Some Questions

[WEI Wei]

Dear all,

I have some questions:

1) In a "yambo -g n -p p" calculation, the "NGsBlkXp" is set at 400. However, it prints "NGsBlkXp= 403" in the report file "r_HF_and_locXC_em1d_ppa_gw0". What is the reason?

2) Also in a "yambo -g n -p p" calculation, the coordinate of k-point is given as
*X* K [1] : 0.000000 0.000000 0.000000 (iku) * Comp.s 9695 * weight 0.0069
*X* K [2] : 0.08333 0.166667 0.000000 (iku) * Comp.s 9743 * weight 0.02778
*X* K [3] : 0.166667 0.333333 0.000000 (iku) * Comp.s 9733 * weight 0.02778
... ...
but on the DFT level (PWscf), it gives as:
k( 1) = ( 0.0000000 0.0000000 0.0000000), wk = 0.0138889
k( 2) = ( 0.0096811 0.0167681 0.0000000), wk = 0.0555556
k( 3) = ( 0.0193621 0.0335362 0.0000000), wk = 0.0555556
I want to know how to transform the coordinate from one to the other one? I mean I want to plot a quasiparticle band structure as that in attachment.

3) From the resulting file of a "yambo -o c" calculation, it shows:
E/ev[1] eps /Im[2] eps /Re[3] eps0/Im[4] eps0/Re[5]
#
0.00000 0.00350 2.36745 0.00505 2.58112
0.02002 0.00351 2.36746 0.00505 2.58114
0.04004 0.00351 2.36750 0.00505 2.58120
0.06006 0.00351 2.36756 0.00506 2.58130
0.08008 0.00351 2.36765 0.00506 2.58144
0.10010 0.00352 2.36776 0.00506 2.58161
... ...
What is the difference between "eps/Im[2]" and "eps0/Im[4]"?

4) In a BSE "yambo -b -o b -y h" calculation WITHOUT QP correction, it gives:
E/ev[1] eps /Im[2] eps /Re[3] eps0/Im[4] eps0/Re[5] eps`/Im[6] eps`/Re[7]
#
0.0000 0.0056 1.3383 0.0042 1.2942 0.0056 1.3383
0.0200 0.0057 1.3395 0.0043 1.2950 0.0057 1.3395
0.0400 0.0057 1.3406 0.0043 1.2959 0.0057 1.3406
0.0601 0.0058 1.3417 0.0043 1.2967 0.0058 1.3417
... ...
The fourth volume of "eps0/Im[4]" means independent-particel result, i.e., LDA-RPA, and if QP correction is taken into account, this volume gives GW-RPA result. Am I right? It seems that QP correction must be considered when solve the BSE calculation.

5) When do a "yambo -b -o b -y h" calculation WITH QP correction (GW-BSE) and a "yambo -o c" also WITH QP correction (GW-RPA), what is difference between "eps0/Im[4]" from GW-BSE and "eps/Im[2]" form GW-RPA?

6) In the o.qp file, the Fermi energy is set at zero. Right?

Many many thanks in advance!!!

[Daniele Varsano]

Dear Wei Wei,
1) You need to fill the shell of G-vectors, so Yambo gives you the nearest number of the G-vectors you indicate in input that has closed shells.
2) You can switch from iku (integer reduced lattice units) coordinate to other units once you know the lattice constant, you can aslo use ypp:
please have a look here
3)

What is the difference between "eps/Im[2]" and "eps0/Im[4]"?

The first one has local field effects included \eps_M=1/eps^-1_00, eps0 is the non-interacting epsilon (Fermi Golden Rule) eps_00.
4) Yes, right if for RPA you mean independent-particle. Of course QP corrections have to be included in BSE calculation, otherwise the theory
it is not consistent.
5) The local field effects as stated above in point 3.
6) Yes

Cheers,

Daniele

[WEI Wei]

Dear Daniele,

Thank you very much for your reply!
I know more and more.