[gpaw-users] Hybrid Functionals
Mikkel Strange
mikst at fysik.dtu.dk
Fri Jan 15 10:46:50 CET 2016
Dear Joan,
I think the problem is that forces are not implemented for hybrid functionals yet.
Does it run if you only ask for the energy?
Best Regards,
Mikkel
________________________________
From: gpaw-users-bounces at listserv.fysik.dtu.dk [gpaw-users-bounces at listserv.fysik.dtu.dk] on behalf of Joan Jones [joanj700 at gmail.com]
Sent: Wednesday, January 13, 2016 11:39 PM
To: gpaw-users at listserv.fysik.dtu.dk
Subject: [gpaw-users] Hybrid Functionals
Dear users,
I've tried to work with B3LYP exchange-correlation functional, but it doesn't work, and I got the error
File "gpaw/gpaw/forces.py", line 17, in calculate
assert not isinstance(ham.xc, HybridXCBase)
AssertionError
AssertionError
assert not isinstance(ham.xc, HybridXCBase)
GPAW CLEANUP (node 7): <type 'exceptions.AssertionError'> occurred. Calling MPI_Abort!
GPAW CLEANUP (node 6): <type 'exceptions.AssertionError'> occurred. Calling MPI_Abort!
AssertionError
GPAW CLEANUP (node 5): <type 'exceptions.AssertionError'> occurred. Calling MPI_Abort!
assert not isinstance(ham.xc, HybridXCBase)
AssertionError
This is part of my input:
from ase.lattice import bulk
from ase.io<http://ase.io> import read,write, PickleTrajectory
from ase.visualize import view
from gpaw import GPAW, PW, Mixer,FermiDirac
from gpaw.eigensolvers import RMM_DIIS
from ase.optimize import BFGS,FIRE,QuasiNewton
from gpaw.xc.hybridk import HybridXC
calc = GPAW(mode=PW(280),
mixer=Mixer(nmaxold=5, beta=0.1, weight=75),
nbands=n,
convergence={'energy': 0.0005,
'density': 1.0e-4,
'eigenstates': 5.0e-8,
'bands': 'occupied'},
kpts=(16,16,16),
occupations=FermiDirac(0.1),
txt='calc{0}.out'.format(str(i)),
xc=HybridXC('B3LYP'),
eigensolver='rmm-diis',
)
I would like to know if it is necessary another parameter to work well the input.
Also, in GPAW's website I found a section titled "HybridFunctionals(https://wiki.fysik.dtu.dk/gpaw/devel/projects/hybrids.html?highlight=hybrid%20functionals)", but the files there aren't. It woud be fine if files would be shared with the users.
Best regards,
Joan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/gpaw-users/attachments/20160115/7c629797/attachment.html>
More information about the gpaw-users
mailing list