Exciton Band Structure Finite Q

[alexmoratalla]

I am trying to use excitondb to plot exciton band structures of finite-q BSE. However, I get the plot at the same k-point for conduction and valence band. I guess this is not implemented, so I will work in the code for that.

[sangallidavide]

Ciao Alejandro, I guess you are using the same implementation for q=0

In ypp when we coded the finite-q BSE we needed to get the k of the hole as well.
So at q=0 you just have the one k

     ikbz = BSS_eh_table(S_indx(neh),1)
     iv = BSS_eh_table(S_indx(neh),2)
     ic = BSS_eh_table(S_indx(neh),3)
     i_spin= spin(BSS_eh_table(S_indx(neh),:))

and one extra line to get the k in the IBZ if needed

      ikibz = Xk%sstar(ikbz,1)

At finite momentum you also need the k of the hole (first in the BZ and then eventually in the IBZ)

     ikpbz = qindx_X(iq,ikbz,1)
     ikp  = Xk%sstar(ikpbz,1)

Please notice the issue pointed out by Fulvio for correctly getting the k-points ... yambo-code/yambopy-devel#29
Also qindx_X in inside ndb.kindx. So it might be useful to read it also for this reason.

[palful]

I think this is actually a duplicate of issue #20. I assigned that to you as well @alexmoratalla. @sangallidavide in issue #20 it is pointed out what the problem is.

Fixing a single plot manually is quite easy, but it would be nice to implement matrix element rotation in yambopy. Issue yambo-code/yambopy-devel#29 may be relevant because of the indexing inconsistency as you point out. The main problem is that the code that interpolates the weights expects always a k-point in the IBZ because it was done with only the q=0 in mind.