Dear Duy, I have no experience with hexagonal systems. but the first three results seem reasonable. maybe some array is not big enough to contain your system? in my experience it's usually a problem of this kind causing errors.
Matteo Duy Le wrote: > Dear Matteo and all, > > I am trying to calculated the U with resp_mat.f90 code for colbalt. > > The primitive cell has 16 Co. > The code extrapolated to 2x2x2, 3x3x3 super cell successfully. > > However, when it came to 4x4x4 super cell, the code was initially > doing pretty fine, and in the end it gave an U1, but the code did not > finish normally. Then error came (see below). > > The worse thing is that the value of U1 given in 4x4x4 extrapolation > is nonsense. > > 16 3.30938654705739 > 128 3.59144217433734 > 432 3.61280367128581 > 1024 -0.703549684908895 > > I did a quick search and found that you have mentioned the error of > resp_mat.f90 code with hexagonal lattice (which my system is). Are the > error you was talking about and the error I am facing with identical? > Do you have the resp_mat.f90 fixed? > > Thank you. > > *-----------------------------------ERROR----------------------------* > forrtl: severe (174): SIGSEGV, segmentation fault occurred > Image PC Routine Line > Source > libpthread.so.0 000000397480DE80 Unknown Unknown > Unknown > libc.so.6 0000003973C71083 Unknown Unknown > Unknown > libc.so.6 0000003973C74CCC Unknown Unknown > Unknown > r.x 00000000006496CD Unknown Unknown > Unknown > r.x 000000000041111C Unknown Unknown > Unknown > r.x 00000000004065C2 Unknown Unknown > Unknown > libc.so.6 0000003973C1D8B4 Unknown Unknown > Unknown > r.x 00000000004064E9 Unknown Unknown > Unknown > > > *-----------------------------------OLD > MESSAGE---------------------------- > Matteo Cococcioni* matteo at umn.edu > <mailto:pw_forum%40pwscf.org?Subject=%5BPw_forum%5D%20resp_mat.f90&In-Reply-To=> > /Fri Dec 5 16:56:00 CET 2008/ > > * Previous message: [Pw_forum] Strange things when calculating > bands using pw.x > <http://www.democritos.it/pipermail/pw_forum/2008-December/010926.html> > * Next message: [Pw_forum] Scf can converge, But Band can not > converge > <http://www.democritos.it/pipermail/pw_forum/2008-December/010881.html> > * *Messages sorted by:* [ date ] > > <http://www.democritos.it/pipermail/pw_forum/2008-December/date.html#10880> > [ thread ] > > <http://www.democritos.it/pipermail/pw_forum/2008-December/thread.html#10880> > [ subject ] > > <http://www.democritos.it/pipermail/pw_forum/2008-December/subject.html#10880> > [ author ] > > <http://www.democritos.it/pipermail/pw_forum/2008-December/author.html#10880> > > > Dear all, > > I have recently realized that the code resp_mat.f90, contained in some > tutorials on LDA+U calculations (e.g., > from http://vlab.msi.umn.edu/events/lecture.shtml) and used to compute > > > U, has some problems and doesn't work as expected. > I have realized this on a hexagonal system but I'm not sure yet if the > problem manifests itself also with other crystal structures. > > Hoping that not many people were using it I apologize for the > > > inconvenient. I also hope to provide a fix in short time. > > regards, > > Matteo > -------------- 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/20081205/7359fd01/attachment.vcf > > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > -- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Matteo Cococcioni Department of Chemical Engineering and Materials Science, University of Minnesota 421 Washington Av. SE Minneapolis, MN 55455 Tel. +1 612 624 9056 Fax +1 612 626 7246 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
