Page 1 of 2

GW Bandstructure

Posted: Sun Dec 01, 2013 11:22 pm
by fatimazahra
Dear all
I want to plot the GW bandstructure, I did the calculation GW approximation by yambo -g n -p p.
I got a file with the name o.qp.
1) Can I get bandstructure from that file ? if yes, how?
2) Can I know the value of gap without plotting bandstructure ?
Thanks
Fatima Zahra

Re: GW Bandstructure

Posted: Mon Dec 02, 2013 12:25 am
by Daniele Varsano
Dear Fatima,
in the o.qp you have all you need for the band structure. As you can read in the legend it is ordered as:

Code: Select all

 K-point    Band       Eo         E-Eo       Sc(Eo)
Eo is the KS energy, E-Eo is the GW correction, Sc is the correlation part of the self-energy and you not need it for the band structure.
In order to plot the band structure, e.g. with gnuplot you just have to edit a little bit the file in order to put in the form:

Code: Select all

K-point1  E1 E2 E3 ...
K-point2  E1 E2 ...
If you know the index (K-point and Band) corresponding to the valance maximum and conduction minimum at KS level you can directly look at the corresponding GW correction in the o.qp file.

Cheers,
Daniele

Re: GW Bandstructure

Posted: Mon Dec 02, 2013 2:33 am
by fatimazahra
Thank you very match
I have another question, What is the role of ypp ?
I want to know is that the the bandstructure, that I found in o.qp, calculate along the Gamma-X-Gamma ?
Thanks for all

Re: GW Bandstructure

Posted: Mon Dec 02, 2013 9:14 am
by Daniele Varsano
Dear Fatima,
I have another question, What is the role of ypp ?
Ypp is a post processing tool, for plot and analysis of results generated by yambo.
you can find its functionality here.
A short description on how generate input files can be found by typing ypp -H
I want to know is that the the bandstructure, that I found in o.qp, calculate along the Gamma-X-Gamma ?
There is not an o.qp in attachment, anyway the o.qp reports indexes, K-points and bands are assigned via the variable
QPkrange, for example:

Code: Select all

%QPkrange 				
1| 5| 1| 32|		#	(GW)	QP generalized Kpoint/Band indices
%
with these settings you are asking to yambo to calculate QP energies for 32 bands for the first 5 K-points. As the K-points are indicated by indexes you have to check in tour report to which K-points do they correspond.

Hope it helps.
Daniele

Re: GW Bandstructure

Posted: Mon Dec 02, 2013 10:28 am
by fatimazahra
Thanks a lot
If I want to calculate the bandstructure in gamma-X-Gamma, I can use point K crystale in SCF and NSCF for specify the points.

Re: GW Bandstructure

Posted: Mon Dec 02, 2013 10:54 am
by Daniele Varsano
Right,
you have to select your K-points in the quantum espresso ground state calculations first. Yambo will calculate QP correction in that points.

Best,

Daniele

Re: GW Bandstructure

Posted: Mon Dec 02, 2013 10:28 pm
by fatimazahra
Hi
I plotted graphene bandstructure but the results are not good, I used Kpoint automatic,
After I run scf and nscf with Kpoints crystale but yambo doesn't run with crystal points k. I read in ypp tutorial that ypp can plot the structureband and dos.
Can I use this command for plotting bandstructure for specify the points or dos?
Thanks

Re: GW Bandstructure

Posted: Mon Dec 02, 2013 10:45 pm
by Daniele Varsano
Dear Fatima,
sorry but I do not understand what is exactly the problem you are reporting. Yambo needs an uniform k-point grid, and it is quite independent on how you generate it with quantum espresso. What it is important is to set in the input of qe the flag:

Code: Select all

force_symmorphic=.true.
collect_wf=.true.
Next, yes you can plot dos, interpolate band structures and plot it with ypp, see the examples in the webpage.
Best,

Daniele

Re: GW Bandstructure

Posted: Tue Dec 03, 2013 1:48 am
by fatimazahra
I used force_symmorphic=.true. and collect_wf=.true. in input scf and nscf, but the problem in input yambo, I have only one Gvector in input yambo. I changed this variable manually but it doesn't work.
I'll plot dos by using ypp and I will try again for plotting bandstructure.
thank you for your help

Re: GW Bandstructure

Posted: Tue Dec 03, 2013 4:30 am
by Daniele Varsano
Dear Fatima,
if you do not post your inputs/outputs/reports it is nearly impossible to help you.
Cheers,
Daniele