The ypp -k

[sdwang]

Dear developers:
I want to kown the what does the option ypp -k mean? Namely, -k <opt> :BZ Grid generator [(k)pt,(q)pt,(l)ongitudinal]?
Thanks!

[claudio]

Dear Shudong Wang

ypp -k k is used to generated shifted regular grid of k-point.

The idea it that if you have performed a GW calculation on a given k-point grid,
and then you want to calculated GW correction to a point not in this grid, you can shift the k-point grid
in such a way to include the new point and preserve the same q vectors (the differences from the k points q=k-k')
In this way you can use the same dielectric constant, that depends from q, and without recalculate it.

Claudio

[sdwang]

Thanks!

Dear Shudong Wang

ypp -k k is used to generated shifted regular grid of k-point.

The idea it that if you have performed a GW calculation on a given k-point grid,
and then you want to calculated GW correction to a point not in this grid, you can shift the k-point grid

Claudio

I have three questions about your interpretation:
1)
>ypp -k k , I get
#
bzgrids # [R] BZ Grid
K_grid # [R] K-grid analysis
cooIn= "iku" # Points coordinates (in) cc/rlu/iku
cooOut= "rlu" # Points coordinates (out) cc/rlu/iku
% GWKpts
0.000000 | 0.00000 | 0.000000 | # GW K-point grid
%
#NoWeights # Do not print K-points weight
#ForceUserPts # Do not check the correcteness of the user points
inthe GWkpts, what K-ponits should I input?
2)
In this way, I got the shifted k-point grid, but how this shifted k grid can be used in my GW calculation? In the input file of GW , the k-point grid can be altered according to the shifted ones?
3)
In yambo,the Brilouin zone used the reqular point.But I want to get the GW band structures of 2D hexagonal unit cell like graphene.How can I do this in yambo??
Thanks!

[claudio]

Dear Shudong Wang

about your questions, check this page:

http://www.yambo-code.org/doc/inputs/ypp_kk.php

where it is explained how to used shifted grids to calculate GW correcions in generic k-points.

Claudio

[damao4361556]

Hi All

According to the tutorials, first we need to the Yambo on the uniform k, the ypp -k k option shift the k to another k' point. Then we need to put the new k' point into the input file of dft codes, i.e., pwscf. Finally, run the yambo again.

My question is, why don't we run the nscf with the desired K (including k and k') and get the GW correction on all the k points, thus avoid running the dft code and YAMBO again?

I am still unclear the purpose of ypp -k k option.

thanks

[Daniele Varsano]

Dear Eric,
not always you know exactly in advance the kpoints you want to calculate the QP corrections,
it may happen you realize there are interesting features in K points you did not sample. The utility permit to calculate them avoiding to re-calculate the screening.
Best,
Daniele