bsahu/pwscf/iotk' > BS> make: *** [libiotk] Error 2 > BS> > BS> It says -- k is an illegal option. But the > BS> The portion of make.sys from where the error seems to come is: > BS> > BS> LIBS = -L$(BGLSYS)/lib -lmpich.rts -lmsglayer.rts -lrts.rts > BS> -ldevices.rts -L/home/bekas/LIBS/BGL -lmass -lmassvbgl -ldcopy -lb-d > BS> -lnewdczed.rts -llapack_440n -lnewdczed.rts -ldcopy -lb-d -lfftwn > BS> -allow-multiple-definition > BS> > BS> # > BS> LD = $(F77) > BS> LIBOBJS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a > BS> ../iotk/src/libiotk.a > BS> LDFLAGS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a > BS> $(LIBOBJS) $(LIBS) -qarch=440d > BS> > BS> AR = ar > BS> ARFLAGS = ruv > BS> RANLIB = ranlib > BS> > BS> > BS> where I do not see any -- or k after "ar". > BS> > BS> What could be wrong? > BS> > BS> > BS> > BS> Bhagawan Sahu > BS> Microelectronics Research Center > BS> University of Texas at Austin > BS> 10100 Burnet Road, Bldg. 160 > BS> MER 1.606H, Mail code R9900, TX 78758-4445 > BS> e-mail: brsahu at physics.utexas.edu > BS> Phone: (512)-471-1344 FAX: (512)-471-8575 > BS> _______________________________________________ > BS> Pw_forum mailing list > BS> Pw_forum at pwscf.org > BS> http://www.democritos.it/mailman/listinfo/pw_forum > BS> > > -- > ======================================================================= > Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu > Center for Molecular Modeling -- University of Pennsylvania > Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 > tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 > ======================================================================= > If you make something idiot-proof, the universe creates a better idiot. > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > Bhagawan Sahu Microelectronics Research Center University of Texas at Austin 10100 Burnet Road, Bldg. 160 MER 1.606H, Mail code R9900, TX 78758-4445 e-mail: brsahu at physics.utexas.edu Phone: (512)-471-1344 FAX: (512)-471-8575 From akohlmey at cmm.chem.upenn.edu Sat Sep 1 18:44:46 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sat, 1 Sep 2007 12:44:46 -0400 (EDT) Subject: [Pw_forum] compilation on Blue Gene/L In-Reply-To: <20070901121318.ntygzbhn0ncwcooc@mail.ph.utexas.edu> Message-ID: On Sat, 1 Sep 2007 brsahu at physics.utexas.edu wrote: BS> Axel, BS> BS> I added to AR and ARFLAGS the -k option as why??? this makes no sense at all. you already know that 'k' is not a supported flag to ar, so why adding it explicitely. BS> BS> AR = ar -k BS> ARFLAGS = ruv -k BS> BS> but it still gives the same "illegal option" error. please look at the output of the make, where you see all the commands it is executing. you will see that when trying to create the libiotk.a library, there are _no_ flags, i.e. not 'ruv'. thus ar interprets the name of the library 'libiotk.a' as the flags to ar and hence the error message of 'k' not being a supported flag. you could try setting AR= ar $(ARFLAGS) in the hope that at least $(AR) is imported. worst case, you'll have to hack the makefile in the iotk subdirectory or its src subdirectory. axel. BS> BS> Bhagawan BS> BS> Quoting Axel Kohlmeyer : BS> BS> > On Fri, 31 Aug 2007 brsahu at physics.utexas.edu wrote: BS> > BS> > this is an error in iotk, the 'k' error is due to the BS> > fact, that ARFLAGS is not properly imported. you could BS> > try adding them to AR. BS> > BS> > good luck (you'll be needing it), BS> > axel. BS> > BS> > BS> > BS> BS> > BS> Dear PWSCF users, BS> > BS> BS> > BS> I am trying to compile the latest version of pwscf code on Blue BS> > BS> Gene/L. I used the BS> > BS> Make.bgl file provided with the distribution and checked the BS> > BS> "correctness" of all the entries in it w.r.t Blue Gene here. There is BS> > BS> a compilation error at the preprocessing stage: BS> > BS> BS> > BS> 1501-510 Compilation successful for file iotk_write.f90. BS> > BS> /opt/ibmcmp/xlf/9.1/bin/blrts_xlf90 -qalias=noaryovrlp -I../include BS> > BS> -O3 -qstrict BS> > BS> -qarch=440d -qtune=440 -qsuffix=cpp=f90 -qdpc -qalias=nointptr -Q -I. BS> > BS> -I../include BS> > BS> -I../Modules -I../PW -I../PH -I/bgl/BlueLight/ppcfloor/bglsys/include BS> > BS> -I../iotk/src BS> > BS> -qfree=f90 -I../include BS> > BS> BS> > -WF,-D__AIX,-D__XLF,-D__MPI,-D__PARA,-D__MASS,-D__FFTW,-D__BGL,-D__TRUE_BGL BS> > BS> -I. BS> > BS> -I../include -I../Modules -I../PW -I../PH BS> > BS> -I/bgl/BlueLight/ppcfloor/bglsys/include BS> > BS> BS> > -I../iotk/src,-D__AIX,-D__XLF,-D__MPI,-D__PARA,-D__MASS,-D__FFTW,-D__BGL,-D__TRUE_BGL BS> > BS> -I../Modules -I../PW -I../PH -c iotk_xtox.f90 BS> > BS> ** iotk_atol_x === End of Compilation 1 === BS> > BS> ** iotk_atoi1 === End of Compilation 2 === BS> > BS> ** iotk_itoa1 === End of Compilation 3 === BS> > BS> ** iotk_ltoa1 === End of Compilation 4 === BS> > BS> 1501-510 Compilation successful for file iotk_xtox.f90. BS> > BS> ar libiotk.a iotk_attr+CHARACTER1_0.o iotk_attr+COMPLEX1_0.o BS> > BS> iotk_attr+COMPLEX1_3.o BS> > BS> iotk_attr+COMPLEX2_0.o iotk_attr+COMPLEX2_3.o iotk_attr+INTEGER1_0.o BS> > BS> iotk_attr+INTEGER1_3.o iotk_attr+INTEGER2_0.o iotk_attr+INTEGER2_3.o BS> > BS> iotk_attr+LOGICAL1_0.o iotk_attr+LOGICAL1_3.o iotk_attr+LOGICAL2_0.o BS> > BS> iotk_attr+LOGICAL2_3.o iotk_attr+REAL1_0.o iotk_attr+REAL1_3.o BS> > BS> iotk_attr+REAL2_0.o BS> > BS> iotk_attr+REAL2_3.o iotk_attr.o iotk_attr_interf.o iotk_base.o BS> > BS> iotk_dat+CHARACTER1_0.o BS> > BS> iotk_dat+CHARACTER1_3.o iotk_dat+COMPLEX1_0.o iotk_dat+COMPLEX1_3.o BS> > BS> iotk_dat+COMPLEX2_0.o BS> > BS> iotk_dat+COMPLEX2_3.o iotk_dat+INTEGER1_0.o iotk_dat+INTEGER1_3.o BS> > BS> iotk_dat+INTEGER2_0.o BS> > BS> iotk_dat+INTEGER2_3.o iotk_dat+LOGICAL1_0.o iotk_dat+LOGICAL1_3.o BS> > BS> iotk_dat+LOGICAL2_0.o BS> > BS> iotk_dat+LOGICAL2_3.o iotk_dat+REAL1_0.o iotk_dat+REAL1_3.o BS> > iotk_dat+REAL2_0.o BS> > BS> iotk_dat+REAL2_3.o iotk_dat.o iotk_dat_interf.o iotk_error.o BS> > BS> iotk_error_interf.o BS> > BS> iotk_files.o iotk_files_interf.o iotk_fmt.o iotk_fmt_interf.o iotk_misc.o BS> > BS> iotk_misc_interf.o iotk_module.o iotk_scan.o iotk_scan_interf.o BS> > iotk_str.o BS> > BS> iotk_str_interf.o iotk_unit.o iotk_unit_interf.o iotk_write.o BS> > BS> iotk_write_interf.o BS> > BS> iotk_xtox.o iotk_xtox_interf.o BS> > BS> ar: illegal option -- k BS> > BS> Usage: ar [emulation options] [-]{dmpqrstx}[abcfilNoPsSuvV] BS> > BS> [member-name] [count] BS> > BS> archive-file file... BS> > BS> ar -M [ > BS> commands: BS> > BS> d - delete file(s) from the archive BS> > BS> m[ab] - move file(s) in the archive BS> > BS> p - print file(s) found in the archive BS> > BS> q[f] - quick append file(s) to the archive BS> > BS> r[ab][f][u] - replace existing or insert new file(s) into the archive BS> > BS> t - display contents of archive BS> > BS> x[o] - extract file(s) from the archive BS> > BS> command specific modifiers: BS> > BS> [a] - put file(s) after [member-name] BS> > BS> [b] - put file(s) before [member-name] (same as [i]) BS> > BS> [N] - use instance [count] of name BS> > BS> [f] - truncate inserted file names BS> > BS> [P] - use full path names when matching BS> > BS> [o] - preserve original dates BS> > BS> [u] - only replace files that are newer than current BS> > BS> archive contents BS> > BS> generic modifiers: BS> > BS> [c] - do not warn if the library had to be created BS> > BS> [s] - create an archive index (cf. ranlib) BS> > BS> [S] - do not build a symbol table BS> > BS> [v] - be verbose BS> > BS> [V] - display the version number BS> > BS> emulation options: BS> > BS> No emulation specific options BS> > BS> ar: supported targets: elf32-powerpc aixcoff-rs6000 elf32-powerpcle BS> > BS> ppcboot elf32-little BS> > BS> elf32-big elf64-powerpc elf64-powerpcle aixcoff64-rs6000 elf64-little BS> > BS> elf64-big srec BS> > BS> symbolsrec tekhex binary ihex BS> > BS> make[2]: *** [libiotk.a] Error 1 BS> > BS> make[2]: Leaving directory `/home/bsahu/pwscf/iotk/src' BS> > BS> make[1]: *** [libiotk.a] Error 2 BS> > BS> make[1]: Leaving directory `/home/bsahu/pwscf/iotk' BS> > BS> make: *** [libiotk] Error 2 BS> > BS> BS> > BS> It says -- k is an illegal option. But the BS> > BS> The portion of make.sys from where the error seems to come is: BS> > BS> BS> > BS> LIBS = -L$(BGLSYS)/lib -lmpich.rts -lmsglayer.rts -lrts.rts BS> > BS> -ldevices.rts -L/home/bekas/LIBS/BGL -lmass -lmassvbgl -ldcopy -lb-d BS> > BS> -lnewdczed.rts -llapack_440n -lnewdczed.rts -ldcopy -lb-d -lfftwn BS> > BS> -allow-multiple-definition BS> > BS> BS> > BS> # BS> > BS> LD = $(F77) BS> > BS> LIBOBJS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a BS> > BS> ../iotk/src/libiotk.a BS> > BS> LDFLAGS = ../flib/ptools.a ../flib/flib.a ../clib/clib.a BS> > BS> $(LIBOBJS) $(LIBS) -qarch=440d BS> > BS> BS> > BS> AR = ar BS> > BS> ARFLAGS = ruv BS> > BS> RANLIB = ranlib BS> > BS> BS> > BS> BS> > BS> where I do not see any -- or k after "ar". BS> > BS> BS> > BS> What could be wrong? BS> > BS> BS> > BS> BS> > BS> BS> > BS> Bhagawan Sahu BS> > BS> Microelectronics Research Center BS> > BS> University of Texas at Austin BS> > BS> 10100 Burnet Road, Bldg. 160 BS> > BS> MER 1.606H, Mail code R9900, TX 78758-4445 BS> > BS> e-mail: brsahu at physics.utexas.edu BS> > BS> Phone: (512)-471-1344 FAX: (512)-471-8575 BS> > BS> _______________________________________________ BS> > BS> Pw_forum mailing list BS> > BS> Pw_forum at pwscf.org BS> > BS> http://www.democritos.it/mailman/listinfo/pw_forum BS> > BS> BS> > BS> > -- BS> > ======================================================================= BS> > Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu BS> > Center for Molecular Modeling -- University of Pennsylvania BS> > Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 BS> > tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 BS> > ======================================================================= BS> > If you make something idiot-proof, the universe creates a better idiot. BS> > BS> > _______________________________________________ BS> > Pw_forum mailing list BS> > Pw_forum at pwscf.org BS> > http://www.democritos.it/mailman/listinfo/pw_forum BS> > BS> BS> BS> BS> Bhagawan Sahu BS> Microelectronics Research Center BS> University of Texas at Austin BS> 10100 Burnet Road, Bldg. 160 BS> MER 1.606H, Mail code R9900, TX 78758-4445 BS> e-mail: brsahu at physics.utexas.edu BS> Phone: (512)-471-1344 FAX: (512)-471-8575 BS> _______________________________________________ BS> Pw_forum mailing list BS> Pw_forum at pwscf.org BS> http://www.democritos.it/mailman/listinfo/pw_forum BS> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From zdw2000 at gmail.com Sun Sep 2 01:38:34 2007 From: zdw2000 at gmail.com (Wei Zhou) Date: Sun, 2 Sep 2007 07:38:34 +0800 Subject: [Pw_forum] about wannnier center Message-ID: <3e36a8fc0709011638t42bd4de0l4067aaf2a5d154f5@mail.gmail.com> , I just to learn to calcuation *maximally localized wannier function, and meet some new question.i have done scf and nscf calculation without problem, but i donnot know how to choose the win_max and froz_max ,is it accrossling to band calculation results? when I do disentangle calculation or wannier calculation ,it always appear error result as below *** from PE : 0 *** in routine : card_wannier_centers *** error msg. : reading line I *** error code : 1 *** aborting *** * it seems that wannier center is wrong ,for my system is sodium , I take the s p orbits into accout, then I set the center as follow atomic 1 0 1 atomic 1 -1 1 atomic 1 -1 2 Is it right ? for wannier calculation ,I am not sure how to control the input date for "wannier_Na.in",would you help me .give some important suggestion? another question is when I calcuate MLWF of the structure with 54 atoms calculated from molecular dyanmics , the symmetry is P1 ,then how I should choose which atom as the wannier center so that the internal atomphere is represent WLWF is a new method for me , so the question maybe is stupid ,excuse me for my unkown. thank you in adviance -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070902/b80f930d/attachment.htm From ezadshojaee at hotmail.com Sun Sep 2 10:25:09 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Sun, 2 Sep 2007 08:25:09 +0000 Subject: [Pw_forum] IFC's Message-ID: i want to know that are we able to use the 'IFC's calculated with two grids' together? for example we calculate them by a 3*3*3 grid, and then 4*4*4 , then use all of them in calculating C(q) for an arbitrary point in the Brillouin zone? _________________________________________________________________ Discover the new Windows Vista http://search.msn.com/results.aspx?q=windows+vista&mkt=en-US&form=QBRE -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070902/d019a332/attachment.htm From ferretti.andrea at unimore.it Sun Sep 2 13:41:14 2007 From: ferretti.andrea at unimore.it (Andrea Ferretti) Date: Sun, 2 Sep 2007 13:41:14 +0200 (CEST) Subject: [Pw_forum] compilation on Blue Gene/L In-Reply-To: References: Message-ID: On Sat, 1 Sep 2007, Axel Kohlmeyer wrote: > > this is an error in iotk, the 'k' error is due to the > fact, that ARFLAGS is not properly imported. you could > try adding them to AR. > > good luck (you'll be needing it), > axel. > it seems that different espresso versions load different flags for AR in the iotk Makefile v.3.2: loads ARFLAGS_DYNAMIC in iotk/src/Makefile, so, you can try so properly set this variable (something like ARFLAGS_DYNAMIC=ruv I hope) if it is not set the reason is explained in make.sys as: # ar command and flags - for most architectures: AR = ar, ARFLAGS = ruv # ARFLAGS_DYNAMICS is used in iotk to produce a dynamical library, # for Mac OS-X with PowerPC and xlf compiler. In all other cases # ARFLAGS_DYNAMICS = $(ARFLAGS) current cvs: loads ARFLAGS in iotk/src/Makefile and not ARFLAGS_DYNAMIC as v.3.2 anyway, I think your case is the former -- Andrea Ferretti National Research Center S3, CNR-INFM ( http://s3.infm.it ) Dip. Fisica, Univ. di Modena e Reggio E. (Italy) Tel: +39 059 2055301; Fax: +39 059 374794; Skype: andrea_ferretti URL: http://www.nanoscience.unimo.it Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html From marzari at MIT.EDU Sun Sep 2 14:24:45 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Sun, 02 Sep 2007 08:24:45 -0400 Subject: [Pw_forum] IFC's In-Reply-To: References: Message-ID: <46DAAB8D.8050304@mit.edu> Ezad Shojaee wrote: > i want to know that are we able to use the 'IFC's calculated with two > grids' together? for example we calculate them by a 3*3*3 grid, and then > 4*4*4 , then use all of them in calculating C(q) for an arbitrary point > in the Brillouin zone? Dear Ezad, with the phonon code you calculate the dynamical matrix at any arbitary q. Usually a regular nxnxnx grid of q is chosen, you fourier transform it in real space to get a nxnxn set of IFCs, and you fourier back at any arbitrary q. It's called Fourier interpolation. Details (important, for polar materials) are in the original Giannozzi 1991 PRB, and in the recent Baroni Rev Mod Phys (2004, I believE). Also, do not forget your affiliation in the emails. nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From eyvaz_isaev at yahoo.com Sun Sep 2 14:29:04 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sun, 2 Sep 2007 05:29:04 -0700 (PDT) Subject: [Pw_forum] IFC's In-Reply-To: Message-ID: <124453.5540.qm@web60311.mail.yahoo.com> Dear Ezad, --- Ezad Shojaee wrote: > > i want to know that are we able to use the 'IFC's > calculated with two grids' together? for example we > calculate them by a 3*3*3 grid, and then 4*4*4 , > then use all of them in calculating C(q) for an > arbitrary point in the Brillouin zone? > I do not think so, because of that usually one has to perform FFT in order to get C(R) of C(q) calculated. C(q) ==> C(R) is done by q2r.x which needs the same number of q-points (and dynamical matrices C(q)) which you used in phonon calculations specifying nq1,nq2,nq3. Otherwise I assume you will recieve an error message like "q-point not allowed". By the way, q2r.x first reads Sysname.dyn0 file which contains nq1,nq2,nq3 already used and total number of q-points for this grid, and then q-points. So, if you like to mix these q-points (and C(q)) which nq1,nq2,nq3 you can specify in the dyn0 file? Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Boardwalk for $500? In 2007? Ha! Play Monopoly Here and Now (it's updated for today's economy) at Yahoo! Games. http://get.games.yahoo.com/proddesc?gamekey=monopolyherenow From eyvaz_isaev at yahoo.com Sun Sep 2 14:59:36 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sun, 2 Sep 2007 05:59:36 -0700 (PDT) Subject: [Pw_forum] IFC's In-Reply-To: <46DAAB8D.8050304@mit.edu> Message-ID: <511480.92181.qm@web60313.mail.yahoo.com> Hi Nicola, --- Nicola Marzari wrote: > Dear Ezad, > > Details (important, for polar materials) are in the original Giannozzi 1991 PRB, and in the recent > Baroni Rev Mod Phys (2004, I > believE). > Presumably, you mean S. Baroni, S. de Gironcoli, A. Dal Corso, and P. Giannozzi, Rev. Mod. Phys. \textbf{73}, 515--562 (2001). Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Moody friends. Drama queens. Your life? Nope! - their life, your story. Play Sims Stories at Yahoo! Games. http://sims.yahoo.com/ From giannozz at nest.sns.it Sun Sep 2 16:14:32 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 2 Sep 2007 16:14:32 +0200 Subject: [Pw_forum] Temperature extraction from PW In-Reply-To: References: Message-ID: On Aug 30, 2007, at 19:21 , Zahraa Ibrahim wrote: > I recently changed from the democritus version of PW to the > espresso version. there isn't a "democritos" version distinct from an "espresso" version: there are version numbers (for the stable versions) and the CVS version (under development) Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From ferretti.andrea at unimore.it Sun Sep 2 16:27:25 2007 From: ferretti.andrea at unimore.it (Andrea Ferretti) Date: Sun, 2 Sep 2007 16:27:25 +0200 (CEST) Subject: [Pw_forum] about wannier center In-Reply-To: <3e36a8fc0709010448k6c29efci9080f2690485be43@mail.gmail.com> References: <464953.68630.qm@web60316.mail.yahoo.com> <46D5EA9C.1000203@anl.gov> <3e36a8fc0709010448k6c29efci9080f2690485be43@mail.gmail.com> Message-ID: Hi, On Sat, 1 Sep 2007, Wei Zhou wrote: > thank both of you for your helpful discuss above, I just to learn to > calcuation *maximally localized wannier function, and meet some new > question.i have done scf and nscf calculation without problem, but i donnot > know how to choose the win_max and froz_max ,is it accrossling to band > calculation results? this is question which goes deep in the definition of the disentanglement procedure; first of all, refer to the original paper by Souza, Marzari and Vanderbilt, PRB 65, 035109 (2001), if you didn't already as a brief summary: win_max and win_min are the input flags defining the bloch states entering the disentanglement procedure (all the eigenstates win_min < eps_nk < win_max are taken in the starting bloch manyfold) froz_min and froz_max define a second (or inner) energy window, which should be contained in the previous one: all the eigenstates falling inside this frozen window will be directly taken into the wannier manyfold (subset of the bloch manyfold which will be used later on to localize maxim. loc. WFs) note that the choice of these parameters may be quite sensible on the final results the largest the energy window (win_min, win_max) you take, the better invariant spread you obtain since you are enlarging the number of dregrees of freedom of the problem... of course, it will take longer the smaller the frozen_window you take, the less constrained your minimization is, and the better minimization you get (but take care, the constrain is useful to take the lowest states as much similar to the origial ones (nscf calculation) as possible) the basic ideas about how to chose these windows are stated in the original PRB paper above... some tricks can also be found in http://www.wannier-transport.org/PDF/want_manual.pdf > when I do disentangle calculation or wannier > calculation ,it always appear error result as below > > *** from PE : 0 > *** in routine : card_wannier_centers > *** error msg. : reading line I > *** error code : 1 > *** aborting *** > > it seems that wannier center is wrong ,for my system is sodium , I take > the s p orbits into accout, then I set the center as follow > atomic 1 0 1 > atomic 1 -1 1 > atomic 1 -1 2 there is a problem with the parameters you provided the syntax is: atomic atomic_index ang_mom_L_number ang_mom_M number the first integer is the index defining the atom you want to consider... the second integer is the L angular momentum quantum number which should be >= 0 (and hence the error given by the code, since you set -1), the third integer is M ( -L,..., +L) to get an sp manyfold, use eg > atomic 1 0 0 > atomic 1 1 0 > atomic 1 1 1 > Is it right ? for wannier calculation ,I am not sure how to control the > input date for "wannier_Na.in",would you help me .give some important > suggestion? > > another question is when I calcuate MLWF of the structure with 54 > atoms calculated from molecular dyanmics , the symmetry is P1 ,then how I > should choose which atom as the wannier center so that the internal > atomphere is represent if I understand correctly, at the moment there is no way of imposing a certain symmetry of the final subset of WFs... especially when you perform a disentanglement procedure. typically the symmetry (or some lower symmetry, if the case) is only recovered during the minimization procedure of max.loc. WFs andrea -- Andrea Ferretti National Research Center S3, CNR-INFM ( http://s3.infm.it ) Dip. Fisica, Univ. di Modena e Reggio E. (Italy) Tel: +39 059 2055301; Fax: +39 059 374794; Skype: andrea_ferretti URL: http://www.nanoscience.unimo.it Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html From giannozz at nest.sns.it Sun Sep 2 16:33:47 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 2 Sep 2007 16:33:47 +0200 Subject: [Pw_forum] IFC's In-Reply-To: References: Message-ID: <22275BBB-B56E-4236-A816-C6CEB9538DA0@nest.sns.it> On Sep 2, 2007, at 10:25 , Ezad Shojaee wrote: > i want to know that are we able to use the 'IFC's calculated with > two grids' together? for example we calculate them by a 3*3*3 grid, > and then 4*4*4 , then use all of them in calculating C(q) for an > arbitrary point in the Brillouin zone? no, it is not: you can use only one specific grid. In principle one could start with a coarser grid and then add the missing points needed to have a denser grid, I think, but it would be a mess Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From ezadshojaee at hotmail.com Sun Sep 2 18:21:08 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Sun, 2 Sep 2007 16:21:08 +0000 Subject: [Pw_forum] IFC's Message-ID: consider one special element of the dynamical matrix, if we calculate IFC's by a 3*3*3 grid, it seems that the code generates 27 real space IFC's to be used in calculating the same element of the dynamical matrix for an arbitrary point, as the file containing IFC's for example "###.fc" is 1 1 3 6 ( indicating the special element) 1 1 1 C(R) 2 1 1 ..... .. 3 3 3 C(R') the code calculates C(q) by sum {C(R)exp(iq.R)}, so does the code use 27 R's for the summation to calculate the element? and does the code defines the 'R' with the 3 numbers in front of the C(R)'s in ###.fc? if it is so, imagine we calculated by a 4*4*4 grid and the ###.fc file containing IFC's in the real space is 1 1 3 6 ( indicating the same element of DM) 1 1 1 C(R'') ... 4 4 4 C(R''') if the 'R' used in the first file to calculate C(R) is the same used to calculate C(R'') in the second file, should we sum the C(R) and C(R''=R) to obtain more accurate figure for IFC's (27 of them) because with this, we have used more q points to calculate C(R)'s in the 27 first among 64 ? _________________________________________________________________ Discover the new Windows Vista http://search.msn.com/results.aspx?q=windows+vista&mkt=en-US&form=QBRE -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070902/2c4e46d1/attachment.htm From ezadshojaee at hotmail.com Sun Sep 2 18:21:09 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Sun, 2 Sep 2007 16:21:09 +0000 Subject: [Pw_forum] IFC's Message-ID: consider one special element of the dynamical matrix, if we calculate IFC's by a 3*3*3 grid, it seems that the code generates 27 real space IFC's to be used in calculating the same element of the dynamical matrix for an arbitrary point, as the file containing IFC's for example "###.fc" is 1 1 3 6 ( indicating the special element) 1 1 1 C(R) 2 1 1 ..... .. 3 3 3 C(R') the code calculates C(q) by sum {C(R)exp(iq.R)}, so does the code use 27 R's for the summation to calculate the element? and does the code defines the 'R' with the 3 numbers in front of the C(R)'s in ###.fc? if it is so, imagine we calculated by a 4*4*4 grid and the ###.fc file containing IFC's in the real space is 1 1 3 6 ( indicating the same element of DM) 1 1 1 C(R'') ... 4 4 4 C(R''') if the 'R' used in the first file to calculate C(R) is the same used to calculate C(R'') in the second file, should we sum the C(R) and C(R''=R) to obtain more accurate figure for IFC's (27 of them) because with this, we have used more q points to calculate C(R)'s in the 27 first among 64 ? _________________________________________________________________ Discover the new Windows Vista http://search.msn.com/results.aspx?q=windows+vista&mkt=en-US&form=QBRE -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070902/643674df/attachment-0001.htm From dkorotin at gmail.com Sun Sep 2 21:03:02 2007 From: dkorotin at gmail.com (Dmitry Korotin) Date: Sun, 2 Sep 2007 21:03:02 +0200 Subject: [Pw_forum] Potassium PBE pseudopotential Message-ID: <2fd252650709021203v4279fd47v41271a3e73eedae5@mail.gmail.com> Dear ESPRESSO users, does anyone have potassium pseudopotential created with PBE exchange-correlation functional for share? Thank you in advance, Dmitry Korotin PhD student, Institute of Metal Physics, Ural branch of RAS, Russia From marzari at MIT.EDU Sun Sep 2 21:17:03 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Sun, 02 Sep 2007 15:17:03 -0400 Subject: [Pw_forum] Potassium PBE pseudopotential In-Reply-To: <2fd252650709021203v4279fd47v41271a3e73eedae5@mail.gmail.com> References: <2fd252650709021203v4279fd47v41271a3e73eedae5@mail.gmail.com> Message-ID: <46DB0C2F.5030108@mit.edu> Dmitry Korotin wrote: > Dear ESPRESSO users, > does anyone have potassium pseudopotential created with PBE > exchange-correlation functional for share? > > Thank you in advance, > Dmitry Korotin > PhD student, > Institute of Metal Physics, > Ural branch of RAS, > Russia Hi Dmitry, I do not have one (if you get one, do send it to the group, so we can update the page). If you do not get any, the simplest would be to generate one using David Vanderbilt's code. The input for LDA is on the web, so it's just a matter of switching the flag from LDA to PBE, and testing it http://www.physics.rutgers.edu/~dhv/uspp/uspp-cur/Work/019-K/019-K-ca-sp-vgrp/ Let us know, nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From zdw2000 at gmail.com Mon Sep 3 04:41:35 2007 From: zdw2000 at gmail.com (Wei Zhou) Date: Mon, 3 Sep 2007 10:41:35 +0800 Subject: [Pw_forum] about wannier center In-Reply-To: References: <464953.68630.qm@web60316.mail.yahoo.com> <46D5EA9C.1000203@anl.gov> <3e36a8fc0709010448k6c29efci9080f2690485be43@mail.gmail.com> Message-ID: <3e36a8fc0709021941x38729ceev7c29d0386549066c@mail.gmail.com> thank you for your reply in details. It help me a lot In fact ,I have read the references you adviced above in the rough,and after have read your reply ,I know the effects of the rang in out windows and inner windows ,but perhaps because my limited knowedge,I still donnot understand how to choose the inner windows according to what to chose the inner windows. for my system Na,my psoeduo potential inclued the 2s,2p,3s orbits, what I am interested is the 3s ,2p orbits, so the outer windows is choosed to included the 3s ,2p orbits,then that is enough . but what is the criterion of choose the inner windows,what orbit it must include , it puzzled me. thank you best regards -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070903/2ed96cc9/attachment.htm From akohlmey at cmm.chem.upenn.edu Mon Sep 3 04:14:40 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sun, 2 Sep 2007 22:14:40 -0400 (EDT) Subject: [Pw_forum] Potassium PBE pseudopotential In-Reply-To: <46DB0C2F.5030108@mit.edu> Message-ID: On Sun, 2 Sep 2007, Nicola Marzari wrote: hi nicola, dmitry, it is usually not that easy. most of the -ca- (=LDA) potentials use ifqopt=2, which is not allowed for GGA functionals. as a consequence of changing that one frequently has to change a few parameters more and pay especially attention to negative densities during pseudization. the potassium potential should be ok with changing ifqopt to 3 only. it may help transferability to generate the all-electron calculation in slightly oxidazed config (+0.5). cheers and good luck, axel. NM> Hi Dmitry, NM> NM> I do not have one (if you get one, do send it to the group, so NM> we can update the page). NM> NM> If you do not get any, the simplest would be to generate NM> one using David Vanderbilt's code. NM> NM> The input for LDA is on the web, so it's just a matter of NM> switching the flag from LDA to PBE, and testing it NM> http://www.physics.rutgers.edu/~dhv/uspp/uspp-cur/Work/019-K/019-K-ca-sp-vgrp/ NM> NM> Let us know, NM> NM> nicola NM> NM> --------------------------------------------------------------------- NM> Prof Nicola Marzari Department of Materials Science and Engineering NM> 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA NM> tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu NM> _______________________________________________ NM> Pw_forum mailing list NM> Pw_forum at pwscf.org NM> http://www.democritos.it/mailman/listinfo/pw_forum NM> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From marzari at MIT.EDU Mon Sep 3 06:20:18 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Mon, 03 Sep 2007 00:20:18 -0400 Subject: [Pw_forum] Potassium PBE pseudopotential In-Reply-To: References: Message-ID: <46DB8B82.4010503@mit.edu> Thanks Axel - I always forget this... The choice 3 8 10.0 ifqopt,nqf,qtryc almost always works. The tutorial does mention the work "black art", though. nicola Axel Kohlmeyer wrote: > On Sun, 2 Sep 2007, Nicola Marzari wrote: > > hi nicola, dmitry, > > it is usually not that easy. most of the -ca- (=LDA) > potentials use ifqopt=2, which is not allowed for > GGA functionals. as a consequence of changing that > one frequently has to change a few parameters more > and pay especially attention to negative densities > during pseudization. the potassium potential should > be ok with changing ifqopt to 3 only. it may help > transferability to generate the all-electron > calculation in slightly oxidazed config (+0.5). > > cheers and good luck, > axel. > --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From brsahu at physics.utexas.edu Mon Sep 3 06:31:56 2007 From: brsahu at physics.utexas.edu (brsahu at physics.utexas.edu) Date: Sun, 2 Sep 2007 23:31:56 -0500 Subject: [Pw_forum] inconsistent number of sticks Message-ID: <20070902233156.x0s9vgbcgurooc4g@mail.ph.utexas.edu> Dear PWSCF users, I get an error message for a run on a IBM P655 machine (about 400 atoms and 200 processors in the job script) 0: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%% 0: from fft_dlay_set : error # 7 0: inconsistent number of sticks 0: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%% 0: 0: stopping ... I checked the archive for similar messages. what causes this message to appear. I can provide the input if asked. Bhagawan Sahu Microelectronics Research Center University of Texas at Austin 10100 Burnet Road, Bldg. 160 MER 1.606H, Mail code R9900, TX 78758-4445 e-mail: brsahu at physics.utexas.edu Phone: (512)-471-1344 FAX: (512)-471-8575 From katalin.gaal-nagy at physik.uni-regensburg.de Mon Sep 3 10:43:43 2007 From: katalin.gaal-nagy at physik.uni-regensburg.de (Katalin Gaal-Nagy) Date: Mon, 3 Sep 2007 10:43:43 +0200 (CEST) Subject: [Pw_forum] IFC's Message-ID: Dear all, the message I send just half an hour ago is held because of the size ... However, I resend it without the attachments. If anyone is interested in the files with the routines, please tell me! Best wishes, Katalin ---------- Forwarded message ---------- Date: Mon, 3 Sep 2007 10:01:59 +0200 (CEST) From: Katalin Gaal-Nagy To: PWSCF Forum Subject: Re: [Pw_forum] IFC's Dear Ezad, dear Paolo, well ... you can do it, with a trick. If I understood correctly, you have the q points on a 3x3x3 grid and the q points on a 4x4x4 grid. A grid containing both sets of q points is the 12x12x12 grid. You can do the following: You calculate the force constants of the 4x4x4 grid and do a back Fourier transform and calculate the dynamical matrices on a 12x12x12 grid. In this set of dynamical matrices, the ones corresponding to the 4x4x4 grid are identical (up to numerical noise, which is little). Then you take these 12x12x12 interpolated dynamical matrices, throw away the interpolated ones corresponding to the 3x3x3 mesh and put instead the 3x3x3 ones you have calculated with ph.x. You again have a full set of dynamical matrices and you can calculate from these the force constants, which correspond to a mixing of a 4x4x4 grid and a 3x3x3 grid. Note, they DON'T correspond to a 12x12x12 grid, since there are still interpolated dynamical matrices left. With the resulting dynamical matrices also no physical meaning like range of corresponding forces can be connected any more. The idea of this scheme can be found under condmat arXiv:0707.0384. For the use of this scheme I modified the postprocessing routines matdyn.f90 of ESPRESSO (I am not sure about the version, if should have been espresso-3.1). I don't remember, if I was putting it in a nice, commented shape (for sure tha Makefile not ... ). The main modification was to tell the code to write out the dynamical matrices for all q in star(q) in a way, that q2r.x can read it again. The modified routines and a makefile are attached. I would apprechiate if you could test, if the compilation is working. However, you have to use the routines q2r.x and matdyn.x ... The matdyn.x has now an additional flag and the input reads as &input asr='simple', amass(1)=28.086, flfrc='si.q888.force', flfrq='si.q16.16.16.freq.interpol', dynmatout=.true. / 349 .... list of k points created with kpoints.x ... It creates (in this case) 349 files containing the interpolated dynamical matrices dynint.$number where $number is the number of the q point given in the list ... Paolo, in case you want to include it to the ESPRESSO paackage and there are problems, please let me know. Good luck, Katalin On Sun, 2 Sep 2007, Paolo Giannozzi wrote: > > On Sep 2, 2007, at 10:25 , Ezad Shojaee wrote: > >> i want to know that are we able to use the 'IFC's calculated with >> two grids' together? for example we calculate them by a 3*3*3 grid, >> and then 4*4*4 , then use all of them in calculating C(q) for an >> arbitrary point in the Brillouin zone? > > no, it is not: you can use only one specific grid. In principle one > could > start with a coarser grid and then add the missing points needed to have > a denser grid, I think, but it would be a mess > > Paolo > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From giannozz at nest.sns.it Mon Sep 3 10:48:44 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 3 Sep 2007 10:48:44 +0200 Subject: [Pw_forum] inconsistent number of sticks In-Reply-To: <20070902233156.x0s9vgbcgurooc4g@mail.ph.utexas.edu> References: <20070902233156.x0s9vgbcgurooc4g@mail.ph.utexas.edu> Message-ID: <8F2EB455-03D5-4C00-A2BE-FC31CEC1ACF5@nest.sns.it> On Sep 3, 2007, at 6:31 , brsahu at physics.utexas.edu wrote: > I can provide the input if asked. please provide the input AND the output Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From ferretti.andrea at unimore.it Mon Sep 3 13:28:25 2007 From: ferretti.andrea at unimore.it (Andrea Ferretti) Date: Mon, 3 Sep 2007 13:28:25 +0200 (CEST) Subject: [Pw_forum] about wannier center In-Reply-To: <3e36a8fc0709021941x38729ceev7c29d0386549066c@mail.gmail.com> References: <464953.68630.qm@web60316.mail.yahoo.com> <46D5EA9C.1000203@anl.gov> <3e36a8fc0709010448k6c29efci9080f2690485be43@mail.gmail.com> <3e36a8fc0709021941x38729ceev7c29d0386549066c@mail.gmail.com> Message-ID: Hi Wei Zhou, to my experience, the inner window should contain more or less all the states you want to represent properly (with some degree of flexibility, of course)... instead the outer window should typically be quite large, including states up to several eV above the inner upper edge. Anyway, I suspect that no general rule can be drawn, and you should perform some tests (eg, selecting one outer window and enlarging it if WFs are not properly localized) andrea On Mon, 3 Sep 2007, Wei Zhou wrote: > thank you for your reply in details. It help me a lot > In fact ,I have read the references you adviced above in the > rough,and after have read your reply ,I know the effects of the rang in > out windows and inner windows ,but perhaps because my limited knowedge,I > still donnot understand how to choose the inner windows according to what to > chose the inner windows. > for my system Na,my psoeduo potential inclued the 2s,2p,3s > orbits, what I am interested is the 3s ,2p orbits, so the outer windows is > choosed to included the 3s ,2p orbits,then that is enough . but what is the > criterion of choose the inner windows,what orbit it must include , it > puzzled me. > thank you > best regards > -- Andrea Ferretti National Research Center S3, CNR-INFM ( http://s3.infm.it ) Dip. Fisica, Univ. di Modena e Reggio E. (Italy) Tel: +39 059 2055301; Fax: +39 059 374794; Skype: andrea_ferretti URL: http://www.nanoscience.unimo.it Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html From dkorotin at gmail.com Mon Sep 3 14:27:27 2007 From: dkorotin at gmail.com (Dmitry Korotin) Date: Mon, 3 Sep 2007 14:27:27 +0200 Subject: [Pw_forum] Potassium PBE pseudopotential In-Reply-To: <46DB8B82.4010503@mit.edu> References: <46DB8B82.4010503@mit.edu> Message-ID: <2fd252650709030527v185cd8acka9f269d381451d6d@mail.gmail.com> Dear all, thank you for your suggestions and tutorials 2007/9/3, Nicola Marzari : > The tutorial does mention the work "black art", though. > that's because I prefer to ask in mailing-list for pseudopotential against to generate my own :) (all my previous attempts to generate one, for example Cu or Pu were unsuccessful). But, I think I'll try to do it this time. Dear Eyvaz, the subject of our studies are lattice distortions in KCuF3. Best regards, Dmitry Korotin, PhD student, Institute of Metal Physics, Ural branch of RAS, Russia From katalin.gaal-nagy at physik.uni-regensburg.de Mon Sep 3 10:01:59 2007 From: katalin.gaal-nagy at physik.uni-regensburg.de (Katalin Gaal-Nagy) Date: Mon, 3 Sep 2007 10:01:59 +0200 (CEST) Subject: [Pw_forum] IFC's In-Reply-To: <22275BBB-B56E-4236-A816-C6CEB9538DA0@nest.sns.it> References: <22275BBB-B56E-4236-A816-C6CEB9538DA0@nest.sns.it> Message-ID: Dear Ezad, dear Paolo, well ... you can do it, with a trick. If I understood correctly, you have the q points on a 3x3x3 grid and the q points on a 4x4x4 grid. A grid containing both sets of q points is the 12x12x12 grid. You can do the following: You calculate the force constants of the 4x4x4 grid and do a back Fourier transform and calculate the dynamical matrices on a 12x12x12 grid. In this set of dynamical matrices, the ones corresponding to the 4x4x4 grid are identical (up to numerical noise, which is little). Then you take these 12x12x12 interpolated dynamical matrices, throw away the interpolated ones corresponding to the 3x3x3 mesh and put instead the 3x3x3 ones you have calculated with ph.x. You again have a full set of dynamical matrices and you can calculate from these the force constants, which correspond to a mixing of a 4x4x4 grid and a 3x3x3 grid. Note, they DON'T correspond to a 12x12x12 grid, since there are still interpolated dynamical matrices left. With the resulting dynamical matrices also no physical meaning like range of corresponding forces can be connected any more. The idea of this scheme can be found under condmat arXiv:0707.0384. For the use of this scheme I modified the postprocessing routines matdyn.f90 of ESPRESSO (I am not sure about the version, if should have been espresso-3.1). I don't remember, if I was putting it in a nice, commented shape (for sure tha Makefile not ... ). The main modification was to tell the code to write out the dynamical matrices for all q in star(q) in a way, that q2r.x can read it again. The modified routines and a makefile are attached. I would apprechiate if you could test, if the compilation is working. However, you have to use the routines q2r.x and matdyn.x ... The matdyn.x has now an additional flag and the input reads as &input asr='simple', amass(1)=28.086, flfrc='si.q888.force', flfrq='si.q16.16.16.freq.interpol', dynmatout=.true. / 349 .... list of k points created with kpoints.x ... It creates (in this case) 349 files containing the interpolated dynamical matrices dynint.$number where $number is the number of the q point given in the list ... Paolo, in case you want to include it to the ESPRESSO paackage and there are problems, please let me know. Good luck, Katalin On Sun, 2 Sep 2007, Paolo Giannozzi wrote: > > On Sep 2, 2007, at 10:25 , Ezad Shojaee wrote: > >> i want to know that are we able to use the 'IFC's calculated with >> two grids' together? for example we calculate them by a 3*3*3 grid, >> and then 4*4*4 , then use all of them in calculating C(q) for an >> arbitrary point in the Brillouin zone? > > no, it is not: you can use only one specific grid. In principle one > could > start with a coarser grid and then add the missing points needed to have > a denser grid, I think, but it would be a mess > > Paolo > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- # Makefile for tools include ../make.sys PHOBJS = \ ../PH/write_dyn_on_file.o \ ../PH/star_q.o #../PH/dyndia.o \ #../PW/cdiagh.o \ #../Modules/clocks.o #../PH/dynmatrix.o \ #../PH/symdyn_munu.o \ #../PH/stop_ph.o \ #../PH/q2qstar_ph.o \ #../PH/write_ramtns.o \ #../PH/trntnsc.o \ #../PH/symdynph_gq.o \ #../PH/print_clock_ph.o \ #../PH/rotate_and_add_dyn.o \ #../PH/write_epsilon_and_zeu.o \ #../PH/deallocate_part.o PWOBJS = \ ../PW/checksym.o \ ../PW/coset.o \ ../PW/cryst_to_car.o \ ../PW/cubicsym.o \ ../PW/eqvect.o \ ../PW/error_handler.o \ ../PW/hexsym.o \ ../PW/irrek.o \ ../PW/inverse_s.o \ ../PW/kpoint_grid.o \ ../PW/mode_group.o \ ../PW/multable.o \ ../PW/noncol.o \ ../PW/pwcom.o \ ../PW/sgam_at.o \ ../PW/sgam_at_mag.o \ ../PW/sgam_ph.o \ ../PW/sgama.o \ ../PW/smallg_q.o \ ../PW/trnvecc.o \ ../PW/w0gauss.o \ ../PW/wsweight.o MODULES = \ ../Modules/basic_algebra_routines.o \ ../Modules/cell_base.o \ ../Modules/constants.o \ ../Modules/control_flags.o \ ../Modules/io_global.o \ ../Modules/io_files.o \ ../Modules/ions_base.o \ ../Modules/fft_types.o \ ../Modules/fft_scalar.o \ ../Modules/kind.o \ ../Modules/mp.o \ ../Modules/mp_global.o \ ../Modules/parallel_include.o \ ../Modules/parameters.o \ ../Modules/parser.o \ ../Modules/path_formats.o \ ../Modules/printout_base.o \ ../Modules/random_numbers.o \ ../Modules/recvec.o \ ../Modules/splinelib.o \ ../Modules/timestep.o \ ../Modules/uspp.o TLDEPS= bindir mods libs pw all : tldeps dynmat.x kpoints.x matdyn.x \ q2r.x q2r.old.x # path_int.x pwi2xsf.x q2r.x q2r.old.x bands_FS.x kvecs_FS.x metadyn_pp.x dynmat.x : dynmat.o rigid.o $(PWOBJS) $(MODULES) $(LIBOBJS) $(MPIF90) $(LDFLAGS) -o $@ \ dynmat.o rigid.o $(PWOBJS) $(MODULES) $(LIBOBJS) $(LIBS) - ( cd ../bin ; ln -fs ../pwtools/$@ . ) kpoints.x : kpoints.o $(PWOBJS) $(MODULES) $(LIBOBJS) $(MPIF90) $(LDFLAGS) -o $@ \ kpoints.o $(PWOBJS) $(MODULES) $(LIBOBJS) $(LIBS) - ( cd ../bin ; ln -fs ../pwtools/$@ . ) matdyn.x : matdyn.o rigid.o writemodes_dyn.o $(PWOBJS) $(PHOBJS) $(MODULES) $(LIBOBJS) $(MPIF90) $(LDFLAGS) -o $@ \ matdyn.o rigid.o writemodes_dyn.o $(PWOBJS) $(PHOBJS) $(MODULES) $(LIBOBJS) $(LIBS) - ( cd ../bin ; ln -fs ../pwtools/$@ . ) q2r.x : q2r.o rigid.o $(PWOBJS) $(MODULES) $(LIBOBJS) $(MPIF90) $(LDFLAGS) -o $@ \ q2r.o rigid.o $(PWOBJS) $(MODULES) $(LIBOBJS) $(LIBS) - ( cd ../bin ; ln -fs ../pwtools/$@ . ) q2r.old.x : q2r.pw.2.0.2.o rigid.o $(PWOBJS) $(MODULES) $(LIBOBJS) $(MPIF90) $(LDFLAGS) -o $@ \ q2r.pw.2.0.2.o rigid.o $(PWOBJS) $(MODULES) $(LIBOBJS) $(LIBS) - ( cd ../bin ; ln -fs ../pwtools/$@ . ) tldeps: test -n "$(TLDEPS)" && ( cd .. ; $(MAKE) $(MFLAGS) $(TLDEPS) || exit 1) || : clean : - /bin/rm -f pwi2xsf pwi2xsf_old *.x *.o *~ *.F90 *.mod *.d *.i work.pc include make.depend -------------- next part -------------- ! Copyright (C) 2001-2004 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 . ! ! ! Modifications Katalin Gaal-Nagy (search KG) November/December 2006 ! Adding an input parameter dynmatout = .true./.false ! for writing out the dynamical matrix file for a given q_in for ! all q in star(q_in). This file should be readable from dynmat.x ! The files with the dynamical matrices are called dynint.