Page 1 of 1
denser K-point grid for BSE
Posted: Fri Sep 12, 2025 3:47 pm
by xjxiao
Dear all,
After calculating GW and obtaining ndb.qp, I want to use a denser K-point grid than that for calculating GW when calculating BSE. How should I proceed? How to modify the K-point grid?
Thanks!
Xiao
Re: denser K-point grid for BSE
Posted: Mon Sep 15, 2025 7:19 am
by Daniele Varsano
Dear Xiangjun,
you need to perform a new nscf calculation with a denser k point grids.
Best,
Daniele
Re: denser K-point grid for BSE
Posted: Mon Sep 15, 2025 1:54 pm
by xjxiao
Daniele Varsano wrote: ↑Mon Sep 15, 2025 7:19 am
Dear Xiangjun,
you need to perform a new nscf calculation with a denser k point grids.
Best,
Daniele
Dear Daniele,
Thanks for your help!
If the grid density is changed in this way, can the ndb.QP obtained from the GW calculation in BSE be used for energy correction? KfnQPdb=" E < ./SAVE/ndb.QP"
Xiao
Re: denser K-point grid for BSE
Posted: Mon Sep 15, 2025 2:26 pm
by Daniele Varsano
Dear Xiangjun,
You can have a try, but I'm not sure you can use the previous database. What you can do is:
1) Extract scissor and stretching parameter from the old calculation and apply it to the new one (if scissor approximation does apply)
2) Repeat GW calculations with the new grid. You do not need to calculate for all the K points. A subset is enough and Yambo will perform an interpolation for the rest.
Best,
Daniele