interpretation of the out file

[marsusi]

Dear Daniele,

I have run the yambo tutorial Solid_Si/YAMBO/k_grid_xxx.
1-What we have set in input file is the max number of G vector (EXXRLvcs=7), and QP band indices (1:15). I don't understand how we insert for calculation of the energy of, or 32 un-occupied bands is given to yambo? Are these eigenvalue energies from LDA? These are appeard in 40 bands (which must be degenerated for spin-up and -down, since we supposed for non-polarized system) in section "[02.05]:Energies and Occpations" from r-HF_7_HF_and_localXC out file.
2-Why are HF and LDA band energies appeared in section [04], so different from those energies appeared in section [02.05]?
3-What physical quantity (very small) are shown after calculated HF and LDA energies at each band of each k-point in section[04]?

Thanks in advance,

Farah Marsusi.
Department of physics,
Amirkabir University of Technology,
Tehran, Iran.

[Daniele Varsano]

Dear Farah,

how we insert for calculation of the energy of, or 32 un-occupied bands is given to yambo

The number of the bands included in the calculation are controlled by the GbndRnge and BndsRnXp, respectively
in the GW summation and in the calculation of the polarizability.

Are these eigenvalue energies from LDA? These are appeard in 40 bands (which must be degenerated for spin-up and -down, since we supposed for non-polarized system) in section

In the GPL version of the code, the calculation are not spin-polarized, so you have bands with double occupation.

2-Why are HF and LDA band energies appeared in section [04], so different from those energies appeared in section [02.05]?

IN section 02.05 you have the KS eigenvalues, in 04 you do not have the eigenvalues but the matrix element of the Hartree-Fock and the Vxc. i.e. <n|HF|n> and <n|LDA|n>

3-What physical quantity (very small) are shown after calculated HF and LDA energies at each band of each k-point in section[04]?

For each matrix element you have the real part and imaginary part. The Im part here should be zero, i.e. the very small number you see.

Hope it helps.

Cheers,

Daniele

[marsusi]

Dear Daniele,

The number of the bands included in the calculation are controlled by the GbndRnge and BndsRnXp, respectively
in the GW summation and in the calculation of the polarizability.

Do you mean this number is its default? But, I see in documentation the default is between (1,50).

Thank you very much. You help me so much.


Farah.
Amirkabir University of Technology,
Tehran, Iran

[Daniele Varsano]

Dear Farah,
I'm not sure I understood your question.
The number of bands, both GbndRnge and BndsRnXp, have to be set in input, and
it is needed to raise them until the convergences is reached, so there is not a default.
This is extremely system-dependent. By default, when you generate the input files by command
line, yambo includes in the generated input, all the bands you calculated in your ground state
calculations.

Cheers,

Daniele

[marsusi]

Dear Daniele,

Sorry take your time, my question is exactly here that I have only:
EXXRLvxc=7
%QPkrange 1|16|1|15
%QPerange 1|16|0.0|-1.0
in my input.

Thanks a lot,

Farah.
Amirkabir University of Technology,
Tehran, Iran.

[Daniele Varsano]

Ok I see,
Your input is surely incomplete, at least I can't see the runlevel, I presume you should have in the first line the
runlevel HF_and_locXC.
if this is the case, in this way you are not calculating quasiparticle level, but just the exchange part of the
self energy needed for the GW. So here in this input you are asking to the code to evaluate the
<n|HF|n> and <n|Vxc|n>. So, no correlation part.
Using this variable you are calculating these matrix element for k=1,16 and for the band 1 to 15 (QPkrange).
QPerange is alternative to the general index QPkrange and here is ignored as a negative value in the
energy appears. This is not a problem as you are using QPkrange.
The calculation are done considering a number of G-vector for the HF up to 7 Ry (Here I presume you
have a Ry in the input file, otherwise you are considering only 7 Gvectors (RL)).
In this calculations the unoccupied bands does not enter, as for the HF, only the occupied bands are needed.
In order to perform a full GW calculations, you should generate your input by typing:
yambo -x -p p -g n
if you want to use the plasmon pole approximations. You will see that GbndRnge and BndsRnXp variables
will appears, as well as the other variables needed for the calculations.
Please have a look to the yambo web page instructions for an explanation on how to generate the input files
for the different runlevels:
http://www.yambo-code.org/doc/runlevels.php
and take your time to follow up the tutorials step by step.

Hope it helps,

Cheers,

Daniele

[marsusi]

Dear Daniele,

Thank you very much. You explained everything so much well.

Thanks a lot,

Farah.

Farah Marsusi,
Amirkabir University of Technology,
Tehran, Iran.