[gpaw-users] Ground state calculation in GW

[Jingjing SHAO]

Dear gpaw-user,

I have a question in regards to the ground state calculation step in GW. As
described in the tutorial, there are two ways to get the unoccupied bands:
1. considering the unocc. bands in the SCF cycles already, without the
diagonalization of the Hamiltonian. 2. only taking the occ.bands (default)
into account for the ground state and then using the diagonalization of the
Hamiltonian to get the unoccupied bands (with ecut or nbands keywords to
save the 'full' Hamiltonian.)
In principle, these two options should give the same results.? However,
when I compare those two procedures, the resulting bands have somehow small
energetic shifts according to their own fermi energy level.
So, in practice is one favourable than the other? Or the small energetic
shifts are only the numerical artefact?

Thank you very much for your kind help and I look forward to your reply.

Best regards,
Jingjing
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/gpaw-users/attachments/20200624/e4228783/attachment.html>