[xcrysden] Runtime error in rendering the Fermi surface
Martin Kroeker
martin at ruby.chemie.uni-freiburg.de
Wed Dec 9 09:27:28 CET 2015
Glad to know you got it working somehow. It could be interesting to
compare the files if you still have the one from the mpi calculation:
At least it occurs to me that the current wn_readbakgen code makes some rather
optimistic assumptions about the layout of the file it tries to read,
skipping a line here and there because it "knows" that there will be a
table heading like "weights of k-points" on it without actually checking.
Any additional output like "hi, this is your friendly mpi version" is
sure to throw it off balance.
Regards,
Martin
>
> Yep. The error seems coming from the way that I put the output1 files
> together after a mpi+k point paralleled run. I tried the post calculation
> using k point parallelization only and put the files together as cat
> case.output1_$i>>case.output1. Everything works fine now.
>
> Thanks for all your suggestions.
>
> Regards,
> Fermin
>
> 2015-12-08 23:33 GMT+08:00 Martin Kroeker <
> martin at ruby.chemie.uni-freiburg.de>:
>
> > > I have searched around the line "weight of sum" but haven't discovered
> > any
> > > stars. It looks like this:
> > Then it would appear (to me at least) that the program is expecting to
> > read more than 354 entries and hits the word "sum" while still trying to
> > obtain numbers. The number of irreducible k-point indices in the file is
> > read from one of the first lines however. Not sure if I am confused or
> > the program is, perhaps something went wrong when you joined the output
> > files ?
> > >
> > >
> > > weights of k-points:
> > > 1 1 1.000000 0.002841
> > > 2 2 2.000000 0.005682
> > > 3 3 2.000000 0.005682
> > > 4 4 2.000000 0.005682
> > > 5 5 2.000000 0.005682
> > > 6 6 2.000000 0.005682
> > > ........
> > > 349 481 1.000000 0.002841
> > > 350 482 2.000000 0.005682
> > > 351 483 2.000000 0.005682
> > > 352 484 2.000000 0.005682
> > > 353 485 2.000000 0.005682
> > > 354 486 2.000000 0.005682
> > > sum of weights: 2.000000
> > > internal and cartesian k-vectors:
> > > 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
> > > 0.00000 0.00000 0.09091 0.00000 0.00000 0.03085
> > > 0.00000 0.00000 0.18182 0.00000 0.00000 0.06170
> > > 0.00000 0.00000 0.27273 0.00000 0.00000 0.09256
> > > ........
> > >
> > > I also browsed through the whole file and it looks like all the numbers
> > are
> > > properly printed.
> > >
> > > Regards,
> > > Fermin
> > >
> > > 2015-12-08 17:54 GMT+08:00 Martin Kroeker <
> > > martin at ruby.chemie.uni-freiburg.de>:
> > >
> > > > Sorry, I did not check the code at first - at line 125 of
> > wn_readbakgen.f
> > > > it
> > > > tries to read the table that follows the "weights of k-points:" line in
> > > > the case.outputkgen, so the problem area is probably many lines from
> > the
> > > > end of the file. Try searching the file for the "sum of weights:" label
> > > > that follows the table - perhaps either the values in the k-point index
> > > > column
> > > > above it are replaced by stars or the "sum of weights:" line appears
> > > > earlier than expected (too few lines in the table for whatever reason)
> > > >
> > > > > Hi Dr. Kroeker,
> > > > >
> > > > > Thanks for the reply.
> > > > > Do you mean the file "case.outputkgen"? I checked this one and it
> > looks
> > > > ok.
> > > > > Here is some lines by the end of the file:
> > > > > 0.50000 0.37500 0.90909 703.99998
> > 527.99999
> > > > > 1279.99996
> > > > > 0.50000 0.50000 0.00000 703.99998
> > 703.99998
> > > > > 0.00000
> > > > > 0.50000 0.50000 0.09091 703.99998
> > 703.99998
> > > > > 128.00000
> > > > > 0.50000 0.50000 0.18182 703.99998
> > 703.99998
> > > > > 255.99999
> > > > > 0.50000 0.50000 0.27273 703.99998
> > 703.99998
> > > > > 383.99999
> > > > > 0.50000 0.50000 0.36364 703.99998
> > 703.99998
> > > > > 511.99999
> > > > > 0.50000 0.50000 0.45455 703.99998
> > 703.99998
> > > > > 639.99998
> > > > >
> > > > > Regards,
> > > > > Fermin
> > > > >
> > > > >
> > > > > 2015-12-08 16:31 GMT+08:00 Martin Kroeker <
> > > > > martin at ruby.chemie.uni-freiburg.de>:
> > > > >
> > > > > > > *"At line 125 of file wn_readbakgen.f (unit=8,
> > > > file='case.outputkgen').
> > > > > > > Fortran runtime error: Bad integer for item 1 in list input"*
> > > > > >
> > > > > > Can you check the (end of the) file for anything non-numeric in the
> > > > first
> > > > > > column ?
> > > > > > Old-style fortran programs have the nasty habit of printing stars
> > > > instead
> > > > > > of
> > > > > > digits if the number is bigger than the fixed field width
> > allocated to
> > > > > > it in the code, e.g. if the program tries to print the band number
> > with
> > > > > > a maximum of five characters you will see
> > 99998...99999...*****...*****
> > > > > >
> > > > > >
> > > > > > --
> > > > > > Dr. Martin Kroeker martin at ruby.chemie.uni-freiburg.de
> > > > > > c/o Prof.Dr. Caroline Roehr
> > > > > > Institut fuer Anorganische und Analytische Chemie der Universitaet
> > > > Freiburg
> > > > > >
> > > > > > _______________________________________________
> > > > > > XCrySDen mailing list
> > > > > > XCrySDen at democritos.it
> > > > > > http://www.democritos.it/mailman/listinfo/xcrysden
> > > > > >
> > > >
> > > > > _______________________________________________
> > > > > XCrySDen mailing list
> > > > > XCrySDen at democritos.it
> > > > > http://www.democritos.it/mailman/listinfo/xcrysden
> > > >
> > > > --
> > > > Dr. Martin Kroeker martin at ruby.chemie.uni-freiburg.de
> > > > c/o Prof.Dr. Caroline Roehr
> > > > Institut fuer Anorganische und Analytische Chemie der Universitaet
> > Freiburg
> > > >
> > > > _______________________________________________
> > > > XCrySDen mailing list
> > > > XCrySDen at democritos.it
> > > > http://www.democritos.it/mailman/listinfo/xcrysden
> > > >
> >
> > > _______________________________________________
> > > XCrySDen mailing list
> > > XCrySDen at democritos.it
> > > http://www.democritos.it/mailman/listinfo/xcrysden
> >
> > --
> > Dr. Martin Kroeker martin at ruby.chemie.uni-freiburg.de
> > c/o Prof.Dr. Caroline Roehr
> > Institut fuer Anorganische und Analytische Chemie der Universitaet Freiburg
> >
> > _______________________________________________
> > XCrySDen mailing list
> > XCrySDen at democritos.it
> > http://www.democritos.it/mailman/listinfo/xcrysden
> >
> _______________________________________________
> XCrySDen mailing list
> XCrySDen at democritos.it
> http://www.democritos.it/mailman/listinfo/xcrysden
--
Dr. Martin Kroeker martin at ruby.chemie.uni-freiburg.de
c/o Prof.Dr. Caroline Roehr
Institut fuer Anorganische und Analytische Chemie der Universitaet Freiburg
More information about the XCrySDen
mailing list