questions about the results of BSE calculations?

[pyadav]

Dear team Yambo,

Can someone please comment on the following two points?

1. How can I convert the RL numbers (wavevectors for energy ) into Rydberg. Any formula or any rough estimation method?
2. The dipole socillator strength reported in file 'o.exc_qpt1_E_sorted' is absolute or relative to the largest dipole oscillator strength?

Thank you,
Pushpendra

[Daniele Varsano]

Dear Pushendra,

1) the relation is Number of RL such that G^2/2 < Ecut. I do not have a rough estimation. Anyway, when you assign an energy cutoff you have the number of RL in the report. Please note that instead of Ha, or Ry, you can also use mHa (mRy).

2) They are normalized and you have the normalization factor at the top of the output file:

Code: Select all

Maximum Residual Value = 

If you want to change it you can modify the source in ./ypp/excitons/excitons_sort_and_report.F
setting:

Code: Select all

R_normalize="no"

Best,
Daniele

[pyadav]

Dear Daniele,

Thank you so much for the reply and clarifications. They are really helpful.

Wishes,
Pushpendra.

[Dean]

Dear Daniele,
After setting "Weight_treshold=0.0 eV", I get all the transitions in the file xxx.exc_qpt1_weights_at_1.
But I find that the weights for one K point in the first part of outfile do not equal the sum of weights in the second part of outfile, such as Gamma point (0.144) vs (0.02+......).
Please see the attached file.
Best wishes,

[Daniele Varsano]

Dear Yimin,

this is just because the weight reported summed up on transitions for each K point and the individual contributions are normalized differently, i.e. the K point contribitions are normalized to its maximum:
see line 154 in
ypp/excitons/excitons_amplitudes.F

Code: Select all

148    K_weight=0._SP
149    do neh = 1,BS_H_dim
150      ikbz  = BSS_eh_table(S_indx(neh),1)
151      ikibz = Xk%sstar(ikbz,1)
152      K_weight(ikibz)=K_weight(ikibz)+A_weight(neh)
153    enddo
154    K_weight=K_weight/maxval(K_weight)
155    titles(1:4)=(/'             ','K-point [iku]','             ','Weight       '/)
156    call msg('o weight','#',titles(:4),INDENT=0,USE_TABS=.true.)
157    do ikibz=1,Xk%nibz
158      if (K_weight(ikibz)>EXCITON_weight_treshold) then
159        call msg('o weight','# ',(/Xk%pt(ikibz,:),K_weight(ikibz)/),INDENT=0,USE_TABS=.true.)
160      endif
161    enddo
162    call msg('o weight','#','',INDENT=0,USE_TABS=.true.)

Indeed, they do not sum-up to 1, but they indicate that:
# : 0.375000000 0.714285672E-1 0.00000000 1.00000000
has the maximum contribution.

Best,
Daniele

[Dean]

Dear Daniele,
Thanks for your reply.
Best,