#q ! where #q is the number of the q point in the input list. ! For having all informations (c/a, etc) which are necessary for dynmat.x, ! also the readfc routine in this ! file has been modificated in order to return the corresponding information ! by adding celldm (it was already read, just not returned to the main). ! #include "f_defs.h" ! Module ifconstants ! ! All variables read from file that need dynamical allocation ! USE kinds, ONLY: DP REAL(DP), ALLOCATABLE :: frc(:,:,:,:,:,:,:), tau_blk(:,:), zeu(:,:,:) ! frc : interatomic force constants in real space ! tau_blk : atomic positions for the original cell ! zeu : effective charges for the original cell INTEGER, ALLOCATABLE :: ityp_blk(:) ! ityp_blk : atomic types for each atom of the original cell ! end Module ifconstants ! !--------------------------------------------------------------------- PROGRAM matdyn !----------------------------------------------------------------------- ! this program calculates the phonon frequencies for a list of generic ! q vectors starting from the interatomic force constants generated ! from the dynamical matrices as written by DFPT phonon code through ! the companion program q2r ! ! matdyn can generate a supercell of the original cell for mass ! approximation calculation. If supercell data are not specified ! in input, the unit cell, lattice vectors, atom types and positions ! are read from the force constant file ! ! Input cards: namelist &input ! flfrc file produced by q2r containing force constants (needed) ! asr (character) indicates the type of Acoustic Sum Rule imposed ! - 'no': no Acoustic Sum Rules imposed (default) ! - 'simple': previous implementation of the asr used ! (3 translational asr imposed by correction of ! the diagonal elements of the force constants matrix) ! - 'crystal': 3 translational asr imposed by optimized ! correction of the force constants (projection). ! - 'one-dim': 3 translational asr + 1 rotational asr ! imposed by optimized correction of the force constants ! (the rotation axis is the direction of periodicity; ! it will work only if this axis considered is one of ! the cartesian axis). ! - 'zero-dim': 3 translational asr + 3 rotational asr ! imposed by optimized correction of the force constants ! Note that in certain cases, not all the rotational asr ! can be applied (e.g. if there are only 2 atoms in a ! molecule or if all the atoms are aligned, etc.). ! In these cases the supplementary asr are cancelled ! during the orthonormalization procedure (see below). ! dos if .true. calculate phonon Density of States (DOS) ! using tetrahedra and a uniform q-point grid (see below) ! NB: may not work properly in noncubic materials ! if .false. calculate phonon bands from the list of q-points ! supplied in input ! nk1,nk2,nk3 uniform q-point grid for DOS calculation (includes q=0) ! deltaE energy step, in cm^(-1), for DOS calculation: from min ! to max phonon energy (default: 1 cm^(-1) if ndos is ! not specified) ! ndos number of energy steps for DOS calculations ! (no default: calculated from deltaE if not specified) ! fldos output file for dos (default: 'matdyn.dos') ! the dos is in states/cm(-1) plotted vs omega in cm(-1) ! and is normalised to 3*nat, i.e. the number of phonons ! flfrq output file for frequencies (default: 'matdyn.freq') ! flvec output file for normal modes (default: 'matdyn.modes') ! at supercell lattice vectors - must form a superlattice of the ! original lattice ! l1,l2,l3 supercell lattice vectors are original cell vectors ! multiplied by l1, l2, l3 respectively ! ntyp number of atom types in the supercell ! amass masses of atoms in the supercell ! readtau read atomic positions of the supercell from input ! (used to specify different masses) ! fltau write atomic positions of the supercell to file "fltau" ! (default: fltau=' ', do not write) ! ! if (readtau) atom types and positions in the supercell follow: ! (tau(i,na),i=1,3), ityp(na) ! Then, if (.not.dos) : ! nq number of q-points ! (q(i,n), i=1,3) nq q-points in 2pi/a units ! If q = 0, the direction qhat (q=>0) for the non-analytic part ! is extracted from the sequence of q-points as follows: ! qhat = q(n) - q(n-1) or qhat = q(n) - q(n+1) ! depending on which one is available and nonzero. ! For low-symmetry crystals, specify twice q = 0 in the list ! if you want to have q = 0 results for two different directions ! USE kinds, ONLY : DP USE mp, ONLY : mp_start, mp_env, mp_end, mp_barrier USE mp_global, ONLY : nproc, mpime, mp_global_start USE ifconstants ! IMPLICIT NONE ! INTEGER :: gid ! ! variables *_blk refer to the original cell, other variables ! to the (super)cell (which may coincide with the original cell) ! INTEGER:: nax, nax_blk INTEGER, PARAMETER:: ntypx=10, nrwsx=200 REAL(DP), PARAMETER :: eps=1.0e-6, rydcm1 = 13.6058*8065.5, & amconv = 1.66042e-24/9.1095e-28*0.5 INTEGER :: nr1, nr2, nr3, nsc, nk1, nk2, nk3, ntetra, ibrav CHARACTER(LEN=256) :: flfrc, flfrq, flvec, fltau, fldos CHARACTER(LEN=10) :: asr CHARACTER(LEN=9) :: symm_type LOGICAL :: dos, has_zstar COMPLEX(DP), ALLOCATABLE :: dyn(:,:,:,:), dyn_blk(:,:,:,:) COMPLEX(DP), ALLOCATABLE :: z(:,:) REAL(DP), ALLOCATABLE:: tau(:,:), q(:,:), w2(:,:), freq(:,:) INTEGER, ALLOCATABLE:: tetra(:,:), ityp(:), itau_blk(:) REAL(DP) :: at(3,3), bg(3,3), omega, alat, &! cell parameters and volume at_blk(3,3), bg_blk(3,3), &! original cell omega_blk, &! original cell volume epsil(3,3), &! dielectric tensor amass(ntypx), &! atomic masses amass_blk(ntypx), &! original atomic masses atws(3,3), &! lattice vector for WS initialization rws(0:3,nrwsx) ! nearest neighbor list, rws(0,*) = norm^2 ! INTEGER :: nat, nat_blk, ntyp, ntyp_blk, & l1, l2, l3, &! supercell dimensions nrws ! number of nearest neighbor ! LOGICAL :: readtau, la2F ! REAL(DP) :: qhat(3), qh, DeltaE, Emin, Emax, E, DOSofE(1) INTEGER :: n, i, j, it, nq, nqx, na, nb, ndos, iout ! ! KG: Additional parameters: ! ! for star_q ! CHARACTER(LEN=3) :: atm(ntypx) ! list of vectors in the star of q logical :: noinv INTEGER, PARAMETER :: natmax=100 integer :: nsym, s (3, 3, 48), invs (48), irt (48, natmax), isq (48), imq, modenum real(DP) :: rtau (3, 48, natmax), sxq (3, 48) ! ! for doing the extra loop ! LOGICAL :: dynmatout INTEGER :: nn, nnn ! running index for the q in the star INTEGER :: nnq ! degeneracy of the star of q ! ! for the dynmat_interpol file ! INTEGER :: fildyn_lent CHARACTER(LEN=8) :: fildyn_int, fildyn_out CHARACTER :: fildyn_d(0:9) REAL(DP) :: celldm(6) data fildyn_d/'0','1','2','3','4','5','6','7','8','9'/ INTEGER :: ios, iudyn CHARACTER(LEN=30) :: fildyn ! End KG ! KG: old namelist ! NAMELIST /input/ flfrc, amass, asr, flfrq, flvec, at, dos, & ! & fldos, nk1, nk2, nk3, l1, l2, l3, ntyp, readtau, fltau, & ! la2F, ndos ! KG: new namelist NAMELIST /input/ flfrc, amass, asr, flfrq, flvec, at, dos, & & fldos, nk1, nk2, nk3, l1, l2, l3, ntyp, readtau, fltau, & la2F, ndos, dynmatout ! ! CALL mp_start() ! CALL mp_env( nproc, mpime, gid ) ! IF ( mpime == 0 ) THEN ! ! ... all calculations are done by the first cpu ! ! set namelist default ! dos = .FALSE. deltaE = 1.0 ndos = 1 nk1 = 0 nk2 = 0 nk3 = 0 asr ='no' readtau=.FALSE. flfrc=' ' fldos='matdyn.dos' flfrq='matdyn.freq' flvec='matdyn.modes' fltau=' ' amass(:) =0.d0 amass_blk(:) =0.d0 at(:,:) = 0.d0 ntyp = 0 l1=1 l2=1 l3=1 la2F=.false. dos=.false. dynmatout=.false. ! CALL input_from_file ( ) ! READ (5,input) ! ! convert masses to atomic units ! amass(:) = amass(:) * amconv ! ! read force constants ! ntyp_blk = ntypx ! avoids fake out-of-bound error CALL readfc ( flfrc, nr1, nr2, nr3, epsil, nat_blk, & ibrav, symm_type, alat, celldm, at_blk, ntyp_blk, & amass_blk, omega_blk, has_zstar, atm) ! CALL recips ( at_blk(1,1),at_blk(1,2),at_blk(1,3), & bg_blk(1,1),bg_blk(1,2),bg_blk(1,3) ) ! ! set up (super)cell ! if (ntyp < 0) then call errore ('matdyn','wrong ntyp ', abs(ntyp)) else if (ntyp == 0) then ntyp=ntyp_blk end if ! ! masses (for mass approximation) ! DO it=1,ntyp IF (amass(it) < 0.d0) THEN CALL errore ('matdyn','wrong mass in the namelist',it) ELSE IF (amass(it) == 0.d0) THEN IF (it.LE.ntyp_blk) THEN WRITE (*,'(a,i3,a,a)') ' mass for atomic type ',it, & & ' not given; uses mass from file ',flfrc amass(it) = amass_blk(it) ELSE CALL errore ('matdyn','missing mass in the namelist',it) END IF END IF END DO ! ! lattice vectors ! IF (SUM(ABS(at(:,:))) == 0.d0) THEN IF (l1.LE.0 .OR. l2.LE.0 .OR. l3.LE.0) CALL & & errore ('matdyn',' wrong l1,l2 or l3',1) at(:,1) = at_blk(:,1)*DBLE(l1) at(:,2) = at_blk(:,2)*DBLE(l2) at(:,3) = at_blk(:,3)*DBLE(l3) END IF ! CALL check_at(at,bg_blk,alat,omega) ! ! the supercell contains "nsc" times the original unit cell ! nsc = NINT(omega/omega_blk) IF (ABS(omega/omega_blk-nsc) > eps) & CALL errore ('matdyn', 'volume ratio not integer', 1) ! ! read/generate atomic positions of the (super)cell ! nat = nat_blk * nsc nax = nat !!! nax_blk = nat_blk !!! ALLOCATE ( tau (3, nat), ityp(nat), itau_blk(nat_blk) ) ! IF (readtau) THEN CALL read_tau & (nat, nat_blk, ntyp, bg_blk, tau, tau_blk, ityp, itau_blk) ELSE CALL set_tau & (nat, nat_blk, at, at_blk, tau, tau_blk, ityp, ityp_blk, itau_blk) ENDIF ! IF (fltau.NE.' ') CALL write_tau (fltau, nat, tau, ityp) ! ! reciprocal lattice vectors ! CALL recips (at(1,1),at(1,2),at(1,3),bg(1,1),bg(1,2),bg(1,3)) ! ! build the WS cell corresponding to the force constant grid ! atws(:,1) = at_blk(:,1)*DBLE(nr1) atws(:,2) = at_blk(:,2)*DBLE(nr2) atws(:,3) = at_blk(:,3)*DBLE(nr3) ! initialize WS r-vectors CALL wsinit(rws,nrwsx,nrws,atws) ! ! end of (super)cell setup ! IF (dos) THEN IF (nk1 < 1 .OR. nk2 < 1 .OR. nk3 < 1) & CALL errore ('matdyn','specify correct q-point grid!',1) ntetra = 6 * nk1 * nk2 * nk3 nqx = nk1*nk2*nk3 ALLOCATE ( tetra(4,ntetra), q(3,nqx) ) CALL gen_qpoints (ibrav, at, bg, nat, tau, ityp, nk1, nk2, nk3, & symm_type, ntetra, nqx, nq, q, tetra) ELSE ! ! read q-point list ! READ (5,*) nq ALLOCATE ( q(3,nq) ) DO n = 1,nq READ (5,*) (q(i,n),i=1,3) END DO END IF ! IF (asr /= 'no') THEN CALL set_asr (asr, nr1, nr2, nr3, frc, zeu, & nat_blk, ibrav, tau_blk) END IF ! IF (flvec.EQ.' ') THEN iout=6 ELSE iout=4 OPEN (unit=iout,file=flvec,status='unknown',form='formatted') END IF ALLOCATE ( dyn(3,3,nat,nat), dyn_blk(3,3,nat_blk,nat_blk) ) ALLOCATE ( z(3*nat,3*nat), w2(3*nat,nq) ) if(la2F) open(300,file='dyna2F',status='unknown') ! ! KG: Modifications here: In the case of dynmatout=.true there is an additional ! loop over the q in star(q_in) ! DO n=1, nq dyn(:,:,:,:) = (0.d0, 0.d0) CALL setupmat (q(1,n), dyn, nat, at, bg, tau, itau_blk, nsc, alat, & dyn_blk, nat_blk, at_blk, bg_blk, tau_blk, omega_blk, & epsil, zeu, frc, nr1,nr2,nr3, has_zstar, rws, nrws) IF (q(1,n)==0.d0 .AND. q(2,n)==0.d0 .AND. q(3,n)==0.d0) THEN ! ! q = 0 : we need the direction q => 0 for the non-analytic part ! IF ( n == 1 ) THEN ! if q is the first point in the list IF ( nq > 1 ) THEN ! one more point qhat(:) = q(:,n) - q(:,n+1) ELSE ! no more points qhat(:) = 0.d0 END IF ELSE IF ( n > 1 ) THEN ! if q is not the first point in the list IF ( q(1,n-1)==0.d0 .AND. & q(2,n-1)==0.d0 .AND. & q(3,n-1)==0.d0 .AND. n < nq ) THEN ! if the preceding q is also 0 : qhat(:) = q(:,n) - q(:,n+1) ELSE ! if the preceding q is npt 0 : qhat(:) = q(:,n) - q(:,n-1) END IF END IF qh = SQRT(qhat(1)**2+qhat(2)**2+qhat(3)**2) ! write(*,*) ' qh, has_zstar ',qh, has_zstar IF (qh /= 0.d0) qhat(:) = qhat(:) / qh IF (qh /= 0.d0 .AND. .NOT. has_zstar) CALL infomsg & ('matdyn','Z* not found in file '//TRIM(flfrc)// & ', TO-LO splitting at q=0 will be absent!', -1) ! CALL nonanal (nat, nat_blk, itau_blk, epsil, qhat, zeu, omega, dyn) ! END IF ! CALL dyndiag(nat,ntyp,amass,ityp,dyn,w2(1,n),z) ! if(la2F) then write(300,*) n do na=1,3*nat write(300,*) (z(na,nb),nb=1,3*nat) end do ! na endif ! CALL writemodes(nax,nat,q(1,n),w2(1,n),z,iout) ! ! KG: here the additional loop, creating an output-file for each q_in, etc ! ! KG: Here starts the calculation of the dynamical matrices for all q in the star(q) ! and its writeout to a file ! IF (dynmatout) THEN ! Open file and write the header ! iudyn = 26 ! ! Creating filename with extention of the number of the point ! fildyn_lent = int(alog10(real(n)))+1 write(fildyn_int,'(i'//fildyn_d(fildyn_lent)//')') n read(fildyn_int,'(a'//fildyn_d(fildyn_lent)//')') fildyn_out fildyn = 'dynint.'//fildyn_out ! OPEN (unit=iudyn, file=fildyn, status='unknown', err=100, iostat=ios) write (iudyn, '("Dynamical matrix file")') write (iudyn, '(a)') ' ' write (iudyn, '(i3,i5,i3,6f11.7)') ntyp, nat, ibrav, celldm write(*,*) alat if (ibrav==0) then write (iudyn,'(a)') symm_type write (iudyn,'(2x,3f15.9)') ((at(i,j),i=1,3),j=1,3) end if do it = 1, ntyp write (iudyn, * ) it, ' ''', atm (it) , ' '' ', amass (it) enddo do na = 1, nat write (iudyn, '(2i5,3f15.7)') na, ityp (na) , (tau (j, na) , j = 1, 3) enddo ! ! Calculate the star of the q given by q(1,n) ! ! KG: email Paolo Giannozzi, pw_forum, Thu, 23 Nov 2006 10:22:23 +0100 ! all modes are calculated: ! modenum=0 noinv=.false. ! sxq=0.0 call star_q (q(1,n),at, bg, ibrav, symm_type, nat, tau, ityp, nr1, & nr2, nr3, & nsym, s, invs, irt, & rtau, nnq, sxq, isq, imq,& noinv, modenum) write(*,*) nnq ! ! Loop over q in star(q_in) ! Do nn=1,nnq dyn(:,:,:,:) = (0.d0, 0.d0) CALL setupmat (sxq(1,nn), dyn, nat, at, bg, tau, itau_blk, nsc, alat, & dyn_blk, nat_blk, at_blk, bg_blk, tau_blk, omega_blk, & epsil, zeu, frc, nr1,nr2,nr3, has_zstar, rws, nrws) IF (q(1,nn)==0.d0 .AND. q(2,nn)==0.d0 .AND. q(3,nn)==0.d0) THEN ! ! q = 0 : we need the direction q => 0 for the non-analytic part ! IF ( nn == 1 ) THEN ! if q is the first point in the list IF ( nnq > 1 ) THEN ! one more point qhat(:) = q(:,nn) - q(:,nn+1) ELSE ! no more points qhat(:) = 0.d0 END IF ELSE IF ( nn > 1 ) THEN ! if q is not the first point in the list IF ( q(1,nn-1)==0.d0 .AND. & q(2,nn-1)==0.d0 .AND. & q(3,nn-1)==0.d0 .AND. nn < nnq ) THEN ! if the preceding q is also 0 : qhat(:) = q(:,nn) - q(:,nn+1) ELSE ! if the preceding q is npt 0 : qhat(:) = q(:,nn) - q(:,nn-1) END IF END IF qh = SQRT(qhat(1)**2+qhat(2)**2+qhat(3)**2) ! write(*,*) ' qh, has_zstar ',qh, has_zstar IF (qh /= 0.d0) qhat(:) = qhat(:) / qh IF (qh /= 0.d0 .AND. .NOT. has_zstar) CALL infomsg & ('matdyn','Z* not found in file '//TRIM(flfrc)// & ', TO-LO splitting at q=0 will be absent!', -1) ! CALL nonanal (nat, nat_blk, itau_blk, epsil, qhat, zeu, omega, dyn) ! END IF ! call write_dyn_on_file (sxq(1, nn), dyn, nat, iudyn) ! END DO ! nnq write(iudyn,*) ' ' CALL writemodes(nax,nat,q(1,n),w2(1,n),z,iudyn) ! ! close the file with the interpolated dynamical matrices ! close (iudyn) ! ENDIF ! END DO !nq ! ! KG: End of modifications ! ! if(la2F) close(300) ! IF(iout .NE. 6) CLOSE(unit=iout) ! ALLOCATE (freq(3*nat, nq)) DO n=1,nq ! freq(i,n) = frequencies in cm^(-1), with negative sign if omega^2 is negative DO i=1,3*nat freq(i,n)= SQRT(ABS(w2(i,n))) * rydcm1 IF (w2(i,n) < 0.0) freq(i,n) = -freq(i,n) END DO END DO ! IF(flfrq.NE.' ') THEN OPEN (unit=2,file=flfrq ,status='unknown',form='formatted') WRITE(2, '(" &plot nbnd=",i4,", nks=",i4," /")') 3*nat, nq DO n=1, nq WRITE(2, '(10x,3f10.6)') q(1,n), q(2,n), q(3,n) WRITE(2,'(6f10.4)') (freq(i,n), i=1,3*nat) END DO CLOSE(unit=2) END IF ! ! IF (dos) THEN Emin = 0.0 Emax = 0.0 DO n=1,nq DO i=1, 3*nat Emin = MIN (Emin, freq(i,n)) Emax = MAX (Emax, freq(i,n)) END DO END DO ! if (ndos > 1) then DeltaE = (Emax - Emin)/(ndos-1) else ndos = NINT ( (Emax - Emin) / DeltaE + 1.51 ) end if OPEN (unit=2,file=fldos,status='unknown',form='formatted') DO n= 1, ndos E = Emin + (n - 1) * DeltaE CALL dos_t(freq, 1, 3*nat, nq, ntetra, tetra, E, DOSofE) ! ! The factor 0.5 corrects for the factor 2 in dos_t, ! that accounts for the spin in the electron DOS. ! !WRITE (2, '(F15.10,F15.2,F15.6,F20.5)') & ! E, E*rydcm1, E*rydTHz, 0.5d0*DOSofE(1) WRITE (2, '(E12.4,E12.4)') E, 0.5d0*DOSofE(1) END DO CLOSE(unit=2) END IF !dos DEALLOCATE (z, w2, dyn, dyn_blk) ! ! for a2F ! IF(la2F) THEN ! ! convert frequencies to Ry ! freq(:,:)= freq(:,:) / rydcm1 Emin = Emin / rydcm1 DeltaE=DeltaE/ rydcm1 ! call a2Fdos (nat, nq, nr1, nr2, nr3, ibrav, at, bg, tau, alat, & nsc, nat_blk, at_blk, bg_blk, itau_blk, omega_blk, & rws, nrws, dos, Emin, DeltaE, ndos, & ntetra, tetra, asr, q, freq) END IF DEALLOCATE ( freq) ! END IF ! CALL mp_barrier() ! CALL mp_end() ! 100 CALL errore ('openfilq', 'opening file'//fildyn, ABS (ios) ) REWIND (iudyn) STOP ! END PROGRAM matdyn ! !----------------------------------------------------------------------- SUBROUTINE readfc ( flfrc, nr1, nr2, nr3, epsil, nat, & ibrav, symm_type, alat, celldm, at, ntyp, amass, omega, has_zstar, atm ) !----------------------------------------------------------------------- ! ! Modification K.GN: ! return also celldm, since it is required for the dynmat files ! USE kinds, ONLY : DP USE ifconstants,ONLY : tau => tau_blk, ityp => ityp_blk, frc, zeu ! IMPLICIT NONE ! I/O variable CHARACTER(LEN=256) flfrc INTEGER ibrav, nr1,nr2,nr3,nat, ntyp REAL(DP) alat, at(3,3), epsil(3,3) LOGICAL has_zstar ! local variables INTEGER i, j, na, nb, m1,m2,m3 INTEGER ibid, jbid, nabid, nbbid, m1bid,m2bid,m3bid REAL(DP) amass(ntyp), amass_from_file, celldm(6), omega INTEGER nt CHARACTER(LEN=3) atm(ntyp) CHARACTER(LEN=9) symm_type ! ! OPEN (unit=1,file=flfrc,status='old',form='formatted') ! ! read cell data ! READ(1,*) ntyp,nat,ibrav,(celldm(i),i=1,6) if (ibrav==0) then read(1,'(a)') symm_type read(1,*) ((at(i,j),i=1,3),j=1,3) end if ! CALL latgen(ibrav,celldm,at(1,1),at(1,2),at(1,3),omega) alat = celldm(1) at = at / alat ! bring at in units of alat CALL volume(alat,at(1,1),at(1,2),at(1,3),omega) ! ! read atomic types, positions and masses ! DO nt = 1,ntyp READ(1,*) i,atm(nt),amass_from_file IF (i.NE.nt) CALL errore ('readfc','wrong data read',nt) IF (amass(nt).EQ.0.d0) THEN amass(nt) = amass_from_file ELSE WRITE(*,*) 'for atomic type',nt,' mass from file not used' END IF END DO ! ALLOCATE (tau(3,nat), ityp(nat), zeu(3,3,nat)) ! DO na=1,nat READ(1,*) i,ityp(na),(tau(j,na),j=1,3) IF (i.NE.na) CALL errore ('readfc','wrong data read',na) END DO ! ! read macroscopic variable ! READ (1,*) has_zstar IF (has_zstar) THEN READ(1,*) ((epsil(i,j),j=1,3),i=1,3) DO na=1,nat READ(1,*) READ(1,*) ((zeu(i,j,na),j=1,3),i=1,3) END DO ELSE zeu (:,:,:) = 0.d0 epsil(:,:) = 0.d0 END IF ! READ (1,*) nr1,nr2,nr3 ! ! read real-space interatomic force constants ! ALLOCATE ( frc(nr1,nr2,nr3,3,3,nat,nat) ) frc(:,:,:,:,:,:,:) = 0.d0 DO i=1,3 DO j=1,3 DO na=1,nat DO nb=1,nat READ (1,*) ibid, jbid, nabid, nbbid IF(i .NE.ibid .OR. j .NE.jbid .OR. & na.NE.nabid .OR. nb.NE.nbbid) & CALL errore ('readfc','error in reading',1) READ (1,*) (((m1bid, m2bid, m3bid, & frc(m1,m2,m3,i,j,na,nb), & m1=1,nr1),m2=1,nr2),m3=1,nr3) END DO END DO END DO END DO ! CLOSE(unit=1) ! RETURN END SUBROUTINE readfc ! !----------------------------------------------------------------------- SUBROUTINE frc_blk(dyn,q,tau,nat,nr1,nr2,nr3,frc,at,bg,rws,nrws) !----------------------------------------------------------------------- ! calculates the dynamical matrix at q from the (short-range part of the) ! force constants ! USE kinds, ONLY : DP ! IMPLICIT NONE INTEGER nr1, nr2, nr3, nat, n1, n2, n3, & ipol, jpol, na, nb, m1, m2, m3, nint, i,j, nrws COMPLEX(DP) dyn(3,3,nat,nat) REAL(DP) frc(nr1,nr2,nr3,3,3,nat,nat), tau(3,nat), q(3), arg, & at(3,3), bg(3,3), r(3), weight, r_ws(3), & total_weight, rws(0:3,nrws) REAL(DP), PARAMETER:: tpi = 2.0*3.14159265358979d0 REAL(DP), EXTERNAL :: wsweight ! DO na=1, nat DO nb=1, nat total_weight=0.0d0 DO n1=-2*nr1,2*nr1 DO n2=-2*nr2,2*nr2 DO n3=-2*nr3,2*nr3 ! ! SUM OVER R VECTORS IN THE SUPERCELL - VERY VERY SAFE RANGE! ! DO i=1, 3 r(i) = n1*at(i,1)+n2*at(i,2)+n3*at(i,3) r_ws(i) = r(i) + tau(i,na)-tau(i,nb) END DO weight = wsweight(r_ws,rws,nrws) IF (weight .GT. 0.0) THEN ! ! FIND THE VECTOR CORRESPONDING TO R IN THE ORIGINAL CELL ! m1 = MOD(n1+1,nr1) IF(m1.LE.0) m1=m1+nr1 m2 = MOD(n2+1,nr2) IF(m2.LE.0) m2=m2+nr2 m3 = MOD(n3+1,nr3) IF(m3.LE.0) m3=m3+nr3 ! ! FOURIER TRANSFORM ! arg = tpi*(q(1)*r(1) + q(2)*r(2) + q(3)*r(3)) DO ipol=1, 3 DO jpol=1, 3 dyn(ipol,jpol,na,nb) = & dyn(ipol,jpol,na,nb) + & frc(m1,m2,m3,ipol,jpol,na,nb) & *CMPLX(COS(arg),-SIN(arg))*weight END DO END DO END IF total_weight=total_weight + weight END DO END DO END DO IF (ABS(total_weight-nr1*nr2*nr3).GT.1.0d-8) THEN WRITE(*,*) total_weight CALL errore ('frc_blk','wrong total_weight',1) END IF END DO END DO ! RETURN END SUBROUTINE frc_blk ! !----------------------------------------------------------------------- SUBROUTINE setupmat (q,dyn,nat,at,bg,tau,itau_blk,nsc,alat, & & dyn_blk,nat_blk,at_blk,bg_blk,tau_blk,omega_blk, & & epsil,zeu,frc,nr1,nr2,nr3,has_zstar,rws,nrws) !----------------------------------------------------------------------- ! compute the dynamical matrix (the analytic part only) ! USE kinds, ONLY : DP ! IMPLICIT NONE REAL(DP), PARAMETER :: tpi=2.d0*3.14159265358979d0 ! ! I/O variables ! INTEGER:: nr1, nr2, nr3, nat, nat_blk, nsc, nrws, itau_blk(nat) REAL(DP) :: q(3), tau(3,nat), at(3,3), bg(3,3), alat, & epsil(3,3), zeu(3,3,nat_blk), rws(0:3,nrws), & frc(nr1,nr2,nr3,3,3,nat_blk,nat_blk) REAL(DP) :: tau_blk(3,nat_blk), at_blk(3,3), bg_blk(3,3), omega_blk COMPLEX(DP) dyn_blk(3,3,nat_blk,nat_blk) COMPLEX(DP) :: dyn(3,3,nat,nat) LOGICAL has_zstar ! ! local variables ! REAL(DP) :: arg COMPLEX(DP) :: cfac(nat) INTEGER :: i,j,k, na,nb, na_blk, nb_blk, iq REAL(DP) qp(3), qbid(3,nsc) ! automatic array ! ! CALL q_gen(nsc,qbid,at_blk,bg_blk,at,bg) ! DO iq=1,nsc ! DO k=1,3 qp(k)= q(k) + qbid(k,iq) END DO ! dyn_blk(:,:,:,:) = (0.d0,0.d0) CALL frc_blk (dyn_blk,qp,tau_blk,nat_blk, & & nr1,nr2,nr3,frc,at_blk,bg_blk,rws,nrws) IF (has_zstar) & CALL rgd_blk(nr1,nr2,nr3,nat_blk,dyn_blk,qp,tau_blk, & epsil,zeu,bg_blk,omega_blk,+1.d0) ! DO na=1,nat na_blk = itau_blk(na) DO nb=1,nat nb_blk = itau_blk(nb) ! arg=tpi* ( qp(1) * ( (tau(1,na)-tau_blk(1,na_blk)) - & (tau(1,nb)-tau_blk(1,nb_blk)) ) + & qp(2) * ( (tau(2,na)-tau_blk(2,na_blk)) - & (tau(2,nb)-tau_blk(2,nb_blk)) ) + & qp(3) * ( (tau(3,na)-tau_blk(3,na_blk)) - & (tau(3,nb)-tau_blk(3,nb_blk)) ) ) ! cfac(nb) = CMPLX(COS(arg),SIN(arg))/nsc ! END DO ! nb ! DO i=1,3 DO j=1,3 ! DO nb=1,nat nb_blk = itau_blk(nb) dyn(i,j,na,nb) = dyn(i,j,na,nb) + cfac(nb) * & dyn_blk(i,j,na_blk,nb_blk) END DO ! nb ! END DO ! j END DO ! i END DO ! na ! END DO ! iq ! RETURN END SUBROUTINE setupmat ! ! !---------------------------------------------------------------------- SUBROUTINE set_asr (asr, nr1, nr2, nr3, frc, zeu, nat, ibrav, tau) !----------------------------------------------------------------------- ! USE kinds, ONLY : DP ! IMPLICIT NONE CHARACTER (LEN=10) :: asr INTEGER :: nr1, nr2, nr3, nat, ibrav REAL(DP) :: frc(nr1,nr2,nr3,3,3,nat,nat), zeu(3,3,nat),tau(3,nat) ! INTEGER :: axis, n, i, j, na, nb, n1,n2,n3, m,p,k,l,q,r, i1,j1,na1 REAL(DP) :: frc_new(nr1,nr2,nr3,3,3,nat,nat), zeu_new(3,3,nat) type vector real(DP),pointer :: vec(:,:,:,:,:,:,:) end type vector ! type (vector) u(6*3*nat) ! These are the "vectors" associated with the sum rules on force-constants ! integer u_less(6*3*nat),n_less,i_less ! indices of the vectors u that are not independent to the preceding ones, ! n_less = number of such vectors, i_less = temporary parameter ! integer ind_v(9*nat*nat*nr1*nr2*nr3,2,7) real(DP) v(9*nat*nat*nr1*nr2*nr3,2) ! These are the "vectors" associated with symmetry conditions, coded by ! indicating the positions (i.e. the seven indices) of the non-zero elements (there ! should be only 2 of them) and the value of that element. We do so in order ! to limit the amount of memory used. ! real(DP) w(nr1,nr2,nr3,3,3,nat,nat), x(nr1,nr2,nr3,3,3,nat,nat) ! temporary vectors and parameters real(DP) :: scal,norm2, sum ! real(DP) zeu_u(6*3,3,3,nat) ! These are the "vectors" associated with the sum rules on effective charges ! integer zeu_less(6*3),nzeu_less,izeu_less ! indices of the vectors zeu_u that are not independent to the preceding ones, ! nzeu_less = number of such vectors, izeu_less = temporary parameter ! real(DP) zeu_w(3,3,nat), zeu_x(3,3,nat) ! temporary vectors ! Initialization. n is the number of sum rules to be considered (if asr.ne.'simple') ! and 'axis' is the rotation axis in the case of a 1D system ! (i.e. the rotation axis is (Ox) if axis='1', (Oy) if axis='2' and (Oz) if axis='3') ! if((asr.ne.'simple').and.(asr.ne.'crystal').and.(asr.ne.'one-dim') & .and.(asr.ne.'zero-dim')) then call errore('matdyn','invalid Acoustic Sum Rule:' // asr, 1) endif if(asr.eq.'crystal') n=3 if(asr.eq.'one-dim') then ! the direction of periodicity is the rotation axis ! It will work only if the crystal axis considered is one of ! the cartesian axis (typically, ibrav=1, 6 or 8, or 4 along the ! z-direction) if (nr1*nr2*nr3.eq.1) axis=3 if ((nr1.ne.1).and.(nr2*nr3.eq.1)) axis=1 if ((nr2.ne.1).and.(nr1*nr3.eq.1)) axis=2 if ((nr3.ne.1).and.(nr1*nr2.eq.1)) axis=3 if (((nr1.ne.1).and.(nr2.ne.1)).or.((nr2.ne.1).and. & (nr3.ne.1)).or.((nr1.ne.1).and.(nr3.ne.1))) then call errore('matdyn','too many directions of & & periodicity in 1D system',axis) endif if ((ibrav.ne.1).and.(ibrav.ne.6).and.(ibrav.ne.8).and. & ((ibrav.ne.4).or.(axis.ne.3)) ) then write(6,*) 'asr: rotational axis may be wrong' endif write(6,'("asr rotation axis in 1D system= ",I4)') axis n=4 endif if(asr.eq.'zero-dim') n=6 ! Acoustic Sum Rule on effective charges ! if(asr.eq.'simple') then do i=1,3 do j=1,3 sum=0.0 do na=1,nat sum = sum + zeu(i,j,na) end do do na=1,nat zeu(i,j,na) = zeu(i,j,na) - sum/nat end do end do end do else ! generating the vectors of the orthogonal of the subspace to project ! the effective charges matrix on ! zeu_u(:,:,:,:)=0.0d0 do i=1,3 do j=1,3 do na=1,nat zeu_new(i,j,na)=zeu(i,j,na) enddo enddo enddo ! p=0 do i=1,3 do j=1,3 ! These are the 3*3 vectors associated with the ! translational acoustic sum rules p=p+1 zeu_u(p,i,j,:)=1.0d0 ! enddo enddo ! if (n.eq.4) then do i=1,3 ! These are the 3 vectors associated with the ! single rotational sum rule (1D system) p=p+1 do na=1,nat zeu_u(p,i,MOD(axis,3)+1,na)=-tau(MOD(axis+1,3)+1,na) zeu_u(p,i,MOD(axis+1,3)+1,na)=tau(MOD(axis,3)+1,na) enddo ! enddo endif ! if (n.eq.6) then do i=1,3 do j=1,3 ! These are the 3*3 vectors associated with the ! three rotational sum rules (0D system - typ. molecule) p=p+1 do na=1,nat zeu_u(p,i,MOD(j,3)+1,na)=-tau(MOD(j+1,3)+1,na) zeu_u(p,i,MOD(j+1,3)+1,na)=tau(MOD(j,3)+1,na) enddo ! enddo enddo endif ! ! Gram-Schmidt orthonormalization of the set of vectors created. ! nzeu_less=0 do k=1,p zeu_w(:,:,:)=zeu_u(k,:,:,:) zeu_x(:,:,:)=zeu_u(k,:,:,:) do q=1,k-1 r=1 do izeu_less=1,nzeu_less if (zeu_less(izeu_less).eq.q) r=0 enddo if (r.ne.0) then call sp_zeu(zeu_x,zeu_u(q,:,:,:),nat,scal) zeu_w(:,:,:) = zeu_w(:,:,:) - scal* zeu_u(q,:,:,:) endif enddo call sp_zeu(zeu_w,zeu_w,nat,norm2) if (norm2.gt.1.0d-16) then zeu_u(k,:,:,:) = zeu_w(:,:,:) / DSQRT(norm2) else nzeu_less=nzeu_less+1 zeu_less(nzeu_less)=k endif enddo ! ! ! Projection of the effective charge "vector" on the orthogonal of the ! subspace of the vectors verifying the sum rules ! zeu_w(:,:,:)=0.0d0 do k=1,p r=1 do izeu_less=1,nzeu_less if (zeu_less(izeu_less).eq.k) r=0 enddo if (r.ne.0) then zeu_x(:,:,:)=zeu_u(k,:,:,:) call sp_zeu(zeu_x,zeu_new,nat,scal) zeu_w(:,:,:) = zeu_w(:,:,:) + scal*zeu_u(k,:,:,:) endif enddo ! ! Final substraction of the former projection to the initial zeu, to get ! the new "projected" zeu ! zeu_new(:,:,:)=zeu_new(:,:,:) - zeu_w(:,:,:) call sp_zeu(zeu_w,zeu_w,nat,norm2) write(6,'("Norm of the difference between old and new effective ", & & "charges: ",F25.20)') SQRT(norm2) ! ! Check projection ! !write(6,'("Check projection of zeu")') !do k=1,p ! zeu_x(:,:,:)=zeu_u(k,:,:,:) ! call sp_zeu(zeu_x,zeu_new,nat,scal) ! if (DABS(scal).gt.1d-10) write(6,'("k= ",I8," zeu_new|zeu_u(k)= ",F15.10)') k,scal !enddo ! do i=1,3 do j=1,3 do na=1,nat zeu(i,j,na)=zeu_new(i,j,na) enddo enddo enddo endif ! ! ! ! ! ! ! Acoustic Sum Rule on force constants in real space ! if(asr.eq.'simple') then do i=1,3 do j=1,3 do na=1,nat sum=0.0 do nb=1,nat do n1=1,nr1 do n2=1,nr2 do n3=1,nr3 sum=sum+frc(n1,n2,n3,i,j,na,nb) end do end do end do end do frc(1,1,1,i,j,na,na) = frc(1,1,1,i,j,na,na) - sum ! write(6,*) ' na, i, j, sum = ',na,i,j,sum end do end do end do else ! generating the vectors of the orthogonal of the subspace to project ! the force-constants matrix on ! do k=1,18*nat allocate(u(k) % vec(nr1,nr2,nr3,3,3,nat,nat)) u(k) % vec (:,:,:,:,:,:,:)=0.0d0 enddo do i=1,3 do j=1,3 do na=1,nat do nb=1,nat do n1=1,nr1 do n2=1,nr2 do n3=1,nr3 frc_new(n1,n2,n3,i,j,na,nb)=frc(n1,n2,n3,i,j,na,nb) enddo enddo enddo enddo enddo enddo enddo ! p=0 do i=1,3 do j=1,3 do na=1,nat ! These are the 3*3*nat vectors associated with the ! translational acoustic sum rules p=p+1 u(p) % vec (:,:,:,i,j,na,:)=1.0d0 ! enddo enddo enddo ! if (n.eq.4) then do i=1,3 do na=1,nat ! These are the 3*nat vectors associated with the ! single rotational sum rule (1D system) p=p+1 do nb=1,nat u(p) % vec (:,:,:,i,MOD(axis,3)+1,na,nb)=-tau(MOD(axis+1,3)+1,nb) u(p) % vec (:,:,:,i,MOD(axis+1,3)+1,na,nb)=tau(MOD(axis,3)+1,nb) enddo ! enddo enddo endif ! if (n.eq.6) then do i=1,3 do j=1,3 do na=1,nat ! These are the 3*3*nat vectors associated with the ! three rotational sum rules (0D system - typ. molecule) p=p+1 do nb=1,nat u(p) % vec (:,:,:,i,MOD(j,3)+1,na,nb)=-tau(MOD(j+1,3)+1,nb) u(p) % vec (:,:,:,i,MOD(j+1,3)+1,na,nb)=tau(MOD(j,3)+1,nb) enddo ! enddo enddo enddo endif ! m=0 do i=1,3 do j=1,3 do na=1,nat do nb=1,nat do n1=1,nr1 do n2=1,nr2 do n3=1,nr3 ! These are the vectors associated with the symmetry constraints q=1 l=1 do while((l.le.m).and.(q.ne.0)) if ((ind_v(l,1,1).eq.n1).and.(ind_v(l,1,2).eq.n2).and. & (ind_v(l,1,3).eq.n3).and.(ind_v(l,1,4).eq.i).and. & (ind_v(l,1,5).eq.j).and.(ind_v(l,1,6).eq.na).and. & (ind_v(l,1,7).eq.nb)) q=0 if ((ind_v(l,2,1).eq.n1).and.(ind_v(l,2,2).eq.n2).and. & (ind_v(l,2,3).eq.n3).and.(ind_v(l,2,4).eq.i).and. & (ind_v(l,2,5).eq.j).and.(ind_v(l,2,6).eq.na).and. & (ind_v(l,2,7).eq.nb)) q=0 l=l+1 enddo if ((n1.eq.MOD(nr1+1-n1,nr1)+1).and.(n2.eq.MOD(nr2+1-n2,nr2)+1) & .and.(n3.eq.MOD(nr3+1-n3,nr3)+1).and.(i.eq.j).and.(na.eq.nb)) q=0 if (q.ne.0) then m=m+1 ind_v(m,1,1)=n1 ind_v(m,1,2)=n2 ind_v(m,1,3)=n3 ind_v(m,1,4)=i ind_v(m,1,5)=j ind_v(m,1,6)=na ind_v(m,1,7)=nb v(m,1)=1.0d0/DSQRT(2.0d0) ind_v(m,2,1)=MOD(nr1+1-n1,nr1)+1 ind_v(m,2,2)=MOD(nr2+1-n2,nr2)+1 ind_v(m,2,3)=MOD(nr3+1-n3,nr3)+1 ind_v(m,2,4)=j ind_v(m,2,5)=i ind_v(m,2,6)=nb ind_v(m,2,7)=na v(m,2)=-1.0d0/DSQRT(2.0d0) endif enddo enddo enddo enddo enddo enddo enddo ! ! Gram-Schmidt orthonormalization of the set of vectors created. ! Note that the vectors corresponding to symmetry constraints are already ! orthonormalized by construction. ! n_less=0 do k=1,p w(:,:,:,:,:,:,:)=u(k) % vec (:,:,:,:,:,:,:) x(:,:,:,:,:,:,:)=u(k) % vec (:,:,:,:,:,:,:) do l=1,m ! call sp2(x,v(l,:),ind_v(l,:,:),nr1,nr2,nr3,nat,scal) do r=1,2 n1=ind_v(l,r,1) n2=ind_v(l,r,2) n3=ind_v(l,r,3) i=ind_v(l,r,4) j=ind_v(l,r,5) na=ind_v(l,r,6) nb=ind_v(l,r,7) w(n1,n2,n3,i,j,na,nb)=w(n1,n2,n3,i,j,na,nb)-scal*v(l,r) enddo enddo if (k.le.(9*nat)) then na1=MOD(k,nat) if (na1.eq.0) na1=nat j1=MOD((k-na1)/nat,3)+1 i1=MOD((((k-na1)/nat)-j1+1)/3,3)+1 else q=k-9*nat if (n.eq.4) then na1=MOD(q,nat) if (na1.eq.0) na1=nat i1=MOD((q-na1)/nat,3)+1 else na1=MOD(q,nat) if (na1.eq.0) na1=nat j1=MOD((q-na1)/nat,3)+1 i1=MOD((((q-na1)/nat)-j1+1)/3,3)+1 endif endif do q=1,k-1 r=1 do i_less=1,n_less if (u_less(i_less).eq.q) r=0 enddo if (r.ne.0) then call sp3(x,u(q) % vec (:,:,:,:,:,:,:), i1,na1,nr1,nr2,nr3,nat,scal) w(:,:,:,:,:,:,:) = w(:,:,:,:,:,:,:) - scal* u(q) % vec (:,:,:,:,:,:,:) endif enddo call sp1(w,w,nr1,nr2,nr3,nat,norm2) if (norm2.gt.1.0d-16) then u(k) % vec (:,:,:,:,:,:,:) = w(:,:,:,:,:,:,:) / DSQRT(norm2) else n_less=n_less+1 u_less(n_less)=k endif enddo ! ! Projection of the force-constants "vector" on the orthogonal of the ! subspace of the vectors verifying the sum rules and symmetry contraints ! w(:,:,:,:,:,:,:)=0.0d0 do l=1,m call sp2(frc_new,v(l,:),ind_v(l,:,:),nr1,nr2,nr3,nat,scal) do r=1,2 n1=ind_v(l,r,1) n2=ind_v(l,r,2) n3=ind_v(l,r,3) i=ind_v(l,r,4) j=ind_v(l,r,5) na=ind_v(l,r,6) nb=ind_v(l,r,7) w(n1,n2,n3,i,j,na,nb)=w(n1,n2,n3,i,j,na,nb)+scal*v(l,r) enddo enddo do k=1,p r=1 do i_less=1,n_less if (u_less(i_less).eq.k) r=0 enddo if (r.ne.0) then x(:,:,:,:,:,:,:)=u(k) % vec (:,:,:,:,:,:,:) call sp1(x,frc_new,nr1,nr2,nr3,nat,scal) w(:,:,:,:,:,:,:) = w(:,:,:,:,:,:,:) + scal*u(k)%vec(:,:,:,:,:,:,:) endif deallocate(u(k) % vec) enddo ! ! Final substraction of the former projection to the initial frc, to get ! the new "projected" frc ! frc_new(:,:,:,:,:,:,:)=frc_new(:,:,:,:,:,:,:) - w(:,:,:,:,:,:,:) call sp1(w,w,nr1,nr2,nr3,nat,norm2) write(6,'("Norm of the difference between old and new force-constants:",& & F25.20)') SQRT(norm2) ! ! Check projection ! !write(6,'("Check projection IFC")') !do l=1,m ! call sp2(frc_new,v(l,:),ind_v(l,:,:),nr1,nr2,nr3,nat,scal) ! if (DABS(scal).gt.1d-10) write(6,'("l= ",I8," frc_new|v(l)= ",F15.10)') l,scal !enddo !do k=1,p ! x(:,:,:,:,:,:,:)=u(k) % vec (:,:,:,:,:,:,:) ! call sp1(x,frc_new,nr1,nr2,nr3,nat,scal) ! if (DABS(scal).gt.1d-10) write(6,'("k= ",I8," frc_new|u(k)= ",F15.10)') k,scal ! deallocate(u(k) % vec) !enddo ! do i=1,3 do j=1,3 do na=1,nat do nb=1,nat do n1=1,nr1 do n2=1,nr2 do n3=1,nr3 frc(n1,n2,n3,i,j,na,nb)=frc_new(n1,n2,n3,i,j,na,nb) enddo enddo enddo enddo enddo enddo enddo endif ! ! return end subroutine set_asr ! !---------------------------------------------------------------------- subroutine sp_zeu(zeu_u,zeu_v,nat,scal) !----------------------------------------------------------------------- ! ! does the scalar product of two effective charges matrices zeu_u and zeu_v ! (considered as vectors in the R^(3*3*nat) space, and coded in the usual way) ! USE kinds, ONLY: DP implicit none integer i,j,na,nat real(DP) zeu_u(3,3,nat) real(DP) zeu_v(3,3,nat) real(DP) scal ! ! scal=0.0d0 do i=1,3 do j=1,3 do na=1,nat scal=scal+zeu_u(i,j,na)*zeu_v(i,j,na) enddo enddo enddo ! return ! end subroutine sp_zeu ! ! !---------------------------------------------------------------------- subroutine sp1(u,v,nr1,nr2,nr3,nat,scal) !----------------------------------------------------------------------- ! ! does the scalar product of two force-constants matrices u and v (considered as ! vectors in the R^(3*3*nat*nat*nr1*nr2*nr3) space, and coded in the usual way) ! USE kinds, ONLY: DP implicit none integer nr1,nr2,nr3,i,j,na,nb,n1,n2,n3,nat real(DP) u(nr1,nr2,nr3,3,3,nat,nat) real(DP) v(nr1,nr2,nr3,3,3,nat,nat) real(DP) scal ! ! scal=0.0d0 do i=1,3 do j=1,3 do na=1,nat do nb=1,nat do n1=1,nr1 do n2=1,nr2 do n3=1,nr3 scal=scal+u(n1,n2,n3,i,j,na,nb)*v(n1,n2,n3,i,j,na,nb) enddo enddo enddo enddo enddo enddo enddo ! return ! end subroutine sp1 ! !---------------------------------------------------------------------- subroutine sp2(u,v,ind_v,nr1,nr2,nr3,nat,scal) !----------------------------------------------------------------------- ! ! does the scalar product of two force-constants matrices u and v (considered as ! vectors in the R^(3*3*nat*nat*nr1*nr2*nr3) space). u is coded in the usual way ! but v is coded as explained when defining the vectors corresponding to the ! symmetry constraints ! USE kinds, ONLY: DP implicit none integer nr1,nr2,nr3,i,nat real(DP) u(nr1,nr2,nr3,3,3,nat,nat) integer ind_v(2,7) real(DP) v(2) real(DP) scal ! ! scal=0.0d0 do i=1,2 scal=scal+u(ind_v(i,1),ind_v(i,2),ind_v(i,3),ind_v(i,4),ind_v(i,5),ind_v(i,6), & ind_v(i,7))*v(i) enddo ! return ! end subroutine sp2 ! !---------------------------------------------------------------------- subroutine sp3(u,v,i,na,nr1,nr2,nr3,nat,scal) !----------------------------------------------------------------------- ! ! like sp1, but in the particular case when u is one of the u(k)%vec ! defined in set_asr (before orthonormalization). In this case most of the ! terms are zero (the ones that are not are characterized by i and na), so ! that a lot of computer time can be saved (during Gram-Schmidt). ! USE kinds, ONLY: DP implicit none integer nr1,nr2,nr3,i,j,na,nb,n1,n2,n3,nat real(DP) u(nr1,nr2,nr3,3,3,nat,nat) real(DP) v(nr1,nr2,nr3,3,3,nat,nat) real(DP) scal ! ! scal=0.0d0 do j=1,3 do nb=1,nat do n1=1,nr1 do n2=1,nr2 do n3=1,nr3 scal=scal+u(n1,n2,n3,i,j,na,nb)*v(n1,n2,n3,i,j,na,nb) enddo enddo enddo enddo enddo ! return ! end subroutine sp3 ! !----------------------------------------------------------------------- SUBROUTINE q_gen(nsc,qbid,at_blk,bg_blk,at,bg) !----------------------------------------------------------------------- ! generate list of q (qbid) that are G-vectors of the supercell ! but not of the bulk ! USE kinds, ONLY : DP ! IMPLICIT NONE INTEGER :: nsc REAL(DP) qbid(3,nsc), at_blk(3,3), bg_blk(3,3), at(3,3), bg(3,3) ! INTEGER, PARAMETER:: nr1=4, nr2=4, nr3=4, & nrm=(2*nr1+1)*(2*nr2+1)*(2*nr3+1) REAL(DP), PARAMETER:: eps=1.0e-7 INTEGER :: i, j, k,i1, i2, i3, idum(nrm), iq REAL(DP) :: qnorm(nrm), qbd(3,nrm) ,qwork(3), delta LOGICAL lbho ! i = 0 DO i1=-nr1,nr1 DO i2=-nr2,nr2 DO i3=-nr3,nr3 i = i + 1 DO j=1,3 qwork(j) = i1*bg(j,1) + i2*bg(j,2) + i3*bg(j,3) END DO ! j ! qnorm(i) = qwork(1)**2 + qwork(2)**2 + qwork(3)**2 ! DO j=1,3 ! qbd(j,i) = at_blk(1,j)*qwork(1) + & at_blk(2,j)*qwork(2) + & at_blk(3,j)*qwork(3) END DO ! j ! idum(i) = 1 ! END DO ! i3 END DO ! i2 END DO ! i1 ! DO i=1,nrm-1 IF (idum(i).EQ.1) THEN DO j=i+1,nrm IF (idum(j).EQ.1) THEN lbho=.TRUE. DO k=1,3 delta = qbd(k,i)-qbd(k,j) lbho = lbho.AND. (ABS(NINT(delta)-delta).LT.eps) END DO ! k IF (lbho) THEN IF(qnorm(i).GT.qnorm(j)) THEN qbd(1,i) = qbd(1,j) qbd(2,i) = qbd(2,j) qbd(3,i) = qbd(3,j) qnorm(i) = qnorm(j) END IF idum(j) = 0 END IF END IF END DO ! j END IF END DO ! i ! iq = 0 DO i=1,nrm IF (idum(i).EQ.1) THEN iq=iq+1 qbid(1,iq)= bg_blk(1,1)*qbd(1,i) + & bg_blk(1,2)*qbd(2,i) + & bg_blk(1,3)*qbd(3,i) qbid(2,iq)= bg_blk(2,1)*qbd(1,i) + & bg_blk(2,2)*qbd(2,i) + & bg_blk(2,3)*qbd(3,i) qbid(3,iq)= bg_blk(3,1)*qbd(1,i) + & bg_blk(3,2)*qbd(2,i) + & bg_blk(3,3)*qbd(3,i) END IF END DO ! i ! IF (iq.NE.nsc) CALL errore('q_gen',' probably nr1,nr2,nr3 too small ', iq) RETURN END SUBROUTINE q_gen ! !----------------------------------------------------------------------- SUBROUTINE check_at(at,bg_blk,alat,omega) !----------------------------------------------------------------------- ! USE kinds, ONLY : DP ! IMPLICIT NONE ! REAL(DP) :: at(3,3), bg_blk(3,3), alat, omega REAL(DP) :: work(3,3) INTEGER :: i,j REAL(DP), PARAMETER :: small=1.d-6 ! work(:,:) = at(:,:) CALL cryst_to_cart(3,work,bg_blk,-1) ! DO j=1,3 DO i =1,3 IF ( ABS(work(i,j)-NINT(work(i,j))) > small) THEN WRITE (*,'(3f9.4)') work(:,:) CALL errore ('check_at','at not multiple of at_blk',1) END IF END DO END DO ! omega =alat**3 * ABS(at(1,1)*(at(2,2)*at(3,3)-at(3,2)*at(2,3))- & at(1,2)*(at(2,1)*at(3,3)-at(2,3)*at(3,1))+ & at(1,3)*(at(2,1)*at(3,2)-at(2,2)*at(3,1))) ! RETURN END SUBROUTINE check_at ! !----------------------------------------------------------------------- SUBROUTINE set_tau (nat, nat_blk, at, at_blk, tau, tau_blk, & ityp, ityp_blk, itau_blk) !----------------------------------------------------------------------- ! USE kinds, ONLY : DP ! IMPLICIT NONE INTEGER nat, nat_blk,ityp(nat),ityp_blk(nat_blk), itau_blk(nat) REAL(DP) at(3,3),at_blk(3,3),tau(3,nat),tau_blk(3,nat_blk) ! REAL(DP) bg(3,3), r(3) ! work vectors INTEGER i,i1,i2,i3,na,na_blk REAL(DP) small INTEGER NN1,NN2,NN3 PARAMETER (NN1=8, NN2=8, NN3=8, small=1.d-8) ! CALL recips (at(1,1),at(1,2),at(1,3),bg(1,1),bg(1,2),bg(1,3)) ! na = 0 ! DO i1 = -NN1,NN1 DO i2 = -NN2,NN2 DO i3 = -NN3,NN3 r(1) = i1*at_blk(1,1) + i2*at_blk(1,2) + i3*at_blk(1,3) r(2) = i1*at_blk(2,1) + i2*at_blk(2,2) + i3*at_blk(2,3) r(3) = i1*at_blk(3,1) + i2*at_blk(3,2) + i3*at_blk(3,3) CALL cryst_to_cart(1,r,bg,-1) ! IF ( r(1).GT.-small .AND. r(1).LT.1.d0-small .AND. & r(2).GT.-small .AND. r(2).LT.1.d0-small .AND. & r(3).GT.-small .AND. r(3).LT.1.d0-small ) THEN CALL cryst_to_cart(1,r,at,+1) ! DO na_blk=1, nat_blk na = na + 1 IF (na.GT.nat) CALL errore('set_tau','too many atoms',na) tau(1,na) = tau_blk(1,na_blk) + r(1) tau(2,na) = tau_blk(2,na_blk) + r(2) tau(3,na) = tau_blk(3,na_blk) + r(3) ityp(na) = ityp_blk(na_blk) itau_blk(na) = na_blk END DO ! END IF ! END DO END DO END DO ! IF (na.NE.nat) CALL errore('set_tau','too few atoms: increase NNs',na) ! RETURN END SUBROUTINE set_tau ! !----------------------------------------------------------------------- SUBROUTINE read_tau & (nat, nat_blk, ntyp, bg_blk, tau, tau_blk, ityp, itau_blk) !--------------------------------------------------------------------- ! USE kinds, ONLY : DP ! IMPLICIT NONE ! INTEGER nat, nat_blk, ntyp, ityp(nat),itau_blk(nat) REAL(DP) bg_blk(3,3),tau(3,nat),tau_blk(3,nat_blk) ! REAL(DP) r(3) ! work vectors INTEGER i,na,na_blk ! REAL(DP) small PARAMETER ( small = 1.d-6 ) ! DO na=1,nat READ(*,*) (tau(i,na),i=1,3), ityp(na) IF (ityp(na).LE.0 .OR. ityp(na) .GT. ntyp) & CALL errore('read_tau',' wrong atomic type', na) DO na_blk=1,nat_blk r(1) = tau(1,na) - tau_blk(1,na_blk) r(2) = tau(2,na) - tau_blk(2,na_blk) r(3) = tau(3,na) - tau_blk(3,na_blk) CALL cryst_to_cart(1,r,bg_blk,-1) IF (ABS( r(1)-NINT(r(1)) ) .LT. small .AND. & ABS( r(2)-NINT(r(2)) ) .LT. small .AND. & ABS( r(3)-NINT(r(3)) ) .LT. small ) THEN itau_blk(na) = na_blk go to 999 END IF END DO CALL errore ('read_tau',' wrong atomic position ', na) 999 CONTINUE END DO ! RETURN END SUBROUTINE read_tau ! !----------------------------------------------------------------------- SUBROUTINE write_tau(fltau,nat,tau,ityp) !----------------------------------------------------------------------- ! USE kinds, ONLY : DP ! IMPLICIT NONE ! INTEGER nat, ityp(nat) REAL(DP) tau(3,nat) CHARACTER(LEN=*) fltau ! INTEGER i,na ! OPEN (unit=4,file=fltau, status='new') DO na=1,nat WRITE(4,'(3(f12.6),i3)') (tau(i,na),i=1,3), ityp(na) END DO CLOSE (4) ! RETURN END SUBROUTINE write_tau ! !----------------------------------------------------------------------- SUBROUTINE gen_qpoints (ibrav, at, bg, nat, tau, ityp, nk1, nk2, nk3, & symm_type, ntetra, nqx, nq, q, tetra) !----------------------------------------------------------------------- ! USE kinds, ONLY : DP ! IMPLICIT NONE ! input INTEGER :: ibrav, nat, nk1, nk2, nk3, ntetra, ityp(*) REAL(DP) :: at(3,3), bg(3,3), tau(3,nat) character(LEN=9) :: symm_type ! output INTEGER :: nqx, nq, tetra(4,ntetra) REAL(DP) :: q(3,nqx) ! local INTEGER :: nrot, nsym, s(3,3,48), ftau(3,48), irt(48,nat) INTEGER :: t_rev = 0 LOGICAL :: minus_q, invsym REAL(DP) :: xqq(3), wk(nqx), mdum(3,nat) CHARACTER(LEN=45) :: sname(48) ! xqq (:) =0.d0 IF (ibrav == 4 .OR. ibrav == 5) THEN ! ! hexagonal or trigonal bravais lattice ! CALL hexsym (at, s, sname, nrot) ELSEIF (ibrav >= 1 .AND. ibrav <= 14) THEN ! ! cubic bravais lattice ! CALL cubicsym (at, s, sname, nrot) ELSEIF (ibrav == 0) THEN if (symm_type=='cubic') then CALL cubicsym (at, s, sname, nrot) elseif (symm_type=='hexagonal') then CALL hexsym (at, s, sname, nrot) else CALL infomsg ('gen_qpoints', 'symm_type missing: assuming cubic symmetry', -1) CALL cubicsym (at, s, sname, nrot) end if ELSE CALL errore ('gen_qpoints', 'wrong ibrav', 1) ENDIF ! CALL kpoint_grid ( nrot, s, bg, nqx, 0,0,0, nk1,nk2,nk3, nq, q, wk) ! CALL sgama (nrot, nat, s, sname, t_rev, at, bg, tau, ityp, nsym, 6, & 6, 6, irt, ftau, nqx, nq, q, wk, invsym, minus_q, xqq, & 0, 0, .FALSE., mdum) IF (ntetra /= 6 * nk1 * nk2 * nk3) & CALL errore ('gen_qpoints','inconsistent ntetra',1) CALL tetrahedra (nsym, s, minus_q, at, bg, nqx, 0, 0, 0, & nk1, nk2, nk3, nq, q, wk, ntetra, tetra) ! RETURN END SUBROUTINE gen_qpoints ! !--------------------------------------------------------------------- SUBROUTINE a2Fdos & (nat, nq, nr1, nr2, nr3, ibrav, at, bg, tau, alat, & nsc, nat_blk, at_blk, bg_blk, itau_blk, omega_blk, rws, nrws, & dos, Emin, DeltaE, ndos, ntetra, tetra, asr, q, freq ) !