[Wien] vec2old_lapw

[Tomohiko Tasaka]

Hi,

thanks for your reply. this fail's occurred in non iterative calculation.
Umm, I think that vec2old within "lapw2 -p -qtl" always fails in the system which
has a local scratch setting(i.e. the value of SCRATCH is not "./"). because vec2old
only copies the vector files into the (same) scratch directory of each client node...

* here is a invoking part of vec2old in lapw2para script:
420 qtl:
421 echo "calculating QTL's from parallel vectors"
422 vec2old_lapw -p -local -$updn
423 $exe $def.def $maxproc
424 rm $tmp* >&/dev/null
425 rm .lapw2${cmplx}para.$$.`hostname` >&/dev/null
426 echo "DONE" >.lapw2para
427 exit 0

is there anyone who ran into the same problem?

Thanks again, with best regards,
Tomo


-----Original Message-----
From: wien-bounces at zeus.theochem.tuwien.ac.at [mailto:wien-bounces at zeus.theochem.tuwien.ac.at] On Behalf Of Laurence Marks
Sent: Monday, March 07, 2011 10:31 PM
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] vec2old_lapw

The vec2old_lapw script is used in a number of places in afm and
iterative calculations. If you want a special version just for a qtl
calculation I suggest that you change it to vec2old_mine and add an
extry in x_lapw, while retaining the original.

I may have misunderstood -- does vec2old fail in a normal iterative
calculation (e.g. -it or -noHinv)?

2011/3/7 Tomohiko Tasaka <tasaka at affinity-science.com>:
> Dear All,
>
> when executing "x lapw2 -qtl -p" calculation on PC cluster
> system(with local SCRATCH), I encounter the following error:
>
> Error in LAPW2
>  'l2main' - error reading parallel vectors
>
> then, I found that it's because that vec2old_lapw script fails to copy
> the vector files of other machines. to deal with this problem, I modified
> vec2old_lapw as below(you can find attached):
>
> //-----------------------------------------------------------
> $ diff vec2old_lapw vec2old_lapw.orig
> 9d8
> < set remotecopy = rcp
> 12d10
> < set exechost=`hostname -f`
> 84c82
> <           set status=`$remote $machines[$i] "cd $PWD;$remotecopy $SCRATCH/$file.vector${updn}_$i
> ${exechost}:${SCRATCH}/$file.vector${updn}_$i"`
> ---
>>           set status=`$remote $machines[$i] "cd $PWD;cp $SCRATCH/$file.vector${updn}_$i
> $file.vector${updn}_$i"`
> //-----------------------------------------------------------
>
> so, what do you think about this modification?
> if there are any problems, incompatibilities, or better solutions, please advise.
>
> Thanks in advance for your time and consideration.
>
> With best regards,
> Tomohiko
>
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>
>



-- 
Laurence Marks
Department of Materials Science and Engineering
MSE Rm 2036 Cook Hall
2220 N Campus Drive
Northwestern University
Evanston, IL 60208, USA
Tel: (847) 491-3996 Fax: (847) 491-7820
email: L-marks at northwestern dot edu
Web: www.numis.northwestern.edu
Chair, Commission on Electron Crystallography of IUCR
www.numis.northwestern.edu/
Electron crystallography is the branch of science that uses electron
scattering and imaging to study the structure of matter.
_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien