Page 1 of 1

vef.py and vef.pl

Posted: Thu Nov 14, 2019 7:15 am
by ibrahim
Dear All,

Recently I found something strange related to vef.py and .pl, the two scripts provide different results for forces, attached is the OUTCAR(1). What happened in that case and which one should I use? With the same vasp version but different installation everything is OK (OUTCAR(2)).

Moreover, when I use newer vasp version (5.4.4), I didn't find "FORCES: max atom, RMS" written in OUTCAR, why?

Many thanks in advance,
Ibrahim

Re: vef.py and vef.pl

Posted: Thu Nov 14, 2019 4:43 pm
by graeme
The newer version of vasp was not built with our vtstcode, which is why that force information is not printed. You can grep the OUTCAR for VTST to check if that code was compiled with your version of vasp.

The difference between vef.pl and vef.py is subtle. The vef.pl script is fine. The vef.py script uses methods in ASE. The issue is that ASE reads any constraints (e.g. frozen atoms) from the POSCAR or CONTCAR file and then applies those constraints to the forces (zeros those components) in the OUTCAR file. If you include a POSCAR or CONTCAR file in the directory with your OUTCAR file, the forces from the two scripts should be the same.

Re: vef.py and vef.pl

Posted: Fri Nov 15, 2019 12:56 am
by ibrahim
Thanks Prof. Graeme.