merged wannier center and coordinate output file

Axel akoh... at
Mon Jul 9 16:03:48 UTC 2007

thanks teo and fawzi for your explanations.

unfortunately, it raises another question,
that i'll better ask in a separate thread.

it is obvious that i can fix everything up with
postprocessing later, but i'd rather have everything
merged in one file from within the code so that i don't
have to synchronize multiple sets of parameters all the time,
preferably binary as postprocessing text format files can
become very time consuming (e.g. in QM/MM!) and also
for the kind of calculations that i have in mind the
text files would be _huge_ and writing them would also
affect parallel performance.

so if cp2k cannot do it as it is right now, can you give
me any hints, on how i could implement this without having
to read through all of the source code? the way i see it,
i'd need to allocate some memory somewhere and potentially
have a proxy topology for it, copy the coordinates into the
allocated memory, then the wannier centers, apply PBC and
then write using the normal write functionality (so one can
have either .xyz, .pdb, .dcd or whatever else format will
be supported in the future).

apropos parallel performance, would there be a way to tell
cp2k to keep files open during a run? on several machines
that we are running on, frequent open/close of files can
have serious impact on (parallel) performance.


More information about the CP2K-user mailing list