----------------------------------------------------------------------- ! USE kinds, ONLY : DP USE ifconstants ! IMPLICIT NONE ! INTEGER, INTENT(in) :: nat, nq, nr1, nr2, nr3, ibrav, ndos, ntetra, & tetra(4, ntetra) LOGICAL, INTENT(in) :: dos CHARACTER(LEN=*), INTENT(IN) :: asr REAL(DP), INTENT(in) :: freq(3*nat,nq), q(3,nq), at(3,3), bg(3,3), & tau(3,nat), alat, Emin, DeltaE ! INTEGER, INTENT(in) :: nsc, nat_blk, itau_blk, nrws REAL(DP), INTENT(in) :: rws(0:3,nrws), at_blk(3,3), bg_blk(3,3), omega_blk ! REAL(DP), ALLOCATABLE :: gamma(:,:), frcg(:,:,:,:,:,:,:) COMPLEX(DP), ALLOCATABLE :: gam(:,:,:,:), gam_blk(:,:,:,:), z(:,:) real(DP) :: lambda, dos_a2F(50), temp, dos_ee(10), dos_tot, & deg(10), fermi(10), E real(DP), parameter :: rydTHz = 3289.828d0, eps_w2 = 0.0000001d0 integer :: isig, ifn, n, m, na, nb, nc, nu, nmodes, & i,j,k, ngauss, jsig, p1, p2, p3, filea2F character(len=14) :: name ! ! nmodes = 3*nat do isig=1,10 filea2F = 60 + isig write(name,"(A10,I2)") 'a2Fmatdyn.',filea2F open(filea2F, file=name, STATUS = 'unknown') READ(filea2F,*) deg(isig), fermi(isig), dos_ee(isig) enddo ! IF(dos) then open(400,file='lambda',status='unknown') write(400,*) write(400,*) ' Electron-phonon coupling constant, lambda ' write(400,*) ELSE open (20,file='gam.lines' ,status='unknown') write(20,*) write(20,*) ' Gamma lines for all modes [THz] ' write(20,*) write(6,*) write(6,*) ' Gamma lines for all modes [Rydberg] ' write(6,*) ENDIF ! ALLOCATE ( frcg(nr1,nr2,nr3,3,3,nat,nat) ) ALLOCATE ( gamma(3*nat,nq), gam(3,3,nat,nat), gam_blk(3,3,nat_blk,nat_blk) ) ALLOCATE ( z(3*nat,3*nat) ) ! frcg(:,:,:,:,:,:,:) = 0.d0 DO isig = 1, 10 filea2F = 60 + isig CALL readfg ( filea2F, nr1, nr2, nr3, nat, frcg ) ! if ( asr /= 'no') then CALL set_asr (asr, nr1, nr2, nr3, frcg, zeu, nat_blk, ibrav, tau_blk) endif ! open(300,file='dyna2F',status='old') ! do n = 1 ,nq gam(:,:,:,:) = (0.d0, 0.d0) read(300,*) do na=1,nmodes read(300,*) (z(na,m),m=1,nmodes) end do ! na ! CALL setgam (q(1,n), gam, nat, at, bg, tau, itau_blk, nsc, alat, & gam_blk, nat_blk, at_blk,bg_blk,tau_blk, omega_blk, & frcg, nr1,nr2,nr3, rws, nrws) ! ! here multiply dyn*gam*dyn for gamma and divide by w2 for lambda at given q ! do nc = 1, nat do k =1, 3 p1 = (nc-1)*3+k nu = p1 gamma(nu,n) = 0.0d0 do i=1,3 do na=1,nat p2 = (na-1)*3+i do j=1,3 do nb=1,nat p3 = (nb-1)*3+j gamma(nu,n) = gamma(nu,n) + DBLE(conjg(z(p2,p1)) * & gam(i,j,na,nb) * z(p3,p1)) enddo ! nb enddo ! j enddo ! na enddo !i gamma(nu,n) = gamma(nu,n) * 3.1415926d0 / 2.0d0 enddo ! k enddo !nc ! ! EndDo !nq all points in BZ close(300) ! file with dyn vectors ! ! after we know gamma(q) and lambda(q) calculate DOS(omega) for spectrum a2F ! if(dos) then ! if(isig.le.9) then write(name,'(A8,I1.1)') 'a2F.dos.',isig else write(name,'(A8,I2.2)') 'a2F.dos.',isig endif ifn = 200 + isig open (ifn,file=name,status='unknown',form='formatted') write(ifn,*) write(ifn,*) ' Eliashberg function a2F (per both spin)' write(ifn,*) ' frequencies in Rydberg ' write(ifn,*) ' DOS normalized to E in Rydberg: a2F_total, a2F(mode) ' write(ifn,*) ! ! correction for small frequencies ! do n = 1, nq do i = 1, nmodes if (freq(i,n).LE.eps_w2) then gamma(i,n) = 0.0d0 endif enddo enddo ! lambda = 0.0d0 do n= 1, ndos ! E = Emin + (n-1)*DeltaE + 0.5d0*DeltaE dos_tot = 0.0d0 do j=1,nmodes ! dos_a2F(j) = 0.0d0 CALL dos_gam(nmodes, nq, j, ntetra, tetra, & gamma, freq, E, dos_a2F(j)) dos_a2F(j) = dos_a2F(j) / dos_ee(isig) dos_tot = dos_tot + dos_a2F(j) ! enddo lambda = lambda + dos_tot/E * DeltaE / 3.1415926d0 write (ifn, 1050) E, dos_tot, (dos_a2F(j),j=1,nmodes) enddo !ndos write(ifn,*) " lambda =",lambda,' Delta = ',DeltaE close (ifn) write(400,'(" Broadening ",F8.4," lambda ",F12.4," dos_el ",F8.4)') & deg(isig),lambda, dos_ee(isig) ! endif !dos ! ! OUTPUT ! if(.not.dos) then write(20,'(" Broadening ",F8.4)') deg(isig) write(6,'(" Broadening ",F8.4)') deg(isig) do n=1, nq write(20,1030) n,(gamma(i,n)*rydTHz,i=1,3*nat) write(6,1040) n, (gamma(i,n),i=1,3*nat) end do endif ! ENDDO !isig ! DEALLOCATE (z, frcg, gamma, gam, gam_blk ) ! close(400) !lambda close(20) ! 1030 FORMAT( 3x,I5,' ',9F8.4 ) 1040 FORMAT( 3x,I5,' ',6F12.9 ) 1050 FORMAT( 3x,F12.6,6F16.8 ) ! RETURN END SUBROUTINE a2Fdos ! !----------------------------------------------------------------------- subroutine setgam (q, gam, nat, at,bg,tau,itau_blk,nsc,alat, & & gam_blk, nat_blk, at_blk,bg_blk,tau_blk,omega_blk, & & frcg, nr1,nr2,nr3, rws,nrws) !----------------------------------------------------------------------- ! compute the dynamical matrix (the analytic part only) ! USE kinds, ONLY : DP implicit none real(DP), parameter :: tpi=2.d0*3.14159265358979d0 ! ! I/O variables ! integer :: nr1, nr2, nr3, nat, nat_blk, & nsc, nrws, itau_blk(nat) real(DP) :: q(3), tau(3,nat), at(3,3), bg(3,3), alat, rws(0:3,nrws) real(DP) :: tau_blk(3,nat_blk), at_blk(3,3), bg_blk(3,3), omega_blk, & frcg(nr1,nr2,nr3,3,3,nat_blk,nat_blk) COMPLEX(DP) :: gam_blk(3,3,nat_blk,nat_blk) COMPLEX(DP) :: gam(3,3,nat,nat) ! ! local variables ! real(DP) :: arg complex(DP) :: cfac(nat) integer :: i,j,k, na,nb, na_blk, nb_blk, iq real(DP) :: qp(3), qbid(3,nsc) ! automatic array ! ! call q_gen(nsc,qbid,at_blk,bg_blk,at,bg) ! do iq=1,nsc ! do k=1,3 qp(k)= q(k) + qbid(k,iq) end do ! gam_blk(:,:,:,:) = (0.d0,0.d0) CALL frc_blk (gam_blk,qp,tau_blk,nat_blk, & nr1,nr2,nr3,frcg,at_blk,bg_blk,rws,nrws) ! do na=1,nat na_blk = itau_blk(na) do nb=1,nat nb_blk = itau_blk(nb) ! arg = tpi * ( qp(1) * ( (tau(1,na)-tau_blk(1,na_blk)) - & (tau(1,nb)-tau_blk(1,nb_blk)) ) + & qp(2) * ( (tau(2,na)-tau_blk(2,na_blk)) - & (tau(2,nb)-tau_blk(2,nb_blk)) ) + & qp(3) * ( (tau(3,na)-tau_blk(3,na_blk)) - & (tau(3,nb)-tau_blk(3,nb_blk)) ) ) ! cfac(nb) = cmplx(cos(arg),sin(arg))/nsc ! end do ! nb do nb=1,nat do i=1,3 do j=1,3 nb_blk = itau_blk(nb) gam(i,j,na,nb) = gam(i,j,na,nb) + cfac(nb) * & gam_blk(i,j,na_blk,nb_blk) end do ! j end do ! i end do ! nb end do ! na ! end do ! iq ! return end subroutine setgam ! !-------------------------------------------------------------------- subroutine dos_gam (nbndx, nq, jbnd, ntetra, & tetra, gamma, et, ef, dos_a2F) !-------------------------------------------------------------------- ! calculates weights with the tetrahedron method (Bloechl version) ! this subroutine is based on tweights.f90 belonging to PW ! it calculates a2F on the surface of given frequency <=> histogram ! Band index means the frequency mode here ! and "et" means the frequency(mode,q-point) ! USE kinds, ONLY: DP use parameters ! USE ifconstants, ONLY : gamma implicit none ! integer :: nq, nbndx, ntetra, tetra(4,ntetra), jbnd real(DP) :: et(nbndx,nq), gamma(nbndx,nq), func real(DP) :: ef, dos_a2F real(DP) :: e1, e2, e3, e4, c1, c2, c3, c4, etetra(4) integer :: ik, ibnd, nt, nk, ns, i, ik1, ik2, ik3, ik4, itetra(4) real(DP) :: f12,f13,f14,f23,f24,f34, f21,f31,f41,f42,f32,f43 real(DP) :: P1,P2,P3,P4, G, Z, o13, Y1,Y2,Y3,Y4, WW, eps,vol, Tint Tint = 0.0d0 eps = 1.0d-14 vol = 1.0d0/ntetra do nt = 1, ntetra ibnd = jbnd ! ! etetra are the energies at the vertexes of the nt-th tetrahedron ! do i = 1, 4 etetra(i) = et(ibnd, tetra(i,nt)) enddo itetra(1) = 0 call hpsort (4,etetra,itetra) ! ! ...sort in ascending order: e1 < e2 < e3 < e4 ! e1 = etetra (1) e2 = etetra (2) e3 = etetra (3) e4 = etetra (4) ! ! kp1-kp4 are the irreducible k-points corresponding to e1-e4 ! ik1 = tetra(itetra(1),nt) ik2 = tetra(itetra(2),nt) ik3 = tetra(itetra(3),nt) ik4 = tetra(itetra(4),nt) Y1 = gamma(ibnd,ik1)/et(ibnd,ik1) Y2 = gamma(ibnd,ik2)/et(ibnd,ik2) Y3 = gamma(ibnd,ik3)/et(ibnd,ik3) Y4 = gamma(ibnd,ik4)/et(ibnd,ik4) f12 = (ef-e2)/(e1-e2) f13 = (ef-e3)/(e1-e3) f14 = (ef-e4)/(e1-e4) f23 = (ef-e3)/(e2-e3) f24 = (ef-e4)/(e2-e4) f34 = (ef-e4)/(e3-e4) f21 = 1.0d0 - f12 f31 = 1.0d0 - f13 f41 = 1.0d0 - f14 f32 = 1.0d0 - f23 f42 = 1.0d0 - f24 f43 = 1.0d0 - f34 o13 = 1.0d0/3.0d0 G = 0.0d0 P1 = 0.0d0 P2 = 0.0d0 P3 = 0.0d0 P4 = 0.0d0 IF(e1.gt.ef.or.e4.lt.ef) then goto 500 ENDIF IF(e3.lt.ef.and.ef.lt.e4) THEN Z = (1.0d0 - f14 * f24 * f34) G = 3.0d0 * (1.0d0 - Z) / (e4-ef) P1 = f14 * o13 P2 = f24 * o13 P3 = f34 * o13 P4 = (f41 + f42 + f43) * o13 goto 200 ENDIF IF(e2.lt.ef.and.ef.lt.e3) THEN G = 3.0d0 * (f23*f31 + f32*f24) P1 = f14 * o13 + f13*f31*f23 / G P2 = f23 * o13 + f24*f24*f32 / G P3 = f32 * o13 + f31*f31*f23 / G P4 = f41 * o13 + f42*f24*f32 / G G = G / (e4-e1) goto 200 ENDIF IF(e1.lt.ef.and.ef.lt.e2) THEN Z = f21 * f31 * f41 G = 3.0d0 * Z / (ef-e1) P1 = o13 * (f12 + f12 + f14) P2 = o13 * f21 P3 = o13 * f31 P4 = o13 * f41 goto 200 ENDIF 200 WW = G * (Y1*P1 + Y2*P2 + Y3*P3 + Y4*P4) Tint = Tint + WW * vol 500 continue enddo ! ntetra dos_a2F = Tint !2 because DOS_ee is per 1 spin return end subroutine dos_gam ! ! !----------------------------------------------------------------------- subroutine readfg ( ifn, nr1, nr2, nr3, nat, frcg ) !----------------------------------------------------------------------- ! USE kinds, ONLY : DP implicit none ! I/O variable integer, intent(in) :: nr1,nr2,nr3, nat real(DP), intent(out) :: frcg(nr1,nr2,nr3,3,3,nat,nat) ! local variables integer i, j, na, nb, m1,m2,m3, ifn integer ibid, jbid, nabid, nbbid, m1bid,m2bid,m3bid ! ! READ (ifn,*) m1, m2, m3 if ( m1 /= nr1 .or. m2 /= nr2 .or. m3 /= nr3) & call errore('readfG','inconsistent nr1, nr2, nr3 read',1) do i=1,3 do j=1,3 do na=1,nat do nb=1,nat read (ifn,*) ibid, jbid, nabid, nbbid if(i.ne.ibid.or.j.ne.jbid.or.na.ne.nabid.or.nb.ne.nbbid) then write(*,*) i,j,na,nb,' <> ', ibid, jbid, nabid, nbbid call errore ('readfG','error in reading',1) else read (ifn,*) (((m1bid, m2bid, m3bid, & frcg(m1,m2,m3,i,j,na,nb), & m1=1,nr1),m2=1,nr2),m3=1,nr3) endif end do end do end do end do ! close(ifn) ! return end subroutine readfg From giannozz at nest.sns.it Mon Sep 3 15:30:19 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 3 Sep 2007 15:30:19 +0200 Subject: [Pw_forum] IFC's In-Reply-To: References: Message-ID: <46292ADB-7DAF-4CF1-9A3E-2B50926CD249@nest.sns.it> On Sep 3, 2007, at 10:43 , Katalin Gaal-Nagy wrote: > the message I send just half an hour ago is held because of the > size ... > However, I resend it without the attachments. If anyone is > interested in > the files with the routines, please tell me! thank you. I just released the previous message (the one with the attachment), it was still there (messages "held for moderator approval" are automaticaly deleted after a while) Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From brsahu at physics.utexas.edu Mon Sep 3 18:22:15 2007 From: brsahu at physics.utexas.edu (Bhagawan Sahu) Date: Mon, 3 Sep 2007 11:22:15 -0500 (CDT) Subject: [Pw_forum] inconsistent number of sticks In-Reply-To: <8F2EB455-03D5-4C00-A2BE-FC31CEC1ACF5@nest.sns.it> Message-ID: Hi Paolo, Here is the input (as the number of atoms are about 400, I trimmed it at the atomic position list in order not to make the mail long) &control calculation='scf' restart_mode='from_scratch', prefix='pwscf' pseudo_dir = '.', outdir='tmp/' / &system ibrav= 0, celldm(1)= 4.648728, nat= 388, ntyp= 3, ecutwfc =30.0, ecutrho =300.0 occupations='smearing', smearing='methfessel-paxton', degauss=0.02 nspin = 2, starting_magnetization(1)= 1, starting_magnetization(2)= -1, starting_magnetization(3)= 0.1, report=1 / &electrons conv_thr = 1.0d-7 mixing_beta = 0.1 / CELL_PARAMETERS {cubic} 1 0 0 0 5.42683 0 0 0 87.2035 ATOMIC_SPECIES C1 12.0107 Cpw91.vdb C2 12.0107 Cpw91.vdb H 1.00794 Hpw91.vdb ATOMIC_POSITIONS {crystal} C1 0.5 0.124532 0.00505313 C2 0 0.124532 0.00836349 C1 0 0.124532 0.0149842 C2 0.5 0.124532 0.0182946 ........... and the output is Program PWSCF v.3.2 starts ... 0: Today is 2Sep2007 at 9: 0: 5 0: 0: Parallel version (MPI) 0: 0: Number of processors in use: 200 0: K-points division: npool = 20 0: R & G space division: proc/pool = 10 0: 0: Ultrasoft (Vanderbilt) Pseudopotentials 0: 0: Current dimensions of program pwscf are: 0: 0: ntypx = 10 npk = 40000 lmax = 3 0: nchix = 6 ndmx = 2000 nbrx = 14 nqfx = 8 0: 0: ============================================================ 0: | pseudopotential report for atomic species: 1 | 0: | pseudo potential version 7 3 6 | 0: ------------------------------------------------------------ 0: | C (US d-loc) PW91 exchange-corr | 0: | z = 6. zv( 1) = 4. exfact = 4.00000 | 0: | ifpcor = 0 atomic energy = -10.74612 Ry | 0: | index orbital occupation energy | 0: | 1 200 2.00 -1.01 | 0: | 2 210 2.00 -0.39 | 0: | rinner = 0.6000 0.6000 0.6000 | 0: | new generation scheme: | 0: | nbeta = 4 kkbeta = 569 rcloc = 1.5400 | 0: | ibeta l epsilon rcut | 0: | 1 0 -1.01 1.50 | 0: | 2 0 -0.39 1.50 | 0: | 3 1 -1.01 1.54 | 0: | 4 1 -0.39 1.54 | 0: ============================================================ 0: ============================================================ 0: | pseudopotential report for atomic species: 2 | 0: | pseudo potential version 7 3 6 | 0: ------------------------------------------------------------ 0: | C (US d-loc) PW91 exchange-corr | 0: | z = 6. zv( 2) = 4. exfact = 4.00000 | 0: | ifpcor = 0 atomic energy = -10.74612 Ry | 0: | index orbital occupation energy | 0: | 1 200 2.00 -1.01 | 0: | 2 210 2.00 -0.39 | 0: | rinner = 0.6000 0.6000 0.6000 | 0: | new generation scheme: | 0: | nbeta = 4 kkbeta = 569 rcloc = 1.5400 | 0: | ibeta l epsilon rcut | 0: | 1 0 -1.01 1.50 | 0: | 2 0 -0.39 1.50 | 0: | 3 1 -1.01 1.54 | 0: | 4 1 -0.39 1.54 | 0: ============================================================ 0: 0: ============================================================ 0: | pseudopotential report for atomic species: 3 | 0: | pseudo potential version 7 3 6 | 0: ------------------------------------------------------------ 0: | hydrogen PW91 exchange-corr | 0: | z = 1. zv( 3) = 1. exfact = 4.00000 | 0: | ifpcor = 0 atomic energy = -0.92034 Ry | 0: | index orbital occupation energy | 0: | 1 100 1.00 -0.48 | 0: | rinner = 0.7000 | 0: | new generation scheme: | 0: | nbeta = 2 kkbeta = 395 rcloc = 0.9000 | 0: | ibeta l epsilon rcut | 0: | 1 0 -0.48 0.90 | 0: | 2 0 0.90 0.90 | 0: ============================================================ 0: * 1 * 11 /= 35 0: * 2 * 11 /= 35 0: * 3 * 11 /= 35 0: * 4 * 11 /= 35 0: * 5 * 11 /= 35 0: * 6 * 11 /= 35 0: * 7 * 11 /= 35 0: * 8 * 11 /= 35 0: * 9 * 11 /= 35 0: * 10 * 11 /= 34 0: 0: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 0: from fft_dlay_set : error # 7 0: inconsistent number of sticks 0: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% The job is run on the IBM-SP5 machine. Bhagawan Sahu Universityy of Texas at Austin Austin TX 78758 USA On Mon, 3 Sep 2007, Paolo Giannozzi wrote: > > On Sep 3, 2007, at 6:31 , brsahu at physics.utexas.edu wrote: > > > I can provide the input if asked. > > please provide the input AND the output > > Paolo > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From shaposh at isp.nsc.ru Tue Sep 4 07:07:41 2007 From: shaposh at isp.nsc.ru (Alexander Shaposhnikov) Date: Tue, 4 Sep 2007 12:07:41 +0700 Subject: [Pw_forum] About "crystal" coordinates In-Reply-To: <1188553023.2576.9.camel@localhost.localdomain> References: <200708311520.33959.shaposh@isp.nsc.ru> <1188553023.2576.9.camel@localhost.localdomain> Message-ID: <200709041207.42079.shaposh@isp.nsc.ru> Hello Tone, sorry for delay -i was out from my institution for several days. The patch you sent is working, the structure is displayed correctly now. Here is the sample input with structure. /////////////////////////////////////////////////// &CONTROL title = Ta2O5 , calculation = 'scf' , restart_mode = 'from_scratch' , outdir = '/home/alex/scr/mono_bc' , pseudo_dir = '/home/alex/pseudo/' , prefix = Ta2O5mbc , disk_io = 'default' , verbosity = 'high' , / &SYSTEM ibrav = 13, celldm(1) = 24.16, celldm(2) = 0.43, celldm(3) = 0.38, celldm(4) = -0.246, nat = 14, ntyp = 2, ecutwfc = 55 , / &ELECTRONS diagonalization = 'david' , / ATOMIC_SPECIES Ta 1.00000 Ta.pbe-nsp-van.UPF O 1.00000 O.pbe-rrkjus.UPF ATOMIC_POSITIONS crystal Ta 0.892500000 0.262300000 0.386500000 Ta 0.107500000 0.737700000 0.613500000 Ta 0.613500000 0.237700000 0.107500000 Ta 0.386500000 0.762300000 0.892500000 O 0.898000000 0.250000000 0.102000000 O 0.102000000 0.750000000 0.898000000 O 0.338000000 0.990000000 0.440000000 O 0.662000000 0.010000000 0.560000000 O 0.560000000 0.510000000 0.662000000 O 0.440000000 0.490000000 0.338000000 O 0.868000000 0.355000000 0.730000000 O 0.132000000 0.645000000 0.270000000 O 0.270000000 0.145000000 0.132000000 O 0.730000000 0.855000000 0.868000000 K_POINTS gamma //////////////////////////////////////////////////// Thank you for the help and your program. Regards, Alex On Friday 31 August 2007 16:37, Tone Kokalj wrote: > On Fri, 2007-08-31 at 15:20 +0700, Alexander Shaposhnikov wrote: > > Compiled it (with ifort) and swaped, but the things become worse :) > > The distances are now on the order of magnitude more than should be. > > I was too fast with the fix. Can you retry with this. Tone From scandolo at ictp.it Tue Sep 4 13:51:45 2007 From: scandolo at ictp.it (Sandro Scandolo) Date: Tue, 04 Sep 2007 13:51:45 +0200 Subject: [Pw_forum] About "crystal" coordinates In-Reply-To: <200709041207.42079.shaposh@isp.nsc.ru> References: <200708311520.33959.shaposh@isp.nsc.ru> <1188553023.2576.9.camel@localhost.localdomain> <200709041207.42079.shaposh@isp.nsc.ru> Message-ID: <46DD46D1.5000405@ictp.it> For the sake of clarity (and for the benefit of our colleagues crystallographers), let me rephrase the original reply to this thread: When using PWSCF with non-primitive lattices, "crystal" coordinates refer to the primitive cell, not to the conventional cell (the latter being the common practice in crystallography). Sandro Alexander Shaposhnikov wrote: > Hello Tone, > > sorry for delay -i was out from my institution for several days. > The patch you sent is working, the structure is displayed correctly now. Here > is the sample input with structure. > > /////////////////////////////////////////////////// > &CONTROL > title = Ta2O5 , > calculation = 'scf' , > restart_mode = 'from_scratch' , > outdir = '/home/alex/scr/mono_bc' , > pseudo_dir = '/home/alex/pseudo/' , > prefix = Ta2O5mbc , > disk_io = 'default' , > verbosity = 'high' , > / > &SYSTEM > ibrav = 13, > celldm(1) = 24.16, > celldm(2) = 0.43, > celldm(3) = 0.38, > celldm(4) = -0.246, > nat = 14, > ntyp = 2, > ecutwfc = 55 , > / > &ELECTRONS > diagonalization = 'david' , > / > ATOMIC_SPECIES > Ta 1.00000 Ta.pbe-nsp-van.UPF > O 1.00000 O.pbe-rrkjus.UPF > ATOMIC_POSITIONS crystal > Ta 0.892500000 0.262300000 0.386500000 > Ta 0.107500000 0.737700000 0.613500000 > Ta 0.613500000 0.237700000 0.107500000 > Ta 0.386500000 0.762300000 0.892500000 > O 0.898000000 0.250000000 0.102000000 > O 0.102000000 0.750000000 0.898000000 > O 0.338000000 0.990000000 0.440000000 > O 0.662000000 0.010000000 0.560000000 > O 0.560000000 0.510000000 0.662000000 > O 0.440000000 0.490000000 0.338000000 > O 0.868000000 0.355000000 0.730000000 > O 0.132000000 0.645000000 0.270000000 > O 0.270000000 0.145000000 0.132000000 > O 0.730000000 0.855000000 0.868000000 > K_POINTS gamma > > //////////////////////////////////////////////////// > > Thank you for the help and your program. > > Regards, > Alex > > On Friday 31 August 2007 16:37, Tone Kokalj wrote: > >> On Fri, 2007-08-31 at 15:20 +0700, Alexander Shaposhnikov wrote: >> >>> Compiled it (with ifort) and swaped, but the things become worse :) >>> The distances are now on the order of magnitude more than should be. >>> >> I was too fast with the fix. Can you retry with this. Tone >> > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > From giannozz at nest.sns.it Tue Sep 4 15:22:20 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 4 Sep 2007 15:22:20 +0200 Subject: [Pw_forum] inconsistent number of sticks In-Reply-To: References: Message-ID: <74A8F861-E60A-4C11-B2FE-E5DC4CF60155@nest.sns.it> On Sep 3, 2007, at 18:22 , Bhagawan Sahu wrote: > Here is the input an interesting bug. Carlo Cavazzoni found it: an integer overflow, due to the very long z axis and consequently to a large value of the 3rd FFT grid dimension (nr3). The fix is simple. If you are using the CVS version, update; otherwise, replace Modules/stick_base.f90 with the attached one Paolo -------------- next part -------------- A non-text attachment was scrubbed... Name: stick_base.f90 Type: application/octet-stream Size: 27427 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20070904/ae4be2e4/attachment-0001.obj -------------- next part -------------- --- Paolo Giannozzi, Democritos and University of Udine, Italy From eyvaz_isaev at yahoo.com Tue Sep 4 22:03:22 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 4 Sep 2007 13:03:22 -0700 (PDT) Subject: [Pw_forum] Potassium PBE pseudopotential In-Reply-To: <2fd252650709030527v185cd8acka9f269d381451d6d@mail.gmail.com> Message-ID: <734197.75828.qm@web60314.mail.yahoo.com> Dear Dmitry, --- Dmitry Korotin wrote: > Dear Eyvaz, > the subject of our studies are lattice distortions > in KCuF3. > OK, I see. I will send K Psp personally to you, and if it works then I can submit it to the website. Bests, Eyvaz. ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Got a little couch potato? Check out fun summer activities for kids. http://search.yahoo.com/search?fr=oni_on_mail&p=summer+activities+for+kids&cs=bz From amosleffler at sbcglobal.net Wed Sep 5 02:46:00 2007 From: amosleffler at sbcglobal.net (Amos Leffler) Date: Tue, 04 Sep 2007 17:46:00 -0700 Subject: [Pw_forum] Compiler Problems Message-ID: <46DDFC48.1020700@sbcglobal.net> Dear Forum, I have been trying to compile espresso-3.2 and with the espresso-3.2.1.diff patch and the results are similar. As downloaded it stops in Modules/input_parameters.f90 with a compiler error. I traced this to line 873 which reads: n_blockocc(2) = (1,1). If it is changed to: n_blockocc = (1) compilation continues until the CPV directory. (I dont know whether this is the correct change but it works.) In the CPV/init_run.f90 file line 257 the first entry in the readfile is a "1". This gives an error which says it must match the dummy INTENT = OUT/INOUT. In line 75 of the file readfile refers back to cp_interface. There does not seem to be any OUT/INOUT entries in this file so I do not know what the first entry in readfile needs. Hopefully someone has encountered these problems and has some answers. Any help would be much appreciated. Amos Leffler From Zahraa.Ibrahim at uoit.ca Wed Sep 5 04:22:13 2007 From: Zahraa.Ibrahim at uoit.ca (Zahraa Ibrahim) Date: Tue, 4 Sep 2007 22:22:13 -0400 Subject: [Pw_forum] Temperature extraction from PW References: Message-ID: Hello Paolo The old pw version was 2.0.3. The new version I am using is 3.2.2. Temperature seems to be defined differently in both versions. Zahraa UOIT -----Original Message----- From: pw_forum-bounces at pwscf.org on behalf of Paolo Giannozzi Sent: Sun 9/2/2007 10:14 AM To: PWSCF Forum Cc: Subject: Re: [Pw_forum] Temperature extraction from PW On Aug 30, 2007, at 19:21 , Zahraa Ibrahim wrote: > I recently changed from the democritus version of PW to the > espresso version. there isn't a "democritos" version distinct from an "espresso" version: there are version numbers (for the stable versions) and the CVS version (under development) Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From giannozz at nest.sns.it Wed Sep 5 12:43:57 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 5 Sep 2007 12:43:57 +0200 Subject: [Pw_forum] Compiler Problems In-Reply-To: <46DDFC48.1020700@sbcglobal.net> References: <46DDFC48.1020700@sbcglobal.net> Message-ID: <57000753-72C4-4FD7-9B3A-6F7CC3862538@nest.sns.it> On Sep 5, 2007, at 2:46 , Amos Leffler wrote: > I have been trying to compile espresso-3.2 and with the > espresso-3.2.1.diff > patch and the results are similar. compiler ? > As downloaded it stops in Modules/input_parameters.f90 with a > compiler error. > I traced this to line 873 which reads: > n_blockocc(2) = (1,1). I think this is a nonstandard syntax that most compilers (but not all) accept. Change it to n_blockocc(2) = 1 > In the CPV/init_run.f90 file line 257 the first entry in the > readfile is a "1". this is actually incorrect and was already fixed in the CVS version. Add a new variable i, set i=1, call readfile (i, .... Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From zucco at dipteris.unige.it Wed Sep 5 15:32:31 2007 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Wed, 5 Sep 2007 15:32:31 +0200 Subject: [Pw_forum] Stack size error on CVS version Message-ID: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> Hello to all, I downloaded the CVS version few minutes ago and after compilation of pwall suite on a MacOSX 10.4.10-intel (with ifort and gcc-4.2, and compiled not as root) I receive an error related to "stack memory size" during execution of example01 fusi:/usr/local/sh/espresso/examples/example01 zucco$ ./run_example /usr/local/sh/espresso/examples/example01 : starting This example shows how to use pw.x to calculate the total energy and the band structure of four simple systems: Si, Al, Cu, Ni. executables directory: /usr/local/softwarehouse/espresso/bin pseudo directory: /usr/local/softwarehouse/espresso/pseudo temporary directory: /Users/zucco/tmp checking that needed directories and files exist... done running pw.x as: /usr/local/softwarehouse/espresso/bin/pw.x running bands.x as: /usr/local/softwarehouse/espresso/bin/bands.x cleaning /Users/zucco/tmp... done running the scf calculation for Si...Cannot set stack size to infinity: Operation not permitted done running the band-structure calculation for Si...Cannot set stack size to infinity: Operation not permitted ... done the ulimit -a (if it is useful) prints: core file size (blocks, -c) 0 data seg size (kbytes, -d) 6144 file size (blocks, -f) unlimited max locked memory (kbytes, -l) unlimited max memory size (kbytes, -m) unlimited open files (-n) 256 pipe size (512 bytes, -p) 1 stack size (kbytes, -s) 65536 cpu time (seconds, -t) unlimited max user processes (-u) 266 virtual memory (kbytes, -v) unlimited BUT. If I run example01 as root all is ok. Maybe this is a problem related by my system but espresso-3.2 compiled few weeks ago, never incur in this problem. Many thanks to all m. ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry http://www.dipteris.unige.it/geochimica DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From giannozz at nest.sns.it Wed Sep 5 16:56:42 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 5 Sep 2007 16:56:42 +0200 Subject: [Pw_forum] Stack size error on CVS version In-Reply-To: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> References: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> Message-ID: On Sep 5, 2007, at 15:32 , Marino Vetuschi Zuccolini wrote: > I downloaded the CVS version few minutes ago and after compilation of > pwall suite on a MacOSX 10.4.10-intel (with ifort and gcc-4.2, and > compiled not as root) I receive an error related to "stack memory > size" no Sir, you received a WARNING, not an error. The code keeps working. It just warns you that it cannot remove the stack size limit (the stack is a part of the RAM. Don't ask me why the RAM is split into "stack", "heap", and whatnot, I don't know). If everything works, never mind. If it annoys you, modify the clib/stack.c routine so that it doesn't write anything, or throw out the standard error output (e.g.: pw.x 2> /dev/ null). Since your stack size seems to be quite big (65Mb) I don't expect any problem. Should the code require more stack, you are left with the following options: - cursing Intel because their compiler eats up so much stack (useless) - cursing Apple because Mac OS-X does not allow users to remove the stack size limit (also useless) - modifying the clib/stack.c routine so that it determines the maximum limit and sets the stack size accordingly (65536kb seems to be the maximum for users on Mac OS-X, so it might be useless as well) - running with root privilege (dangerous) See also here: http://www.democritos.it:8888/O-sesame/chngview?cn=4710 Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From zucco at dipteris.unige.it Wed Sep 5 17:09:52 2007 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Wed, 5 Sep 2007 17:09:52 +0200 Subject: [Pw_forum] Stack size error on CVS version In-Reply-To: References: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> Message-ID: <31539652-BB35-4514-99E5-5030507D35B8@dipteris.unige.it> Thanks for reply, On 5 Sep 2007, at 16:56, Paolo Giannozzi wrote: > On Sep 5, 2007, at 15:32 , Marino Vetuschi Zuccolini wrote: > >> I downloaded the CVS version few minutes ago and after compilation of >> pwall suite on a MacOSX 10.4.10-intel (with ifort and gcc-4.2, and >> compiled not as root) I receive an error related to "stack memory >> size" > > no Sir, you received a WARNING, not an error. The code keeps working. > It just warns you that it cannot remove the stack size limit (the > stack is a > part of the RAM. Don't ask me why the RAM is split into "stack", > "heap", > and whatnot, I don't know). If everything works, never mind. > unfortunately on my mac doesn't run a all, it crash suddenly. > If it annoys you, modify the clib/stack.c routine so that it doesn't > write > anything, or throw out the standard error output (e.g.: pw.x 2> /dev/ > null). > Since your stack size seems to be quite big (65Mb) I don't expect any > problem. > Should the code require more stack, you are left with the following > options: > - cursing Intel because their compiler eats up so much stack (useless) > - cursing Apple because Mac OS-X does not allow users to remove the > stack size limit (also useless) > - modifying the clib/stack.c routine so that it determines the maximum > limit and sets the stack size accordingly (65536kb seems to be the > maximum for users on Mac OS-X, so it might be useless as well) > - running with root privilege (dangerous) This choice at the moment is only a way to understand IF I'm able to solve with some tricks. I found an error in compilation in CPV/read_pseudo.f90 if (rgrid(is)%mesh > SIZE(rgrid%r) ) call errore ('read_pseudo','increase ndmx',rgrid(is)%mesh) maybe is if (rgrid(is)%mesh > SIZE(rgrid(is)%r) ) call errore ('read_pseudo','increase ndmx',rgrid(is)%mesh) Thanks m. > > See also here: > http://www.democritos.it:8888/O-sesame/chngview?cn=4710 > > Paolo > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry http://www.dipteris.unige.it/geochimica Visit our BAXEICO computing cluster homepage http://qed.dipteris.unige.it/ganglia -- out of service DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From giannozz at nest.sns.it Wed Sep 5 17:37:10 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 5 Sep 2007 17:37:10 +0200 Subject: [Pw_forum] Stack size error on CVS version In-Reply-To: <31539652-BB35-4514-99E5-5030507D35B8@dipteris.unige.it> References: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> <31539652-BB35-4514-99E5-5030507D35B8@dipteris.unige.it> Message-ID: On Sep 5, 2007, at 17:09 , Marino Vetuschi Zuccolini wrote: > unfortunately on my mac doesn't run a all, it crash suddenly. pw.x? all tests I did (yesterday) with Intel compiler (9.1.038), serial compilation, on Mac OS-X (10.4.10) were successful. Do tests work in your case? > I found an error in compilation in CPV/read_pseudo.f90 oops...fixed, thank you P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From akohlmey at cmm.chem.upenn.edu Wed Sep 5 17:21:06 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 5 Sep 2007 11:21:06 -0400 (EDT) Subject: [Pw_forum] Stack size error on CVS version In-Reply-To: Message-ID: On Wed, 5 Sep 2007, Paolo Giannozzi wrote: PG> Should the code require more stack, you are left with the following PG> options: PG> - cursing Intel because their compiler eats up so much stack (useless) true. it is annoying, but using the stack for temporary variables has a few advantages: speed, simplicity, and no memory leaks (you don't have to free(3) memory that you allocate with alloca(3) instead of malloc(3)). for as long as you pass large arrays as arguments or keep them in common blocks (as in old style fortran programs) you'll get very good performance with that. PG> - cursing Apple because Mac OS-X does not allow users to remove the PG> stack size limit (also useless) PG> - modifying the clib/stack.c routine so that it determines the maximum PG> limit and sets the stack size accordingly (65536kb seems to be the PG> maximum for users on Mac OS-X, so it might be useless as well) PG> - running with root privilege (dangerous) there is one more option: - if you have an intel compiler version 10.0 or above, you can use the flag -heap-arrays which will put temporary arrays (the major reason for excessive stack growth) on the heap instead of the stack. cheers, axel. PG> PG> See also here: PG> http://www.democritos.it:8888/O-sesame/chngview?cn=4710 PG> PG> Paolo PG> --- PG> Paolo Giannozzi, Democritos and University of Udine, Italy PG> PG> PG> _______________________________________________ PG> Pw_forum mailing list PG> Pw_forum at pwscf.org PG> http://www.democritos.it/mailman/listinfo/pw_forum PG> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From zucco at dipteris.unige.it Thu Sep 6 10:46:02 2007 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Thu, 6 Sep 2007 10:46:02 +0200 Subject: [Pw_forum] Stack size error on CVS version In-Reply-To: References: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> <31539652-BB35-4514-99E5-5030507D35B8@dipteris.unige.it> Message-ID: <1661A694-F802-4B19-94A1-55B65BE66310@dipteris.unige.it> Hello to all, On 5 Sep 2007, at 17:37, Paolo Giannozzi wrote: > pw.x? all tests I did (yesterday) with Intel compiler (9.1.038), > serial compilation, on Mac OS-X (10.4.10) were successful. > Do tests work in your case Well, you are right (in the future I will remember to myself to check one more time before email to the forum..) although a particular behavior happens. If I run the complete cycle of runs on top of all tests the "annoying" message doesn't appear, but if I run a single example from its directory, the "stack" message magically appear. Anyway I think this is not a problem at all, and thanks to all for replies. One more thing. Doing complete test cycle some tests using pw.x/cp.x crashes. Here in the following the messages: ex17 running the NEB calculation with cp.x for H2+H => H+H2...Error condition encountered during test: exit status = 180 Aborting ex18 running the calculation with fixed ions...Error condition encountered during test: exit status = 180 Aborting ex23 running the Wannier functions calculation...Error condition encountered during test: exit status = 180 Aborting ex27 running the pw.x SCF calculation...Error condition encountered during test: exit status = 2 Aborting ex28 running meta-dynamics calculation ( T = 0 K ) with pw.x ...Error condition encountered during test: exit status = 2 Aborting ex29 running the calculation BO-MD for Si-dimer...Error condition encountered during test: exit status = 2 Aborting ex30 running the CG calculation for bulk MgO E_field=0.001 a.u. ...Error condition encountered during test: exit status = 180 Aborting I hope this helps m. ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry http://www.dipteris.unige.it/geochimica DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From giannozz at nest.sns.it Thu Sep 6 11:26:22 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 6 Sep 2007 11:26:22 +0200 Subject: [Pw_forum] Stack size error on CVS version In-Reply-To: <1661A694-F802-4B19-94A1-55B65BE66310@dipteris.unige.it> References: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> <31539652-BB35-4514-99E5-5030507D35B8@dipteris.unige.it> <1661A694-F802-4B19-94A1-55B65BE66310@dipteris.unige.it> Message-ID: <8449FC2D-5DAD-46D8-98FA-158C0C3373FF@nest.sns.it> On Sep 6, 2007, at 10:46 , Marino Vetuschi Zuccolini wrote: > One more thing. Doing complete test cycle some tests using pw.x/cp.x > crashes. Here in the following the messages: > [...] > I hope this helps not really: this just says that an error occurred, not why. What happened is written into the output files of each example. Notice that the following examples presently do not work in the CVS version: 24, 26, 29, 32 Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From zucco at dipteris.unige.it Thu Sep 6 11:51:41 2007 From: zucco at dipteris.unige.it (Marino Vetuschi Zuccolini) Date: Thu, 6 Sep 2007 11:51:41 +0200 Subject: [Pw_forum] Tests failed in CVS version In-Reply-To: <8449FC2D-5DAD-46D8-98FA-158C0C3373FF@nest.sns.it> References: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> <31539652-BB35-4514-99E5-5030507D35B8@dipteris.unige.it> <1661A694-F802-4B19-94A1-55B65BE66310@dipteris.unige.it> <8449FC2D-5DAD-46D8-98FA-158C0C3373FF@nest.sns.it> Message-ID: <02440F36-21DC-4E49-BCCA-D6DEF98F5A82@dipteris.unige.it> Hi Paolo, On 6 Sep 2007, at 11:26, Paolo Giannozzi wrote: > not really: this just says that an error occurred, not why. > What happened is written into the output files of each example. > ex 17: crash and prints these lines ------------------------------ iteration 1 ------------------------------ tcpu = 0.0 self-consistency for image 1 ex 18, 23, 30: crashes suddenly with the first call to formf: eself=..... ex 27, 28: show the error from rdiaghg : error # 24 but if you need the outputs, please let me know > Notice that the following examples presently do not work in the > CVS version: 24, 26, 29, 32 > Yes I see that some packages are not compiled. m. ******************************************************* Marino Vetuschi Zuccolini zucco at dipteris.unige.it Researcher / Geochemist Laboratory of Geochemistry http://www.dipteris.unige.it/geochimica DIPartimento per lo studio della TErra e delle sue RISorse - Universit? di Genova Tel. ++39 010 3538136 Fax. ++39 010 352169 Corso Europa 26, 16132 - Genova - Italy From giannozz at nest.sns.it Thu Sep 6 13:54:32 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 6 Sep 2007 13:54:32 +0200 Subject: [Pw_forum] Tests failed in CVS version In-Reply-To: <02440F36-21DC-4E49-BCCA-D6DEF98F5A82@dipteris.unige.it> References: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> <31539652-BB35-4514-99E5-5030507D35B8@dipteris.unige.it> <1661A694-F802-4B19-94A1-55B65BE66310@dipteris.unige.it> <8449FC2D-5DAD-46D8-98FA-158C0C3373FF@nest.sns.it> <02440F36-21DC-4E49-BCCA-D6DEF98F5A82@dipteris.unige.it> Message-ID: <30886720-DAF1-4F9E-89BE-F6BF6BD2D020@nest.sns.it> On Sep 6, 2007, at 11:51 , Marino Vetuschi Zuccolini wrote: > ex 17: crash and prints these lines > [...] > ex 18, 23, 30: crashes suddenly with the first call to > formf: eself=..... > > ex 27, 28: show the error > from rdiaghg : error # 24 try to load fortran blas instead of the system-supplied blas. All weird and unexplained crashes that I get on Intel Mac OS-X seem to disappear if I replace LAPACK_LIBS = -llapack with LAPACK_LIBS = ../flib/blas.a -llapack in make.sys, remove the *.x file, relink (ignore the warnings). This works for g95 and gfortran; I cannot try right now with the Intel compiler. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From tone.kokalj at ijs.si Thu Sep 6 15:02:38 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Thu, 06 Sep 2007 15:02:38 +0200 Subject: [Pw_forum] Tests failed in CVS version In-Reply-To: <30886720-DAF1-4F9E-89BE-F6BF6BD2D020@nest.sns.it> References: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> <31539652-BB35-4514-99E5-5030507D35B8@dipteris.unige.it> <1661A694-F802-4B19-94A1-55B65BE66310@dipteris.unige.it> <8449FC2D-5DAD-46D8-98FA-158C0C3373FF@nest.sns.it> <02440F36-21DC-4E49-BCCA-D6DEF98F5A82@dipteris.unige.it> <30886720-DAF1-4F9E-89BE-F6BF6BD2D020@nest.sns.it> Message-ID: <1189083758.2543.17.camel@localhost.localdomain> Paola, I am working on the Cl-paper. In chapter of Cl2 molecular adsorption, you specified that Leeuw has computationally predicted very small dissociation barrier: I don't find where in the paper. I find only that he guesses on the basis of experiments that the barrier is small. Can you reffer me the page of the Leeuw, where he calculates the barrier. Tone From tone.kokalj at ijs.si Thu Sep 6 15:12:51 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Thu, 06 Sep 2007 15:12:51 +0200 Subject: [Pw_forum] Tests failed in CVS version In-Reply-To: <1189083758.2543.17.camel@localhost.localdomain> References: <7930E192-2F46-4AD3-9CE0-A2C79A1A61DF@dipteris.unige.it> <31539652-BB35-4514-99E5-5030507D35B8@dipteris.unige.it> <1661A694-F802-4B19-94A1-55B65BE66310@dipteris.unige.it> <8449FC2D-5DAD-46D8-98FA-158C0C3373FF@nest.sns.it> <02440F36-21DC-4E49-BCCA-D6DEF98F5A82@dipteris.unige.it> <30886720-DAF1-4F9E-89BE-F6BF6BD2D020@nest.sns.it> <1189083758.2543.17.camel@localhost.localdomain> Message-ID: <1189084371.2543.20.camel@localhost.localdomain> I apologise for the last message: apparently, I pressed the reply button on the wrong email. Tone From xianghjun at gmail.com Thu Sep 6 17:58:27 2007 From: xianghjun at gmail.com (Hongjun Xiang) Date: Thu, 06 Sep 2007 09:58:27 -0600 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? Message-ID: <46E023A3.5060206@gmail.com> Dear all, As mentioned in the paper [Phys. Rev. Lett. 94, 145501 (2005)] by Cococcioni et al., this feature is implemented in PWSCF. Could someone tell me if this is included in the current version of PWSCF (v. 3.2)? Thank you very much. Best regards, Hongjun Xiang ============================================================= H. J. Xiang Postdoctoral Research Associate National Renewable Energy Laboratory http://www4.ncsu.edu/~hxiang/ ============================================================= From okunoyukihiro2000 at yahoo.co.jp Thu Sep 6 17:50:48 2007 From: okunoyukihiro2000 at yahoo.co.jp (Yukihiro Okuno) Date: Fri, 07 Sep 2007 00:50:48 +0900 Subject: [Pw_forum] Atom Mass in vc-relax or relaxation Message-ID: <200709061550.AA00338@Einstein.yahoo.co.jp> Dear PW Users. I want to optimize the lattice parameter and internal atomic position, so I choose vc-relax calculation. In the relaxation , can I change atomic mass arbitrary instead of true atom mass? In the test calculation, I had changed the mass of the system and get almost same lattice parameter and atomic internal position. Does not relative relation of crystal coordinate of atoms depend on their mass ? And if some atoms have heavy mass and does not move, and then they cause the difficulties of conversence, is it one way to set same masses for all atoms to improve the conversence? Sincerely, Yukihiro Okuno. ---- Yukihiro Okuno okunoyukihiro2000 at yahoo.co.jp From xianghjun at gmail.com Thu Sep 6 18:01:37 2007 From: xianghjun at gmail.com (Hongjun Xiang) Date: Thu, 06 Sep 2007 10:01:37 -0600 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? Message-ID: <46E02461.6030104@gmail.com> Dear all, As mentioned in the paper [Phys. Rev. Lett. 94, 145501 (2005)] by Cococcioni et al., this feature is implemented in PWSCF. Could someone tell me if this is included in the current version of PWSCF (v. 3.2)? Thank you very much. Best regards, Hongjun Xiang ============================================================= H. J. Xiang Postdoctoral Research Associate National Renewable Energy Laboratory http://www4.ncsu.edu/~hxiang/ ============================================================= From matteo at umn.edu Thu Sep 6 18:13:04 2007 From: matteo at umn.edu (Matteo Cococcioni) Date: Thu, 06 Sep 2007 11:13:04 -0500 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? In-Reply-To: <46E02461.6030104@gmail.com> References: <46E02461.6030104@gmail.com> Message-ID: <46E02710.80303@umn.edu> Dear Hongjun, the electronic enthalpy is not contained in the current official version of the code (v 3.2). However it's implemented in the CVS version (README.cvs for instruction on how to download it) of the package. Precisely it's contained in the CP part (directory CPV). Instructions contained in INPUT_CP should be enough to run a calculation with the electronic enthalpy. Otherwise feel free to ask for more help. Regards, Matteo Hongjun Xiang wrote: > Dear all, > As mentioned in the paper [Phys. Rev. Lett. 94, 145501 (2005)] by > Cococcioni et al., > this feature is implemented in PWSCF. Could someone tell me if this is > included in the current version of PWSCF (v. 3.2)? > > Thank you very much. > > Best regards, > Hongjun Xiang > > ============================================================= > H. J. Xiang > Postdoctoral Research Associate > National Renewable Energy Laboratory > http://www4.ncsu.edu/~hxiang/ > ============================================================= > > > > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- A non-text attachment was scrubbed... Name: matteo.vcf Type: text/x-vcard Size: 294 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20070906/02ceb461/attachment.vcf From giannozz at nest.sns.it Thu Sep 6 18:42:15 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 6 Sep 2007 18:42:15 +0200 Subject: [Pw_forum] Atom Mass in vc-relax or relaxation In-Reply-To: <200709061550.AA00338@Einstein.yahoo.co.jp> References: <200709061550.AA00338@Einstein.yahoo.co.jp> Message-ID: <415A3396-E6F7-484D-8E69-E9F53DCBB765@nest.sns.it> On Sep 6, 2007, at 17:50 , Yukihiro Okuno wrote: > In the relaxation , can I change atomic mass arbitrary instead of > true atom mass? you can > Does not relative relation of crystal coordinate of atoms depend on > their mass ? not necessarily; there are algorithms that make usage of atomic masses for structural optimization, but most algorithms I know don't > And if some atoms have heavy mass and does not move, and then they > cause the difficulties of conversence, is it one way to set same > masses for all atoms > to improve the conversence? it is: when damped dynamics is used to achieve structural relaxation, this is a simple and effective way to "precondition" the dynamics, i.e. to avoid having both "slow" and "fast" degrees of freedom, with wildly different speed of approach to the minimum Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From xianghjun at gmail.com Thu Sep 6 18:46:30 2007 From: xianghjun at gmail.com (Hongjun Xiang) Date: Thu, 06 Sep 2007 10:46:30 -0600 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? In-Reply-To: <389095204.26878@ustc.edu.cn> References: <46E02461.6030104@gmail.com> <389095204.26878@ustc.edu.cn> Message-ID: <46E02EE6.4040105@gmail.com> Dear Dr. Cococcioni, Thank you very much for your reply. I will try to use the CVS version. PS, I am sorry for sending duplicating e-mails. Best regards, Hongjun Xiang ============================================================= H. J. Xiang Postdoctoral Research Associate National Renewable Energy Laboratory http://www4.ncsu.edu/~hxiang/ ============================================================= Matteo Cococcioni wrote: > > Dear Hongjun, > > the electronic enthalpy is not contained in the current official > version of the code (v 3.2). > However it's implemented in the CVS version (README.cvs for > instruction on how to download it) of > the package. Precisely it's contained in the CP part (directory CPV). > Instructions contained in > INPUT_CP should be enough to run a calculation with the electronic > enthalpy. Otherwise feel free to ask for more help. > > Regards, > > Matteo > From cesards at msi.umn.edu Thu Sep 6 20:57:38 2007 From: cesards at msi.umn.edu (Cesar R.S. da Silva) Date: Thu, 6 Sep 2007 13:57:38 -0500 (CDT) Subject: [Pw_forum] Atom Mass in vc-relax or relaxation In-Reply-To: References: Message-ID: <54083.209.162.14.191.1189105058.squirrel@www.msi.umn.edu> Dear Yukihiro, The answer to your question "can I change atomic mass arbitrary instead of true atom mass?" is YES! Fastest relaxation is obtained if you set all atomic masses to the mass average. Be aware that cell mass is not the sum of the atomic masses. This is the default and is a terrible choice. This is the most likely reason you have not noticed andy change in the cell. My suggestion is to start with wmass==2*10**(-3) and dt=80 and see what happens. Yours, Cesar. > > Message: 8 > Date: Fri, 07 Sep 2007 00:50:48 +0900 > From: Yukihiro Okuno > Subject: [Pw_forum] Atom Mass in vc-relax or relaxation > To: pw_forum at pwscf.org > Message-ID: <200709061550.AA00338 at Einstein.yahoo.co.jp> > Content-Type: text/plain; charset=us-ascii > > Dear PW Users. > > I want to optimize the lattice parameter and internal atomic position, so > I choose > vc-relax calculation. > > In the relaxation , can I change atomic mass arbitrary instead of true > atom mass? > In the test calculation, I had changed the mass of the system and get > almost same > lattice parameter and atomic internal position. > > Does not relative relation of crystal coordinate of atoms depend on their > mass ? > > And if some atoms have heavy mass and does not move, and then they cause > the difficulties of conversence, > is it one way to set same masses for all atoms to improve the conversence? > > Sincerely, > > Yukihiro Okuno. > > ---- > Yukihiro Okuno okunoyukihiro2000 at yahoo.co.jp > From suza.rri at gmail.com Fri Sep 7 18:50:28 2007 From: suza.rri at gmail.com (Suza W) Date: Fri, 7 Sep 2007 22:20:28 +0530 Subject: [Pw_forum] Convergence Message-ID: Dear all, I'm running the scf calculation for a single charged layer. I find that the chemical potential is very sensitive with the vacuum length. It does not get converged even after using a large vacuum of 100 a. u. What should one do to get a converged value? Yours sincerely, Suza Raman Research Institute -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070907/63732718/attachment.htm From baroni at sissa.it Fri Sep 7 20:15:54 2007 From: baroni at sissa.it (Stefano Baroni) Date: Fri, 7 Sep 2007 20:15:54 +0200 Subject: [Pw_forum] Convergence In-Reply-To: References: Message-ID: Dear Suza: I am not sure I understand clearly your question, but it seems to me that an isolated infinite and charged layer cannot have any well defined work function, because it would generate a uniform electric field in the vacuum, with a corresponding potential energy that diverges linearly far from the layer. The lack of convergence of any quantities related to the electrostatic potential energy in the vacuum (such as the work function) is exactly what I would expect. Let alone the extra complications arising from periodic boundary conditions ... Cheers - Stefano Baroni On Sep 7, 2007, at 6:50 PM, Suza W wrote: > > > Dear all, > > I'm running the scf calculation for a single charged > layer. I find that the chemical potential is very sensitive > with the vacuum length. It does not get > converged even after using a large vacuum of 100 a. u. > What should one do to get a converged value? > > Yours sincerely, > Suza > Raman Research Institute > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070907/b9b192d8/attachment.htm From suza.rri at gmail.com Sat Sep 8 16:19:03 2007 From: suza.rri at gmail.com (Suza W) Date: Sat, 8 Sep 2007 19:49:03 +0530 Subject: [Pw_forum] Convergence In-Reply-To: References: Message-ID: Dear Sir, Thank you for your helpful reply. Yours sincerely, Suza -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070908/eed99098/attachment.htm From okunoyukihiro2000 at yahoo.co.jp Sat Sep 8 18:02:49 2007 From: okunoyukihiro2000 at yahoo.co.jp (Yukihiro Okuno) Date: Sun, 09 Sep 2007 01:02:49 +0900 Subject: [Pw_forum] conversence problem in vc-relax (Atom Mass in relaxation) Message-ID: <200709081602.AA00341@Einstein.yahoo.co.jp> Dear PWScf Users. I'm Yukihiro Okuno. I had posted about the Atom Mass in vc-relax, and thank you for the responces. We set same atom mass 20.0 ( average of atom mass), and do vc-relax calculation as the following input file, &control calculation = 'vc-relax', restart_mode = 'from_scratch', pseudo_dir = '/pseudo/', outdir = './', prefix='vc-test' nstep = 100 etot_conv_thr = 1.0D-5 forc_conv_thr = 1.0D-4 dt = 80.0 tstress = .true. tprnfor = .true. / &system ibrav=6 celldm(1)=6.83994507291 celldm(3)=1.09801321247 nat=5 ntyp=3 nbnd=26 ecutwfc=60.0 ecutrho = 600.0 occupations = 'fixed' degauss=0.00 / &electrons conv_thr = 1e-10, mixing_beta=0.3, / &IONS ion_dynamics='damp' bfgs_ndim=3 pot_extrapolation = "second_order" wfc_extrapolation = "second_order" / &CELL cell_dynamics = 'damp-w', press = 0.0 wmass = 0.02 / But the calculation is not conversed and oscillate. The system has only five atoms and simple P4mm perovskite structure. I had change wmass as default value, but the convergence is not achieved. Is it too hard to converse to set forc_conv_thr = 1.0D-4 in VC-relax calculation ? ( I want to get severe internal atomic coordinate, and forc_conv_thr=1.0D-3 is not sufficient.) Simple atomic internal relaxation, I can get converged result. The stress and total force of output file is as follow. stress values ,, entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.13 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -2.09 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -1.17 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.60 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.07 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.37 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.84 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.95 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 1.07 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 1.22 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 1.36 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 1.47 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 1.55 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.96 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.65 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.39 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.20 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.04 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.08 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.13 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.13 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.63 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.89 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -1.05 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -1.04 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -1.13 and The total force, Total force = 0.000512 Total SCF correction = 0.000057 Total force = 0.000499 Total SCF correction = 0.000063 Total force = 0.000472 Total SCF correction = 0.000024 Total force = 0.000411 Total SCF correction = 0.000022 Total force = 0.000385 Total SCF correction = 0.000053 Total force = 0.000527 Total SCF correction = 0.000012 Total force = 0.000487 Total SCF correction = 0.000188 Total force = 0.000572 Total SCF correction = 0.000010 Total force = 0.000537 Total SCF correction = 0.000016 Total force = 0.000484 Total SCF correction = 0.000016 Total force = 0.000440 Total SCF correction = 0.000007 Total force = 0.000407 Total SCF correction = 0.000017 Total force = 0.000369 Total SCF correction = 0.000016 Total force = 0.000391 Total SCF correction = 0.000045 Total force = 0.000552 Total SCF correction = 0.000015 Total force = 0.000569 Total SCF correction = 0.000130 Total force = 0.000683 Total SCF correction = 0.000026 Total force = 0.000623 Total SCF correction = 0.000025 Total force = 0.000544 Total SCF correction = 0.000020 Total force = 0.000404 Total SCF correction = 0.000015 Total force = 0.000280 Total SCF correction = 0.000028 Total force = 0.000219 Total SCF correction = 0.000011 Total force = 0.000348 Total SCF correction = 0.000039 Total force = 0.000552 Total SCF correction = 0.000015 Total force = 0.000688 Total SCF correction = 0.000019 Total force = 0.000741 Total SCF correction = 0.000013 Total force = 0.000723 Total SCF correction = 0.000013 Total force = 0.000638 Total SCF correction = 0.000018 Total force = 0.000504 Total SCF correction = 0.000017 Total force = 0.000368 Total SCF correction = 0.000029 Total force = 0.000256 Total SCF correction = 0.000020 Total force = 0.001173 Total SCF correction = 0.000016 Total force = 0.001841 Total SCF correction = 0.000271 Total force = 0.001924 Total SCF correction = 0.000220 Total force = 0.002328 Total SCF correction = 0.000057 Total force = 0.002448 Total SCF correction = 0.000019 Total force = 0.002318 Total SCF correction = 0.000016 Total force = 0.002004 Total SCF correction = 0.000018 Total force = 0.001547 Total SCF correction = 0.000022 Total force = 0.001029 Total SCF correction = 0.000026 Total force = 0.000669 Total SCF correction = 0.000015 Total force = 0.002834 Total SCF correction = 0.000206 Total force = 0.003057 Total SCF correction = 0.000170 Total force = 0.001864 Total SCF correction = 0.000273 Total force = 0.001107 Total SCF correction = 0.000101 Total force = 0.000870 Total SCF correction = 0.000021 Total force = 0.000355 Total SCF correction = 0.000068 Total force = 0.000613 Total SCF correction = 0.000114 Total force = 0.000484 Total SCF correction = 0.000038 Total force = 0.000338 Total SCF correction = 0.000180 Total force = 0.000372 Total SCF correction = 0.000125 Total force = 0.000460 Total SCF correction = 0.000028 Total force = 0.000482 Total SCF correction = 0.000014 Total force = 0.000528 Total SCF correction = 0.000018 Total force = 0.000652 Total SCF correction = 0.000008 Total force = 0.000729 Total SCF correction = 0.000017 Total force = 0.000740 Total SCF correction = 0.000013 Total force = 0.000684 Total SCF correction = 0.000005 Total force = 0.000573 Total SCF correction = 0.000005 Total force = 0.000446 Total SCF correction = 0.000012 Total force = 0.000369 Total SCF correction = 0.000010 Total force = 0.000408 Total SCF correction = 0.000005 Total force = 0.000602 Total SCF correction = 0.000044 Total force = 0.000897 Total SCF correction = 0.000015 Total force = 0.001145 Total SCF correction = 0.000020 Total force = 0.001247 Total SCF correction = 0.000004 Total force = 0.001348 Total SCF correction = 0.000050 Sincerely. ---- Yukihiro Okuno okunoyukihiro2000 at yahoo.co.jp From emenendez at macul.ciencias.uchile.cl Sat Sep 8 21:53:46 2007 From: emenendez at macul.ciencias.uchile.cl (Eduardo Ariel Menendez P) Date: Sat, 8 Sep 2007 15:53:46 -0400 (CLT) Subject: [Pw_forum] Bug in dynmat Message-ID: Hi, I found a bug in program dynmat. It appears when ibrav=0. espresso-3.2/bin-opt/dynmat.x References: Message-ID: On Sep 8, 2007, at 21:53 , Eduardo Ariel Menendez P wrote: > I found a bug in program dynmat. It appears when ibrav=0. > from latgen : error # 1 > wrong at for ibrav=0 > I think the problem is this > In dynmat.f90: > call latgen(ibrav,celldm,at(1,1),at(1,2),at(1,3),omega) > In flib/latgen.f90: > subroutine latgen(ibrav,celldm,a1,a2,a3,omega) > ......... > real(DP), intent(inout) :: a1(3), a2(3), a3(3) > > dynmat passes real numers a(i,j) instead of arrays. no, it is not this. The above syntax is perfectly valid: (old) fortran passes the address of a variable (a pointer, in C language) to a subroutine. The error means that you have ibrav=0 .and. zero vectors at. > Moreover, in one package of version 3.2, it seems that the following > lines were lost in soubroutine readmat. > > if (ibrav==0) then > read(1,'(a)') symm_type > read(1,*) ((at(i,j),i=1,3),j=1,3) > end if this is a more likely reason for the error above! I though case ibrav=0 had been fixed, but apparently it had not. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From cesards at msi.umn.edu Mon Sep 10 19:33:29 2007 From: cesards at msi.umn.edu (Cesar R.S. da Silva) Date: Mon, 10 Sep 2007 12:33:29 -0500 (CDT) Subject: [Pw_forum] conversence problem in vc-relax (Atom Mass in In-Reply-To: References: Message-ID: <55669.209.162.14.191.1189445609.squirrel@www.msi.umn.edu> Hi Yukihiro, if you are using > etot_conv_thr = 1.0D-5 > forc_conv_thr = 1.0D-4 you should set a tighter convergence for SCF. My suggestion is conv_thr = 1.0d-8 Yours, Cesar. > > Message: 2 > Date: Sun, 09 Sep 2007 01:02:49 +0900 > From: Yukihiro Okuno > Subject: [Pw_forum] conversence problem in vc-relax (Atom Mass in > relaxation) > To: pw_forum at pwscf.org > Message-ID: <200709081602.AA00341 at Einstein.yahoo.co.jp> > Content-Type: text/plain; charset=us-ascii > > Dear PWScf Users. > > I'm Yukihiro Okuno. > > I had posted about the Atom Mass in vc-relax, and thank you for the > responces. > We set same atom mass 20.0 ( average of atom mass), and do vc-relax > calculation as > the following input file, > > > &control > calculation = 'vc-relax', > restart_mode = 'from_scratch', > pseudo_dir = '/pseudo/', > outdir = './', > prefix='vc-test' > nstep = 100 > etot_conv_thr = 1.0D-5 > forc_conv_thr = 1.0D-4 > dt = 80.0 > tstress = .true. > tprnfor = .true. > / > &system > ibrav=6 > celldm(1)=6.83994507291 > celldm(3)=1.09801321247 > nat=5 > ntyp=3 > nbnd=26 > ecutwfc=60.0 > ecutrho = 600.0 > occupations = 'fixed' > degauss=0.00 > / > &electrons > conv_thr = 1e-10, > mixing_beta=0.3, > / > &IONS > ion_dynamics='damp' > bfgs_ndim=3 > pot_extrapolation = "second_order" > wfc_extrapolation = "second_order" > / > &CELL > cell_dynamics = 'damp-w', > press = 0.0 > wmass = 0.02 > / > > > But the calculation is not conversed and oscillate. The system has only > five atoms > and simple P4mm perovskite structure. > I had change wmass as default value, but the convergence is not achieved. > Is it too hard to converse to set forc_conv_thr = 1.0D-4 in VC-relax > calculation ? > ( I want to get severe internal atomic coordinate, and > forc_conv_thr=1.0D-3 is not > sufficient.) Simple atomic internal relaxation, I can get converged > result. > > > The stress and total force of output file is as follow. > > stress values ,, > > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -0.13 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -2.09 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -1.17 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -0.60 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 0.07 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 0.37 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 0.84 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 0.95 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 1.07 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 1.22 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 1.36 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 1.47 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 1.55 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 0.96 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 0.65 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 0.39 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 0.20 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > 0.04 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -0.08 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -0.13 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -0.13 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -0.63 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -0.89 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -1.05 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -1.04 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= > -1.13 > > > and The total force, > > Total force = 0.000512 Total SCF correction = 0.000057 > Total force = 0.000499 Total SCF correction = 0.000063 > Total force = 0.000472 Total SCF correction = 0.000024 > Total force = 0.000411 Total SCF correction = 0.000022 > Total force = 0.000385 Total SCF correction = 0.000053 > Total force = 0.000527 Total SCF correction = 0.000012 > Total force = 0.000487 Total SCF correction = 0.000188 > Total force = 0.000572 Total SCF correction = 0.000010 > Total force = 0.000537 Total SCF correction = 0.000016 > Total force = 0.000484 Total SCF correction = 0.000016 > Total force = 0.000440 Total SCF correction = 0.000007 > Total force = 0.000407 Total SCF correction = 0.000017 > Total force = 0.000369 Total SCF correction = 0.000016 > Total force = 0.000391 Total SCF correction = 0.000045 > Total force = 0.000552 Total SCF correction = 0.000015 > Total force = 0.000569 Total SCF correction = 0.000130 > Total force = 0.000683 Total SCF correction = 0.000026 > Total force = 0.000623 Total SCF correction = 0.000025 > Total force = 0.000544 Total SCF correction = 0.000020 > Total force = 0.000404 Total SCF correction = 0.000015 > Total force = 0.000280 Total SCF correction = 0.000028 > Total force = 0.000219 Total SCF correction = 0.000011 > Total force = 0.000348 Total SCF correction = 0.000039 > Total force = 0.000552 Total SCF correction = 0.000015 > Total force = 0.000688 Total SCF correction = 0.000019 > Total force = 0.000741 Total SCF correction = 0.000013 > Total force = 0.000723 Total SCF correction = 0.000013 > Total force = 0.000638 Total SCF correction = 0.000018 > Total force = 0.000504 Total SCF correction = 0.000017 > Total force = 0.000368 Total SCF correction = 0.000029 > Total force = 0.000256 Total SCF correction = 0.000020 > Total force = 0.001173 Total SCF correction = 0.000016 > Total force = 0.001841 Total SCF correction = 0.000271 > Total force = 0.001924 Total SCF correction = 0.000220 > Total force = 0.002328 Total SCF correction = 0.000057 > Total force = 0.002448 Total SCF correction = 0.000019 > Total force = 0.002318 Total SCF correction = 0.000016 > Total force = 0.002004 Total SCF correction = 0.000018 > Total force = 0.001547 Total SCF correction = 0.000022 > Total force = 0.001029 Total SCF correction = 0.000026 > Total force = 0.000669 Total SCF correction = 0.000015 > Total force = 0.002834 Total SCF correction = 0.000206 > Total force = 0.003057 Total SCF correction = 0.000170 > Total force = 0.001864 Total SCF correction = 0.000273 > Total force = 0.001107 Total SCF correction = 0.000101 > Total force = 0.000870 Total SCF correction = 0.000021 > Total force = 0.000355 Total SCF correction = 0.000068 > Total force = 0.000613 Total SCF correction = 0.000114 > Total force = 0.000484 Total SCF correction = 0.000038 > Total force = 0.000338 Total SCF correction = 0.000180 > Total force = 0.000372 Total SCF correction = 0.000125 > Total force = 0.000460 Total SCF correction = 0.000028 > Total force = 0.000482 Total SCF correction = 0.000014 > Total force = 0.000528 Total SCF correction = 0.000018 > Total force = 0.000652 Total SCF correction = 0.000008 > Total force = 0.000729 Total SCF correction = 0.000017 > Total force = 0.000740 Total SCF correction = 0.000013 > Total force = 0.000684 Total SCF correction = 0.000005 > Total force = 0.000573 Total SCF correction = 0.000005 > Total force = 0.000446 Total SCF correction = 0.000012 > Total force = 0.000369 Total SCF correction = 0.000010 > Total force = 0.000408 Total SCF correction = 0.000005 > Total force = 0.000602 Total SCF correction = 0.000044 > Total force = 0.000897 Total SCF correction = 0.000015 > Total force = 0.001145 Total SCF correction = 0.000020 > Total force = 0.001247 Total SCF correction = 0.000004 > Total force = 0.001348 Total SCF correction = 0.000050 > > Sincerely. > > > ---- > Yukihiro Okuno okunoyukihiro2000 at yahoo.co.jp > > From konstantin_kudin at yahoo.com Mon Sep 10 21:55:49 2007 From: konstantin_kudin at yahoo.com (Konstantin Kudin) Date: Mon, 10 Sep 2007 12:55:49 -0700 (PDT) Subject: [Pw_forum] Bug in dynmat In-Reply-To: Message-ID: <522187.36076.qm@web57703.mail.re3.yahoo.com> --- Paolo Giannozzi wrote: > > On Sep 8, 2007, at 21:53 , Eduardo Ariel Menendez P wrote: > > Moreover, in one package of version 3.2, it seems that the > following > > lines were lost in soubroutine readmat. > > > > if (ibrav==0) then > > read(1,'(a)') symm_type > > read(1,*) ((at(i,j),i=1,3),j=1,3) > > end if > > this is a more likely reason for the error above! I though case > ibrav=0 > had been fixed, but apparently it had not. This has been fixed, but after the version 3.2 : http://www.democritos.it:8888/O-sesame/chngview?cn=4429 Kostya ____________________________________________________________________________________ Yahoo! oneSearch: Finally, mobile search that gives answers, not web links. http://mobile.yahoo.com/mobileweb/onesearch?refer=1ONXIC From scandolo at ictp.it Mon Sep 10 22:23:14 2007 From: scandolo at ictp.it (Sandro Scandolo) Date: Mon, 10 Sep 2007 22:23:14 +0200 Subject: [Pw_forum] conversence problem in vc-relax (Atom Mass in relaxation) In-Reply-To: <200709081602.AA00341@Einstein.yahoo.co.jp> References: <200709081602.AA00341@Einstein.yahoo.co.jp> Message-ID: <46E5A7B2.1010306@ictp.it> Dear Yukihiro, while a higher accuracy may be required in some exceptional cases, converging the stress tensor (or the average pressure) within 1 kbar of the target value is already a physically sound achievement. The error DeltaV in volume that one is making with a DeltaP~+-1 kbar error in the pressure is given by DeltaV/V=DeltaP/Bulk_modulus. With the exception of molecular or very soft solids, the bulk modulus of the vast majority of solids is typically > 10 GPa, or 100 kbar, so your volume is likely to be converged to better than 1% already.... This of course should not prevent you from asking a higher accuracy, if required, in which case Cesar's comments are likely to be helpful. Regards, Sandro Yukihiro Okuno wrote: > Dear PWScf Users. > > I'm Yukihiro Okuno. > > I had posted about the Atom Mass in vc-relax, and thank you for the responces. > We set same atom mass 20.0 ( average of atom mass), and do vc-relax calculation as > the following input file, > > > &control > calculation = 'vc-relax', > restart_mode = 'from_scratch', > pseudo_dir = '/pseudo/', > outdir = './', > prefix='vc-test' > nstep = 100 > etot_conv_thr = 1.0D-5 > forc_conv_thr = 1.0D-4 > dt = 80.0 > tstress = .true. > tprnfor = .true. > / > &system > ibrav=6 > celldm(1)=6.83994507291 > celldm(3)=1.09801321247 > nat=5 > ntyp=3 > nbnd=26 > ecutwfc=60.0 > ecutrho = 600.0 > occupations = 'fixed' > degauss=0.00 > / > &electrons > conv_thr = 1e-10, > mixing_beta=0.3, > / > &IONS > ion_dynamics='damp' > bfgs_ndim=3 > pot_extrapolation = "second_order" > wfc_extrapolation = "second_order" > / > &CELL > cell_dynamics = 'damp-w', > press = 0.0 > wmass = 0.02 > / > > > But the calculation is not conversed and oscillate. The system has only five atoms > and simple P4mm perovskite structure. > I had change wmass as default value, but the convergence is not achieved. > Is it too hard to converse to set forc_conv_thr = 1.0D-4 in VC-relax calculation ? > ( I want to get severe internal atomic coordinate, and forc_conv_thr=1.0D-3 is not > sufficient.) Simple atomic internal relaxation, I can get converged result. > > > The stress and total force of output file is as follow. > > stress values ,, > > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -0.13 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -2.09 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -1.17 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -0.60 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 0.07 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 0.37 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 0.84 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 0.95 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 1.07 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 1.22 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 1.36 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 1.47 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 1.55 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 0.96 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 0.65 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 0.39 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 0.20 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= 0.04 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -0.08 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -0.13 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -0.13 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -0.63 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -0.89 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -1.05 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -1.04 > entering subroutine stress ... > total stress (Ry/bohr**3) (kbar) P= -1.13 > > > and The total force, > > Total force = 0.000512 Total SCF correction = 0.000057 > Total force = 0.000499 Total SCF correction = 0.000063 > Total force = 0.000472 Total SCF correction = 0.000024 > Total force = 0.000411 Total SCF correction = 0.000022 > Total force = 0.000385 Total SCF correction = 0.000053 > Total force = 0.000527 Total SCF correction = 0.000012 > Total force = 0.000487 Total SCF correction = 0.000188 > Total force = 0.000572 Total SCF correction = 0.000010 > Total force = 0.000537 Total SCF correction = 0.000016 > Total force = 0.000484 Total SCF correction = 0.000016 > Total force = 0.000440 Total SCF correction = 0.000007 > Total force = 0.000407 Total SCF correction = 0.000017 > Total force = 0.000369 Total SCF correction = 0.000016 > Total force = 0.000391 Total SCF correction = 0.000045 > Total force = 0.000552 Total SCF correction = 0.000015 > Total force = 0.000569 Total SCF correction = 0.000130 > Total force = 0.000683 Total SCF correction = 0.000026 > Total force = 0.000623 Total SCF correction = 0.000025 > Total force = 0.000544 Total SCF correction = 0.000020 > Total force = 0.000404 Total SCF correction = 0.000015 > Total force = 0.000280 Total SCF correction = 0.000028 > Total force = 0.000219 Total SCF correction = 0.000011 > Total force = 0.000348 Total SCF correction = 0.000039 > Total force = 0.000552 Total SCF correction = 0.000015 > Total force = 0.000688 Total SCF correction = 0.000019 > Total force = 0.000741 Total SCF correction = 0.000013 > Total force = 0.000723 Total SCF correction = 0.000013 > Total force = 0.000638 Total SCF correction = 0.000018 > Total force = 0.000504 Total SCF correction = 0.000017 > Total force = 0.000368 Total SCF correction = 0.000029 > Total force = 0.000256 Total SCF correction = 0.000020 > Total force = 0.001173 Total SCF correction = 0.000016 > Total force = 0.001841 Total SCF correction = 0.000271 > Total force = 0.001924 Total SCF correction = 0.000220 > Total force = 0.002328 Total SCF correction = 0.000057 > Total force = 0.002448 Total SCF correction = 0.000019 > Total force = 0.002318 Total SCF correction = 0.000016 > Total force = 0.002004 Total SCF correction = 0.000018 > Total force = 0.001547 Total SCF correction = 0.000022 > Total force = 0.001029 Total SCF correction = 0.000026 > Total force = 0.000669 Total SCF correction = 0.000015 > Total force = 0.002834 Total SCF correction = 0.000206 > Total force = 0.003057 Total SCF correction = 0.000170 > Total force = 0.001864 Total SCF correction = 0.000273 > Total force = 0.001107 Total SCF correction = 0.000101 > Total force = 0.000870 Total SCF correction = 0.000021 > Total force = 0.000355 Total SCF correction = 0.000068 > Total force = 0.000613 Total SCF correction = 0.000114 > Total force = 0.000484 Total SCF correction = 0.000038 > Total force = 0.000338 Total SCF correction = 0.000180 > Total force = 0.000372 Total SCF correction = 0.000125 > Total force = 0.000460 Total SCF correction = 0.000028 > Total force = 0.000482 Total SCF correction = 0.000014 > Total force = 0.000528 Total SCF correction = 0.000018 > Total force = 0.000652 Total SCF correction = 0.000008 > Total force = 0.000729 Total SCF correction = 0.000017 > Total force = 0.000740 Total SCF correction = 0.000013 > Total force = 0.000684 Total SCF correction = 0.000005 > Total force = 0.000573 Total SCF correction = 0.000005 > Total force = 0.000446 Total SCF correction = 0.000012 > Total force = 0.000369 Total SCF correction = 0.000010 > Total force = 0.000408 Total SCF correction = 0.000005 > Total force = 0.000602 Total SCF correction = 0.000044 > Total force = 0.000897 Total SCF correction = 0.000015 > Total force = 0.001145 Total SCF correction = 0.000020 > Total force = 0.001247 Total SCF correction = 0.000004 > Total force = 0.001348 Total SCF correction = 0.000050 > > Sincerely. > > > ---- > Yukihiro Okuno okunoyukihiro2000 at yahoo.co.jp > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > From sei.ou0722 at gmail.com Tue Sep 11 16:03:04 2007 From: sei.ou0722 at gmail.com (sei ou) Date: Tue, 11 Sep 2007 23:03:04 +0900 Subject: [Pw_forum] Spherical integration of spin Message-ID: <161e5bd50709110703w6c898075nf761ca7aa969efa1@mail.gmail.com> Dear PWscf users, Anyone know how to perform spherical integration of spin by PWscf? Or is there any other methods in PWscf can be used to determine oxidation state of an ion? Thank you in advance. Sei Ou Simulatio Co. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070911/eeaec424/attachment.htm From levuser at gmail.com Tue Sep 11 21:37:05 2007 From: levuser at gmail.com (Sergey Levchenko) Date: Tue, 11 Sep 2007 12:37:05 -0700 Subject: [Pw_forum] "Wrong atomic coordinates" message Message-ID: Hi! I've encountered a peculiar problem. I want to run PWSCF 3.2 for the following atomic system: Li 0.5 0.5 0.5 H 0.5 0.0 0.0 lattice vectors (fcc): 0.0 0.5 0.5 0.5 0.0 0.5 0.5 0.5 0.0 alat=7.555754 bohr I specify ibrav = 0, and nosym true. But the program crashes right at the beginning with the message: from setup : error # 1 Wrong atomic coordinates I get the same message with ibrav=2. What could be the problem? Thank you. Sincerely, Sergey Levchenko, University of Pennsylvania. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070911/d5e77498/attachment.htm From baroni at sissa.it Tue Sep 11 21:39:14 2007 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 11 Sep 2007 21:39:14 +0200 Subject: [Pw_forum] Spherical integration of spin In-Reply-To: <161e5bd50709110703w6c898075nf761ca7aa969efa1@mail.gmail.com> References: <161e5bd50709110703w6c898075nf761ca7aa969efa1@mail.gmail.com> Message-ID: <985A6CCB-7FB0-4A10-9F10-C6F1BAFEDC62@sissa.it> Dear Sei: > Anyone know how to perform spherical integration of spin by PWscf? can you explain more precisely what you mean? the integration of the spin polarization over atom-centered sphere, o something more esoteric? > Or is there any other methods in PWscf can be used to determine > oxidation state of an ion? charge population analysis? Stefano Baroni --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070911/a8840bba/attachment.htm From eyvaz_isaev at yahoo.com Tue Sep 11 22:02:54 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 11 Sep 2007 13:02:54 -0700 (PDT) Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: Message-ID: <738802.18110.qm@web60311.mail.yahoo.com> Hi Sergey, It seems something wrong in your input file. Could you please post your input file? I can suggest that you specify only ireeducible atomic positions while all atomic positions have to be listed. Bests, Eyvaz. --- Sergey Levchenko wrote: > Hi! > > I've encountered a peculiar problem. I want to run > PWSCF 3.2 for the > following atomic system: > > Li 0.5 0.5 0.5 > H 0.5 0.0 0.0 > > lattice vectors (fcc): > > 0.0 0.5 0.5 > 0.5 0.0 0.5 > 0.5 0.5 0.0 > > alat=7.555754 bohr > > I specify ibrav = 0, and nosym true. But the program > crashes right at the > beginning with the message: > > from setup : error # 1 > Wrong atomic coordinates > > I get the same message with ibrav=2. > > What could be the problem? > > Thank you. > > Sincerely, > > Sergey Levchenko, > University of Pennsylvania. > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Got a little couch potato? Check out fun summer activities for kids. http://search.yahoo.com/search?fr=oni_on_mail&p=summer+activities+for+kids&cs=bz From levuser at gmail.com Wed Sep 12 01:03:51 2007 From: levuser at gmail.com (Sergey Levchenko) Date: Tue, 11 Sep 2007 19:03:51 -0400 Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: <738802.18110.qm@web60311.mail.yahoo.com> References: <738802.18110.qm@web60311.mail.yahoo.com> Message-ID: Hi, Eyvaz, I think I know what my mistake was. I specified atomic_positions card with default option which is alat, but that is still in cartesian coordinates, right? I wanted to specify in the basis of the lattice vectors. Do I have to use option crystal? Here is my input: LiH &CONTROL calculation = 'scf' restart_mode = 'from_scratch' prefix = 'lih' pseudo_dir = '/workspace/lev/PWSCF/LiH' outdir = '/workspace/lev/PWSCF/LiH' iprint = 2 verbosity = 'high' tprnfor = .true. / &SYSTEM ibrav = 0, celldm(1) = 7.5557540587, nat = 2, ntyp = 2, ecutwfc = 70.0, occupations = 'fixed', nosym = .true. / &ELECTRONS conv_thr = 1.D-6, mixing_beta = 0.3D0, / ATOMIC_SPECIES Li 1.0 li-3.ncpp H 1.0 h.ncpp ATOMIC_POSITIONS Li 0.5 0.5 0.5 H 0.6 0.0 0.0 K_POINTS {automatic} 4 4 4 0 0 0 CELL_PARAMETERS 0.0 0.5 0.5 0.5 0.0 0.5 0.5 0.5 0.0 Thanks! Sergey. On 9/11/07, Eyvaz Isaev wrote: > > Hi Sergey, > > It seems something wrong in your input file. Could you > please post your input file? > I can suggest that you specify only ireeducible atomic > positions while all atomic positions have to be > listed. > > Bests, > Eyvaz. > > --- Sergey Levchenko wrote: > > > Hi! > > > > I've encountered a peculiar problem. I want to run > > PWSCF 3.2 for the > > following atomic system: > > > > Li 0.5 0.5 0.5 > > H 0.5 0.0 0.0 > > > > lattice vectors (fcc): > > > > 0.0 0.5 0.5 > > 0.5 0.0 0.5 > > 0.5 0.5 0.0 > > > > alat=7.555754 bohr > > > > I specify ibrav = 0, and nosym true. But the program > > crashes right at the > > beginning with the message: > > > > from setup : error # 1 > > Wrong atomic coordinates > > > > I get the same message with ibrav=2. > > > > What could be the problem? > > > > Thank you. > > > > Sincerely, > > > > Sergey Levchenko, > > University of Pennsylvania. > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > ------------------------------------------------------------------- > Prof. Eyvaz Isaev, > Theoretical Physics Department, Moscow State Institute of Steel & Alloys, > Russia, and > Condensed Matter Theory Group, Uppsala University, Sweden > Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com > > > > > ____________________________________________________________________________________ > Got a little couch potato? > Check out fun summer activities for kids. > > http://search.yahoo.com/search?fr=oni_on_mail&p=summer+activities+for+kids&cs=bz > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070911/cd78bb2d/attachment-0001.htm From marzari at MIT.EDU Wed Sep 12 02:05:25 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Tue, 11 Sep 2007 20:05:25 -0400 Subject: [Pw_forum] Spherical integration of spin In-Reply-To: <985A6CCB-7FB0-4A10-9F10-C6F1BAFEDC62@sissa.it> References: <161e5bd50709110703w6c898075nf761ca7aa969efa1@mail.gmail.com> <985A6CCB-7FB0-4A10-9F10-C6F1BAFEDC62@sissa.it> Message-ID: <46E72D45.2040800@mit.edu> Stefano Baroni wrote: >> Or is there any other methods in PWscf can be used to determine >> oxidation state of an ion? > > charge population analysis? > Exactly - you want to look at the projecwfc postprocessing tool, so that you can e.g. project onto d-atomic orbitals, and see how much your ground state overlaps with those. nicola -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From sei.ou0722 at gmail.com Wed Sep 12 03:16:34 2007 From: sei.ou0722 at gmail.com (sei ou) Date: Wed, 12 Sep 2007 10:16:34 +0900 Subject: [Pw_forum] Spherical integration of spin In-Reply-To: <985A6CCB-7FB0-4A10-9F10-C6F1BAFEDC62@sissa.it> References: <161e5bd50709110703w6c898075nf761ca7aa969efa1@mail.gmail.com> <985A6CCB-7FB0-4A10-9F10-C6F1BAFEDC62@sissa.it> Message-ID: <161e5bd50709111816m7a4c64e7k221654fb03cec479@mail.gmail.com> Dear Stefano and Nicola, Thank you very much for your reply. I do want to do integration of the spin polarization over atom-centered sphere. We specially pay attention to radial range at 0~2 angstrom. Regards, Sei > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070912/e14adaff/attachment.htm From eyvaz_isaev at yahoo.com Wed Sep 12 10:28:44 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 12 Sep 2007 01:28:44 -0700 (PDT) Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: Message-ID: <715563.24301.qm@web60313.mail.yahoo.com> Hi Sergei, --- Sergey Levchenko wrote: > I think I know what my mistake was. I specified > atomic_positions card with default option which is > alat, but that is still in cartesian coordinates, > right? I wanted to specify in the basis of the > lattice vectors. Do I have to use option crystal? No, the error is not related to specifing atomic coordinates either in cartesian or wrt lattice vectors. Of course, you will find incorrect symmetry if you setup atomic positions improperly. This will result in incorrect total energy. Actually I used your input file, except Pseudopotentials name, but did not observe any message like you mentioned (more strictly, no message). So, your input file looks correct. Check your own file in your computer whether it contains any extra symbol at the end of a line (usually ^M due to DOS, Windows). Use "vi" editor for this purpose. It might also be you need an extra blank at the beginning of these lines. Bests, Eyvaz. > > Here is my input: > > LiH > &CONTROL > calculation = 'scf' > restart_mode = 'from_scratch' > prefix = 'lih' > pseudo_dir = '/workspace/lev/PWSCF/LiH' > outdir = '/workspace/lev/PWSCF/LiH' > iprint = 2 > verbosity = 'high' > tprnfor = .true. > / > &SYSTEM > ibrav = 0, > celldm(1) = 7.5557540587, > nat = 2, > ntyp = 2, > ecutwfc = 70.0, > occupations = 'fixed', > nosym = .true. > / > &ELECTRONS > conv_thr = 1.D-6, > mixing_beta = 0.3D0, > / > ATOMIC_SPECIES > Li 1.0 li-3.ncpp > H 1.0 h.ncpp > ATOMIC_POSITIONS > Li 0.5 0.5 0.5 > H 0.6 0.0 0.0 > K_POINTS {automatic} > 4 4 4 0 0 0 > CELL_PARAMETERS > 0.0 0.5 0.5 > 0.5 0.0 0.5 > 0.5 0.5 0.0 > > Thanks! > > Sergey. > > On 9/11/07, Eyvaz Isaev > wrote: > > > > Hi Sergey, > > > > It seems something wrong in your input file. Could > you > > please post your input file? > > I can suggest that you specify only ireeducible > atomic > > positions while all atomic positions have to be > > listed. > > > > Bests, > > Eyvaz. > > > > --- Sergey Levchenko wrote: > > > > > Hi! > > > > > > I've encountered a peculiar problem. I want to > run > > > PWSCF 3.2 for the > > > following atomic system: > > > > > > Li 0.5 0.5 0.5 > > > H 0.5 0.0 0.0 > > > > > > lattice vectors (fcc): > > > > > > 0.0 0.5 0.5 > > > 0.5 0.0 0.5 > > > 0.5 0.5 0.0 > > > > > > alat=7.555754 bohr > > > > > > I specify ibrav = 0, and nosym true. But the > program > > > crashes right at the > > > beginning with the message: > > > > > > from setup : error # 1 > > > Wrong atomic coordinates > > > > > > I get the same message with ibrav=2. > > > > > > What could be the problem? > > > > > > Thank you. > > > > > > Sincerely, > > > > > > Sergey Levchenko, > > > University of Pennsylvania. > > > > > _______________________________________________ > > > Pw_forum mailing list > > > Pw_forum at pwscf.org > > > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > > > > ------------------------------------------------------------------- > > Prof. Eyvaz Isaev, > > Theoretical Physics Department, Moscow State > Institute of Steel & Alloys, > > Russia, and > > Condensed Matter Theory Group, Uppsala University, > Sweden > > Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com > > > > > > > > > > > ____________________________________________________________________________________ > > Got a little couch potato? > > Check out fun summer activities for kids. > > > > > http://search.yahoo.com/search?fr=oni_on_mail&p=summer+activities+for+kids&cs=bz > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Pinpoint customers who are looking for what you sell. http://searchmarketing.yahoo.com/ From ding at sissa.it Wed Sep 12 11:00:48 2007 From: ding at sissa.it (Xunlei Ding) Date: Wed, 12 Sep 2007 11:00:48 +0200 Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: <715563.24301.qm@web60313.mail.yahoo.com> References: <715563.24301.qm@web60313.mail.yahoo.com> Message-ID: <46E7AAC0.7020906@sissa.it> Dear Eyvaz, Yes, I did the same test as you, and no error like "Wrong atomic coordinates" was found. But when I used xcrysten to see the input file, it was found that the distance between H and Li is about 0.4A. I am not sure whether some versions of espresso will think this too close distance as "Wrong atomic coordinates". My suggestion to Sergey is: at first, check what Eyvaz has said , ie "Check your own file in your computer whether it contains any extra symbol at the end of a line (usually ^M due to DOS, Windows).". If no problem, then add "crystal" to the line "ATOMIC_POSITIONS" to see what will happen. Best regards, Ding Eyvaz Isaev wrote: >Hi Sergei, > >--- Sergey Levchenko wrote: > > >>I think I know what my mistake was. I specified >>atomic_positions card with default option which is > >> >> >alat, but that is still in cartesian coordinates, > > >>right? I wanted to specify in the basis of the >>lattice vectors. Do I have to use option crystal? >> >> > >No, the error is not related to specifing atomic >coordinates either in cartesian or wrt lattice >vectors. Of course, you will find incorrect symmetry >if you setup atomic positions improperly. This will >result in incorrect total energy. > >Actually I used your input file, except >Pseudopotentials name, but did not observe any message >like you mentioned (more strictly, no message). So, >your input file looks correct. Check your own file in >your computer whether it contains any extra symbol at >the end of a line (usually ^M due to DOS, Windows). >Use "vi" editor for this purpose. >It might also be you need an extra blank at the >beginning of these lines. > >Bests, >Eyvaz. > > >>Here is my input: >> >>LiH >>&CONTROL >> calculation = 'scf' >> restart_mode = 'from_scratch' >> prefix = 'lih' >> pseudo_dir = '/workspace/lev/PWSCF/LiH' >> outdir = '/workspace/lev/PWSCF/LiH' >> iprint = 2 >> verbosity = 'high' >> tprnfor = .true. >>/ >>&SYSTEM >> ibrav = 0, >> celldm(1) = 7.5557540587, >> nat = 2, >> ntyp = 2, >> ecutwfc = 70.0, >> occupations = 'fixed', >> nosym = .true. >>/ >>&ELECTRONS >> conv_thr = 1.D-6, >> mixing_beta = 0.3D0, >>/ >>ATOMIC_SPECIES >>Li 1.0 li-3.ncpp >>H 1.0 h.ncpp >>ATOMIC_POSITIONS >>Li 0.5 0.5 0.5 >>H 0.6 0.0 0.0 >>K_POINTS {automatic} >>4 4 4 0 0 0 >>CELL_PARAMETERS >>0.0 0.5 0.5 >>0.5 0.0 0.5 >>0.5 0.5 0.0 >> >>Thanks! >> >>Sergey. >> >>On 9/11/07, Eyvaz Isaev >>wrote: >> >> >>>Hi Sergey, >>> >>>It seems something wrong in your input file. Could >>> >>> >>you >> >> >>>please post your input file? >>>I can suggest that you specify only ireeducible >>> >>> >>atomic >> >> >>>positions while all atomic positions have to be >>>listed. >>> >>>Bests, >>>Eyvaz. >>> >>>--- Sergey Levchenko wrote: >>> >>> >>> >>>>Hi! >>>> >>>>I've encountered a peculiar problem. I want to >>>> >>>> >>run >> >> >>>>PWSCF 3.2 for the >>>>following atomic system: >>>> >>>>Li 0.5 0.5 0.5 >>>>H 0.5 0.0 0.0 >>>> >>>>lattice vectors (fcc): >>>> >>>>0.0 0.5 0.5 >>>>0.5 0.0 0.5 >>>>0.5 0.5 0.0 >>>> >>>>alat=7.555754 bohr >>>> >>>>I specify ibrav = 0, and nosym true. But the >>>> >>>> >>program >> >> >>>>crashes right at the >>>>beginning with the message: >>>> >>>>from setup : error # 1 >>>>Wrong atomic coordinates >>>> >>>>I get the same message with ibrav=2. >>>> >>>>What could be the problem? >>>> >>>>Thank you. >>>> >>>>Sincerely, >>>> >>>>Sergey Levchenko, >>>>University of Pennsylvania. >>>> >>>> >>_______________________________________________ >> >> >>>>Pw_forum mailing list >>>>Pw_forum at pwscf.org >>>> >>>> >>>> >>http://www.democritos.it/mailman/listinfo/pw_forum >> >> >>> >>> >>> >------------------------------------------------------------------- > > >>>Prof. Eyvaz Isaev, >>>Theoretical Physics Department, Moscow State >>> >>> >>Institute of Steel & Alloys, >> >> >>>Russia, and >>>Condensed Matter Theory Group, Uppsala University, >>> >>> >>Sweden >> >> >>>Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com >>> >>> >>> >>> >>> >>> >>> >____________________________________________________________________________________ > > >>>Got a little couch potato? >>>Check out fun summer activities for kids. >>> >>> >>> >>> >http://search.yahoo.com/search?fr=oni_on_mail&p=summer+activities+for+kids&cs=bz > > >>>_______________________________________________ >>>Pw_forum mailing list >>>Pw_forum at pwscf.org >>>http://www.democritos.it/mailman/listinfo/pw_forum >>> >>>_______________________________________________ >>> >>> >>Pw_forum mailing list >>Pw_forum at pwscf.org >>http://www.democritos.it/mailman/listinfo/pw_forum >> >> >> > > >------------------------------------------------------------------- >Prof. Eyvaz Isaev, >Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and >Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com > > > >____________________________________________________________________________________ >Pinpoint customers who are looking for what you sell. >http://searchmarketing.yahoo.com/ >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum > > > From levuser at gmail.com Wed Sep 12 14:46:21 2007 From: levuser at gmail.com (Sergey Levchenko) Date: Wed, 12 Sep 2007 08:46:21 -0400 Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: <46E7AAC0.7020906@sissa.it> References: <715563.24301.qm@web60313.mail.yahoo.com> <46E7AAC0.7020906@sissa.it> Message-ID: Dear friends, I am terribly sorry, I made a mistake in my input file. With the coordinates specified in that file (Li 0.5 0.5 0.5, H 0.6 0.0 0.0) I also don't get the error message. The problem seems to occur only for atomic coordinates (Li 0.5 0.5 0.5, H 0.5 0.0 0.0). Thank you for your patience. Sincerely, Sergey Levchenko, University of Penn. On 9/12/07, Xunlei Ding wrote: > > Dear Eyvaz, > > Yes, I did the same test as you, and no error like "Wrong atomic > coordinates" was found. > But when I used xcrysten to see the input file, > it was found that the distance between H and Li is about 0.4A. > I am not sure whether some versions of espresso will think this too > close distance as "Wrong atomic coordinates". > > My suggestion to Sergey is: > at first, check what Eyvaz has said , ie "Check your own file in your > computer whether it contains any extra symbol at the end of a line > (usually ^M due to DOS, Windows).". > If no problem, then add "crystal" to the line "ATOMIC_POSITIONS" to see > what will happen. > > Best regards, > Ding > > Eyvaz Isaev wrote: > > >Hi Sergei, > > > >--- Sergey Levchenko wrote: > > > > > >>I think I know what my mistake was. I specified > >>atomic_positions card with default option which is > > >> > >> > >alat, but that is still in cartesian coordinates, > > > > > >>right? I wanted to specify in the basis of the > >>lattice vectors. Do I have to use option crystal? > >> > >> > > > >No, the error is not related to specifing atomic > >coordinates either in cartesian or wrt lattice > >vectors. Of course, you will find incorrect symmetry > >if you setup atomic positions improperly. This will > >result in incorrect total energy. > > > >Actually I used your input file, except > >Pseudopotentials name, but did not observe any message > >like you mentioned (more strictly, no message). So, > >your input file looks correct. Check your own file in > >your computer whether it contains any extra symbol at > >the end of a line (usually ^M due to DOS, Windows). > >Use "vi" editor for this purpose. > >It might also be you need an extra blank at the > >beginning of these lines. > > > >Bests, > >Eyvaz. > > > > > >>Here is my input: > >> > >>LiH > >>&CONTROL > >> calculation = 'scf' > >> restart_mode = 'from_scratch' > >> prefix = 'lih' > >> pseudo_dir = '/workspace/lev/PWSCF/LiH' > >> outdir = '/workspace/lev/PWSCF/LiH' > >> iprint = 2 > >> verbosity = 'high' > >> tprnfor = .true. > >>/ > >>&SYSTEM > >> ibrav = 0, > >> celldm(1) = 7.5557540587, > >> nat = 2, > >> ntyp = 2, > >> ecutwfc = 70.0, > >> occupations = 'fixed', > >> nosym = .true. > >>/ > >>&ELECTRONS > >> conv_thr = 1.D-6, > >> mixing_beta = 0.3D0, > >>/ > >>ATOMIC_SPECIES > >>Li 1.0 li-3.ncpp > >>H 1.0 h.ncpp > >>ATOMIC_POSITIONS > >>Li 0.5 0.5 0.5 > >>H 0.6 0.0 0.0 > >>K_POINTS {automatic} > >>4 4 4 0 0 0 > >>CELL_PARAMETERS > >>0.0 0.5 0.5 > >>0.5 0.0 0.5 > >>0.5 0.5 0.0 > >> > >>Thanks! > >> > >>Sergey. > >> > >>On 9/11/07, Eyvaz Isaev > >>wrote: > >> > >> > >>>Hi Sergey, > >>> > >>>It seems something wrong in your input file. Could > >>> > >>> > >>you > >> > >> > >>>please post your input file? > >>>I can suggest that you specify only ireeducible > >>> > >>> > >>atomic > >> > >> > >>>positions while all atomic positions have to be > >>>listed. > >>> > >>>Bests, > >>>Eyvaz. > >>> > >>>--- Sergey Levchenko wrote: > >>> > >>> > >>> > >>>>Hi! > >>>> > >>>>I've encountered a peculiar problem. I want to > >>>> > >>>> > >>run > >> > >> > >>>>PWSCF 3.2 for the > >>>>following atomic system: > >>>> > >>>>Li 0.5 0.5 0.5 > >>>>H 0.5 0.0 0.0 > >>>> > >>>>lattice vectors (fcc): > >>>> > >>>>0.0 0.5 0.5 > >>>>0.5 0.0 0.5 > >>>>0.5 0.5 0.0 > >>>> > >>>>alat=7.555754 bohr > >>>> > >>>>I specify ibrav = 0, and nosym true. But the > >>>> > >>>> > >>program > >> > >> > >>>>crashes right at the > >>>>beginning with the message: > >>>> > >>>>from setup : error # 1 > >>>>Wrong atomic coordinates > >>>> > >>>>I get the same message with ibrav=2. > >>>> > >>>>What could be the problem? > >>>> > >>>>Thank you. > >>>> > >>>>Sincerely, > >>>> > >>>>Sergey Levchenko, > >>>>University of Pennsylvania. > >>>> > >>>> > >>_______________________________________________ > >> > >> > >>>>Pw_forum mailing list > >>>>Pw_forum at pwscf.org > >>>> > >>>> > >>>> > >>http://www.democritos.it/mailman/listinfo/pw_forum > >> > >> > >>> > >>> > >>> > >------------------------------------------------------------------- > > > > > >>>Prof. Eyvaz Isaev, > >>>Theoretical Physics Department, Moscow State > >>> > >>> > >>Institute of Steel & Alloys, > >> > >> > >>>Russia, and > >>>Condensed Matter Theory Group, Uppsala University, > >>> > >>> > >>Sweden > >> > >> > >>>Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com > >>> > >>> > >>> > >>> > >>> > >>> > >>> > > >____________________________________________________________________________________ > > > > > >>>Got a little couch potato? > >>>Check out fun summer activities for kids. > >>> > >>> > >>> > >>> > > > http://search.yahoo.com/search?fr=oni_on_mail&p=summer+activities+for+kids&cs=bz > > > > > >>>_______________________________________________ > >>>Pw_forum mailing list > >>>Pw_forum at pwscf.org > >>>http://www.democritos.it/mailman/listinfo/pw_forum > >>> > >>>_______________________________________________ > >>> > >>> > >>Pw_forum mailing list > >>Pw_forum at pwscf.org > >>http://www.democritos.it/mailman/listinfo/pw_forum > >> > >> > >> > > > > > >------------------------------------------------------------------- > >Prof. Eyvaz Isaev, > >Theoretical Physics Department, Moscow State Institute of Steel & Alloys, > Russia, and > >Condensed Matter Theory Group, Uppsala University, Sweden > Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com > > > > > > > > >____________________________________________________________________________________ > >Pinpoint customers who are looking for what you sell. > >http://searchmarketing.yahoo.com/ > >_______________________________________________ > >Pw_forum mailing list > >Pw_forum at pwscf.org > >http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070912/a5ecd400/attachment-0001.htm From degironc at sissa.it Wed Sep 12 15:45:05 2007 From: degironc at sissa.it (degironc) Date: Wed, 12 Sep 2007 15:45:05 +0200 Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: References: <715563.24301.qm@web60313.mail.yahoo.com> <46E7AAC0.7020906@sissa.it> Message-ID: <46E7ED61.7040101@sissa.it> Dear Sergey, 0.5 0.5 0.5 and 0.5 0.0 0.0 are connected by a Bravais lattice vector. You are putting two atoms in the same position. stefano Sergey Levchenko wrote: > Dear friends, > > I am terribly sorry, I made a mistake in my input file. With the > coordinates specified in that file (Li 0.5 0.5 0.5, H 0.6 0.0 0.0) I > also don't get the error message. The problem seems to occur only for > atomic coordinates (Li 0.5 0.5 0.5, H 0.5 0.0 0.0). > > Thank you for your patience. > > Sincerely, > > Sergey Levchenko, > University of Penn. > > > On 9/12/07, *Xunlei Ding* > wrote: > > Dear Eyvaz, > > Yes, I did the same test as you, and no error like "Wrong atomic > coordinates" was found. > But when I used xcrysten to see the input file, > it was found that the distance between H and Li is about 0.4A. > I am not sure whether some versions of espresso will think this too > close distance as "Wrong atomic coordinates". > > My suggestion to Sergey is: > at first, check what Eyvaz has said , ie "Check your own file in your > computer whether it contains any extra symbol at the end of a line > (usually ^M due to DOS, Windows).". > If no problem, then add "crystal" to the line "ATOMIC_POSITIONS" > to see > what will happen. > > Best regards, > Ding > > Eyvaz Isaev wrote: > > >Hi Sergei, > > > >--- Sergey Levchenko < levuser at gmail.com > > wrote: > > > > > >>I think I know what my mistake was. I specified > >>atomic_positions card with default option which is > > >> > >> > >alat, but that is still in cartesian coordinates, > > > > > >>right? I wanted to specify in the basis of the > >>lattice vectors. Do I have to use option crystal? > >> > >> > > > >No, the error is not related to specifing atomic > >coordinates either in cartesian or wrt lattice > >vectors. Of course, you will find incorrect symmetry > >if you setup atomic positions improperly. This will > >result in incorrect total energy. > > > >Actually I used your input file, except > >Pseudopotentials name, but did not observe any message > >like you mentioned (more strictly, no message). So, > >your input file looks correct. Check your own file in > >your computer whether it contains any extra symbol at > >the end of a line (usually ^M due to DOS, Windows). > >Use "vi" editor for this purpose. > >It might also be you need an extra blank at the > >beginning of these lines. > > > >Bests, > >Eyvaz. > > > > > >>Here is my input: > >> > >>LiH > >>&CONTROL > >> calculation = 'scf' > >> restart_mode = 'from_scratch' > >> prefix = 'lih' > >> pseudo_dir = '/workspace/lev/PWSCF/LiH' > >> outdir = '/workspace/lev/PWSCF/LiH' > >> iprint = 2 > >> verbosity = 'high' > >> tprnfor = .true. > >>/ > >>&SYSTEM > >> ibrav = 0, > >> celldm(1) = 7.5557540587, > >> nat = 2, > >> ntyp = 2, > >> ecutwfc = 70.0, > >> occupations = 'fixed', > >> nosym = .true. > >>/ > >>&ELECTRONS > >> conv_thr = 1.D-6, > >> mixing_beta = 0.3D0, > >>/ > >>ATOMIC_SPECIES > >>Li 1.0 li-3.ncpp > >>H 1.0 h.ncpp > >>ATOMIC_POSITIONS > >>Li 0.5 0.5 0.5 > >>H 0.6 0.0 0.0 > >>K_POINTS {automatic} > >>4 4 4 0 0 0 > >>CELL_PARAMETERS > >>0.0 0.5 0.5 > >>0.5 0.0 0.5 > >>0.5 0.5 0.0 > >> > >>Thanks! > >> > >>Sergey. > >> > >>On 9/11/07, Eyvaz Isaev < eyvaz_isaev at yahoo.com > > > >>wrote: > >> > >> > >>>Hi Sergey, > >>> > >>>It seems something wrong in your input file. Could > >>> > >>> > >>you > >> > >> > >>>please post your input file? > >>>I can suggest that you specify only ireeducible > >>> > >>> > >>atomic > >> > >> > >>>positions while all atomic positions have to be > >>>listed. > >>> > >>>Bests, > >>>Eyvaz. > >>> > >>>--- Sergey Levchenko < levuser at gmail.com > > wrote: > >>> > >>> > >>> > >>>>Hi! > >>>> > >>>>I've encountered a peculiar problem. I want to > >>>> > >>>> > >>run > >> > >> > >>>>PWSCF 3.2 for the > >>>>following atomic system: > >>>> > >>>>Li 0.5 0.5 0.5 > >>>>H 0.5 0.0 0.0 > >>>> > >>>>lattice vectors (fcc): > >>>> > >>>>0.0 0.5 0.5 > >>>>0.5 0.0 0.5 > >>>>0.5 0.5 0.0 > >>>> > >>>>alat= 7.555754 bohr > >>>> > >>>>I specify ibrav = 0, and nosym true. But the > >>>> > >>>> > >>program > >> > >> > >>>>crashes right at the > >>>>beginning with the message: > >>>> > >>>>from setup : error # 1 > >>>>Wrong atomic coordinates > >>>> > >>>>I get the same message with ibrav=2. > >>>> > >>>>What could be the problem? > >>>> > >>>>Thank you. > >>>> > >>>>Sincerely, > >>>> > >>>>Sergey Levchenko, > >>>>University of Pennsylvania. > >>>> > >>>> > >>_______________________________________________ > >> > >> > >>>>Pw_forum mailing list > >>>> Pw_forum at pwscf.org > >>>> > >>>> > >>>> > >>http://www.democritos.it/mailman/listinfo/pw_forum > > >> > >> > >>> > >>> > >>> > >------------------------------------------------------------------- > > > > > >>>Prof. Eyvaz Isaev, > >>>Theoretical Physics Department, Moscow State > >>> > >>> > >>Institute of Steel & Alloys, > >> > >> > >>>Russia, and > >>>Condensed Matter Theory Group, Uppsala University, > >>> > >>> > >>Sweden > >> > >> > >>>Eyvaz.Isaev at fysik.uu.se , > eyvaz_isaev at yahoo.com > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >____________________________________________________________________________________ > > > > > >>>Got a little couch potato? > >>>Check out fun summer activities for kids. > >>> > >>> > >>> > >>> > > > http://search.yahoo.com/search?fr=oni_on_mail&p=summer+activities+for+kids&cs=bz > > > > > > >>>_______________________________________________ > >>>Pw_forum mailing list > >>> Pw_forum at pwscf.org > >>>http://www.democritos.it/mailman/listinfo/pw_forum > >>> > >>>_______________________________________________ > >>> > >>> > >>Pw_forum mailing list > >>Pw_forum at pwscf.org > >>http://www.democritos.it/mailman/listinfo/pw_forum > > >> > >> > >> > > > > > >------------------------------------------------------------------- > >Prof. Eyvaz Isaev, > >Theoretical Physics Department, Moscow State Institute of Steel & > Alloys, Russia, and > >Condensed Matter Theory Group, Uppsala University, Sweden > Eyvaz.Isaev at fysik.uu.se , > eyvaz_isaev at yahoo.com > > > > > > > >____________________________________________________________________________________ > >Pinpoint customers who are looking for what you sell. > >http://searchmarketing.yahoo.com/ > >_______________________________________________ > >Pw_forum mailing list > >Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From eyvaz_isaev at yahoo.com Wed Sep 12 15:51:44 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 12 Sep 2007 06:51:44 -0700 (PDT) Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: <46E7AAC0.7020906@sissa.it> Message-ID: <664749.85017.qm@web60317.mail.yahoo.com> Hi Ding, --- Xunlei Ding wrote: > Yes, I did the same test as you, and no error >like "Wrong atomic coordinates" was found. But >when I used xcrysten to see the input file, it was >found that the distance between H and Li is about >0.4A. I am not sure whether some versions of >espresso will think this too close distance as >"Wrong atomic coordinates". As on can see from the 4 lines provided below setup.f90 actually checks whether two atoms have the same coordinates. But what you suggested is not the case. ##### from setup.f90 ! ... Test that atoms do not overlap ! IF ( .NOT. ( lchk_tauxk( nat, tau, bg ) ) ) & CALL errore( 'setup', 'Wrong atomic coordinates ', 1 ) If atoms are not overlapped (i.e. have different coordinates) then lchk_tauxk is true. Bests, Eyvaz. > > My suggestion to Sergey is: > at first, check what Eyvaz has said , ie "Check your > own file in your > computer whether it contains any extra symbol at the > end of a line > (usually ^M due to DOS, Windows).". > If no problem, then add "crystal" to the line > "ATOMIC_POSITIONS" to see > what will happen. > > Best regards, > Ding > > Eyvaz Isaev wrote: > > >Hi Sergei, > > > >--- Sergey Levchenko wrote: > > > > > >>I think I know what my mistake was. I specified > >>atomic_positions card with default option which is > > > >> > >> > >alat, but that is still in cartesian coordinates, > > > > > >>right? I wanted to specify in the basis of the > >>lattice vectors. Do I have to use option crystal? > >> > >> > > > >No, the error is not related to specifing atomic > >coordinates either in cartesian or wrt lattice > >vectors. Of course, you will find incorrect > symmetry > >if you setup atomic positions improperly. This will > >result in incorrect total energy. > > > >Actually I used your input file, except > >Pseudopotentials name, but did not observe any > message > >like you mentioned (more strictly, no message). So, > >your input file looks correct. Check your own file > in > >your computer whether it contains any extra symbol > at > >the end of a line (usually ^M due to DOS, Windows). > >Use "vi" editor for this purpose. > >It might also be you need an extra blank at the > >beginning of these lines. > > > >Bests, > >Eyvaz. > > > > > >>Here is my input: > >> > >>LiH > >>&CONTROL > >> calculation = 'scf' > >> restart_mode = 'from_scratch' > >> prefix = 'lih' > >> pseudo_dir = '/workspace/lev/PWSCF/LiH' > >> outdir = '/workspace/lev/PWSCF/LiH' > >> iprint = 2 > >> verbosity = 'high' > >> tprnfor = .true. > >>/ > >>&SYSTEM > >> ibrav = 0, > >> celldm(1) = 7.5557540587, > >> nat = 2, > >> ntyp = 2, > >> ecutwfc = 70.0, > >> occupations = 'fixed', > >> nosym = .true. > >>/ > >>&ELECTRONS > >> conv_thr = 1.D-6, > >> mixing_beta = 0.3D0, > >>/ > >>ATOMIC_SPECIES > >>Li 1.0 li-3.ncpp > >>H 1.0 h.ncpp > >>ATOMIC_POSITIONS > >>Li 0.5 0.5 0.5 > >>H 0.6 0.0 0.0 > >>K_POINTS {automatic} > >>4 4 4 0 0 0 > >>CELL_PARAMETERS > >>0.0 0.5 0.5 > >>0.5 0.0 0.5 > >>0.5 0.5 0.0 > >> > >>Thanks! > >> > >>Sergey. > >> > >>On 9/11/07, Eyvaz Isaev > >>wrote: > >> > >> > >>>Hi Sergey, > >>> > >>>It seems something wrong in your input file. > Could > >>> > >>> > >>you > >> > >> > >>>please post your input file? > >>>I can suggest that you specify only ireeducible > >>> > >>> > >>atomic > >> > >> > >>>positions while all atomic positions have to be > >>>listed. > >>> > >>>Bests, > >>>Eyvaz. > >>> > >>>--- Sergey Levchenko wrote: > >>> > >>> > >>> > >>>>Hi! > >>>> > >>>>I've encountered a peculiar problem. I want to > >>>> > >>>> > >>run > >> > >> > >>>>PWSCF 3.2 for the > >>>>following atomic system: > >>>> > >>>>Li 0.5 0.5 0.5 > >>>>H 0.5 0.0 0.0 > >>>> > >>>>lattice vectors (fcc): > >>>> > >>>>0.0 0.5 0.5 > >>>>0.5 0.0 0.5 > >>>>0.5 0.5 0.0 > >>>> > >>>>alat=7.555754 bohr > >>>> > >>>>I specify ibrav = 0, and nosym true. But the > >>>> > >>>> > >>program > >> > >> > >>>>crashes right at the > >>>>beginning with the message: > >>>> > >>>>from setup : error # 1 > >>>>Wrong atomic coordinates > >>>> > >>>>I get the same message with ibrav=2. > >>>> > >>>>What could be the problem? > >>>> > >>>>Thank you. > >>>> > >>>>Sincerely, > >>>> > >>>>Sergey Levchenko, > >>>>University of Pennsylvania. > >>>> > >>>> > >>_______________________________________________ > >> > >> > >>>>Pw_forum mailing list > >>>>Pw_forum at pwscf.org > >>>> > >>>> > >>>> > >>http://www.democritos.it/mailman/listinfo/pw_forum > >> > >> > >>> > >>> > === message truncated === ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Check out the hottest 2008 models today at Yahoo! Autos. http://autos.yahoo.com/new_cars.html From eyvaz_isaev at yahoo.com Wed Sep 12 16:28:36 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Wed, 12 Sep 2007 07:28:36 -0700 (PDT) Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: Message-ID: <786830.44196.qm@web60313.mail.yahoo.com> Dear Sergei, --- Sergey Levchenko wrote: > Dear friends, > > I am terribly sorry, I made a mistake in my input > file. With the coordinates> specified in that file >(Li 0.5 0.5 0.5, H 0.6 0.00.0) I also don't >get the error message. The problem seems to occur >only for atomic coordinates (Li > 0.5 0.5 0.5, H 0.5 0.0 0.0). Exactly, because these two positions are equivalent in combination with basis vectors you used. Just draw an FCC lattice (cubic cell) on a piece of a paper and you immediately can see that they are equivalent if you shift atom (0,0,1/2) along [110] direction. If you use basis vectors for simple cubic lattice, no error message appeares. So, check your atomic positions again or basis vectors. Bests, Eyvaz. > > Thank you for your patience. > > Sincerely, > > Sergey Levchenko, > University of Penn. > > > On 9/12/07, Xunlei Ding wrote: > > > > Dear Eyvaz, > > > > Yes, I did the same test as you, and no error like > "Wrong atomic > > coordinates" was found. > > But when I used xcrysten to see the input file, > > it was found that the distance between H and Li is > about 0.4A. > > I am not sure whether some versions of espresso > will think this too > > close distance as "Wrong atomic coordinates". > > > > My suggestion to Sergey is: > > at first, check what Eyvaz has said , ie "Check > your own file in your > > computer whether it contains any extra symbol at > the end of a line > > (usually ^M due to DOS, Windows).". > > If no problem, then add "crystal" to the line > "ATOMIC_POSITIONS" to see > > what will happen. > > > > Best regards, > > Ding > > > > Eyvaz Isaev wrote: > > > > >Hi Sergei, > > > > > >--- Sergey Levchenko wrote: > > > > > > > > >>I think I know what my mistake was. I specified > > >>atomic_positions card with default option which > is > > > >> > > >> > > >alat, but that is still in cartesian coordinates, > > > > > > > > >>right? I wanted to specify in the basis of the > > >>lattice vectors. Do I have to use option > crystal? > > >> > > >> > > > > > >No, the error is not related to specifing atomic > > >coordinates either in cartesian or wrt lattice > > >vectors. Of course, you will find incorrect > symmetry > > >if you setup atomic positions improperly. This > will > > >result in incorrect total energy. > > > > > >Actually I used your input file, except > > >Pseudopotentials name, but did not observe any > message > > >like you mentioned (more strictly, no message). > So, > > >your input file looks correct. Check your own > file in > > >your computer whether it contains any extra > symbol at > > >the end of a line (usually ^M due to DOS, > Windows). > > >Use "vi" editor for this purpose. > > >It might also be you need an extra blank at the > > >beginning of these lines. > > > > > >Bests, > > >Eyvaz. > > > > > > > > >>Here is my input: > > >> > > >>LiH > > >>&CONTROL > > >> calculation = 'scf' > > >> restart_mode = 'from_scratch' > > >> prefix = 'lih' > > >> pseudo_dir = '/workspace/lev/PWSCF/LiH' > > >> outdir = '/workspace/lev/PWSCF/LiH' > > >> iprint = 2 > > >> verbosity = 'high' > > >> tprnfor = .true. > > >>/ > > >>&SYSTEM > > >> ibrav = 0, > > >> celldm(1) = 7.5557540587, > > >> nat = 2, > > >> ntyp = 2, > > >> ecutwfc = 70.0, > > >> occupations = 'fixed', > > >> nosym = .true. > > >>/ > > >>&ELECTRONS > > >> conv_thr = 1.D-6, > > >> mixing_beta = 0.3D0, > > >>/ > > >>ATOMIC_SPECIES > > >>Li 1.0 li-3.ncpp > > >>H 1.0 h.ncpp > > >>ATOMIC_POSITIONS > > >>Li 0.5 0.5 0.5 > > >>H 0.6 0.0 0.0 > > >>K_POINTS {automatic} > > >>4 4 4 0 0 0 > > >>CELL_PARAMETERS > > >>0.0 0.5 0.5 > > >>0.5 0.0 0.5 > > >>0.5 0.5 0.0 > > >> > > >>Thanks! > > >> > > >>Sergey. > > >> > > >>On 9/11/07, Eyvaz Isaev > > >>wrote: > > >> > > >> > > >>>Hi Sergey, > > >>> > > >>>It seems something wrong in your input file. > Could > > >>> > > >>> > > >>you > > >> > > >> > > >>>please post your input file? > > >>>I can suggest that you specify only ireeducible > > >>> > > >>> > > >>atomic > > >> > > >> > > >>>positions while all atomic positions have to be > > >>>listed. > > >>> > > >>>Bests, > > >>>Eyvaz. > > >>> > > >>>--- Sergey Levchenko wrote: > > >>> > > >>> > > >>> > > >>>>Hi! > > >>>> > > >>>>I've encountered a peculiar problem. I want to > > >>>> > > >>>> > > >>run > > >> > > >> > > >>>>PWSCF 3.2 for the > > >>>>following atomic system: > > >>>> > > >>>>Li 0.5 0.5 0.5 > > >>>>H 0.5 0.0 0.0 > > >>>> > > >>>>lattice vectors (fcc): > > >>>> > > >>>>0.0 0.5 0.5 > > >>>>0.5 0.0 0.5 > > >>>>0.5 0.5 0.0 > > >>>> > > >>>>alat=7.555754 bohr > > >>>> > > >>>>I specify ibrav = 0, and nosym true. But the > > >>>> > > >>>> > > >>program > > >> > > >> > > >>>>crashes right at the > > >>>>beginning with the message: > > >>>> > > >>>>from setup : error # 1 > > >>>>Wrong atomic coordinates > > >>>> > > >>>>I get the same message with ibrav=2. > === message truncated ===> _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Catch up on fall's hot new shows on Yahoo! TV. Watch previews, get listings, and more! http://tv.yahoo.com/collections/3658 From xianghjun at gmail.com Thu Sep 13 00:41:02 2007 From: xianghjun at gmail.com (Hongjun Xiang) Date: Wed, 12 Sep 2007 16:41:02 -0600 Subject: [Pw_forum] How to calculate unoccupied states using cp.x In-Reply-To: <389095204.26878@ustc.edu.cn> References: <46E02461.6030104@gmail.com> <389095204.26878@ustc.edu.cn> Message-ID: <46E86AFE.1050609@gmail.com> Dear all, I would like to obtain the band gap of a system. As is well-known, pw.x can give eigenvalues of unoccupied states. When I tried to use cp.x with the "damp" method to calculate the band gap of water molecule by increasing the nbnd value to 8, all eigenvalues of unoccupied states are zero. My input file: =============================== # Electronic damped dynamics &CONTROL title = ' Water Molecule ', calculation = 'cp', restart_mode = 'from_scratch', ndr = 51, ndw = 51, nstep = 100, iprint = 10, isave = 100, tstress = .FALSE., tprnfor = .TRUE., dt = 4.0d0, etot_conv_thr = 1.d-9, ekin_conv_thr = 1.d-4, prefix = 'h2o_mol' pseudo_dir = './', outdir = './tmp/' / &SYSTEM ibrav = 14, celldm(1) = 10.0, celldm(2) = 1.0, celldm(3) = 1.0, celldm(4) = 0.0, celldm(5) = 0.0, celldm(6) = 0.0, nat = 3, ntyp = 2, ecutwfc = 70.0, nbnd=8 / &ELECTRONS emass = 300.d0, emass_cutoff = 2.5d0, orthogonalization = 'ortho', electron_dynamics = 'damp', / &IONS ion_dynamics = 'none', ion_temperature = 'not_controlled', / ATOMIC_SPECIES O 16.0d0 O.BLYP.UPF H 1.00d0 H.fpmd.UPF ATOMIC_POSITIONS (bohr) O 5.0099 5.0099 5.0000 0 0 0 H 6.8325 4.7757 4.9999 1 1 1 H 4.7757 6.8325 4.9998 1 1 1 ================================================== The output file reads: ... Eigenvalues (eV), kp = 1 , spin = 1 -24.48 -12.17 -8.43 -6.36 0.00 0.00 0.00 0.00 ... ================================================= I also tried the CG method: &ELECTRONS electron_dynamics = 'damp', tcg=.true. / ====================================== The output file in case of nbnd=8 reads: ... Eigenvalues (eV), kp = 1 , spin = 1 -24.35 -12.22 -8.40 -6.38 0.00 0.00 0.00 0.00 ... ====================================== The output file in case of nbnd=4 reads: ... Eigenvalues (eV), kp = 1 , spin = 1 -24.48 -12.18 -8.43 -6.36 ... ===================================== It seems that when nbnd is larger than the number of the occupied states, CG gives even wrongs occupied eigenvalues. Would someone please tell me if it is possible to obtain correct eigenvalues of unoccupied states using cp.x? If yes, how? P.S., I am using espresso-3.2. Thank you very much. Best regards, Hongjun Xiang ============================================================= H. J. Xiang Postdoctoral Research Associate National Renewable Energy Laboratory http://www4.ncsu.edu/~hxiang/ ============================================================= From marzari at MIT.EDU Thu Sep 13 01:39:51 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 12 Sep 2007 19:39:51 -0400 Subject: [Pw_forum] How to calculate unoccupied states using cp.x In-Reply-To: <46E86AFE.1050609@gmail.com> References: <46E02461.6030104@gmail.com> <389095204.26878@ustc.edu.cn> <46E86AFE.1050609@gmail.com> Message-ID: <46E878C7.9030301@mit.edu> > Would someone please tell me if it is possible to obtain correct > eigenvalues of unoccupied states using cp.x? If yes, how? > > P.S., I am using espresso-3.2. > > Thank you very much. > > Best regards, > Hongjun Xiang > I haven't done this in while, but, broadly speaking, these should be the two possibilities (I suppose you are looking at an insulator - actually, let's consider the water molecule - i.e. 8 electrons, four bands). 1) You do a ground state calculation with CP, using 4 bands. You get the right ground state, the right occupied eigenvalues. You restart the code (option 'nscf') with a larger, arbitrary number of states. In 'nscf' the charge density is not updated (so your Hamiltonian is frozen forever), and you minimize with damped CP dynamics a total energy functional in which the Hamiltonian is the frozen one above, and in which both the occupied and the unoccupied states are fully occupied. This gives you correctly both the occupied and unoccupied eigenstates and eigenvalues (if you think at it, you are using the correct H[frozen rho]|psi_i> to evolve all the psi_i to the overall ground state). In standard CP wouldn't work basically for the same reasons you cannot deal with a metal. The ortho constraint mixes all the states in the manifold together, and there is no dynamics on the occupation numbers (or matrix), while the system is not invariant for a unitary transformation in the extended manifold (empty + full). 2) a second option, for which you need the latest CVS, is to use CP in "ensemble-DFT" mode (there is a PRL of ours from 1997, and Paolo Umari has recently finished coding all the various parts). Basically, this is composed of a conj-grad minimization on the orbitals (this can also be used for pure insulators, and it is faster in reaching the ground state than damped dynamics) and an optimization of the occupation matrix. In practice, it evolves orbitals and occupations at the same time, so it would allow you to do a single calculation with filled and empty states at the same time. For an insulator, it is probably not worth, and reciped 1) is more efficient. For a metal, it is the only possibility, although it will still struggle (I think) to give very high, completely empty orbitals correctly (since those do not affect the total energy you are minimizing - we have a precondition on the occupations to speed things up). Note that only the 'cs' smearing option works, at this stage, and the code requires a bit of insider infos to work smoothly. We'll improve the documentation. Bottom line - for an insulator, especially if you need a lot of empty states, use recipe 1). For a metal, use recipe 2) to converge to the ground state. If you need only few empty orbitals, 2) will probably give them correctly already. If you need a lot, use 2) with only a few empty ones, and then run 1) after 2), with option 'nscf' and a lot of states (that will be all filled up, and optimized properly). Best luck - let us know, nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From raveiga at yahoo.com Thu Sep 13 04:12:01 2007 From: raveiga at yahoo.com (Roberto G. A. Veiga) Date: Wed, 12 Sep 2007 19:12:01 -0700 (PDT) Subject: [Pw_forum] "Wrong atomic coordinates" message Message-ID: <322506.65900.qm@web33502.mail.mud.yahoo.com> Hello, all: I'm trying to do a variable cell geometry optimization of a carbon nanotube with "pw.x". My input follows: --------------------------------------------------------------------------------------- &CONTROL title = 'nanotubo' , calculation = 'vc-relax' , restart_mode = 'from_scratch' , wf_collect = .true. , outdir = '~/pw-calc/nanotubos/8x0/outdir/' , wfcdir = '~/pw-calc/nanotubos/8x0/wfcdir/' , pseudo_dir = '~/pw-calc/pseudo/' , prefix = 'nanotubo' , disk_io = 'default' , verbosity = 'default' , etot_conv_thr = 1D-8 , forc_conv_thr = 1D-4 , ekin_conv_thr = 1D-4 , nstep = 200 , / &SYSTEM ibrav = 8, A = 4.32 , B = 13.00 , C = 16.0 , nat = 64, ntyp = 1, ecutwfc = 40.0 , / &ELECTRONS electron_maxstep = 100, conv_thr = 1D-4 , diagonalization = 'cg' , diago_cg_maxiter = 200, / &IONS ion_dynamics = 'damp' , / &CELL cell_dynamics = 'damp-pr' , wmass = 400 , / ATOMIC_SPECIES C 12.01070 C.pbe.uspp ATOMIC_POSITIONS angstrom C 0.0000000 0.0000000 3.1756544 C 0.7200000 1.2152703 2.9339221 C 0.0000000 2.2455268 2.2455268 C 0.7200000 2.9339221 1.2152703 C 0.0000000 3.1756544 0.0000000 C 0.7200000 2.9339221 1.2152703 C 0.0000000 2.2455268 2.2455268 C 0.7200000 1.2152703 2.9339221 C 0.0000000 0.0000000 3.1756544 C 0.7200000 -1.2152703 2.9339221 C 0.0000000 -2.2455268 2.2455268 C 0.7200000 -2.9339221 1.2152703 C 0.0000000 -3.1756544 0.0000000 C 0.7200000 -2.9339221 1.2152703 C 0.0000000 -2.2455268 2.2455268 C 0.7200000 -1.2152703 2.9339221 C 2.1600000 1.2152703 2.9339221 C 2.8800000 2.2455268 2.2455268 C 2.1600000 2.9339221 1.2152703 C 2.8800000 3.1756544 0.0000000 C 2.1600000 2.9339221 1.2152703 C 2.8800000 2.2455268 2.2455268 C 2.1600000 1.2152703 2.9339221 C 2.8800000 0.0000000 3.1756544 C 2.1600000 -1.2152703 2.9339221 C 2.8800000 -2.2455268 2.2455268 C 2.1600000 -2.9339221 1.2152703 C 2.8800000 -3.1756544 0.0000000 C 2.1600000 -2.9339221 1.2152703 C 2.8800000 -2.2455268 2.2455268 C 2.1600000 -1.2152703 2.9339221 C 2.8800000 0.0000000 3.1756544 C 4.3200000 2.2455268 2.2455268 C 5.0400000 2.9339221 1.2152703 C 4.3200000 3.1756544 0.0000000 C 5.0400000 2.9339221 1.2152703 C 4.3200000 2.2455268 2.2455268 C 5.0400000 1.2152703 2.9339221 C 4.3200000 0.0000000 3.1756544 C 5.0400000 -1.2152703 2.9339221 C 4.3200000 -2.2455268 2.2455268 C 5.0400000 -2.9339221 1.2152703 C 4.3200000 -3.1756544 0.0000000 C 5.0400000 -2.9339221 1.2152703 C 4.3200000 -2.2455268 2.2455268 C 5.0400000 -1.2152703 2.9339221 C 4.3200000 0.0000000 3.1756544 C 5.0400000 1.2152703 2.9339221 C 6.4800000 2.9339221 1.2152703 C 7.2000000 3.1756544 0.0000000 C 6.4800000 2.9339221 1.2152703 C 7.2000000 2.2455268 2.2455268 C 6.4800000 1.2152703 2.9339221 C 7.2000000 0.0000000 3.1756544 C 6.4800000 -1.2152703 2.9339221 C 7.2000000 -2.2455268 2.2455268 C 6.4800000 -2.9339221 1.2152703 C 7.2000000 -3.1756544 0.0000000 C 6.4800000 -2.9339221 1.2152703 C 7.2000000 -2.2455268 2.2455268 C 6.4800000 -1.2152703 2.9339221 C 7.2000000 0.0000000 3.1756544 C 6.4800000 1.2152703 2.9339221 C 7.2000000 2.2455268 2.2455268 K_POINTS automatic 8 1 1 0 0 0 --------------------------------------------------------------------------------------- With this input, I'm getting the following error message: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from setup : error # 1 Wrong atomic coordinates %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% What's going on? Regards, Roberto --------------------------------- Yahoo! oneSearch: Finally, mobile search that gives answers, not web links. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070912/4c85ea28/attachment.htm From marzari at MIT.EDU Thu Sep 13 05:14:33 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 12 Sep 2007 23:14:33 -0400 Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: <322506.65900.qm@web33502.mail.mud.yahoo.com> References: <322506.65900.qm@web33502.mail.mud.yahoo.com> Message-ID: <46E8AB19.2050809@mit.edu> Dear Roberto, it's *just* been discussed today on the mailing list. Alternatively, google it, and you'll find ! ... Test that atoms do not overlap ! IF ( .NOT. ( lchk_tauxk( nat, tau, bg ) ) ) & CALL errore( 'setup', 'Wrong atomic coordinates ', 1 ) We are all happy to help, but it's truly a matter of courtesy trying to avoid using the bandwidth and attention of everyone. Maybe you have discovered an important bug, but then say so (i.e. that the atoms do not overlap). nicola PS: as a second matter of courtesy, an academic/research/professional signature is asked for in our postings. I believe everyone that subscribes to the mailing list now receives this alert - let us know if you didn't receive it, and we'll fix it > With this input, I'm getting the following error message: > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from setup : error # 1 > Wrong atomic coordinates > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > What's going on? > > Regards, > > Roberto -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From hyello25 at gmail.com Thu Sep 13 06:23:14 2007 From: hyello25 at gmail.com (hyello25 Qgrui) Date: Thu, 13 Sep 2007 12:23:14 +0800 Subject: [Pw_forum] the problem of scf convergence Message-ID: hello, everyone. These days, I performed the convergence of scf calculation with some USPP, but never see a good ecut convergence. The system only contains 3 atoms, with 12x12x12 kpoints automatic generated in MP grids. Why does the press differ each so mush? Is 200ev Ry too high for USPP? And which press is more convictional? Can anyone give me some suggestion? Thanks in advance. ecut=80 : ! total energy = -61.89999202 Ry total stress (Ry/bohr**3) (kbar) P= 0.59 ecut=100 : ! total energy = -61.90651146 Ry total stress (Ry/bohr**3) (kbar) P= - 5.39 ecut=140 : ! total energy = -61.90331161 Ry total stress (Ry/bo hr**3) (kbar) P= 0.97 ecut=180 : ! total energy = -61.90482904 Ry total stress (Ry/bohr**3) (kbar) P= 10.64 ecut=200 : ! total energy = -61.90542701 Ry total stress (Ry/bohr**3) (kbar) P= - 4.55 ecut=240 : ! total energy = -61.90875043 Ry total stress (Ry/bohr**3) (kbar) P= 5.92 ecut=280 : ! total energy = -61.90854533 Ry total stress (Ry/bohr**3) (kbar) P= 1.68 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070913/7ecef9a0/attachment.htm From giannozz at nest.sns.it Thu Sep 13 08:49:48 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 13 Sep 2007 08:49:48 +0200 Subject: [Pw_forum] How to calculate unoccupied states using cp.x In-Reply-To: <46E86AFE.1050609@gmail.com> References: <46E02461.6030104@gmail.com> <389095204.26878@ustc.edu.cn> <46E86AFE.1050609@gmail.com> Message-ID: <03D656EF-A255-401E-88F4-B72A8E655928@nest.sns.it> On Sep 13, 2007, at 24:41 , Hongjun Xiang wrote: > It seems that when nbnd is larger than the number of the occupied > states, > CG gives even wrongs occupied eigenvalues. before stating that eigenvalues are wrong, you should carefully verify that they are well converged. Eigenvalues are more sensitive than total energies to convergence, since the latter is a variational quantity (i.e. has a minimum for the ground state charge density), the former aren't. There is an option (I don't remember which one) that allows calculating empty Kohn-Sham eigenvalues in cp.x . It is described in the input documentation, file Doc/INPUT_CP . Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From baroni at sissa.it Thu Sep 13 08:50:37 2007 From: baroni at sissa.it (Stefano Baroni) Date: Thu, 13 Sep 2007 08:50:37 +0200 Subject: [Pw_forum] the problem of scf convergence In-Reply-To: References: Message-ID: <75EF0C56-8357-427D-BB5A-5FC16C81D1D5@sissa.it> Dear unnamed friend: system? input file? name? affiliation? thanks - Stefano Baroni On Sep 13, 2007, at 6:23 AM, hyello25 Qgrui wrote: > hello, everyone. > These days, I performed the convergence of scf calculation with > some USPP, but never see a good ecut convergence. The system only > contains 3 atoms, with 12x12x12 kpoints automatic generated in MP > grids. > Why does the press differ each so mush? Is 200ev Ry too high for > USPP? And which press is more convictional? > Can anyone give me some suggestion? Thanks in advance. > ecut=80 : > ! total energy = -61.89999202 Ry > total stress (Ry/bohr**3) (kbar) > P= 0.59 > ecut=100 : > ! total energy = -61.90651146 Ry > total stress (Ry/bohr**3) (kbar) > P= -5.39 > ecut=140 : > ! total energy = -61.90331161 Ry > total stress (Ry/bo hr**3) > (kbar) P= 0.97 > ecut=180 : > ! total energy = -61.90482904 Ry > total stress (Ry/bohr**3) (kbar) > P= 10.64 > ecut=200 : > ! total energy = -61.90542701 Ry > total stress (Ry/bohr**3) (kbar) > P= -4.55 > ecut=240 : > ! total energy = -61.90875043 Ry > total stress (Ry/bohr**3) (kbar) > P= 5.92 > ecut=280 : > ! total energy = -61.90854533 Ry > total stress (Ry/bohr**3) (kbar) > P= 1.68 > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070913/09df9107/attachment.htm From giannozz at nest.sns.it Thu Sep 13 09:20:41 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 13 Sep 2007 09:20:41 +0200 Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: <322506.65900.qm@web33502.mail.mud.yahoo.com> References: <322506.65900.qm@web33502.mail.mud.yahoo.com> Message-ID: <2210FA47-306D-4F96-B5C4-F3FBB0827C76@nest.sns.it> On Sep 13, 2007, at 4:12 , Roberto G. A. Veiga wrote: > ibrav = 8, > A = 4.32 , > B = 13.00 , > C = 16.0 simple orthorhombic cell, x side = 4.32A, y side = 13A, z side =16A (Amgstrom, NOT Bohr radii) > C 0.0000000 0.0000000 3.1756544 > .... > C 4.3200000 0.0000000 3.1756544 > from setup : error # 1 > Wrong atomic coordinates of course: there are (at least) two overlapping atoms P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Thu Sep 13 09:27:46 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 13 Sep 2007 09:27:46 +0200 Subject: [Pw_forum] the problem of scf convergence In-Reply-To: References: Message-ID: <72644968-00AD-4567-BF45-E957D66FCB47@nest.sns.it> On Sep 13, 2007, at 6:23 , hyello25 Qgrui wrote: > Is 200ev Ry too high for USPP? "200ev" or "200Ry" ? there are two cutoffs for US PP, one for KS orbitals (ecutrho), one for the augmentation charges (ecutrho >= 4*ecutwfc). 200Ry for the former is a lot, for the latter it is not that much P. --- Paolo Giannozzi, Democritos and University of Udine, Italy From raveiga at yahoo.com Thu Sep 13 14:30:02 2007 From: raveiga at yahoo.com (Roberto G. A. Veiga) Date: Thu, 13 Sep 2007 05:30:02 -0700 (PDT) Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: <2210FA47-306D-4F96-B5C4-F3FBB0827C76@nest.sns.it> Message-ID: <26069.64979.qm@web33505.mail.mud.yahoo.com> Dear Paolo: yes, I saw that. I don't know what happened, since I have had a look at that structure with RASMOL and it was ok. It's likely I made some stupid mistake when I put it inside the PW input file. Anyway, I apologyze for any inconvenience. Regards, Roberto Paolo Giannozzi wrote: On Sep 13, 2007, at 4:12 , Roberto G. A. Veiga wrote: > ibrav = 8, > A = 4.32 , > B = 13.00 , > C = 16.0 simple orthorhombic cell, x side = 4.32A, y side = 13A, z side =16A (Amgstrom, NOT Bohr radii) > C 0.0000000 0.0000000 3.1756544 > .... > C 4.3200000 0.0000000 3.1756544 > from setup : error # 1 > Wrong atomic coordinates of course: there are (at least) two overlapping atoms P. --- Paolo Giannozzi, Democritos and University of Udine, Italy _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum --------------------------------- Luggage? GPS? Comic books? Check out fitting gifts for grads at Yahoo! Search. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070913/6e0340ab/attachment-0001.htm From daijiayu at nudt.edu.cn Thu Sep 13 14:23:20 2007 From: daijiayu at nudt.edu.cn (=?gb2312?B?tPe80e7a?=) Date: Thu, 13 Sep 2007 20:23:20 +0800 Subject: [Pw_forum] a phcg.x error. Message-ID: <389686200.10227@nudt.edu.cn> Hi, everyone, I wanted to use the phcg.x code to calculate the IR and Raman spectra. At first, I used pw.x to do scf calculation. The K_POINTS was put as (gamma). Then, when i run the phcg.x, it told me that: from cg_readin: error double grid not implemented The file *.nm.in is similar to that in example09, and there was no error when i run the example. So, please tell me what is the reason possibly? Thank you. Dai From tone.kokalj at ijs.si Thu Sep 13 14:50:09 2007 From: tone.kokalj at ijs.si (Tone Kokalj) Date: Thu, 13 Sep 2007 14:50:09 +0200 Subject: [Pw_forum] "Wrong atomic coordinates" message In-Reply-To: <26069.64979.qm@web33505.mail.mud.yahoo.com> References: <26069.64979.qm@web33505.mail.mud.yahoo.com> Message-ID: <1189687809.2573.41.camel@localhost.localdomain> On Thu, 2007-09-13 at 05:30 -0700, Roberto G. A. Veiga wrote: > Dear Paolo: > > yes, I saw that. I don't know what happened, since I have had a look > at that structure with RASMOL and it was ok. It's likely I made some > stupid mistake when I put it inside the PW input file. With RASMOL you saw the "molecule": read this as "a chunck of nanotube that is larger than the unitcell". And there is nothing wrong with your molecule. Note that in your input A=4.32, but the x coordinates in your input span the range between 0 -- 7.2 angstroms (why: see above). Therefore either "cut the atoms with large x values" or enlarge the A parameter thus making the supercell. Regards, Tone From giannozz at nest.sns.it Thu Sep 13 15:05:50 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 13 Sep 2007 15:05:50 +0200 Subject: [Pw_forum] a phcg.x error. In-Reply-To: <389686200.10227@nudt.edu.cn> References: <389686200.10227@nudt.edu.cn> Message-ID: <6ACDA50E-AF8D-4D0F-98ED-132D85D6C82B@nest.sns.it> On Sep 13, 2007, at 14:23 , ??? wrote: > when I run the phcg.x, it told me that: > from cg_readin: error > double grid not implemented the phcg.x code works only for norm-conserving pseudopotentials (and it doesn't calculate Raman coefficients). In addition, it doesn't work with double grid (i.e. different FFT grids for Kohn-Sham orbitals and for the charge density) because this is useful only for ultrasoft pseudopotentials. Most likely you are using a cutoff for charge density (ecutrho) that is larger than four times the cutoff for orbitals (ecutwfc). And, by the way: I know who you are, but please let everybody else know by signing your post with a name and an affiliation Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From daijiayu at nudt.edu.cn Thu Sep 13 15:19:51 2007 From: daijiayu at nudt.edu.cn (=?gb2312?B?tPe80e7a?=) Date: Thu, 13 Sep 2007 21:19:51 +0800 Subject: [Pw_forum] a phcg.x error. Message-ID: <389689591.11220@nudt.edu.cn> >the phcg.x code works only for norm-conserving pseudopotentials >(and it doesn't calculate Raman coefficients). In addition, it doesn't >work with double grid (i.e. different FFT grids for Kohn-Sham >orbitals and for the charge density) because this is useful only for >ultrasoft pseudopotentials. Most likely you are using a cutoff for >charge density (ecutrho) that is larger than four times the cutoff >for orbitals (ecutwfc). Thank you for your introduction. And i will correct my input file right now. >And, by the way: I know who you are, but please let everybody >else know by signing your post with a name and an affiliation OK, thank you for your reminding me to set the signing of my e-mail box. I forgot it before. It is the same as that below? :) Jiayu Dai ------------------------------ Jiayu Dai, National University of Defense Technology, P R China From hyello25 at gmail.com Thu Sep 13 15:29:34 2007 From: hyello25 at gmail.com (hyello25 Qgrui) Date: Thu, 13 Sep 2007 21:29:34 +0800 Subject: [Pw_forum] the problem of scf convergence Message-ID: The system is Li2O, and the input file and the result is : ---! convergence scf caculation for Li2O !--- ecut=80 : ! total energy = -61.89999202 Ry total stress (Ry/bohr**3) (kbar) P= 0.59 ecut=100 : ! total energy = -61.90651146 Ry total stress (Ry/bohr**3) (kbar) P= - 5.39 ecut=140 : ! total energy = -61.90331161 Ry total stress (Ry/bohr**3) (kbar) P= 0.97 ecut=180 : ! total energy = -61.90482904 Ry total stress (Ry/bohr**3) (kbar) P= 10.64 ecut=200 : ! total energy = -61.90542701 Ry total stress (Ry/bohr**3) (kbar) P= - 4.55 ecut=240 : ! total energy = -61.90875043 Ry total stress (Ry/bohr**3) (kbar) P= 5.92 ecut=280 : ! total energy = -61.90854533 Ry total stress (Ry/bohr**3) (kbar) P= 1.68 ecut=300 : ! total energy = -61.90926981 Ry total stress (Ry/bohr**3) (kbar) P= - 1.18 ecut=320 : ! total energy = -61.90990283 Ry total stress (Ry/bohr**3) (kbar) P= 2.17 ecut=350 : ! total energy = -61.91013980 Ry total stress (Ry/bohr**3) (kbar) P= 5.51 ecut=380 : ! total energy = -61.91023915 Ry total stress (Ry/bohr**3) (kbar) P= 4.47 --------------------------------------------------------- #!