Page 2 of 3
Re: Abinit to yambo
Posted: Wed May 26, 2021 9:53 am
by Daniele Varsano
Dear Farzaneh,
what I was asking is the config.log of the yambo compilation.
Best,
Daniele
Re: Abinit to yambo
Posted: Wed May 26, 2021 4:03 pm
by FGH
Dear Daniele,
Here is the file you want.
Re: Abinit to yambo
Posted: Fri May 28, 2021 5:32 pm
by Davide Sangalli
Dear Farzaneh,
I inspected the problem.
There is a bug in a2y which is triggered by the very low cutoff you are using.
Replace the file interfaces/a2y/a2y_db1.F
with the attached one and recompile a2y.
D.
2021/06/03 edit: file interfaces/a2y/a2y_db1.F updated
Re: Abinit to yambo
Posted: Sun May 30, 2021 12:24 pm
by FGH
Dear Davide,
Could you please to tell me what should I do step by step? and I should to say that I install the yambo code on the cluster.
Sincerely yours
Re: Abinit to yambo
Posted: Sun May 30, 2021 1:45 pm
by Daniele Varsano
Dear Farzaneh,
1. download the file Davide posted (a2y_db1.F.txt)
2. copy it in the yambo directory
2. go to the yambo directory
3. cd ./interfaces/a2y
4. cp ../../a2y_db1.F.txt a2y_db1.F
5. recompile the code
Best,
Daniele
Re: Abinit to yambo
Posted: Sun May 30, 2021 2:50 pm
by FGH
Dear Daniele,
Thank you so much.
Re: Abinit to yambo
Posted: Tue Jun 01, 2021 11:58 am
by Marzieh
Dear developers,
I tried to re-compile yambo with the a2y_db1.F which Davide has attached.
but it seems it does not work.
I wil make an example to understand what I mean:
I used ecut=30 , ecut =40 and ecut=50 in abinit input for BeS
After launch
yambo_rt -i -V RL -F setup.in (setup input file for real time)
I got the same value for MaxGvecs in all of them(17635):
setup # [R] Initialization
MaxGvecs= 17635 RL # [INI] Max number of G-vectors planned to use
I did not see any change in MaxGvecs when I change ecut in abinit input.
Could you please tell me how to fix it?
Best,
Marzieh
Re: Abinit to yambo
Posted: Tue Jun 01, 2021 12:07 pm
by Laura Caputo
Dear developers,
I have to say that I have the same problem as Marzieh. With two different cut off from the Abinit calculation, I obtain the same value of the number of G-vectors. I tried with different cut off values but does not change anything. I add that I had higher values of cut-off energy compared to the one that Farzaneh used (like 800eV).
Re: Abinit to yambo
Posted: Thu Jun 03, 2021 4:28 pm
by Davide Sangalli
Dear all,
the a2y interface was generating many more G-vectors than needed to be sure that all k-centered spheres fit in the gamma-centered sphere.
The calculations with Yambo are independent of such number, since the WFs are defined only on the needed G-vectors and zero outside.
However, as a consequence, Yambo has to load many more g-vectors than needed.
I've further refined the previous fix. Now everything should work fine.
Attached the new file (I've also upgraded the file in my previous reply to this last version).
Best,
D.
Re: Abinit to yambo
Posted: Sat Jun 05, 2021 11:12 am
by FGH
Dear Davide,
I think there is another problem. Now I get the gap but every two gap gain the same. I have attached the files. In order to compare these file I send the outputs through three posts.