Page 1 of 1
Electron-phonon coupling
Posted: Mon Feb 15, 2021 6:44 pm
by niloufar
Hello,
I know this might be irrelevant to this forum but I thought you might also have experience in this part. I have problem with phonon step in Quantum espresso and I couldn't find the origin of the error. I want to get phonon-electron coupling in my system with Yambo and in the ph.x step in QE, after calculating the K points and in the middle of electric field part, I get this error:
task # 1
from solve_e : error # 1
called in the wrong case
task # 0
from solve_e : error # 1
called in the wrong case
Do you have any idea what the problem is?
Thanks in advance for your help.
Re: Electron-phonon coupling
Posted: Tue Feb 16, 2021 8:52 am
by claudio
No idea about this error, did you try to post it on the QE forum?
Try also to change the parallelization and see if it helps.
Re: Electron-phonon coupling
Posted: Tue Feb 16, 2021 9:37 am
by palful
Hi Niloufar,
Would you mind posting your input files for the pw.x and ph.x steps?
I will take a look and run a small example to see if I can reproduce the issue.
Cheers,
Fulvio
Re: Electron-phonon coupling
Posted: Tue Feb 16, 2021 10:57 am
by niloufar
Thanks a lot.
That would be nice of you .
Re: Electron-phonon coupling
Posted: Tue Feb 16, 2021 11:07 am
by niloufar
palful wrote: ↑Tue Feb 16, 2021 9:37 am
Hi Niloufar,
Would you mind posting your input files for the pw.x and ph.x steps?
I will take a look and run a small example to see if I can reproduce the issue.
Cheers,
Fulvio
I posted my input files. Thanks a lot for helping me.
niloufar wrote: ↑Tue Feb 16, 2021 10:57 am
Thanks a lot.
That would be nice of you .
Re: Electron-phonon coupling
Posted: Tue Feb 16, 2021 1:52 pm
by palful
Dear Niloufar,
Judging from you ph.txt file, I believe you are in the case described
here for the variable
epsil:
If .true. in a q=0 calculation for a non metal the macroscopic dielectric constant of the system is computed. Do not set epsil to .true. if you have a metallic system or q/=0: the code will complain and stop.
Just add the Gamma point at the beginning of the qpoint list in your ph.txt input and the calculation of the macroscopic electric field should complete correctly:
Code: Select all
/
36
0. 0. 0. 1
-0.05 -0.05 -0.25 1
...
Cheers,
Fulvio
Re: Electron-phonon coupling
Posted: Tue Feb 16, 2021 4:50 pm
by niloufar
palful wrote: ↑Tue Feb 16, 2021 1:52 pm
Dear Niloufar,
Judging from you ph.txt file, I believe you are in the case described
here for the variable
epsil:
If .true. in a q=0 calculation for a non metal the macroscopic dielectric constant of the system is computed. Do not set epsil to .true. if you have a metallic system or q/=0: the code will complain and stop.
Just add the Gamma point at the beginning of the qpoint list in your ph.txt input and the calculation of the macroscopic electric field should complete correctly:
Code: Select all
/
36
0. 0. 0. 1
-0.05 -0.05 -0.25 1
...
Cheers,
Fulvio
Thank you very much for your recommendation, I will do them and see what happens.
Good luck,
Re: Electron-phonon coupling
Posted: Tue Mar 02, 2021 8:33 am
by niloufar
palful wrote: ↑Tue Feb 16, 2021 1:52 pm
Dear Niloufar,
Fulvio
Hello dear Palful,
as you helped, I fixed the problem which was using gamma point. but I have another question. ph.x is running for days and it seems that it's in a never-ending loop
. It doesn't end. It calculates K points , modes and self consistent calculation ... and then repeats these steps! Is it normal?? I use the latest version of QE. Could it be a bug or something?
Thanks
Re: Electron-phonon coupling
Posted: Tue Mar 02, 2021 9:54 am
by palful
Hi Niloufar,
Ummmm... in general ph.x is quite slow, and considering the size of your system, I would expect it to take days if not more.
In general the code finds modes, kpoints and performs a self-consistent calculation for each q point that you gave in input. If you are seeing this (if you look in the output for "q = ", you should see that it changes when the steps are repeated), then it's the normal functioning. In this case - and if you have the computational resources - you may try to speed it up by submitting different jobs for each q point (or for a subgroup of q points), so that each one completes faster.
If instead you're observing something different than what I described above, then there might be some issues but for that I would need to see input and output files.
Cheers,
Fulvio
Re: Electron-phonon coupling
Posted: Tue Mar 02, 2021 11:30 am
by niloufar
Actually the report in the terminal is too long and I couldn't find the "q" as you said. But according what you said I think that's OK because I have 36 q points and that should take some time.
Thanks again for your help