/bin/sh #run from directory where this script is cd `echo $0 | sed 's/\(.*\)\/.*/\1/'` # extract pathname # check whether echo has the -e option if test "`echo -e`" = "-e" ; then ECHO=echo ; else ECHO="echo -e" ; fi touch totResult.out cat > totResult.out < li2O.ecut$i.scf.in < li2O.ecut$i.scf.out cat >> totResult.out<> totResult.out grep -e 'P=' < li2O.ecut$i.scf.out >> totResult.out done $ECHO "------END TIME: " date $ECHO $ECHO -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070913/a02762b1/attachment.htm From akohlmey at cmm.chem.upenn.edu Thu Sep 13 16:00:33 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Thu, 13 Sep 2007 10:00:33 -0400 (EDT) Subject: [Pw_forum] the problem of scf convergence In-Reply-To: Message-ID: On Thu, 13 Sep 2007, hyello25 Qgrui wrote: HQ> The system is Li2O, and the input file and the result is : HQ> HQ> ATOMIC_SPECIES HQ> Li 6.941 Li.pw91-s-van_ak.UPF HQ> O 15.9994 O.pw91-van_ak.UPF please try a different Li pseudopotential (i.e. a non-ak one). the one you are using seems to have some not yet understood problems for higher-order properties: lattice constant from EOS fit, MD and geometry optimizations were fine, phonons were bad. cheers, axel. -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From xianghjun at gmail.com Thu Sep 13 17:49:04 2007 From: xianghjun at gmail.com (Hongjun Xiang) Date: Thu, 13 Sep 2007 09:49:04 -0600 Subject: [Pw_forum] How to calculate unoccupied states using cp.x In-Reply-To: <46E878C7.9030301@mit.edu> References: <46E02461.6030104@gmail.com> <389095204.26878@ustc.edu.cn> <46E86AFE.1050609@gmail.com> <46E878C7.9030301@mit.edu> Message-ID: <46E95BF0.4070807@gmail.com> Dear Prof. Marzari, Thanks a lot for your detailed answer. I tried the first method, it works well. One minor point is that one should use the same "nbnd" in the ground state calculation as that in nscf calculation. Otherwise, it will complain: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from cp_readfile : error # 30 electron, bands or spin do not match %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Regarding the second method, I will try later. Have a good day. Best regards, Hongjun Xiang ============================================================= H. J. Xiang Postdoctoral Research Associate National Renewable Energy Laboratory http://www4.ncsu.edu/~hxiang/ ============================================================= Nicola Marzari wrote: > >> Would someone please tell me if it is possible to obtain correct >> eigenvalues of unoccupied states using cp.x? If yes, how? >> >> P.S., I am using espresso-3.2. >> >> Thank you very much. >> >> Best regards, >> Hongjun Xiang >> > > > I haven't done this in while, but, broadly speaking, these should > be the two possibilities (I suppose you are looking at an insulator - > actually, let's consider the water molecule - i.e. 8 electrons, four > bands). > > 1) You do a ground state calculation with CP, using 4 bands. You get > the right ground state, the right occupied eigenvalues. You restart > the code (option 'nscf') with a larger, arbitrary number of states. > In 'nscf' the charge density is not updated (so your Hamiltonian > is frozen forever), and you minimize with damped CP dynamics > a total energy functional in which the Hamiltonian is the frozen > one above, and in which both the occupied and the unoccupied states are > fully occupied. This gives you correctly both the occupied and > unoccupied eigenstates and eigenvalues (if you think at it, you > are using the correct H[frozen rho]|psi_i> to evolve all the psi_i > to the overall ground state). In standard CP wouldn't work basically > for the same reasons you cannot deal with a metal. The ortho constraint > mixes all the states in the manifold together, and there is no > dynamics on the occupation numbers (or matrix), while the system > is not invariant for a unitary transformation in the extended > manifold (empty + full). > > 2) a second option, for which you need the latest CVS, is to > use CP in "ensemble-DFT" mode (there is a PRL of ours from 1997, and > Paolo Umari has recently finished coding all the various parts). > Basically, this is composed of a conj-grad minimization on the > orbitals (this can also be used for pure insulators, and it is > faster in reaching the ground state than damped dynamics) and an > optimization of the occupation matrix. In practice, it evolves > orbitals and occupations at the same time, so it would allow you to do > a single > calculation with filled and empty states at the same time. For an > insulator, it is probably not worth, and reciped 1) is more efficient. > For a metal, it is the only possibility, although it will still > struggle (I think) to give very high, completely empty orbitals > correctly (since those do not affect the total energy you are > minimizing - we have a precondition on the occupations to speed > things up). Note that only the 'cs' smearing option works, at this > stage, and the code requires a bit of insider infos to work smoothly. > We'll improve the documentation. > > Bottom line - for an insulator, especially if you need a lot of > empty states, use recipe 1). For a metal, use recipe 2) to converge > to the ground state. If you need only few empty orbitals, 2) will > probably give them correctly already. If you need a lot, use 2) > with only a few empty ones, and then run 1) after 2), with option > 'nscf' and a lot of states (that will be all filled up, and > optimized properly). > > Best luck - let us know, > > nicola > > > --------------------------------------------------------------------- > Prof Nicola Marzari Department of Materials Science and Engineering > 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA > tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From xianghjun at gmail.com Thu Sep 13 19:07:06 2007 From: xianghjun at gmail.com (Hongjun Xiang) Date: Thu, 13 Sep 2007 11:07:06 -0600 Subject: [Pw_forum] How to calculate unoccupied states using cp.x In-Reply-To: <389666229.32510@ustc.edu.cn> References: <46E02461.6030104@gmail.com> <389095204.26878@ustc.edu.cn> <46E86AFE.1050609@gmail.com> <389666229.32510@ustc.edu.cn> Message-ID: <46E96E3A.5010602@gmail.com> Dear Prof. Giannozzi, Thank you for your reply. I obtain correct occupied and unoccupied eigenvalues using the following setting: &SYSTEM ... nbnd=4 / &ELECTRONS emass = 300.d0, emass_cutoff = 2.5d0, orthogonalization = 'ortho', electron_dynamics = 'damp', tcg=.true. empty_states_nbnd=4 / Have a good day. Best regards, Hongjun Xiang ============================================================= H. J. Xiang Postdoctoral Research Associate National Renewable Energy Laboratory http://www4.ncsu.edu/~hxiang/ ============================================================= Paolo Giannozzi wrote: > On Sep 13, 2007, at 24:41 , Hongjun Xiang wrote: > > >> It seems that when nbnd is larger than the number of the occupied >> states, >> CG gives even wrongs occupied eigenvalues. >> > > before stating that eigenvalues are wrong, you should carefully verify > that they are well converged. Eigenvalues are more sensitive than > total energies to convergence, since the latter is a variational > quantity (i.e. has a minimum for the ground state charge density), > the former aren't. > > There is an option (I don't remember which one) that allows calculating > empty Kohn-Sham eigenvalues in cp.x . It is described in the input > documentation, file Doc/INPUT_CP . > > Paolo > --- > Paolo Giannozzi, Democritos and University of Udine, Italy > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > From brad.malone at gmail.com Fri Sep 14 00:04:18 2007 From: brad.malone at gmail.com (Brad Malone) Date: Thu, 13 Sep 2007 15:04:18 -0700 Subject: [Pw_forum] Visualization of local DOS Message-ID: Hi, I have calculated the local DOS but am not sure as to how I should visualize the output of the pp.x calculation. Is there another postprocessing step that I could do similar to chdens.x for the charge density to put it into XCRYSDEN format? Or alternatively, could you please explain to me how the data in the output file is organized so that I may do something myself? Thanks for your time, Brad -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070913/ccffede9/attachment.htm From eyvaz_isaev at yahoo.com Fri Sep 14 10:21:44 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 14 Sep 2007 01:21:44 -0700 (PDT) Subject: [Pw_forum] Visualization of local DOS In-Reply-To: Message-ID: <527847.22154.qm@web60317.mail.yahoo.com> Dear Brad, As it was noticed many times please provide your affiliation. As concerns visualization of DOS, and local DOS you need no extra program. Use gnuplot (www.gnuplot.info), or Grace, former XMgr (http://plasma-gate.weizmann.ac.il/Grace/). These files are suitable for both visualization (plotting) programs. Bests, Eyvaz. --- Brad Malone wrote: > Hi, I have calculated the local DOS but am not sure > as to how I should > visualize the output of the pp.x calculation. Is > there another > postprocessing step that I could do similar to > chdens.x for the charge > density to put it into XCRYSDEN format? Or > alternatively, could you please > explain to me how the data in the output file is > organized so that I may do > something myself? > > Thanks for your time, > Brad > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Shape Yahoo! in your own image. Join our Network Research Panel today! http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7 From brad.malone at gmail.com Fri Sep 14 19:46:57 2007 From: brad.malone at gmail.com (Brad Malone) Date: Fri, 14 Sep 2007 10:46:57 -0700 Subject: [Pw_forum] Visualization of Local DOS Message-ID: Eyvaz, I think the issue was that the machine I did the calculation on had only Espresso v2.1.5 installed. As far as I can tell from the documentation this version of pp.x doesn't include a plotting feature for the local DOS like the one provided in the later releases. But it looks like the intermediate file is still the same, so I can simply transport this file to another machine with a later version on it in order to plot. Thanks for your help, Brad, UC Berkeley -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070914/7432a009/attachment.htm From brad.malone at gmail.com Fri Sep 14 21:51:34 2007 From: brad.malone at gmail.com (Brad Malone) Date: Fri, 14 Sep 2007 12:51:34 -0700 Subject: [Pw_forum] Visualization of Local DOS Message-ID: Actually, when I tried plotting with the pp.x program I got the following error: > task # 0 > from set_fft_dim : error # 1 > input nr1 value not allowed > The top of my local density of states file is: > 45 45 108 45 45 108 24 2 > 0 1.00000000 0.00000000 0.00000000 0.00000000 0.00000000 > 0.00000000 > 10.75226179600000 -5.1673100000000005E-004 > 3.7165372000000003E-002 > -5.1773999999999996E-004 10.75226235600000 - > 3.7164016000000002E-002 > 8.8574731000000000E-002 -8.8575347000000004E-002 25.38119132300000 > > 4.0528473457 4.0000000000 40.0000000000 10 So the progam has an issue with the number '45' it appears. Is there some reason why this shouldn't be allowed? I saw that the requirements for nr1 to be 'allowed' or not is the same in v2 and it is in v3, looking at the set_fft_dim.f90 file. Is there perhaps another difference as to why the FFT value isn't allowed? I could of course do the entire calculation over on the system with v3, but it takes a decent amount of time so I'd prefer to not if I can get away with it, especially since I don't know for sure if that would fix the problem. Thanks again, Brad, UC Berkeley -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070914/1f48c005/attachment.htm From eyvaz_isaev at yahoo.com Fri Sep 14 22:18:08 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 14 Sep 2007 13:18:08 -0700 (PDT) Subject: [Pw_forum] Visualization of Local DOS In-Reply-To: Message-ID: <990267.27370.qm@web60325.mail.yahoo.com> Hi Brad, --- Brad Malone wrote: > Actually, when I tried plotting with the pp.x > program I got the following > error: > > > task # 0 > > from set_fft_dim : error # 1 > > input nr1 value not allowed > > > > The top of my local density of states file is: > My apologies, you are not correct with "local density of states", in fact, you have charge density distribution in the real space. The former is integral of delta(e-e_{nk}) over the Brillouin zone, but the latter is just \rho(x,y,z). Bests, Eyvaz. > > 45 45 108 45 45 108 > 24 2 > > 0 1.00000000 0.00000000 0.00000000 > 0.00000000 0.00000000 > > 0.00000000 > > 10.75226179600000 > -5.1673100000000005E-004 > > 3.7165372000000003E-002 > > -5.1773999999999996E-004 10.75226235600000 > - > > 3.7164016000000002E-002 > > 8.8574731000000000E-002 > -8.8575347000000004E-002 25.38119132300000 > > > > 4.0528473457 4.0000000000 > 40.0000000000 10 > > > So the progam has an issue with the number '45' it > appears. Is there some > reason why this shouldn't be allowed? I saw that > the requirements for nr1 > to be 'allowed' or not is the same in v2 and it is > in v3, looking at the > set_fft_dim.f90 file. Is there perhaps another > difference as to why the FFT > value isn't allowed? > > I could of course do the entire calculation over on > the system with v3, but > it takes a decent amount of time so I'd prefer to > not if I can get away with > it, especially since I don't know for sure if that > would fix the problem. > > Thanks again, > Brad, > UC Berkeley > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Need a vacation? Get great deals to amazing places on Yahoo! Travel. http://travel.yahoo.com/ From giannozz at nest.sns.it Fri Sep 14 22:52:57 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 14 Sep 2007 22:52:57 +0200 Subject: [Pw_forum] Visualization of Local DOS In-Reply-To: References: Message-ID: <69D2AD43-D2AF-4CE4-AA1B-22FA1F10A97A@nest.sns.it> On Sep 14, 2007, at 21:51 , Brad Malone wrote: > So the progam has an issue with the number '45' it appears. > Is there some reason why this shouldn't be allowed? yes, there is one, as usual (there is always a reason in quantum-espresso; it may not be a good reason, but it is one): not all FFT dimensions are allowed... > I saw that the requirements for nr1 to be 'allowed' or not > is the same in v2 and it is in v3, looking at the set_fft_dim.f90 > file. > Is there perhaps another difference as to why the FFT value isn't > allowed? ...and the allowed values differ on different machines. You produced that file on a different machine, or with different FFT's. Try to read it on the same machine you used to produce it. Unfortunately there is no simple general fix, as long as the arrays are written in real space Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Fri Sep 14 22:57:52 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Fri, 14 Sep 2007 22:57:52 +0200 Subject: [Pw_forum] Visualization of Local DOS In-Reply-To: References: Message-ID: <2C71BD94-214B-404C-9CAA-FEF85C51FA23@nest.sns.it> On Sep 14, 2007, at 21:51 , Brad Malone wrote: > from set_fft_dim : error # 1 > input nr1 value not allowed actually there is another possibility: the file format may have changed between versions. Check what is the "input nr1 value" that is not allowed Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From hyello25 at gmail.com Sat Sep 15 07:50:34 2007 From: hyello25 at gmail.com (hyello25 Qgrui) Date: Sat, 15 Sep 2007 13:50:34 +0800 Subject: [Pw_forum] the problem of scf convergence Message-ID: hi, I performed other USPPs tests (blyp-van_ak, pw91-van_ak, pz-van_ak) and found that all these calculations with USPPs were not convergence both in press and Etot.For example, with blyp-van_ak( result is at the bottom of the mail).All these USPPs are downloaded from the homepage of PWSCF/pseudopotential. However, this unconvergence situation disappeared with NCPP (PBE-TM-GGA, converted from fhi format PP using fhi2upf code). The result is at the bottom as well. As Axel Kohlmeyer mentioned: "please try a different Li pseudopotential (i.e. a non-ak one). the one you are using seems to have some not yet understood problems for higher-order properties: lattice constant from EOS fit, MD and geometry optimizations were fine, phonons were bad." And can anyone tell me what does the 'ak'(in blyp-van_ak) mean? What kind of systems are these 'ak' USPPs suitable for calculation? WITH blyp-van_ak USPP:------------------------------------------------------ ---! convergence scf caculation for Li2O !--- ecutwfc = 100 Ry, ecutrho = 320 Ry : ! total energy = -61.82021122 Ry total stress (Ry/bohr**3) (kbar) P= 19.03 ecutwfc = 100 Ry, ecutrho = 360 Ry : ! total energy = -61.82317722 Ry total stress (Ry/bohr**3) (kbar) P= - 6.75 ecutwfc = 100 Ry, ecutrho = 400 Ry : ! total energy = -61.82694596 Ry total stress (Ry/bohr**3) (kbar) P= 4.67 ecutwfc = 100 Ry, ecutrho = 600 Ry : ! total energy = -61.82089409 Ry total stress (Ry/bohr**3) (kbar) P= 4.29 ecutwfc = 100 Ry, ecutrho = 800 Ry : ! total energy = -61.81983678 Ry total stress (Ry/bohr**3) (kbar) P= 19.40 ecutwfc = 100 Ry, ecutrho = 1000 Ry : ! total energy = -61.82031614 Ry total stress (Ry/bohr**3) (kbar) P= 22.33 ecutwfc = 200 Ry, ecutrho = 800 Ry : ! total energy = -61.82432402 Ry total stress (Ry/bohr**3) (kbar) P= 9.13 ecutwfc = 200 Ry, ecutrho = 1000 Ry : ! total energy = -61.82480674 Ry total stress (Ry/bohr**3) (kbar) P= 12.09 ecutwfc = 200 Ry, ecutrho = 1600 Ry : ! total energy = -61.82437317 Ry total stress (Ry/bohr**3) (kbar) P= 6.83 ecutwfc = 200 Ry, ecutrho = 1800 Ry : ! total energy = -61.82423710 Ry total stress (Ry/bohr**3) (kbar) P= 5.72 ------------------------------------------------------------------------------- WITH PBE-MT-GGA NCPP:----------------------------------------------------------- ---! convergence scf caculation for Li2O !--- ecutwfc = ecut = 50 Ry : ! total energy = -33.53719310 Ry total stress (Ry/bohr**3) (kbar) P= - 329.31 ecutwfc = ecut = 80 Ry : ! total energy = -33.60362083 Ry total stress (Ry/bohr**3) (kbar) P= - 46.81 ecutwfc = ecut = 100 Ry : ! total energy = -33.60420031 Ry total stress (Ry/bohr**3) (kbar) P= - 48.44 ecutwfc = ecut = 140 Ry : ! total energy = -33.60518304 Ry total stress (Ry/bohr**3) (kbar) P= - 46.36 ecutwfc = ecut = 180 Ry : ! total energy = -33.60525632 Ry total stress (Ry/bohr**3) (kbar) P= - 45.97 ecutwfc = ecut = 200 Ry : ! total energy = -33.60552604 Ry total stress (Ry/bohr**3) (kbar) P= - 46.30 ecutwfc = ecut = 240 Ry : ! total energy = -33.60553853 Ry total stress (Ry/bohr**3) (kbar) P= - 46.24 ecutwfc = ecut = 280 Ry : ! total energy = -33.60553917 Ry total stress (Ry/bohr**3) (kbar) P= - 46.23 ecutwfc = ecut = 300 Ry : ! total energy = -33.60553970 Ry total stress (Ry/bohr**3) (kbar) P= - 46.24 ecutwfc = ecut = 340 Ry : ! total energy = -33.60553987 Ry total stress (Ry/bohr**3) (kbar) P= - 46.23 ecutwfc = ecut = 400 Ry : ! total energy = -33.60554013 Ry total stress (Ry/bohr**3) (kbar) P= - 46.23 --------------------------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070915/6fa7290d/attachment-0001.htm From degironc at sissa.it Sat Sep 15 08:34:27 2007 From: degironc at sissa.it (degironc) Date: Sat, 15 Sep 2007 08:34:27 +0200 Subject: [Pw_forum] Visualization of Local DOS In-Reply-To: References: Message-ID: <46EB7CF3.3040003@sissa.it> Brad Malone wrote: > Actually, when I tried plotting with the pp.x program I got the > following error: > > task # 0 > from set_fft_dim : error # 1 > input nr1 value not allowed > the complaint comes from the fft part. It clearly says that the first fft dimension you are using (45) is not allowed by the FFT library you are using. Different libraries accept different values for their dimensions. Have you generated the file on one machine (using a given library) and try to analyse it on a different machine (with a different library)? If yes, the problem is likely the incompatibility of the requirements of the two libraries. Use the input variables nr1,nr2,nr3 in pwscf to overwrite the values that the code choose by default so as to find a value that is accepted by both libraries: if 45=3*3*5 is not accepted try 48=3*2*2*2*2 ...(46 or 47 are likely universally bad choices...). If the code and the post-processing have been run on the same machine with the same library then there is something strange. In such a case please provide more information on what you are doing so as to make it reproducible. hope this helps, stefano de Gironcoli, SISSA and DEMOCRITOS > The top of my local density of states file is: > > 45 45 108 45 45 108 24 2 > 0 1.00000000 0.00000000 0.00000000 0.00000000 > 0.00000000 0.00000000 > 10.75226179600000 -5.1673100000000005E-004 > 3.7165372000000003E-002 > -5.1773999999999996E-004 10.75226235600000 > -3.7164016000000002E-002 > 8.8574731000000000E-002 -8.8575347000000004E-002 > 25.38119132300000 > 4.0528473457 4.0000000000 40.0000000000 10 > > > So the progam has an issue with the number '45' it appears. Is there > some reason why this shouldn't be allowed? I saw that the > requirements for nr1 to be 'allowed' or not is the same in v2 and it > is in v3, looking at the set_fft_dim.f90 file. Is there perhaps > another difference as to why the FFT value isn't allowed? > > I could of course do the entire calculation over on the system with > v3, but it takes a decent amount of time so I'd prefer to not if I can > get away with it, especially since I don't know for sure if that would > fix the problem. > > Thanks again, > Brad, > UC Berkeley > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From baroni at sissa.it Sat Sep 15 10:41:38 2007 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 15 Sep 2007 10:41:38 +0200 Subject: [Pw_forum] Visualization of Local DOS In-Reply-To: References: Message-ID: Dear Brad: On Sep 14, 2007, at 7:46 PM, Brad Malone wrote: > I think the issue was that the machine I did the calculation on had > only Espresso v2.1.5 installed. please consider that upgrading to the latest version is a simple way to give back to the community that develops, mantains, and supports scientific software. 1) it helps developers to find out and fix bugs 2) it helps potential users to discover, appreciate, and use new features 3) it helps participants in the forum to keep focused onto up-to-date issues, rather than on issues that have possibly been already tackled and solved with the release of a past version. this being said, we all do appreciate your using Quantum ESPRESSO, regardless of which version is your favorite one. Thanks - Stefano --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070915/d5c7eb67/attachment.htm From eyvaz_isaev at yahoo.com Sat Sep 15 14:11:41 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Sat, 15 Sep 2007 05:11:41 -0700 (PDT) Subject: [Pw_forum] the problem of scf convergence In-Reply-To: Message-ID: <395327.41792.qm@web60325.mail.yahoo.com> Hi, Axel just suggested using a "non-ak", i.e. not generated by Axel Kohlmeyer, pseudopotential. It was quite interesting to test my own pseudopotential for Li which gave a=4.64Ang (Exp. 4.60Ang), and B=83GPa (Exp. 81GPa) for Li20. Phonons at Gamma point look good (without splitting of the T_1u mode): T_1u - 431cm^-1, and T_2g - 524cm^-1. Bests, Eyvaz. --- hyello25 Qgrui wrote: > hi, > I performed other USPPs tests (blyp-van_ak, > pw91-van_ak, pz-van_ak) and > found that all these calculations with USPPs were > not convergence both in > press and Etot.For example, with blyp-van_ak( result > is at the bottom of the > mail).All these USPPs are downloaded from the > homepage of > PWSCF/pseudopotential. > However, this unconvergence situation disappeared > with NCPP (PBE-TM-GGA, > converted from fhi format PP using fhi2upf code). > The result is at the > bottom as well. > As Axel Kohlmeyer mentioned: > "please try a different Li pseudopotential (i.e. a > non-ak one). > the one you are using seems to have some not yet > understood > problems for higher-order properties: lattice > constant from > EOS fit, MD and geometry optimizations were fine, > phonons > were bad." > And can anyone tell me what does the 'ak'(in > blyp-van_ak) mean? What kind > of systems are these 'ak' USPPs suitable for > calculation? > > WITH blyp-van_ak > USPP:------------------------------------------------------ > ---! convergence scf caculation for Li2O > !--- > ecutwfc = 100 Ry, ecutrho = 320 Ry : > ! total energy = -61.82021122 Ry > total stress (Ry/bohr**3) > (kbar) P= > 19.03 > ecutwfc = 100 Ry, ecutrho = 360 Ry : > ! total energy = -61.82317722 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 6.75 > ecutwfc = 100 Ry, ecutrho = 400 Ry : > ! total energy = -61.82694596 Ry > total stress (Ry/bohr**3) > (kbar) P= > 4.67 > ecutwfc = 100 Ry, ecutrho = 600 Ry : > ! total energy = -61.82089409 Ry > total stress (Ry/bohr**3) > (kbar) P= > 4.29 > ecutwfc = 100 Ry, ecutrho = 800 Ry : > ! total energy = -61.81983678 Ry > total stress (Ry/bohr**3) > (kbar) P= > 19.40 > ecutwfc = 100 Ry, ecutrho = 1000 Ry : > ! total energy = -61.82031614 Ry > total stress (Ry/bohr**3) > (kbar) P= > 22.33 > ecutwfc = 200 Ry, ecutrho = 800 Ry : > ! total energy = -61.82432402 Ry > total stress (Ry/bohr**3) > (kbar) P= > 9.13 > ecutwfc = 200 Ry, ecutrho = 1000 Ry : > ! total energy = -61.82480674 Ry > total stress (Ry/bohr**3) > (kbar) P= > 12.09 > ecutwfc = 200 Ry, ecutrho = 1600 Ry : > ! total energy = -61.82437317 Ry > total stress (Ry/bohr**3) > (kbar) P= > 6.83 > ecutwfc = 200 Ry, ecutrho = 1800 Ry : > ! total energy = -61.82423710 Ry > total stress (Ry/bohr**3) > (kbar) P= > 5.72 > ------------------------------------------------------------------------------- > > WITH PBE-MT-GGA > NCPP:----------------------------------------------------------- > ---! convergence scf caculation for Li2O > !--- > ecutwfc = ecut = 50 Ry : > ! total energy = -33.53719310 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 329.31 > ecutwfc = ecut = 80 Ry : > ! total energy = -33.60362083 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 46.81 > ecutwfc = ecut = 100 Ry : > ! total energy = -33.60420031 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 48.44 > ecutwfc = ecut = 140 Ry : > ! total energy = -33.60518304 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 46.36 > ecutwfc = ecut = 180 Ry : > ! total energy = -33.60525632 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 45.97 > ecutwfc = ecut = 200 Ry : > ! total energy = -33.60552604 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 46.30 > ecutwfc = ecut = 240 Ry : > ! total energy = -33.60553853 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 46.24 > ecutwfc = ecut = 280 Ry : > ! total energy = -33.60553917 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 46.23 > ecutwfc = ecut = 300 Ry : > ! total energy = -33.60553970 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 46.24 > ecutwfc = ecut = 340 Ry : > ! total energy = -33.60553987 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 46.23 > ecutwfc = ecut = 400 Ry : > ! total energy = -33.60554013 Ry > total stress (Ry/bohr**3) > (kbar) P= - > 46.23 > --------------------------------------------------------------------------------- > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Boardwalk for $500? In 2007? Ha! Play Monopoly Here and Now (it's updated for today's economy) at Yahoo! Games. http://get.games.yahoo.com/proddesc?gamekey=monopolyherenow From baroni at sissa.it Sat Sep 15 15:31:11 2007 From: baroni at sissa.it (Stefano Baroni) Date: Sat, 15 Sep 2007 15:31:11 +0200 Subject: [Pw_forum] INTERESTING position available soon Message-ID: <5EDA6AE1-B0B8-4852-B608-C0DCE0DBC2D1@sissa.it> Dear Quantum Espressionists, this is to inform you that a "technologist" position is going to be advertised soon by the DEMOCRITOS National Simulation Center in Trieste, in collaboration with the Scuola Internazionale Superiore di Studi Avanzati (SISSA), to work on the development and maintenance of the software distributed with QE (mainly, but not exclusively, PWscf, CP/FPMD, PHONON). The candidate will have the responsability of coordinating the team of developers, initially under the supervision of and later, in collaboration with Paolo Giannozzi. A good deal of scientific independence will be granted to the succesful applicant. By this I mean that it is expected that a fraction of his/her worktime (say, of the order of 1/3) will be devoted to his/her own research in total and complete independence. The salary will be of the order of 1500 ?/month after taxes, including full social security (national public pension funds) and health insurance. The contract will be initially for one year, automatically renewable for two more years. Although no formal commitment can be made at this time, it is expected that the job will become tenurable before the deadline of the third year. I would like to invite potentially interested candidates to submit an informal "expression of interest" with a brief CV and list of publications to: Prof. Paolo Giannozzi p.giannozzi at fisica.uniud.it The post will not be officially advertised until suitable expression of interests are received. More information are of course available upon request either from me or from Paolo. Cheers, Stefano Baroni --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070915/9f16874b/attachment.htm From yukihiro_okuno at fujifilm.co.jp Fri Sep 14 06:38:03 2007 From: yukihiro_okuno at fujifilm.co.jp (yukihiro_okuno at fujifilm.co.jp) Date: Fri, 14 Sep 2007 13:38:03 +0900 Subject: [Pw_forum] Error in vc-relax Message-ID: Dear PWSCF users. I'm now doing vc-relax calculation, but met the difficulties of conversence. My conversence criterio is not too strict but the total force and stress is not conversed, and the calculation stopped by the message (defalt stress criterio is too strict to calculate ?) I had change etot_conv_thr = 1.0D-4 to etot_conv_thr = 1.0D-6 but the situation is not changed. The error message is %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from checkallsym : error # 3 not orthogonal operation %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... The lattice we calaulation is tetragonal but the output CELL_PARAMETRE become like CELL_PARAMETERS (alat) 0.992118220 0.000000000 0.000000000 0.000000000 0.992117943 0.000000000 0.000000000 0.000000000 1.114913401 slightly orhogonal. The force conversence is Total force = 0.001187 Total SCF correction = 0.000008 Total force = 0.001134 Total SCF correction = 0.000012 Total force = 0.001063 Total SCF correction = 0.000005 Total force = 0.000977 Total SCF correction = 0.000002 Total force = 0.000878 Total SCF correction = 0.000005 Total force = 0.000705 Total SCF correction = 0.000056 Total force = 0.000603 Total SCF correction = 0.000006 Total force = 0.000586 Total SCF correction = 0.000012 Total force = 0.000664 Total SCF correction = 0.000020 Total force = 0.000634 Total SCF correction = 0.000110 Total force = 0.000828 Total SCF correction = 0.000245 Total force = 0.000789 Total SCF correction = 0.000015 Total force = 0.000890 Total SCF correction = 0.000014 Total force = 0.001005 Total SCF correction = 0.000007 Total force = 0.001192 Total SCF correction = 0.000292 Total force = 0.001165 Total SCF correction = 0.000021 Total force = 0.001248 Total SCF correction = 0.000021 Total force = 0.001264 Total SCF correction = 0.000044 Total force = 0.001290 Total SCF correction = 0.000049 Total force = 0.001293 Total SCF correction = 0.000054 Total force = 0.001274 Total SCF correction = 0.000057 Total force = 0.001235 Total SCF correction = 0.000057 Total force = 0.001176 Total SCF correction = 0.000055 Total force = 0.001102 Total SCF correction = 0.000051 Total force = 0.001002 Total SCF correction = 0.000041 Total force = 0.000977 Total SCF correction = 0.000074 Total force = 0.001074 Total SCF correction = 0.000038 Total force = 0.001172 Total SCF correction = 0.000072 Total force = 0.001191 Total SCF correction = 0.000165 Total force = 0.001256 Total SCF correction = 0.000016 Total force = 0.001272 Total SCF correction = 0.000018 Total force = 0.001412 Total SCF correction = 0.000320 Total force = 0.001330 Total SCF correction = 0.000040 Total force = 0.001375 Total SCF correction = 0.000018 and The stress is entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -1.83 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -1.63 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -1.39 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -1.11 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.80 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.47 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.53 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.42 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.26 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= -0.09 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.05 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.19 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.40 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.52 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.69 entering subroutine stress ... total stress (Ry/bohr**3) (kbar) P= 0.93 My Input file is like this, &control calculation = 'vc-relax', restart_mode = 'from_scratch', pseudo_dir = '../pseudo/', outdir = './', prefix='SystemName' nstep = 100 etot_conv_thr = 1.0D-4 forc_conv_thr = 1.0D-3 dt = 80.0 tstress = .true. tprnfor = .true. / &system ibrav=6 celldm(1)=6.86225965271 celldm(3)=1.0806385896 nat=5 ntyp=3 nbnd=26 ecutwfc=60.0 ecutrho = 600.0 occupations = 'fixed' degauss=0.00 / &electrons conv_thr = 1e-10, mixing_beta=0.3, / &IONS bfgs_ndim=3 pot_extrapolation = "second_order" wfc_extrapolation = "second_order" / &CELL cell_dynamics = 'damp-w', press = 0.0 wmass = 0.1 / How should treat such a conversence difficulties and erro message? Sincerely. Yukihiro Okuno. From giannozz at nest.sns.it Sat Sep 15 22:08:40 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sat, 15 Sep 2007 22:08:40 +0200 Subject: [Pw_forum] Error in vc-relax In-Reply-To: References: Message-ID: <0F412AD0-7A9A-41A9-96D3-06AEA9A78B38@nest.sns.it> On Sep 14, 2007, at 6:38 , yukihiro_okuno at fujifilm.co.jp wrote: > from checkallsym : error # 3 > not orthogonal operation loss of symmetry during the optimization. The safe procedure is to break the initial symmetry > CELL_PARAMETERS (alat) > 0.992118220 0.000000000 0.000000000 > 0.000000000 0.992117943 0.000000000 > 0.000000000 0.000000000 1.114913401 > > slightly orhogonal. "orthorhombic", I guess. This is exactly the reason for the above error message. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Sun Sep 16 20:40:56 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 16 Sep 2007 20:40:56 +0200 Subject: [Pw_forum] Error in vc-relax In-Reply-To: References: Message-ID: On Sep 14, 2007, at 6:38 , yukihiro_okuno at fujifilm.co.jp wrote: > My Input file is like this, by the way: if you supply an input file that can be run, the probability that someone fixes a problem like yours is very small but different from zero. With an incomplete input, the probability is exactly zero. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From giannozz at nest.sns.it Sun Sep 16 20:42:27 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 16 Sep 2007 20:42:27 +0200 Subject: [Pw_forum] the problem of scf convergence In-Reply-To: References: Message-ID: <6CC78A86-A0F6-4949-913A-3632F4C04D80@nest.sns.it> On Sep 15, 2007, at 7:50 , hyello25 Qgrui wrote: > And can anyone tell me what does the 'ak'(in blyp-van_ak) mean? ak == Axel Kohlmeyer, who was so kind to supply a number of ultrasoft pseudotentials generated using David Vanderbilt's code. Paolo --- Paolo Giannozzi, Democritos and University of Udine, Italy From akohlmey at cmm.chem.upenn.edu Mon Sep 17 02:38:21 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Sun, 16 Sep 2007 20:38:21 -0400 (EDT) Subject: [Pw_forum] the problem of scf convergence In-Reply-To: <6CC78A86-A0F6-4949-913A-3632F4C04D80@nest.sns.it> Message-ID: On Sun, 16 Sep 2007, Paolo Giannozzi wrote: PG> PG> On Sep 15, 2007, at 7:50 , hyello25 Qgrui wrote: PG> PG> > And can anyone tell me what does the 'ak'(in blyp-van_ak) mean? PG> PG> ak == Axel Kohlmeyer, who was so kind to supply a number of PG> ultrasoft pseudotentials generated using David Vanderbilt's PG> code. hi paolo, talking about kindness... would you please be so kind and pull all the lithium "ak" potentials from the pseudopotential library. thanks to this episode, we have now an easier to check parameter in order to correct for the problems with this set of potentials. i still wonder what causes this, but that is probably buried deep in the black magic of pseudopotential creation. sometimes you have it, sometimes you don't. oh well... ciao, axel. PG> PG> Paolo PG> --- PG> Paolo Giannozzi, Democritos and University of Udine, Italy PG> PG> PG> _______________________________________________ PG> Pw_forum mailing list PG> Pw_forum at pwscf.org PG> http://www.democritos.it/mailman/listinfo/pw_forum PG> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From giannozz at nest.sns.it Mon Sep 17 16:06:53 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 17 Sep 2007 16:06:53 +0200 Subject: [Pw_forum] Error in vc-relax In-Reply-To: References: Message-ID: <82AA16D3-44C3-495C-BDF6-4C401A1CA2D5@nest.sns.it> Since complaints about the variable-cell optimization show up more often than not: can anybody provide a complete test job (i.e. one that can be rerun), with complete output, showing a problem in a reproducible way, in a reasonable amount of computer time, in a recent versions of Q-E? Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From akohlmey at cmm.chem.upenn.edu Mon Sep 17 18:29:12 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 17 Sep 2007 12:29:12 -0400 (EDT) Subject: [Pw_forum] Error in vc-relax In-Reply-To: <82AA16D3-44C3-495C-BDF6-4C401A1CA2D5@nest.sns.it> Message-ID: On Mon, 17 Sep 2007, Paolo Giannozzi wrote: hi paolo, PG> Since complaints about the variable-cell optimization PG> show up more often than not: can anybody provide a PG> complete test job (i.e. one that can be rerun), with my impression is more often, that this is a manifestation of vc-relax being a less straightforward to use method and that people don't always understand all the involved subtleties. as a consequence of that, one is tempted to use vc-relax in a way that is not working well, or for calculations (lattice constant determination) where there may be other more efficient and reliable ways to obtain this information. i can sympathize very much with these ideas, since i had them all myself and it was a (sometimes painful) learning process to get a better understanding of where i was going wrong. PG> complete output, showing a problem in a reproducible PG> way, in a reasonable amount of computer time, in a PG> recent versions of Q-E? to some level, we'll just have to get used to people trying things with vc-relax that cannot work (you have to get burned to stay away from the fire!) or do not work in the way they hoped for. other than that a few more extensive explanations, pointers to (more) commented tutorial examples, also in the INPUT_PW file and perhaps an adjusted default for the cell mass (as cesar has pointed out a couple of times), could handle the majority of the cases. cheers, axel. p.s.: i agree with your assessment, that when somebody posts an incomplete input, that the chance of somebody completing it and using it for testings is (almost) zero. posting only an input fragement, should be ok (and saves a lot of bandwidth and storage), when people are not certain about their input flags altogether, but if a problem persists, there is not alternative to making the whole input available, so that somebody can (try to) re-run the calculation elsewhere. PG> PG> Paolo PG> --- PG> Paolo Giannozzi, Dept of Physics, University of Udine PG> via delle Scienze 208, 33100 Udine, Italy PG> Phone +39-0432-558216, fax +39-0432-558222 PG> PG> PG> PG> _______________________________________________ PG> Pw_forum mailing list PG> Pw_forum at pwscf.org PG> http://www.democritos.it/mailman/listinfo/pw_forum PG> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From konstantin_kudin at yahoo.com Mon Sep 17 20:16:12 2007 From: konstantin_kudin at yahoo.com (Konstantin Kudin) Date: Mon, 17 Sep 2007 11:16:12 -0700 (PDT) Subject: [Pw_forum] Error in vc-relax In-Reply-To: Message-ID: <86898.80785.qm@web57705.mail.re3.yahoo.com> Wow, this was like a very entertaining post! I nominate Axel for the appropriate prize !!! Kids, don't play with the fire that is QE ! :-) By setting the entertainment aspect aside, I suspect that most users would rather much prefer to get the job done quickly while remaining for the most part ignorant about the code intrinsics, and move on with their life, as opposed to diving deep into the manuals and PW forum archives in order to figure out how to force the code to do what is needed. So, from this angle, some sort of *automatic* solution would be desirable. I have used vc-relax quite a lot for 2D cell optimizations. The code crashes each time there is a symmetry decrease. I am certain that changes in the cell geometry cause such effects, for example, when the input cell is hexagonal, and the first step makes it non-hexagonal. So, then, for example, the system goes from 8 symmetry operations to 4, and the code crashes. A fix would involve either making the code respect the input symmetry such that the cell changes do not alter it, or, the code could re-initialize with the new symmetry when a problem is detected, sort of like doing an internal restart. My examples are all huge, so I won't post them yet. Kostya --- Axel Kohlmeyer wrote: > On Mon, 17 Sep 2007, Paolo Giannozzi wrote: > > hi paolo, > > PG> Since complaints about the variable-cell optimization > PG> show up more often than not: can anybody provide a > PG> complete test job (i.e. one that can be rerun), with > > my impression is more often, that this is a manifestation > of vc-relax being a less straightforward to use method and > that people don't always understand all the involved subtleties. > as a consequence of that, one is tempted to use vc-relax > in a way that is not working well, or for calculations > (lattice constant determination) where there may be other > more efficient and reliable ways to obtain this information. > > i can sympathize very much with these ideas, since i had > them all myself and it was a (sometimes painful) learning > process to get a better understanding of where i was > going wrong. > > PG> complete output, showing a problem in a reproducible > PG> way, in a reasonable amount of computer time, in a > PG> recent versions of Q-E? > > to some level, we'll just have to get used to people trying > things with vc-relax that cannot work (you have to get > burned to stay away from the fire!) or do not work in the > way they hoped for. other than that a few more extensive > explanations, pointers to (more) commented tutorial examples, > also in the INPUT_PW file and perhaps an adjusted default for > the cell mass (as cesar has pointed out a couple of times), > could handle the majority of the cases. > > cheers, > axel. > > p.s.: i agree with your assessment, that when somebody posts > an incomplete input, that the chance of somebody completing it > and using it for testings is (almost) zero. posting only an > input fragement, should be ok (and saves a lot of bandwidth and > storage), when people are not certain about their input flags > altogether, but if a problem persists, there is not alternative > to making the whole input available, so that somebody can (try to) > re-run the calculation elsewhere. > > PG> > PG> Paolo > PG> --- > PG> Paolo Giannozzi, Dept of Physics, University of Udine > PG> via delle Scienze 208, 33100 Udine, Italy > PG> Phone +39-0432-558216, fax +39-0432-558222 > PG> > PG> > PG> > PG> _______________________________________________ > PG> Pw_forum mailing list > PG> Pw_forum at pwscf.org > PG> http://www.democritos.it/mailman/listinfo/pw_forum > PG> > > -- > ======================================================================= > Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu > http://www.cmm.upenn.edu > Center for Molecular Modeling -- University of Pennsylvania > Department of Chemistry, 231 S.34th Street, Philadelphia, PA > 19104-6323 > tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: > 1-215-898-5425 > ======================================================================= > If you make something idiot-proof, the universe creates a better > idiot. > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ____________________________________________________________________________________ Fussy? Opinionated? Impossible to please? Perfect. Join Yahoo!'s user panel and lay it on us. http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7 From giannozz at nest.sns.it Tue Sep 18 10:49:41 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 18 Sep 2007 10:49:41 +0200 Subject: [Pw_forum] Error in vc-relax In-Reply-To: <86898.80785.qm@web57705.mail.re3.yahoo.com> References: <86898.80785.qm@web57705.mail.re3.yahoo.com> Message-ID: On Sep 17, 2007, at 20:16 , Konstantin Kudin wrote: > A fix would involve either making the code respect the input > symmetry such that the cell changes do not alter it this is something that the code is already supposed to do (at least with the Wentzcovitch lagrangian) Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From ezadshojaee at hotmail.com Tue Sep 18 12:44:28 2007 From: ezadshojaee at hotmail.com (Ezad Shojaee) Date: Tue, 18 Sep 2007 10:44:28 +0000 Subject: [Pw_forum] about phonon by grid Message-ID: Hi when i calculate the IFC's by a (2 2 2) grid, and also (3 3 3) the dispersion curve seems to be reasonable and also converging; but when i try (4 4 4) grid, it comes with a lot of huge negative figures! what may be the problem? _________________________________________________________________ Explore the seven wonders of the world http://search.msn.com/results.aspx?q=7+wonders+world&mkt=en-US&form=QBRE -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070918/60d0f207/attachment.htm From marcel at physik.tu-berlin.de Tue Sep 18 15:02:59 2007 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Tue, 18 Sep 2007 15:02:59 +0200 (CEST) Subject: [Pw_forum] about phonon by grid In-Reply-To: References: Message-ID: <20070918145727.E78913@rosa.physik-pool.tu-berlin.de> Hi On Tue, 18 Sep 2007, Ezad Shojaee wrote: > > Hi > when i calculate the IFC's by a (2 2 2) grid, and also (3 3 3) the dispersion curve seems to be reasonable and also converging; but when i try (4 4 4) grid, it comes with a lot of huge negative figures! what may be the problem? > _________________________________________________________________ > Explore the seven wonders of the world > http://search.msn.com/results.aspx?q=7+wonders+world&mkt=en-US&form=QBRE Hi Ezad 4 4 4 is not so big. I would think small changes will occur if you increase q-grid. What system do you calculate? Are the acoustic frequencies negative, or other ones too. How did you caluclate the dispersion, did you impose acoustic sum rule or similar? Maybe your system is not well relaxed, e.g., you are in a local energy minimum and not the global one? Cheers Marcel ________________________________________________________________________ Marcel Mohr Institut f?r Festk?rperphysik, TU Berlin marcel(at)physik.tu-berlin.de Sekr. PN 5-4 TEL: +49-30-314 24442 Hardenbergstr. 