All displacements in a dataset must run as a single job?

Phonons, DFPT, electron-phonon, electric-field response, mechanical response…

Moderators: mverstra, joaocarloscabreu

Locked
uma
Posts: 33
Joined: Fri Apr 26, 2013 3:43 pm

All displacements in a dataset must run as a single job?

Post by uma » Mon Jul 14, 2014 2:09 pm

I wish to get the DDB file with displacement of 108 atoms. Should I run this as a single job only..? Due to wall time restriction, I submitted the first 50 atoms as one job and then the remaining atoms displaced in the 2nd job. I have all the output files, but only 1 DDB which has only the data(2nd derivative of energy) for the 2nd job. I was able to use the mrgddb tool but I get 'missing information in ddb for 1111' when I use the anaddb tool. Can you suggest a way to get around this..?

Uma,
Senior PostDoc, TUD, Germany

bxu
Posts: 2
Joined: Thu Oct 06, 2011 12:21 pm

Re: All displacements in a dataset must run as a single job?

Post by bxu » Wed Aug 06, 2014 12:48 am

Hi Uma,

If you split the DDB calculation, you can either do individuel calculation in different directories or rename the files before coping back. The idea is to avoid overwritting with the same name. Even with separate directories, you still need to rename DDB files before mrgddb.

Best,
Bin

Boris
Posts: 128
Joined: Tue Feb 16, 2010 10:13 am
Location: France

Re: All displacements in a dataset must run as a single job?

Post by Boris » Wed Aug 06, 2014 7:14 pm

----------------------------------------------------------
Boris Dorado
Atomic Energy Commission
France
----------------------------------------------------------

Locked