36 FAX: +49-30-314 27705 10623 Berlin From wlyim at puccini.che.pitt.edu Tue Sep 18 15:22:48 2007 From: wlyim at puccini.che.pitt.edu (William Yim) Date: Tue, 18 Sep 2007 09:22:48 -0400 (EDT) Subject: [Pw_forum] about phonon by grid In-Reply-To: <20070918145727.E78913@rosa.physik-pool.tu-berlin.de> Message-ID: Hello, This may happen when the calculations are not converged very well. In this case, you may need to check: 1) energy cutoffs for wavefunction and charge density; 2) k-point settings for your scf calculations; 3) convergence threshold. In principle, if you use a denser q-mesh for IFC calculation, you may also need a denser k-mesh for your SCF. Correct me if I am wrong. Best regards, William On Tue, 18 Sep 2007, Marcel Mohr wrote: > Hi > > On Tue, 18 Sep 2007, Ezad Shojaee wrote: > > > > > Hi when i calculate the IFC's by a (2 2 2) grid, and also (3 3 3) the > > dispersion curve seems to be reasonable and also converging; but when > > i try (4 4 4) grid, it comes with a lot of huge negative figures! what > > may be the problem? > > _________________________________________________________________ > > Explore the seven wonders of the world > > http://search.msn.com/results.aspx?q=7+wonders+world&mkt=en-US&form=QBRE > > Hi Ezad > > 4 4 4 is not so big. I would think small changes will occur if you > increase q-grid. > What system do you calculate? > Are the acoustic frequencies negative, or other ones too. How did you > caluclate the dispersion, did you impose acoustic sum rule or similar? > Maybe your system is not well relaxed, e.g., you are in a local > energy minimum and not the global one? > > Cheers Marcel > > ________________________________________________________________________ > Marcel Mohr Institut f?r Festk?rperphysik, TU Berlin > marcel(at)physik.tu-berlin.de Sekr. PN 5-4 > TEL: +49-30-314 24442 Hardenbergstr. 36 > FAX: +49-30-314 27705 10623 Berlin > > -- Dr. Wai-Leung Yim Institut fuer Reine und Angewandte Chemie, Theoretische Chemie, Carl von Ossietzky Universtaet Oldenburg, 26129 Oldenburg, Germany Email: wlyim at puccini.che.pitt.edu Phone: +49-441-798-3950 (office) Fax: +49-441-798-3964 From lanhaiping at gmail.com Tue Sep 18 18:26:21 2007 From: lanhaiping at gmail.com (lan haiping) Date: Wed, 19 Sep 2007 00:26:21 +0800 Subject: [Pw_forum] Question about induced dipole moment Message-ID: Hi, All. I have read a work of Prof Marzari performed by QE package to compute static dielectric properties of nanotube systems : PRL 96, 166801. To extract dielectric quantities, there are 2 methods mentioned in this work. I have a question about the method adopted by the sawtooth potential . In order to obtain transverse polarizability , we need know the dipole moment induced by external sawtooth potential . I have done several tests on similar systems carbon nanotube (6,0), but i have not found how to extract such induced dipole moment . Would anyone please give me a hand on this subject ? Any tips or references are appreciated ! Best wishes Hai-Ping -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070919/9333b7af/attachment.htm From marzari at MIT.EDU Tue Sep 18 22:30:06 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Tue, 18 Sep 2007 16:30:06 -0400 Subject: [Pw_forum] Question about induced dipole moment In-Reply-To: References: Message-ID: <46F0354E.6040908@mit.edu> Dear Hai-Ping, it was done directly with real space integration - i.e. in the transferse direction, the dipole is given by the integral of r rho(r). It could also be done by looking at the change in position of the sum of the centers of the Wannier functions - this is less accurate, since the reciprocal-space expression for the position operator is based on a finite-differences discretization. nicols lan haiping wrote: > Hi, All. > I have read a work of Prof Marzari performed by QE package to > compute static dielectric properties of nanotube systems : PRL 96, 166801. > To extract dielectric quantities, there are 2 methods mentioned in > this work. I have a question about the method adopted by the sawtooth > potential . In order to obtain transverse polarizability , we need know > the dipole moment induced by external sawtooth potential . I have done > several tests on similar systems carbon nanotube (6,0), but i have not > found how to extract such induced dipole moment . > Would anyone please give me a hand on this subject ? Any tips or > references are appreciated ! > > Best wishes > Hai-Ping > -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From lanhaiping at gmail.com Wed Sep 19 12:51:21 2007 From: lanhaiping at gmail.com (lan haiping) Date: Wed, 19 Sep 2007 18:51:21 +0800 Subject: [Pw_forum] Question about induced dipole moment In-Reply-To: <46F0354E.6040908@mit.edu> References: <46F0354E.6040908@mit.edu> Message-ID: Thank you, Prof Marzari. As far as i know, the changes of the WFs' centers can be applied to insulating systems. While for metallic or semi-metallic systems, Is such scheme still useful ? in fact, i still donot know how to define the wfs' centers for metallic systems. i found it is hard to achieve convergence when i applied electric-enthalpy approache to metallic systems no matter which direction of the uniform electric field involves. i do notice the difficulties of scf convergence of systems under electric field. i tried to reduce mixing parameter , or increase the smearing width, but no success. Are there any other parameters i should pay attention to ? With Regards, Hai-Ping On 9/19/07, Nicola Marzari wrote: > > > > Dear Hai-Ping, > > it was done directly with real space integration - i.e. in the > transferse direction, the dipole is given by the integral of r rho(r). > > It could also be done by looking at the change in position of the > sum of the centers of the Wannier functions - this is less accurate, > since the reciprocal-space expression for the position operator is > based on a finite-differences discretization. > > > nicols > > > lan haiping wrote: > > Hi, All. > > I have read a work of Prof Marzari performed by QE package to > > compute static dielectric properties of nanotube systems : PRL 96, > 166801. > > To extract dielectric quantities, there are 2 methods mentioned in > > this work. I have a question about the method adopted by the sawtooth > > potential . In order to obtain transverse polarizability , we need know > > the dipole moment induced by external sawtooth potential . I have done > > several tests on similar systems carbon nanotube (6,0), but i have not > > found how to extract such induced dipole moment . > > Would anyone please give me a hand on this subject ? Any tips or > > references are appreciated ! > > > > Best wishes > > Hai-Ping > > > > -- > --------------------------------------------------------------------- > Prof Nicola Marzari Department of Materials Science and Engineering > 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA > tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070919/fdd8c3b3/attachment.htm From degironc at sissa.it Wed Sep 19 13:03:39 2007 From: degironc at sissa.it (degironc) Date: Wed, 19 Sep 2007 13:03:39 +0200 Subject: [Pw_forum] Question about induced dipole moment In-Reply-To: References: <46F0354E.6040908@mit.edu> Message-ID: <46F1020B.60106@sissa.it> the static dielectric constant of a metal is infinit.... what are you trying to calculate ? stefano lan haiping wrote: > Thank you, Prof Marzari. > > As far as i know, the changes of the WFs' centers can be applied to > insulating systems. > While for metallic or semi-metallic systems, Is such scheme still > useful ? in fact, i still donot know how to define the wfs' centers > for metallic systems. > > i found it is hard to achieve convergence when i applied > electric-enthalpy approache to > metallic systems no matter which direction of the uniform electric > field involves. > i do notice the difficulties of scf convergence of systems under > electric field. i tried to > reduce mixing parameter , or increase the smearing width, but no > success. Are there > any other parameters i should pay attention to ? > > With Regards, > Hai-Ping > > > On 9/19/07, *Nicola Marzari* > wrote: > > > > Dear Hai-Ping, > > it was done directly with real space integration - i.e. in the > transferse direction, the dipole is given by the integral of r rho(r). > > It could also be done by looking at the change in position of the > sum of the centers of the Wannier functions - this is less accurate, > since the reciprocal-space expression for the position operator is > based on a finite-differences discretization. > > > nicols > > > lan haiping wrote: > > Hi, All. > > I have read a work of Prof Marzari performed by QE package to > > compute static dielectric properties of nanotube systems : PRL > 96, 166801. > > To extract dielectric quantities, there are 2 methods > mentioned in > > this work. I have a question about the method adopted by the > sawtooth > > potential . In order to obtain transverse polarizability , we > need know > > the dipole moment induced by external sawtooth potential . I > have done > > several tests on similar systems carbon nanotube (6,0), but i > have not > > found how to extract such induced dipole moment . > > Would anyone please give me a hand on this subject ? Any > tips or > > references are appreciated ! > > > > Best wishes > > Hai-Ping > > > > -- > --------------------------------------------------------------------- > Prof Nicola Marzari Department of Materials Science and Engineering > 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA > tel 617.4522758 fax 2586534 marzari at mit.edu > http://quasiamore.mit.edu > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > -- > Hai-Ping Lan > Department of Electronics , > Peking University , Bejing, 100871 > lanhaiping at gmail.com , hplan at pku.edu.cn > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From lanhaiping at gmail.com Wed Sep 19 13:15:42 2007 From: lanhaiping at gmail.com (lan haiping) Date: Wed, 19 Sep 2007 19:15:42 +0800 Subject: [Pw_forum] Question about induced dipole moment In-Reply-To: <46F1020B.60106@sissa.it> References: <46F0354E.6040908@mit.edu> <46F1020B.60106@sissa.it> Message-ID: Dear Stefano, i notice such problem, but in Prof Marzari's work on cnt, the transverse polarizabilities of metallic nanotubes are given. My confusion may be due to this example. Thank you Regards Hai-Ping On 9/19/07, degironc wrote: > > the static dielectric constant of a metal is infinit.... > what are you trying to calculate ? > stefano > > lan haiping wrote: > > Thank you, Prof Marzari. > > > > As far as i know, the changes of the WFs' centers can be applied to > > insulating systems. > > While for metallic or semi-metallic systems, Is such scheme still > > useful ? in fact, i still donot know how to define the wfs' centers > > for metallic systems. > > > > i found it is hard to achieve convergence when i applied > > electric-enthalpy approache to > > metallic systems no matter which direction of the uniform electric > > field involves. > > i do notice the difficulties of scf convergence of systems under > > electric field. i tried to > > reduce mixing parameter , or increase the smearing width, but no > > success. Are there > > any other parameters i should pay attention to ? > > > > With Regards, > > Hai-Ping > > > > > > On 9/19/07, *Nicola Marzari* > > wrote: > > > > > > > > Dear Hai-Ping, > > > > it was done directly with real space integration - i.e. in the > > transferse direction, the dipole is given by the integral of r > rho(r). > > > > It could also be done by looking at the change in position of the > > sum of the centers of the Wannier functions - this is less accurate, > > since the reciprocal-space expression for the position operator is > > based on a finite-differences discretization. > > > > > > nicols > > > > > > lan haiping wrote: > > > Hi, All. > > > I have read a work of Prof Marzari performed by QE package > to > > > compute static dielectric properties of nanotube systems : PRL > > 96, 166801. > > > To extract dielectric quantities, there are 2 methods > > mentioned in > > > this work. I have a question about the method adopted by the > > sawtooth > > > potential . In order to obtain transverse polarizability , we > > need know > > > the dipole moment induced by external sawtooth potential . I > > have done > > > several tests on similar systems carbon nanotube (6,0), but i > > have not > > > found how to extract such induced dipole moment . > > > Would anyone please give me a hand on this subject ? Any > > tips or > > > references are appreciated ! > > > > > > Best wishes > > > Hai-Ping > > > > > > > -- > > > --------------------------------------------------------------------- > > Prof Nicola Marzari Department of Materials Science and > Engineering > > 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 > USA > > tel 617.4522758 fax 2586534 marzari at mit.edu > > http://quasiamore.mit.edu > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > > > > -- > > Hai-Ping Lan > > Department of Electronics , > > Peking University , Bejing, 100871 > > lanhaiping at gmail.com , hplan at pku.edu.cn > > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070919/fc5631b4/attachment.htm From marzari at MIT.EDU Wed Sep 19 14:40:11 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Wed, 19 Sep 2007 08:40:11 -0400 Subject: [Pw_forum] Question about induced dipole moment In-Reply-To: References: <46F0354E.6040908@mit.edu> <46F1020B.60106@sissa.it> Message-ID: <46F118AB.9090902@mit.edu> lan haiping wrote: > Dear Stefano, > i notice such problem, but in Prof Marzari's work on cnt, the transverse > polarizabilities of metallic nanotubes are given. My confusion may be > due to this example. > Thank you > > Regards > Hai-Ping Dear Hai-Ping, 1) the transverse polarizabilities are finite for both metallic and insulating nanotubes. It's the longitudinal one that is infinite in metallic tubes. Note that 3n,0 tubes at small diameters are semiconducting with small gaps. 2) as mentioned, we did calculate them in real space from the integral of the charge density. This is an appropriate operation for a finite, isolated system - in the transverse direction this is what they are. 3) re. electric enthalpy - not sure what the status of that is. I'll try to discuss it when Paolo Umari visits later in October. 4) you could use WFs - both for truly semiconducting nanotubes, and even for the worse case scenario of (n,n) armchair (metallic) ones. Treat the system as insulating, and you should find a solution with WFs that corresponds to covalent bonds, and 1 p_z-like WF every 2 carbons. The center displacement as a function of field gives you the polarizaton. Best, nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From lanhaiping at gmail.com Thu Sep 20 14:02:33 2007 From: lanhaiping at gmail.com (lan haiping) Date: Thu, 20 Sep 2007 20:02:33 +0800 Subject: [Pw_forum] Question about induced dipole moment In-Reply-To: <46F118AB.9090902@mit.edu> References: <46F0354E.6040908@mit.edu> <46F1020B.60106@sissa.it> <46F118AB.9090902@mit.edu> Message-ID: Thank you very much , Prof Marzari . Why can we safely treat metallic nanotubes with wfs' centers change ? Just because of its feature DOS pattern at Fermi level ? With Regards? Hai-Ping On 9/19/07, Nicola Marzari wrote: > > lan haiping wrote: > > Dear Stefano, > > i notice such problem, but in Prof Marzari's work on cnt, the transverse > > polarizabilities of metallic nanotubes are given. My confusion may be > > due to this example. > > Thank you > > > > Regards > > Hai-Ping > > > Dear Hai-Ping, > > 1) the transverse polarizabilities are finite for both metallic > and insulating nanotubes. It's the longitudinal one that is > infinite in metallic tubes. Note that 3n,0 tubes at small diameters > are semiconducting with small gaps. > > 2) as mentioned, we did calculate them in real space from the integral > of the charge density. This is an appropriate operation for a finite, > isolated system - in the transverse direction this is what they are. > > 3) re. electric enthalpy - not sure what the status of that is. > I'll try to discuss it when Paolo Umari visits later in October. > > 4) you could use WFs - both for truly semiconducting nanotubes, > and even for the worse case scenario of (n,n) armchair (metallic) > ones. Treat the system as insulating, and you should > find a solution with WFs that corresponds to covalent bonds, > and 1 p_z-like WF every 2 carbons. The center displacement as a > function of field gives you the polarizaton. > > Best, > > nicola > > > --------------------------------------------------------------------- > Prof Nicola Marzari Department of Materials Science and Engineering > 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA > tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping at gmail.com, hplan at pku.edu.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070920/e9fb4346/attachment.htm From marzari at MIT.EDU Thu Sep 20 14:42:21 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 20 Sep 2007 08:42:21 -0400 Subject: [Pw_forum] Question about induced dipole moment In-Reply-To: References: <46F0354E.6040908@mit.edu> <46F1020B.60106@sissa.it> <46F118AB.9090902@mit.edu> Message-ID: <46F26AAD.6040802@mit.edu> lan haiping wrote: > Thank you very much , Prof Marzari . > Why can we safely treat metallic nanotubes with wfs' centers change ? > Just because of its feature DOS pattern at Fermi level ? > > With Regards? > Hai-Ping I'd say because the system is not metallic in the tranverse direction. But it's true that even in graphene one can use the WFs center to reach meaningful conclusions about the contribution of the different orbitals to the dielectric response - but let's not start a thread on this. nicola -- --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From wangjunjie1981_0 at 163.com Thu Sep 20 15:44:30 2007 From: wangjunjie1981_0 at 163.com (wangjunjie1981_0) Date: Thu, 20 Sep 2007 21:44:30 +0800 (CST) Subject: [Pw_forum] how to generate the pseudopotential by using virtual.x Message-ID: <1451057135.1346001190295870162.JavaMail.coremail@bj163app128.163.com> Dear allI want to mix the pseudopotentials of two atoms by virtual.x.But when it is done, the error is happen. Through the hint. i know different pseudopotentials have different mesh is the reason of generating the error,but mesh of different pseudopotentials is general disagreement,how can i do to generate correct pseudopotential ? I'm getting following error ,when i mix the pseudopotentials by virtual.x Generate the UPF pseudopotential for a virtual atom combining two pseudopootentials in UPF format Input PP file # 1 in UPF format > Li.cpi_ggapw.UPF IOS= 0 1 4 Reading pseudopotential file in UPF format... ...done Input PP file # 2 in UPF format > N.cpi_ggapw.UPF IOS= 0 2 4 Reading pseudopotential file in UPF format... ...done New Pseudo = x Li.cpi_ggapw.UPF + (1-x) N.cpi_ggapw.UPF mixing parameter x [0 References: <1451057135.1346001190295870162.JavaMail.coremail@bj163app128.163.com> Message-ID: <8136C4C2-A7ED-469D-B49C-4D419D0BD1EE@nest.sns.it> On Sep 20, 2007, at 15:44 , wangjunjie1981_0 wrote: > pseudopotentials have different mesh the logarithmic grids used in atomic codes have the form: r(i) = r_0 * exp ( (i-1) dx ) (grid starting at r(1)>0) or r(i) = r_0 * ( exp( (i-1)*dx ) - 1.0_dp ) (grid starting at r(1)=0) where dx and r_0 are grid parameters. Typically, r_0 is written as r_0 = exp(xmin) / Z where Z is the atomic number, and one chooses xmin. If you want the same grid with different Z, you should choose exp(xmin_1) / Z_1 = exp(xmin_2) / Z_2 . Otherwise, you have to implement some form of interpolation between the two grids: it is not that hard. In any event, you should really think whether mixing Lithium and Nitrogen atoms (to get Litrogen?) using the virtual crystal approximation is a sensible operation. One typically uses the VCA for isovalent atoms (e.g. Si and Ge, G and Al, ...) Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From marcello.rosini at unimore.it Thu Sep 20 17:38:54 2007 From: marcello.rosini at unimore.it (Marcello Rosini) Date: Thu, 20 Sep 2007 17:38:54 +0200 Subject: [Pw_forum] BFGS & Wolfe condition Message-ID: <46F2940E.6010902@unimore.it> I' am performing a bfgs relaxation after 3 scf cycles and one bfg relaxation I get "wolfe condition not staisfied" and the program stops. Does anyone have suggestions? NB: the forces acting on the atom a reasonably small and all the positions are correct. Here you can find the input and the output http://cdm.unimo.it/home/fisica/rosini.marcello/input.txt http://cdm.unimo.it/home/fisica/rosini.marcello/output.txt thanks alot Marcello -- ------------------------------------------------------------------------ *Dr. ROSINI MARCELLO* rosini.marcello at unimo.it --------------------------------------------------- dipartimento di fisica Universit? degli studi di Modena e Reggio Emilia via campi 213/a, 41100 Modena, Italy tel +39 059 2055312 - fax +39 059 2055616 ------------------------------------------------------------------------ From giannozz at nest.sns.it Thu Sep 20 17:56:48 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 20 Sep 2007 17:56:48 +0200 Subject: [Pw_forum] BFGS & Wolfe condition In-Reply-To: <46F2940E.6010902@unimore.it> References: <46F2940E.6010902@unimore.it> Message-ID: <9E0247DC-6AD6-4F56-9AD0-78EE6165AA04@nest.sns.it> On Sep 20, 2007, at 17:38 , Marcello Rosini wrote: > I' am performing a bfgs relaxation > after 3 scf cycles and one bfg relaxation I get > "wolfe condition not satisfied" > and the program stops. are you sure you didn't simply reach the time limit? "wolfe condition not satisfied" is not an error condition Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From marzari at MIT.EDU Thu Sep 20 18:16:07 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 20 Sep 2007 12:16:07 -0400 Subject: [Pw_forum] BFGS & Wolfe condition In-Reply-To: <46F2940E.6010902@unimore.it> References: <46F2940E.6010902@unimore.it> Message-ID: <46F29CC7.8080400@mit.edu> Marcello Rosini wrote: > I' am performing a bfgs relaxation > after 3 scf cycles and one bfg relaxation I get > "wolfe condition not staisfied" > and the program stops. > Does anyone have suggestions? > NB: the forces acting on the atom a reasonably small and all the > positions are correct. > Here you can find the input and the output > http://cdm.unimo.it/home/fisica/rosini.marcello/input.txt > http://cdm.unimo.it/home/fisica/rosini.marcello/output.txt > > thanks alot > Marcello Hi Marcello, no idea, but Google knows everything. It looks like the message error you get is from an old bfgs http://www.democritos.it/pipermail/pw_forum/2004-January/000806.html . The current bfgs does not print that error message, but checks for the Wolfe condition: http://reference.wolfram.com/mathematica/tutorial/UnconstrainedOptimizationLineSearchMethods.html Now, what that probably means is that there has been an accumulation of approximations (e.g. in your hessian) that your predicted steps along the line direction are not going to be very accurate (again, wild guess). To solve the problem just (using a newer version of espresso) restart everything, from scratch, with the current relaxed coordinates, so bfgs starts from scratch with a new hessian. There might be better strategies - mine are just guesses. Moral - if you have a question, google has the answer. It's like the library of Babel, but searchable. nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From marcel at physik.tu-berlin.de Thu Sep 20 20:13:44 2007 From: marcel at physik.tu-berlin.de (Marcel Mohr) Date: Thu, 20 Sep 2007 20:13:44 +0200 (CEST) Subject: [Pw_forum] Which dielectric constant is calculated In-Reply-To: <46F29CC7.8080400@mit.edu> References: <46F2940E.6010902@unimore.it> <46F29CC7.8080400@mit.edu> Message-ID: <20070920194239.P53674@rosa.physik-pool.tu-berlin.de> Dear list-members during a PHONON calculation of a polar crystal, the dielectric constant is calculated and written to the dynamical matrix. With this and dynmat the LO TO splitting can be calculated. Is this dielectric constant the static one, or the high-frequency one (\epsilon_\infty)? I am confused, because in the paper from Gianozzi et al. PRB 43 7231, in formula (18) where the non-analytic part of the dynamical matrix is calculated, a \epsilon\infty enters which is called the high-frequency static dieelectric tensor. But this one is not for frequencies like visible light? (like \epsilon_\infty in experimental books, like, e.g., in Landolt-Boernstein) I am grateful for any hints Best regards Marcel ________________________________________________________________________ Marcel Mohr Institut f?r Festk?rperphysik, TU Berlin marcel(at)physik.tu-berlin.de Sekr. PN 5-4 TEL: +49-30-314 24442 Hardenbergstr. 36 FAX: +49-30-314 27705 10623 Berlin From giannozz at nest.sns.it Thu Sep 20 22:38:46 2007 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 20 Sep 2007 22:38:46 +0200 Subject: [Pw_forum] Which dielectric constant is calculated In-Reply-To: <20070920194239.P53674@rosa.physik-pool.tu-berlin.de> References: <46F2940E.6010902@unimore.it> <46F29CC7.8080400@mit.edu> <20070920194239.P53674@rosa.physik-pool.tu-berlin.de> Message-ID: On Sep 20, 2007, at 20:13 , Marcel Mohr wrote: > Is this dielectric constant the static one, or the high-frequency > one (\epsilon_\infty)? > > I am confused, because in the paper from Gianozzi et al. PRB 43 > 7231, in formula (18) where the non-analytic part of the dynamical > matrix is calculated, a \epsilon\infty enters which is called the > high-frequency static dieelectric tensor. this answers your question: it is both ! Seriously: it is the dielectric tensor at frequencies that are high with respect to phonon modes, but low wrt electronic excitations. So it is "static" wrt the latter and "high-frequency" wrt the former. Paolo --- Paolo Giannozzi, Dept of Physics, University of Udine via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 From wangjunjie1981_0 at 163.com Fri Sep 21 03:18:21 2007 From: wangjunjie1981_0 at 163.com (wangjunjie1981_0) Date: Fri, 21 Sep 2007 09:18:21 +0800 (CST) Subject: [Pw_forum] how to generate the pseudopotential using virtual.x Message-ID: <991226506.76251190337501135.JavaMail.coremail@bj163app104.163.com> DearPaolo Thank you for your help. But i still don't understand when i generate the pseudopotential using the fhi98PP, the pseudopotential has had the different mesh. How can i do to avoid its happen. At the same time, i want to ask whether the pseudopotential from the internet can't directly be used to generate mixing pseudopotential by virtual.x. In addition, i try to shield the sentences about the different mesh in virtual.f90, as a consequence, the program don't have an error and give a mixing pseudopotential. But i do not know whether the mixing pseudopotential is correct or not. Thank you very much in advance. Best regards -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070921/8866a35a/attachment.htm From degironc at sissa.it Fri Sep 21 09:27:36 2007 From: degironc at sissa.it (degironc) Date: Fri, 21 Sep 2007 09:27:36 +0200 Subject: [Pw_forum] how to generate the pseudopotential using virtual.x In-Reply-To: <991226506.76251190337501135.JavaMail.coremail@bj163app104.163.com> References: <991226506.76251190337501135.JavaMail.coremail@bj163app104.163.com> Message-ID: <46F37268.5030900@sissa.it> Dear wangjunjie1981, what is your affililation? wangjunjie1981_0 wrote: > Dear Paolo > Thank you for your help. But i still don't understand when i generate > the pseudopotential using the fhi98PP, the pseudopotential has had the > different mesh. How can i do to avoid its happen. At the same time, i > want to ask whether the pseudopotential from the internet can't > directly be used to generate mixing pseudopotential by virtual.x. In order to mix pseudopotentials with two different meshes one needs to interpolate at least one of the two on a common mesh. This interpolation is not dificult but has not been tested... hence the error message. > In addition, i try to shield the sentences about the different mesh in > virtual.f90, as a consequence, the program don't have an error and > give a mixing pseudopotential. But i do not know whether the mixing > pseudopotential is correct or not. A way to give back to the community would be to help testing/fixing the interpolation scheme... A simple strategy I see for that is building a *fake* virtual pseudopotential with 0% of potential1 and 100% of potential2 ... this forcea the system to use the mesh of potential1 and if the interpolation is fine, should give resuslts equivalent to the original potential2. If this is not the case there is a problem in the interpolation. If/when everything is fine one could safely remove the error messages in virtual.x best regards, Stefano de Gironcoli > Thank you very much in advance. > Best regards > > > ------------------------------------------------------------------------ > ??70??????????????????2008?????? > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From hande at newton.physics.metu.edu.tr Fri Sep 21 15:57:55 2007 From: hande at newton.physics.metu.edu.tr (Hande Ustunel) Date: Fri, 21 Sep 2007 16:57:55 +0300 (EEST) Subject: [Pw_forum] noninteger error in ld1.x Message-ID: Dear PWSCF developers, While accidentally using the cvs version of ld1.x to generate a pseudopotential using the cvs version, I came across an error that I thought I'd draw your attention to. For iswitch=2, i.e. while testing an already generated pseudopotential, it gives the following complaint from ld1_readin : error # 1 found noninteger valence 3.75, if you want this specify zval in inputp for the input &input atom='C', rlderiv=4.0, eminld=-4.0, emaxld= 4.0, deld = 0.02, nld = 3, config='1s2 2s2 2p2.0 3d0.0', iswitch=2, prefix='c' dft='pbe', / &test file_pseudo='C.pbe.uspp-rrkj.UPF' nconf=3, configts(1)=' 2s2.0 2p1.75 ', configts(2)=' 2s2.00 2p2.00 ', configts(3)=' 2s2.00 2p2.00 ', / which shouldn't happen for the test case unless I'm mistaken. My untrained eye tells me that the line which checks for noninteger values for iswitch=3 and which used to be in the scope of the if (iswitch==3) statement is no longer inside but outside this scope and therefore seems to extend to the iswitch=2 case. It works fine with version 3.2. Best wishes, Hande Ustunel -- Hande Ustunel Department of Physics Office 439 Middle East Technical University Ankara 06531, Turkey Tel : +90 312 210 3264 http://www.physics.metu.edu.tr/~hande From dalcorso at sissa.it Fri Sep 21 16:08:26 2007 From: dalcorso at sissa.it (Dal Corso Andrea) Date: Fri, 21 Sep 2007 16:08:26 +0200 Subject: [Pw_forum] noninteger error in ld1.x In-Reply-To: References: Message-ID: <1190383706.3209.8.camel@dhpc-5-07.sissa.it> On Fri, 2007-09-21 at 16:57 +0300, Hande Ustunel wrote: > Dear PWSCF developers, > > While accidentally using the cvs version of ld1.x to generate a > pseudopotential using the cvs version, I came across an error that I > thought I'd draw your attention to. For iswitch=2, i.e. while testing an > already generated pseudopotential, it gives the following complaint > > from ld1_readin : error # 1 > found noninteger valence 3.75, if you want this specify zval in inputp > > for the input > > &input > atom='C', > rlderiv=4.0, > eminld=-4.0, > emaxld= 4.0, > deld = 0.02, > nld = 3, > config='1s2 2s2 2p2.0 3d0.0', > iswitch=2, > prefix='c' > dft='pbe', > / > &test > file_pseudo='C.pbe.uspp-rrkj.UPF' > nconf=3, > configts(1)=' 2s2.0 2p1.75 ', > configts(2)=' 2s2.00 2p2.00 ', > configts(3)=' 2s2.00 2p2.00 ', > / > > > which shouldn't happen for the test case unless I'm mistaken. My untrained > eye tells me that the line which checks for noninteger values for iswitch=3 > and which used to be in the scope of the if (iswitch==3) statement is no > longer inside but outside this scope and therefore seems to extend to the > iswitch=2 case. It works fine with version 3.2. > You are right, thank you. I will correct it. Best Regards, Andrea > Best wishes, > Hande Ustunel > -- Andrea Dal Corso Tel. 0039-040-3787428 SISSA, Via Beirut 2/4 Fax. 0039-040-3787528 34014 Trieste (Italy) e-mail: dalcorso at sissa.it From wangjunjie1981_0 at 163.com Mon Sep 24 05:39:16 2007 From: wangjunjie1981_0 at 163.com (wangjunjie1981_0) Date: Mon, 24 Sep 2007 11:39:16 +0800 (CST) Subject: [Pw_forum] how to generate the pseudopotential using virtual.x In-Reply-To: References: Message-ID: <19280995.489211190605156581.JavaMail.coremail@bj163app45.163.com> Dear Stefano Thank you for your advice, i can understand your ideas.But i have a question how to achive the interpolation.Is it included in PWSCF program or generated by other special program? How can i have or find ? I hope your help again, Thank you very much. I will give back to the community about the testing results. Best regardsJunjie Wang -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070924/30303ba6/attachment.htm From baroni at sissa.it Mon Sep 24 09:15:56 2007 From: baroni at sissa.it (Stefano Baroni) Date: Mon, 24 Sep 2007 09:15:56 +0200 Subject: [Pw_forum] how to generate the pseudopotential using virtual.x In-Reply-To: <19280995.489211190605156581.JavaMail.coremail@bj163app45.163.com> References: <19280995.489211190605156581.JavaMail.coremail@bj163app45.163.com> Message-ID: <8C42541A-99BD-4CB8-8943-6D361E93D1DF@sissa.it> Dear Junjie: I do not know whether the distribution contains a ready-to-use tool. At the very worse, it contains a routine for cubic spline interpolation that, although not commented, is rather self- explanatory. If you do not know what sline interpolation is, the wikipedia is your friend http://en.wikipedia.org/wiki/Spline_%28mathematics%29 in case you do not have access to wikipedia, you can try mathworld: http://mathworld.wolfram.com/Spline.html Stefano (another one!) B. On Sep 24, 2007, at 5:39 AM, wangjunjie1981_0 wrote: > > Dear Stefano > Thank you for your advice, i can understand your ideas.But i have > a question how to achive the interpolation.Is it included in PWSCF > program or generated by other special program? How can i have or > find ? I hope your help again, Thank you very much I will give back > to the community about the testing results. > > > Best regards > Junjie Wang > > > ?70?????????2008??? > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070924/bd7735d4/attachment.htm From daijiayu at nudt.edu.cn Mon Sep 24 15:15:47 2007 From: daijiayu at nudt.edu.cn (=?gb2312?B?tPe80e7a?=) Date: Mon, 24 Sep 2007 21:15:47 +0800 Subject: [Pw_forum] error in the calculation of mpirun ph.x. Message-ID: <390639747.23084@nudt.edu.cn> Dear users, When i use mpirun -np 4 ph.x -npool 2 < *.ph.in > *.ph.out, it is OK at the begining(including the mpirun -np 4 pw.x calculation). But when it began to calculate the dynamic matrix of second q point, it stopped. The error information is: net_send: could not write to fd=5, errno=32. I run the example06, and met the same problem. So, what's the reason? It's the fault of the mpi install or the ph.x install? Cheers, Thank you. Jiayu From daijiayu at nudt.edu.cn Mon Sep 24 15:15:58 2007 From: daijiayu at nudt.edu.cn (=?gb2312?B?tPe80e7a?=) Date: Mon, 24 Sep 2007 21:15:58 +0800 Subject: [Pw_forum] error in the calculation of mpirun ph.x. Message-ID: <390639758.23088@nudt.edu.cn> Dear users, When i use mpirun -np 4 ph.x -npool 2 < *.ph.in > *.ph.out, it is OK at the begining(including the mpirun -np 4 pw.x calculation). But when it began to calculate the dynamic matrix of second q point, it stopped. The error information is: net_send: could not write to fd=5, errno=32. I run the example06, and met the same problem. So, what's the reason? It's the fault of the mpi install or the ph.x install? Cheers, Thank you. Jiayu ------------------------------ Jiayu Dai National University of Defense Technology, P R China From chen_shao_hua197 at yahoo.com.tw Mon Sep 24 19:08:08 2007 From: chen_shao_hua197 at yahoo.com.tw (=?big5?q?=B3=AF=20=A4=D6=B5=D8?=) Date: Tue, 25 Sep 2007 01:08:08 +0800 (CST) Subject: [Pw_forum] error in the q2r.out Message-ID: <703688.33680.qm@web73009.mail.tp2.yahoo.com> Dear forum's members, The input file "q2r.in" content is below: &input zasr='simple', fildyn='doped0.01-inn.dyn', flfrc='doped0.01-inn.fc', la2F=.true. / The q2r.out error messsage as below: ... ... ... reading force constants from file doped0.01-inn.dyn30 nqs= 12 q= 0.25000000 0.57735027 0.24844721 q= -0.25000000 -0.57735027 -0.24844721 q= -0.25000000 -0.57735027 0.24844721 q= 0.25000000 0.57735027 -0.24844721 q= -0.37500000 0.50518149 0.24844721 q= 0.37500000 -0.50518149 -0.24844721 q= 0.62500000 0.07216878 0.24844721 q= -0.62500000 -0.07216878 -0.24844721 q= -0.62500000 -0.0721bm_list_16854: p4_error: listener select: -1 p1_16855: (0.800781) net_recv failed for fd = 3 p1_16855: p4_error: net_recv read, errno = : 104 rm_l_1_16856: (0.800781) net_send: could not write to fd=5, errno = 32 p6_4305: p4_error: net_recv read: probable EOF on socket: 1 p5_16906: p4_error: net_recv read: probable EOF on socket: 1 rm_l_6_4307: (0.796875) net_send: could not write to fd=5, errno = 32 rm_l_5_16908: (0.800781) net_send: could not write to fd=5, errno = 32 p3_16405: p4_error: net_recv read: probable EOF on socket: 1 p7_4306: p4_error: net_recv read: probable EOF on socket: 1 rm_l_7_4308: (0.796875) net_send: could not write to fd=5, errno = 32 rm_l_3_16407: (0.800781) net_send: could not write to fd=5, errno = 32 rm_l_2_16406: p4_error: interrupt SIGx: 15 rm_l_2_16406: (0.812500) net_send: could not write to fd=5, errno = 32 rm_l_4_16907: p4_error: interrupt SIGx: 15 rm_l_4_16907: (0.812500) net_send: could not write to fd=5, errno = 32 p1_16855: (6.824219) net_send: could not write to fd=5, errno = 32 p6_4305: (12.804688) net_send: could not write to fd=5, errno = 32 p5_16906: (12.808594) net_send: could not write to fd=5, errno = 32 p7_4306: (14.812500) net_send: could not write to fd=5, errno = 32 p3_16405: (14.820312) net_send: could not write to fd=5, errno = 32 Does anyone know what it means ? Thanks in advance max Physics Department,National Taiwan University,Taipei,Taiwan ____________________________________________________________________________________ ??????????? Yahoo!??????????? http://tw.info.yahoo.com/seal/index.html From akohlmey at cmm.chem.upenn.edu Mon Sep 24 18:19:41 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Mon, 24 Sep 2007 12:19:41 -0400 (EDT) Subject: [Pw_forum] error in the q2r.out In-Reply-To: <703688.33680.qm@web73009.mail.tp2.yahoo.com> Message-ID: On Tue, 25 Sep 2007, [big5] ?? ???? wrote: CSH> Dear forum's members, CSH> [...] CSH> p5_16906: (12.808594) net_send: could not write to CSH> fd=5, errno = 32 CSH> p7_4306: (14.812500) net_send: could not write to CSH> fd=5, errno = 32 CSH> p3_16405: (14.820312) net_send: could not write to CSH> fd=5, errno = 32 CSH> CSH> Does anyone know what it means ? that means, that all your MPI "slave" processes died. did you check whether your input runs in serial? cheers, axel. CSH> CSH> Thanks in advance CSH> max CSH> Physics Department,National Taiwan CSH> University,Taipei,Taiwan CSH> CSH> CSH> CSH> ____________________________________________________________________________________ CSH> ?]?w?w???????A?O?@?z?? Yahoo!?_???b?????Q?b???s???I CSH> http://tw.info.yahoo.com/seal/index.html CSH> _______________________________________________ CSH> Pw_forum mailing list CSH> Pw_forum at pwscf.org CSH> http://www.democritos.it/mailman/listinfo/pw_forum CSH> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From daijiayu at nudt.edu.cn Tue Sep 25 00:53:14 2007 From: daijiayu at nudt.edu.cn (=?gb2312?B?tPe80e7a?=) Date: Tue, 25 Sep 2007 06:53:14 +0800 Subject: [Pw_forum] error in the q2r.out Message-ID: <390674394.28281@nudt.edu.cn> >CSH> Dear forum's members, >CSH> > >[...] > >CSH> p5_16906: (12.808594) net_send: could not write to >CSH> fd=5, errno = 32 >CSH> p7_4306: (14.812500) net_send: could not write to >CSH> fd=5, errno = 32 >CSH> p3_16405: (14.820312) net_send: could not write to >CSH> fd=5, errno = 32 >CSH> >CSH> Does anyone know what it means ? > >that means, that all your MPI "slave" processes died. >did you check whether your input runs in serial? > >cheers, > axel. I have the same problem when i run mpirun -np 4 ph.x -npool 2 < *.ph.in . But if i run mpirun -np 4 ph.x -npool 1 < *.ph.in or mpirun -np 4 ph.x < *.ph.in, the error would disappear. Is it the meaning of the memory error? what did you mean "in serial"? Thanks, Jiayu ------------------------------ Jiayu Dai National University of Defense Technology, P R China From baroni at sissa.it Tue Sep 25 08:19:36 2007 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 25 Sep 2007 08:19:36 +0200 Subject: [Pw_forum] error in the q2r.out In-Reply-To: <390674394.28281@nudt.edu.cn> References: <390674394.28281@nudt.edu.cn> Message-ID: <6CEE3881-D1B7-4D3B-9431-D4FE77DDFAE5@sissa.it> On Sep 25, 2007, at 12:53 AM, ??? wrote: > >> CSH> Dear forum's members, >> CSH> >> >> [...] >> >> CSH> p5_16906: (12.808594) net_send: could not write to >> CSH> fd=5, errno = 32 >> CSH> p7_4306: (14.812500) net_send: could not write to >> CSH> fd=5, errno = 32 >> CSH> p3_16405: (14.820312) net_send: could not write to >> CSH> fd=5, errno = 32 >> CSH> >> CSH> Does anyone know what it means ? >> >> that means, that all your MPI "slave" processes died. >> did you check whether your input runs in serial? >> >> cheers, >> axel. > > I have the same problem when i run mpirun -np 4 ph.x -npool 2 < > *.ph.in . But if i > run mpirun -np 4 ph.x -npool 1 < *.ph.in or mpirun -np 4 ph.x < > *.ph.in, the > error would disappear. Is it the meaning of the memory error? what > did you mean > "in serial"? "in serial" \equiv "not in parallel", I guess ... SB --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070925/33418ca6/attachment.htm From akohlmey at cmm.chem.upenn.edu Tue Sep 25 13:37:02 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Tue, 25 Sep 2007 07:37:02 -0400 (EDT) Subject: [Pw_forum] error in the q2r.out In-Reply-To: <390674394.28281@nudt.edu.cn> Message-ID: On Tue, 25 Sep 2007, [gb2312] ?????? wrote: JD> >cheers, JD> > axel. JD> JD> I have the same problem when i run mpirun -np 4 ph.x -npool 2 < *.ph.in . But if i please when running with mpirun you should use the -in flag instead of redirecting input from stdin (and using filename globbing, i.e. filename expansion with *, is even more problematic as it may redirect multiple files). not all MPI packages replicate stdin to all nodes. JD> run mpirun -np 4 ph.x -npool 1 < *.ph.in or mpirun -np 4 ph.x < *.ph.in, the JD> error would disappear. Is it the meaning of the memory error? what did you mean do generate a CRASH file? have you looked at it? it probably means, that the calculation that you were asking for, does not support parallelization over k-point pools. JD> "in serial"? "in serial" means "not parallel" there is no memory error. when your remote processes die, your local process cannot write to their corresponding sockets anymore. this is one of the (many) drawbacks of using MPICH: the error reporting of this MPI packages is not very good and frequently it "eats" the last lines of output of the crashing application (which would be essential to understand what is going on). cheers, axel. JD> JD> Thanks, JD> Jiayu JD> JD> JD> JD> JD> ------------------------------ JD> Jiayu Dai JD> National University of Defense Technology, P R China JD> -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From wangjunjie1981_0 at 163.com Wed Sep 26 05:07:41 2007 From: wangjunjie1981_0 at 163.com (wangjunjie1981_0) Date: Wed, 26 Sep 2007 11:07:41 +0800 (CST) Subject: [Pw_forum] How to generate the pseudopotential using virtual.x In-Reply-To: References: Message-ID: <1767719638.825261190776061262.JavaMail.coremail@bj163app127.163.com> Dear all Thank you very much for your help, i basically understand your ideas, but i look the interpolate sentences in the source file virtual.f90. I want to know whether the interpolate can be implemented in PWSCF-self. If the interpolate is operated by hand, the error is easily happen. I can not certain the results is correct or not. Best regards Junjie Wang -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070926/af6a4e96/attachment.htm From baroni at sissa.it Wed Sep 26 08:13:13 2007 From: baroni at sissa.it (Stefano Baroni) Date: Wed, 26 Sep 2007 08:13:13 +0200 Subject: [Pw_forum] How to generate the pseudopotential using virtual.x In-Reply-To: <1767719638.825261190776061262.JavaMail.coremail@bj163app127.163.com> References: <1767719638.825261190776061262.JavaMail.coremail@bj163app127.163.com> Message-ID: Dear Junjie, I am not sure I really get your point. I understand you would like to add virtual-crystal capabilities to PWscf itself? If I remember correctly, it was a deliberate choice to leave PC capabilities out of PWscf itself, and to delegate them to the prior generation of an appropriate pseudopotential, which may require some interpolation or other numerical tweaking. There is nothing magic about it. If the code complains, it is only that the data you have generated are different from what the code expects. By "different" I do not mean "wrong", I mean "in the wrong format", "wrong in number", or the like ... SB On Sep 26, 2007, at 5:07 AM, wangjunjie1981_0 wrote: > Dear all > > Thank you very much for your help, i basically understand your > ideas, but i look the interpolate sentences in the source file > virtual.f90. I want to know whether the interpolate can be > implemented in PWSCF-self. If the interpolate is operated by hand, > the error is easily happen. I can not certain the results is > correct or not. > Best regards > Junjie Wang > > > ?70?????????2008??? > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) Please, if possible, don't send me MS Word or PowerPoint attachments Why? See: http://www.gnu.org/philosophy/no-word-attachments.html -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070926/be6813f3/attachment.htm From zdw2000 at gmail.com Wed Sep 26 10:14:58 2007 From: zdw2000 at gmail.com (Wei Zhou) Date: Wed, 26 Sep 2007 16:14:58 +0800 Subject: [Pw_forum] about metal and molecular dynamics Message-ID: <3e36a8fc0709260114g6ab290c9n75acd03b475a958c@mail.gmail.com> hello every body I have searched the pwscf forum. and found some discuss about metal and molecular dynamics. but I still feel puzzled . the main question is my system is metal ,and I want to do some variable cell dynamics . for QE codes, it has been said there have been two methods to do this ---PW.X and CP. so my question is which is better for my aims is QE3.2 can do this ? thank you in advance. -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070926/9c4f5c72/attachment.htm From degironc at sissa.it Wed Sep 26 15:19:28 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 26 Sep 2007 15:19:28 +0200 Subject: [Pw_forum] how to generate the pseudopotential using virtual.x In-Reply-To: <19280995.489211190605156581.JavaMail.coremail@bj163app45.163.com> References: <19280995.489211190605156581.JavaMail.coremail@bj163app45.163.com> Message-ID: <46FA5C60.6090207@sissa.it> dear Junjie Wang, if you comment out (or replace by warnings) the error messages in virtual.x that make the code stop when interpolate is .true. then the code performs the interpolation and produces a pseudopotential (I think you already tried that). What I was suggesting is that you generate a *virtual* pseudo made by 0% pseudo1 + 100% pseudo2 (which should interpolate pseudo2 on the grid of pseudo1) and check that the results you get for this virtual pseudopotential agree to your satisfaction with what you can calculate directly with pseudo2. If this is the case this means that the interpolation is working fine and you could mix the two pseudopoteianls also in other fractions. If not there is something wrong in the interpolation. In any case please report back on the outcome of your tests. thank you, stefano -- Stefano de Gironcoli - SISSA and DEMOCRITOS wangjunjie1981_0 wrote: > > Dear Stefano > Thank you for your advice, i can understand your ideas.But i have a > question how to achive the interpolation.Is it included in PWSCF > program or generated by other special program? How can i have or find > ? I hope your help again, Thank you very much I will give back to the > community about the testing results. > Best regards > Junjie Wang > > > ------------------------------------------------------------------------ > ??70??????????????????2008?????? > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From akohlmey at cmm.chem.upenn.edu Thu Sep 27 04:50:55 2007 From: akohlmey at cmm.chem.upenn.edu (Axel Kohlmeyer) Date: Wed, 26 Sep 2007 22:50:55 -0400 Subject: [Pw_forum] about metal and molecular dynamics In-Reply-To: <3e36a8fc0709260114g6ab290c9n75acd03b475a958c@mail.gmail.com> References: <3e36a8fc0709260114g6ab290c9n75acd03b475a958c@mail.gmail.com> Message-ID: <7b6913e90709261950m6b542118of540278d77794a82@mail.gmail.com> On 9/26/07, Wei Zhou wrote: > > hello every body > I have searched the pwscf forum. and found some discuss about metal > and molecular dynamics. but I still feel puzzled . the main question is my > system is metal ,and I want to do some variable cell dynamics . for QE > codes, it has been said there have been two methods to do this --- PW.X and > CP. so my question is which is better for my aims hello! there is no simple answer to this. you have a lot of options (between the codes and within each of them), which one is the best depends a lot on what you want to know, what system exactly you want to treat, what accuracy you need and how well you can figure out the best way to run each code. cheers, axel. > is QE3.2 can do this ? > thank you in advance. > -- > > > > > > > > > > > ZhouDawei > JiLin Universiyt ,ChangChun ,China > zdw2000 at gmail.com > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- ======================================================================= Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu Center for Molecular Modeling -- University of Pennsylvania Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From zdw2000 at gmail.com Thu Sep 27 05:09:55 2007 From: zdw2000 at gmail.com (Wei Zhou) Date: Thu, 27 Sep 2007 11:09:55 +0800 Subject: [Pw_forum] about metal and molecular dynamics In-Reply-To: <7b6913e90709261950m6b542118of540278d77794a82@mail.gmail.com> References: <3e36a8fc0709260114g6ab290c9n75acd03b475a958c@mail.gmail.com> <7b6913e90709261950m6b542118of540278d77794a82@mail.gmail.com> Message-ID: <3e36a8fc0709262009y371fa433w924778ee7f70e1ea@mail.gmail.com> thank you for your answer. but I want to know if the pw.x of QE code can do variable cell dynamics 'npt' using nose to control temperature. or the cp of QE code can do this .or both of them can do this . 2007/9/27, Axel Kohlmeyer : > > On 9/26/07, Wei Zhou wrote: > > > > hello every body > > I have searched the pwscf forum. and found some discuss about > metal > > and molecular dynamics. but I still feel puzzled . the main question is > my > > system is metal ,and I want to do some variable cell dynamics . for QE > > codes, it has been said there have been two methods to do this --- PW.Xand > > CP. so my question is which is better for my aims > > hello! > > there is no simple answer to this. > you have a lot of options (between the codes and within each of them), > which one is the best depends a lot on what you want to know, what system > exactly you want to treat, what accuracy you need and how well you > can figure out the best way to run each code. > > cheers, > axel. > > > is QE3.2 can do this ? > > thank you in advance. > > -- > > > > > > > > > > > > > > > > > > > > > > ZhouDawei > > JiLin Universiyt ,ChangChun ,China > > zdw2000 at gmail.com > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > -- > ======================================================================= > Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu http://www.cmm.upenn.edu > Center for Molecular Modeling -- University of Pennsylvania > Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323 > tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425 > ======================================================================= > If you make something idiot-proof, the universe creates a better idiot. > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- ZhouDawei JiLin Universiyt ,ChangChun ,China zdw2000 at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070927/a290cc03/attachment.htm From chen_shao_hua197 at yahoo.com.tw Thu Sep 27 18:02:21 2007 From: chen_shao_hua197 at yahoo.com.tw (=?big5?q?=B3=AF=20=A4=D6=B5=D8?=) Date: Fri, 28 Sep 2007 00:02:21 +0800 (CST) Subject: [Pw_forum] error in the q2r.out In-Reply-To: <390674394.28281@nudt.edu.cn> Message-ID: <242305.29454.qm@web73006.mail.tp2.yahoo.com> Dear Axel and Jiayu, First,the code is parallel version and the PC cluster is 2cpus/per node . last error produced under 4 nodes(8cpus) running. I saw Jiayu's experienced about this running.So I tested the same running but under 1 nodes(2cpus).Even I don't know why Jiayu's experienced can work. The q2r.out seems fine this time. But the matdyn.out.dos showed the error: p1_374: (22.726562) net_recv failed for fd = 3 p1_374: p4_error: net_recv read, errno = : 104 bm_list_373: p4_error: listener select: -1 rm_l_1_375: (22.726562) net_send: could not write to fd=5, errno = 32 And the error.out showed PGFIO-F-223/formatted write/internal file/illegal P, T or B edit descriptor - value missing. In source file matdyn.F90, at line number 1667 p4_error: latest msg from perror: Bad file descriptor mpiexec: Warning: tasks 0-1 exited with status 1. For matdyn.out.freq ,it showed p1_381: (1.125000) net_recv failed for fd = 3 bm_list_380: p4_error: listener select: -1 p1_381: p4_error: net_recv read, errno = : 104 rm_l_1_382: (1.125000) net_send: could not write to fd=5, errno = 32 The error.out showed PGFIO-F-235/formatted write/unit=20/edit descriptor does not match item type. File name = gam.lines formatted, sequential access record = 6 In source file matdyn.F90, at line number 1797 p4_error: latest msg from perror: Bad file descriptor mpiexec: Warning: tasks 0-1 exited with status 1. What those means ? Best Regards max Physics Department,National Taiwan University,Taiwan --- ??? ?? > > >CSH> Dear forum's members, > >CSH> > > > >[...] > > > >CSH> p5_16906: (12.808594) net_send: could not > write to > >CSH> fd=5, errno = 32 > >CSH> p7_4306: (14.812500) net_send: could not write > to > >CSH> fd=5, errno = 32 > >CSH> p3_16405: (14.820312) net_send: could not > write to > >CSH> fd=5, errno = 32 > >CSH> > >CSH> Does anyone know what it means ? > > > >that means, that all your MPI "slave" processes > died. > >did you check whether your input runs in serial? > > > >cheers, > > axel. > > I have the same problem when i run mpirun -np 4 ph.x > -npool 2 < *.ph.in . But if i > run mpirun -np 4 ph.x -npool 1 < *.ph.in or mpirun > -np 4 ph.x < *.ph.in, the > error would disappear. Is it the meaning of the > memory error? what did you mean > "in serial"? > > Thanks, > Jiayu > > > > > ------------------------------ > Jiayu Dai > National University of Defense Technology, P R China > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ____________________________________________________________________________________ ????????????? - ???? Yahoo!??????http://tw.info.yahoo.com/seal/index.html From chen_shao_hua197 at yahoo.com.tw Thu Sep 27 20:50:53 2007 From: chen_shao_hua197 at yahoo.com.tw (=?big5?q?=B3=AF=20=A4=D6=B5=D8?=) Date: Fri, 28 Sep 2007 02:50:53 +0800 (CST) Subject: [Pw_forum] ambient bulk Indium's structure Message-ID: <833137.53101.qm@web73005.mail.tp2.yahoo.com> Dear forum's members, I am not sure that ambient bulk Indium's structure is Tetragonal Primitive(st) or Tetragonal body-centered(bct).Could anyone know it ? Best Regards max Physics Department,National Taiwan University,Taiwan ____________________________________________________________________________________ ????????????? - ???? Yahoo!??????http://tw.info.yahoo.com/seal/index.html From eyvaz_isaev at yahoo.com Thu Sep 27 21:53:33 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Thu, 27 Sep 2007 12:53:33 -0700 (PDT) Subject: [Pw_forum] ambient bulk Indium's structure In-Reply-To: <833137.53101.qm@web73005.mail.tp2.yahoo.com> Message-ID: <273966.16513.qm@web60314.mail.yahoo.com> Dear Max, Before submitting such a kind question please have a look at textbooks, Handbooks, Internet. You will find lots of interesting information. At ambient conditions In is in bct. Bests, Eyvaz. --- ?? ???? wrote: > Dear forum's members, > > I am not sure that ambient bulk Indium's structure > is > Tetragonal Primitive(st) or Tetragonal > body-centered(bct).Could anyone know it ? > > Best Regards > max > Physics Department,National Taiwan University,Taiwan > > > > > > > > ____________________________________________________________________________________ > ?????????b???A?O???b???w?? - ???W?]?w > Yahoo!?_???w??????http://tw.info.yahoo.com/seal/index.html > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Check out the hottest 2008 models today at Yahoo! Autos. http://autos.yahoo.com/new_cars.html From xianghjun at gmail.com Fri Sep 28 00:06:54 2007 From: xianghjun at gmail.com (Hongjun Xiang) Date: Thu, 27 Sep 2007 16:06:54 -0600 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? In-Reply-To: <46E02710.80303@umn.edu> References: <46E02461.6030104@gmail.com> <46E02710.80303@umn.edu> Message-ID: <46FC297E.5060508@gmail.com> Dear Dr. Cococcioni and others, I tried some test calculations and realized that the pressure changes the electronic structure and thus decreases the volume of the system. If I am interested in the band gap of a finite system under pressure, which method of the followings is correct? (1) Relax the structure under the pressure, and calculate the gap using the relaxed structure under the pressure. (2) Relax the structure under the pressure, and calculate the gap using the relaxed structure without the pressure. It turns out the results from the above procedures are different. Thank you very much. Best regards, Hongjun Xiang ============================================================= H. J. Xiang Postdoctoral Research Associate National Renewable Energy Laboratory http://www4.ncsu.edu/~hxiang/ ============================================================= Matteo Cococcioni wrote: > > Dear Hongjun, > > the electronic enthalpy is not contained in the current official > version of the code (v 3.2). > However it's implemented in the CVS version (README.cvs for > instruction on how to download it) of > the package. Precisely it's contained in the CP part (directory CPV). > Instructions contained in > INPUT_CP should be enough to run a calculation with the electronic > enthalpy. Otherwise feel free to ask for more help. > > Regards, > > Matteo > From marzari at MIT.EDU Fri Sep 28 00:57:18 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 27 Sep 2007 18:57:18 -0400 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? In-Reply-To: <46FC297E.5060508@gmail.com> References: <46E02461.6030104@gmail.com> <46E02710.80303@umn.edu> <46FC297E.5060508@gmail.com> Message-ID: <46FC354E.9070605@mit.edu> Hongjun Xiang wrote: > Dear Dr. Cococcioni and others, > I tried some test calculations and realized that the pressure changes > the electronic structure and thus > decreases the volume of the system. If I am interested in the band gap > of a finite system under pressure, which method of the followings is > correct? > (1) Relax the structure under the pressure, and calculate the gap using > the relaxed structure under the pressure. > (2) Relax the structure under the pressure, and calculate the gap using > the relaxed structure without the pressure. > > It turns out the results from the above procedures are different. > > Thank you very much. > > Best regards, > Hongjun Xiang Dear Xiang, 1) is the correct procedure - but be aware that the band gap is not a property that you can address in DFT - not in principle, and largely not in practice. At best, you could look at DFT predictions of, say, the derivative of the band gap in bulk semiconductors with respect to biaxial strain, and see how well you can do to predict the slope. I think the error would be in the 20-30% range, a bit better than the 100% error on the band gap. nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From marzari at MIT.EDU Fri Sep 28 02:17:56 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 27 Sep 2007 20:17:56 -0400 Subject: [Pw_forum] about metal and molecular dynamics In-Reply-To: <3e36a8fc0709262009y371fa433w924778ee7f70e1ea@mail.gmail.com> References: <3e36a8fc0709260114g6ab290c9n75acd03b475a958c@mail.gmail.com> <7b6913e90709261950m6b542118of540278d77794a82@mail.gmail.com> <3e36a8fc0709262009y371fa433w924778ee7f70e1ea@mail.gmail.com> Message-ID: <46FC4834.10605@mit.edu> Wei Zhou wrote: > thank you for your answer. but I want to know if the pw.x of QE code can > do variable cell dynamics 'npt' using nose to control temperature. or > the cp of QE code can do this .or both of them can do this . The cp code, in its current cvs version, can do md in metals using the "ensemble-DFT" algorithm. It's recent and poorly documented (my fault - e.g. the only smearing that works is "cs", even if "fd" is what is reported...), but we keep working on it, and it's already in a stable/robust form. It uses the nose thermostat. We hadn't really given any thought to the npt dynamics earlier on, but we are testing it these very days. nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From matteo at umn.edu Fri Sep 28 02:57:35 2007 From: matteo at umn.edu (Matteo Cococcioni) Date: Thu, 27 Sep 2007 19:57:35 -0500 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? In-Reply-To: <46FC354E.9070605@mit.edu> References: <46E02461.6030104@gmail.com> <46E02710.80303@umn.edu> <46FC297E.5060508@gmail.com> <46FC354E.9070605@mit.edu> Message-ID: <46FC517F.8000403@umn.edu> Dear Xiang, let me add something to what Nicola wrote. when a system is under pressure also its electronic structure is under pressure. In fact pressure is transferred to the system by electrostatic and Pauli repulsion interactions with the electrons of other atoms/molecule in the pressurizing environment of your system. However in the electronic enthalpy method to impose the pressure, the effect of the collisions with external atoms or molecules is mimicked by a smooth potential of the charge density (like if the colliding particles were smeared in a continuum) that is meant to produce the same average effect. This turns out to work quite well for the structure of the system. The surface electronic state however could be distorted (compressed). To answer to your question (not easy) I suppose one should think of what happens to the system between collisions with external atoms/molecules of the environment. do the electrons relax to their instantaneous ground state corresponding to a compressed ionic configuration? I would guess yes, and probably they do before ions have time to do the same. so I have a preference for answer (2). Anyway this is just my guess, I have not much experience with the electronic relaxation once you release pressure in the compressed configuration. Hope this helps, Matteo Nicola Marzari wrote: > Hongjun Xiang wrote: >> Dear Dr. Cococcioni and others, >> I tried some test calculations and realized that the pressure changes >> the electronic structure and thus >> decreases the volume of the system. If I am interested in the band >> gap of a finite system under pressure, which method of the followings >> is correct? >> (1) Relax the structure under the pressure, and calculate the gap >> using the relaxed structure under the pressure. >> (2) Relax the structure under the pressure, and calculate the gap >> using the relaxed structure without the pressure. >> >> It turns out the results from the above procedures are different. >> >> Thank you very much. >> >> Best regards, >> Hongjun Xiang > > > Dear Xiang, > > 1) is the correct procedure - but be aware that the band gap is > not a property that you can address in DFT - not in principle, and > largely not in practice. At best, you could look at DFT predictions > of, say, the derivative of the band gap in bulk semiconductors > with respect to biaxial strain, and see how well you can do to predict > the slope. I think the error would be in the 20-30% range, a bit better > than the 100% error on the band gap. > > nicola > > --------------------------------------------------------------------- > Prof Nicola Marzari Department of Materials Science and Engineering > 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA > tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -------------- next part -------------- A non-text attachment was scrubbed... Name: matteo.vcf Type: text/x-vcard Size: 294 bytes Desc: not available Url : http://www.democritos.it/pipermail/pw_forum/attachments/20070927/d774b3bc/attachment.vcf From chen_shao_hua197 at yahoo.com.tw Fri Sep 28 08:14:20 2007 From: chen_shao_hua197 at yahoo.com.tw (=?big5?q?=B3=AF=20=A4=D6=B5=D8?=) Date: Fri, 28 Sep 2007 14:14:20 +0800 (CST) Subject: [Pw_forum] ambient bulk Indium's structure In-Reply-To: <273966.16513.qm@web60314.mail.yahoo.com> Message-ID: <869932.51400.qm@web73006.mail.tp2.yahoo.com> Dear Eyvaz, I apologize for what I done. I had checked Ashcroft's textbook before.I thought it is bct and book also record lattice constant = 4.59 A .I use it and do relaxation.But too far from real lattice constant( =3.22 A ,get from internet later).I just do this till 3.5 A then quit.So I spent few weeks to find other reason but never doubt about 4.59 A . This experience make me also doubt book's record "TET" structure. Sorry again.I should checked carefully before ask question. max --- Eyvaz Isaev ?? > Dear Max, > > Before submitting such a kind question please have a > look at textbooks, Handbooks, Internet. > You will find lots of interesting information. > > At ambient conditions In is in bct. > > Bests, > Eyvaz. > > --- ? ?? wrote: > > > Dear forum's members, > > > > I am not sure that ambient bulk Indium's structure > > is > > Tetragonal Primitive(st) or Tetragonal > > body-centered(bct).Could anyone know it ? > > > > Best Regards > > max > > Physics Department,National Taiwan > University,Taiwan > > > > > > > > > > > > > > > > > ____________________________________________________________________________________ > > ????????????? - ???? > > > Yahoo!??????http://tw.info.yahoo.com/seal/index.html > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > ------------------------------------------------------------------- > Prof. Eyvaz Isaev, > Theoretical Physics Department, Moscow State > Institute of Steel & Alloys, Russia, and > Condensed Matter Theory Group, Uppsala University, > Sweden Eyvaz.Isaev at fysik.uu.se, > eyvaz_isaev at yahoo.com > > > > ____________________________________________________________________________________ > Check out the hottest 2008 models today at Yahoo! > Autos. > http://autos.yahoo.com/new_cars.html > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ____________________________________________________________________________________ ??????????? Yahoo!??????????? http://tw.info.yahoo.com/seal/index.html From lukethulin at netscape.net Fri Sep 28 14:25:59 2007 From: lukethulin at netscape.net (lukethulin at netscape.net) Date: Fri, 28 Sep 2007 08:25:59 -0400 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? In-Reply-To: <46FC354E.9070605@mit.edu> References: <46E02461.6030104@gmail.com> <46E02710.80303@umn.edu> <46FC297E.5060508@gmail.com> <46FC354E.9070605@mit.edu> Message-ID: <8C9CFE1DC7C0640-52C-88DB@FWM-M39.sysops.aol.com> I have done such a study of biaxial strain on anatase titania to find the derivative of the band gap using both DFT and GW approximation.? For small strain, the error between the two methods was ~5% and grows rapidly to ~15% as the strain increases.? Maybe this information is useful. Luke Thulin Nanoptek Corp. Maynard, MA -----Original Message----- From: Nicola Marzari To: PWSCF Forum Sent: Thu, 27 Sep 2007 5:57 pm Subject: Re: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? Hongjun Xiang wrote:? > Dear Dr. Cococcioni and others,? > I tried some test calculations and realized that the pressure changes > the electronic structure and thus? > decreases the volume of the system. If I am interested in the band gap > of a finite system under pressure, which method of the followings is > correct?? > (1) Relax the structure under the pressure, and calculate the gap using > the relaxed structure under the pressure.? > (2) Relax the structure under the pressure, and calculate the gap using > the relaxed structure without the pressure.? > > It turns out the results from the above procedures are different.? > > Thank you very much.? > > Best regards,? > Hongjun Xiang? ? Dear Xiang,? ? 1) is the correct procedure - but be aware that the band gap is? not a property that you can address in DFT - not in principle, and? largely not in practice. At best, you could look at DFT predictions? of, say, the derivative of the band gap in bulk semiconductors? with respect to biaxial strain, and see how well you can do to predict? the slope. I think the error would be in the 20-30% range, a bit better? than the 100% error on the band gap.? ? ? nicola? ? ---------------------------------------------------------------------? Prof Nicola Marzari Department of Materials Science and Engineering? 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA? tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu? _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum ________________________________________________________________________ Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20070928/3e32e549/attachment.htm From eyvaz_isaev at yahoo.com Fri Sep 28 16:34:52 2007 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Fri, 28 Sep 2007 07:34:52 -0700 (PDT) Subject: [Pw_forum] ambient bulk Indium's structure In-Reply-To: <869932.51400.qm@web73006.mail.tp2.yahoo.com> Message-ID: <910342.319.qm@web60312.mail.yahoo.com> Dear Max, --- ?? ???? wrote: > I apologize for what I done. OK, no problem, taking into your previous effort to account. > I had checked Ashcroft's textbook before.I thought > it is bct and book also record lattice constant > = 4.59A. Experimental lattice parameters (according to webelements.com) are a=b=3.2523A, c=4.9461A. So, this looks like "c". > I use it and do relaxation.But too far from real > lattice constant( =3.22 A ,get from internet > later).I just do this till 3.5 A then quit.So I > spent few weeks to find other reason but never > doubt about 4.59A. But this one is "a". > This experience make me also doubt book's record > "TET" structure. If you use 2 atoms in the unit cell, "tetragonal" is correct without any doubts. See papers of Simak et al. about In, you will find them interesting for yourself, though they are done using the FP-LMTO method. Bests, Eyvaz. > > Sorry again.I should checked carefully before ask > question. > > max > > --- Eyvaz Isaev ???G > > > Dear Max, > > > > Before submitting such a kind question please have > a > > look at textbooks, Handbooks, Internet. > > You will find lots of interesting information. > > > > At ambient conditions In is in bct. > > > > Bests, > > Eyvaz. > > > > --- ?? ???? wrote: > > > > > Dear forum's members, > > > > > > I am not sure that ambient bulk Indium's > structure > > > is > > > Tetragonal Primitive(st) or Tetragonal > > > body-centered(bct).Could anyone know it ? > > > > > > Best Regards > > > max > > > Physics Department,National Taiwan > > University,Taiwan > > > > > > > > > > > > > > > > > > > > > > > > > > > ____________________________________________________________________________________ > > > ?????????b???A?O???b???w?? - ???W?]?w > > > > > > Yahoo!?_???w??????http://tw.info.yahoo.com/seal/index.html > > > _______________________________________________ > > > Pw_forum mailing list > > > Pw_forum at pwscf.org > > > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > > > > ------------------------------------------------------------------- > > Prof. Eyvaz Isaev, > > Theoretical Physics Department, Moscow State > > Institute of Steel & Alloys, Russia, and > > Condensed Matter Theory Group, Uppsala University, > > Sweden Eyvaz.Isaev at fysik.uu.se, > > eyvaz_isaev at yahoo.com > > > > > > > > > ____________________________________________________________________________________ > > Check out the hottest 2008 models today at Yahoo! > > Autos. > > http://autos.yahoo.com/new_cars.html > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > ____________________________________________________________________________________ > ?]?w?w???????A?O?@?z?? Yahoo!?_???b?????Q?b???s???I > http://tw.info.yahoo.com/seal/index.html > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------- Prof. Eyvaz Isaev, Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, and Condensed Matter Theory Group, Uppsala University, Sweden Eyvaz.Isaev at fysik.uu.se, eyvaz_isaev at yahoo.com ____________________________________________________________________________________ Check out the hottest 2008 models today at Yahoo! Autos. http://autos.yahoo.com/new_cars.html From xianghjun at gmail.com Fri Sep 28 17:50:53 2007 From: xianghjun at gmail.com (Hongjun Xiang) Date: Fri, 28 Sep 2007 09:50:53 -0600 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? In-Reply-To: <390942363.13070@ustc.edu.cn> References: <46E02461.6030104@gmail.com> <46E02710.80303@umn.edu> <46FC297E.5060508@gmail.com> <46FC354E.9070605@mit.edu> <390942363.13070@ustc.edu.cn> Message-ID: <46FD22DD.9050607@gmail.com> Dear Prof. Marzari and Dr. Cococcioni, Thank you for your replies. Both opinions seem reasonable. I will think about this more seriously. Best regards, Hongjun Xiang ============================================================= H. J. Xiang Postdoctoral Research Associate National Renewable Energy Laboratory http://www4.ncsu.edu/~hxiang/ ============================================================= Matteo Cococcioni wrote: > > Dear Xiang, > > let me add something to what Nicola wrote. > > when a system is under pressure also its electronic structure is under > pressure. In fact pressure is transferred to the system > by electrostatic and Pauli repulsion interactions with the electrons > of other atoms/molecule in the pressurizing environment of your > system. However in the electronic enthalpy method to impose the > pressure, the effect of the collisions with external atoms or > molecules is mimicked by a smooth potential of the charge density > (like if the colliding particles were smeared in a continuum) > that is meant to produce the same average effect. This turns out to > work quite well for the structure of the system. The surface > electronic state however could be distorted (compressed). > To answer to your question (not easy) I suppose one should think of > what happens to the system between collisions with external > atoms/molecules of the environment. do the electrons relax to their > instantaneous ground state corresponding to a compressed ionic > configuration? I would guess yes, and probably they do before ions > have time to do the same. so I have a preference for answer (2). > Anyway this is just my guess, I have not much experience with the > electronic relaxation once you release pressure in the compressed > configuration. > > Hope this helps, > > Matteo > > > > Nicola Marzari wrote: >> Hongjun Xiang wrote: >>> Dear Dr. Cococcioni and others, >>> I tried some test calculations and realized that the pressure >>> changes the electronic structure and thus >>> decreases the volume of the system. If I am interested in the band >>> gap of a finite system under pressure, which method of the >>> followings is correct? >>> (1) Relax the structure under the pressure, and calculate the gap >>> using the relaxed structure under the pressure. >>> (2) Relax the structure under the pressure, and calculate the gap >>> using the relaxed structure without the pressure. >>> >>> It turns out the results from the above procedures are different. >>> >>> Thank you very much. >>> >>> Best regards, >>> Hongjun Xiang >> >> >> Dear Xiang, >> >> 1) is the correct procedure - but be aware that the band gap is >> not a property that you can address in DFT - not in principle, and >> largely not in practice. At best, you could look at DFT predictions >> of, say, the derivative of the band gap in bulk semiconductors >> with respect to biaxial strain, and see how well you can do to predict >> the slope. I think the error would be in the 20-30% range, a bit better >> than the 100% error on the band gap. >> >> nicola >> >> --------------------------------------------------------------------- >> Prof Nicola Marzari Department of Materials Science and Engineering >> 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA >> tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Pw_forum mailing list >> Pw_forum at pwscf.org >> http://www.democritos.it/mailman/listinfo/pw_forum From xianghjun at gmail.com Fri Sep 28 17:52:16 2007 From: xianghjun at gmail.com (Hongjun Xiang) Date: Fri, 28 Sep 2007 09:52:16 -0600 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? In-Reply-To: <390982687.13028@ustc.edu.cn> References: <46E02461.6030104@gmail.com> <46E02710.80303@umn.edu> <46FC297E.5060508@gmail.com> <46FC354E.9070605@mit.edu> <390982687.13028@ustc.edu.cn> Message-ID: <46FD2330.1020008@gmail.com> Dear Luke, Thanks a lot for the information. It would be useful for my study. Best regards, Hongjun Xiang ============================================================= H. J. Xiang Postdoctoral Research Associate National Renewable Energy Laboratory http://www4.ncsu.edu/~hxiang/ ============================================================= lukethulin at netscape.net wrote: > I have done such a study of biaxial strain on anatase titania to find > the derivative of the band gap using both DFT and GW approximation. > For small strain, the error between the two methods was ~5% and grows > rapidly to ~15% as the strain increases. Maybe this information is useful. > > Luke Thulin > Nanoptek Corp. > Maynard, MA > From degironc at sissa.it Fri Sep 28 18:15:59 2007 From: degironc at sissa.it (Stefano de Gironcoli) Date: Fri, 28 Sep 2007 18:15:59 +0200 Subject: [Pw_forum] Is Electronic-Enthalpy Functional for Finite Systems Under Pressure implemented in espresso-3.2? In-Reply-To: <46FD22DD.9050607@gmail.com> References: <46E02461.6030104@gmail.com> <46E02710.80303@umn.edu> <46FC297E.5060508@gmail.com> <46FC354E.9070605@mit.edu> <390942363.13070@ustc.edu.cn> <46FD22DD.9050607@gmail.com> Message-ID: <46FD28BF.9000104@sissa.it> Dear all, I just want to add my two cents to this very nice discussion. Possibly the answer is different depending on whether the pressurizing medium is in the gas or in the liquid/solid phase. In the former case I would tend to agree with Matteo Cococcioni, at least if the average time between impacts is longer than typical electronic relaxation times (femptosecond), while I would tend to agree with Nicola Marzari for liquid/solid pressurizing media since in that case I would expect the contact to be rather continuous and constant, or at least fluctuating around some average value ... stefano --- Stefano de Gironcoli - SISSA and DEMOCRITOS Hongjun Xiang wrote: > Dear Prof. Marzari and Dr. Cococcioni, > > Thank you for your replies. > Both opinions seem reasonable. I will think about this more seriously. > > Best regards, > Hongjun Xiang > > ============================================================= > H. J. Xiang > Postdoctoral Research Associate > National Renewable Energy Laboratory > http://www4.ncsu.edu/~hxiang/ > ============================================================= > Matteo Cococcioni wrote: >> >> Dear Xiang, >> >> let me add something to what Nicola wrote. >> >> when a system is under pressure also its electronic structure is >> under pressure. In fact pressure is transferred to the system >> by electrostatic and Pauli repulsion interactions with the electrons >> of other atoms/molecule in the pressurizing environment of your >> system. However in the electronic enthalpy method to impose the >> pressure, the effect of the collisions with external atoms or >> molecules is mimicked by a smooth potential of the charge density >> (like if the colliding particles were smeared in a continuum) >> that is meant to produce the same average effect. This turns out to >> work quite well for the structure of the system. The surface >> electronic state however could be distorted (compressed). >> To answer to your question (not easy) I suppose one should think of >> what happens to the system between collisions with external >> atoms/molecules of the environment. do the electrons relax to their >> instantaneous ground state corresponding to a compressed ionic >> configuration? I would guess yes, and probably they do before ions >> have time to do the same. so I have a preference for answer (2). >> Anyway this is just my guess, I have not much experience with the >> electronic relaxation once you release pressure in the compressed >> configuration. >> >> Hope this helps, >> >> Matteo >> >> >> >> Nicola Marzari wrote: >>> Hongjun Xiang wrote: >>>> Dear Dr. Cococcioni and others, >>>> I tried some test calculations and realized that the pressure >>>> changes the electronic structure and thus >>>> decreases the volume of the system. If I am interested in the band >>>> gap of a finite system under pressure, which method of the >>>> followings is correct? >>>> (1) Relax the structure under the pressure, and calculate the gap >>>> using the relaxed structure under the pressure. >>>> (2) Relax the structure under the pressure, and calculate the gap >>>> using the relaxed structure without the pressure. >>>> >>>> It turns out the results from the above procedures are different. >>>> >>>> Thank you very much. >>>> >>>> Best regards, >>>> Hongjun Xiang >>> >>> >>> Dear Xiang, >>> >>> 1) is the correct procedure - but be aware that the band gap is >>> not a property that you can address in DFT - not in principle, and >>> largely not in practice. At best, you could look at DFT predictions >>> of, say, the derivative of the band gap in bulk semiconductors >>> with respect to biaxial strain, and see how well you can do to predict >>> the slope. I think the error would be in the 20-30% range, a bit better >>> than the 100% error on the band gap. >>> >>> nicola >>> >>> --------------------------------------------------------------------- >>> Prof Nicola Marzari Department of Materials Science and Engineering >>> 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA >>> tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu >>> ------------------------------------------------------------------------ >>> >>> >>> _______________________________________________ >>> Pw_forum mailing list >>> Pw_forum at pwscf.org >>> http://www.democritos.it/mailman/listinfo/pw_forum > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From brsahu at physics.utexas.edu Fri Sep 28 18:35:26 2007 From: brsahu at physics.utexas.edu (brsahu at physics.utexas.edu) Date: Fri, 28 Sep 2007 11:35:26 -0500 Subject: [Pw_forum] benchmark calculations for parallel PWSCF Message-ID: <20070928113526.5vxbqd2r9mo0ww0s@mail.ph.utexas.edu> Dear PWSCF users, Is it possibe to get a info on benchmark PWSCF SCF parallel calculations (CPU time versus number of processors) for a particular physical system on any computing machines (linux clusters, IBM (Power series), SGI, HP etc. Is it possible to share the numbers with me if you have such numbers? I am aware that lot depends on what type of processors, what communication channel, what libraries etc. But at this stage I am not looking at a particular set of hardware/software set-up but only a rough idea of how SCF parallel run scales with number of processors for a given system size Bhagawan Sahu Microelectronics Research Center University of Texas at Austin 10100 Burnet Road, Bldg. 160 MER 1.606H, Mail code R9900, TX 78758-4445 e-mail: brsahu at physics.utexas.edu Phone: (512)-471-1344 FAX: (512)-471-8575 From chen_shao_hua197 at yahoo.com.tw Fri Sep 28 20:05:58 2007 From: chen_shao_hua197 at yahoo.com.tw (=?big5?q?=B3=AF=20=A4=D6=B5=D8?=) Date: Sat, 29 Sep 2007 02:05:58 +0800 (CST) Subject: [Pw_forum] ambient bulk Indium's structure In-Reply-To: <910342.319.qm@web60312.mail.yahoo.com> Message-ID: <237513.90842.qm@web73002.mail.tp2.yahoo.com> Dear Eyvaz, Thank you very much for providing the useful website and papers information. max --- Eyvaz Isaev ?? > Dear Max, > > --- ? ?? wrote: > > I apologize for what I done. > > OK, no problem, taking into your previous effort to > account. > > > I had checked Ashcroft's textbook before.I thought > > it is bct and book also record lattice constant > > = > 4.59A. > > Experimental lattice parameters (according to > webelements.com) are a=b=3.2523A, c=4.9461A. > So, this looks like "c". > > > I use it and do relaxation.But too far from real > > > lattice constant( =3.22 A ,get from internet > > later).I just do this till 3.5 A then quit.So I > > spent few weeks to find other reason but never > > doubt about 4.59A. > But this one is "a". > > > This experience make me also doubt book's record > > > "TET" structure. > If you use 2 atoms in the unit cell, "tetragonal" is > correct without any doubts. > > See papers of Simak et al. about In, you will find > them interesting for yourself, though they are done > using the FP-LMTO method. > > Bests, > Eyvaz. > > > > Sorry again.I should checked carefully before ask > > question. > > > > max > > > > --- Eyvaz Isaev ?? > > > > > Dear Max, > > > > > > Before submitting such a kind question please > have > > a > > > look at textbooks, Handbooks, Internet. > > > You will find lots of interesting information. > > > > > > At ambient conditions In is in bct. > > > > > > Bests, > > > Eyvaz. > > > > > > --- ? ?? > wrote: > > > > > > > Dear forum's members, > > > > > > > > I am not sure that ambient bulk Indium's > > structure > > > > is > > > > Tetragonal Primitive(st) or Tetragonal > > > > body-centered(bct).Could anyone know it ? > > > > > > > > Best Regards > > > > max > > > > Physics Department,National Taiwan > > > University,Taiwan > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > ____________________________________________________________________________________ > > > > ????????????? - ???? > > > > > > > > > > Yahoo!??????http://tw.info.yahoo.com/seal/index.html > > > > > _______________________________________________ > > > > Pw_forum mailing list > > > > Pw_forum at pwscf.org > > > > > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > > > > > > > > > > ------------------------------------------------------------------- > > > Prof. Eyvaz Isaev, > > > Theoretical Physics Department, Moscow State > > > Institute of Steel & Alloys, Russia, and > > > Condensed Matter Theory Group, Uppsala > University, > > > Sweden Eyvaz.Isaev at fysik.uu.se, > > > eyvaz_isaev at yahoo.com > > > > > > > > > > > > > > > ____________________________________________________________________________________ > > > Check out the hottest 2008 models today at > Yahoo! > > > Autos. > > > http://autos.yahoo.com/new_cars.html > > > _______________________________________________ > > > Pw_forum mailing list > > > Pw_forum at pwscf.org > > > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > > > > > > > > > ____________________________________________________________________________________ > > ??????????? > Yahoo!??????????? > > http://tw.info.yahoo.com/seal/index.html > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > ------------------------------------------------------------------- > Prof. Eyvaz Isaev, > Theoretical Physics Department, Moscow State > Institute of Steel & Alloys, Russia, and > Condensed Matter Theory Group, Uppsala University, > Sweden Eyvaz.Isaev at fysik.uu.se, > eyvaz_isaev at yahoo.com > > > > ____________________________________________________________________________________ > Check out the hottest 2008 models today at Yahoo! > Autos. > http://autos.yahoo.com/new_cars.html > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > ____________________________________________________________________________________ ?????????????????????? Yahoo!?????????? http://tw.mobile.yahoo.com/texts/mail.php