Some questions about Lead example in 2018 worksop

General discussion around the EPW software

Moderator: stiwari

inoka
Posts: 37
Joined: Sun Oct 16, 2016 5:41 pm
Affiliation:

Some questions about Lead example in 2018 worksop

Post by inoka »

Dear Developers


I am trying to finish the lead example in your recent workshop 2018 for resistivity calculation.
I had finished the same example before from your EPW/Documentation site.


I am sorry that I have gotten few questions and I am really appreciate your answers.

Q1. In your workshop tutorial it is said that automatically lead.res.01 is created along with lead.a2f.01 files. However according to EPW/Documentation we have to run a python script to generate resistivity data. Is this a new feature of a new version or what ? If so wich EPW/Quantum Espresso modules I need to use?

Q2
Also I noted there is no separate location for k-mesh path (meshes/path.dat) given in epw.in. How the code get k-mesh path now?

Q3
I also have a question on what you have said about "n" in Zemans formula. Does this mean any typical metallic material contribute only 8 electron for mobility ?

For my real case I am interested about uranium compounds. Electron configuration of U-metal is [Rn] 5f3 6d1 7s2. How I should get this now.
Also if it is a compound ( U3Si) What it should be.


where n is the number of electron per unit volume and n(!; T) is the Bose-Einstein distribution.
Usually this means the number of electrons that contribute to the mobility and so it is typically 8
(full shell) but not always. This number of electron is hardcoded in EPW/src/a2f.f90 along with
the temperature range (0-1000K). If you want to modify this, you will need to recompile the code.


Q4
Also all the calculations are now running in the same folder (cannot see pp,phonons, epw folders). Is this also a new version?


I am sorry may be I am not updated, I am actually running QE 6.1 module. Please let me know what is the latest module we have now.

Thank you
Inoka

sponce
Site Admin
Posts: 616
Joined: Wed Jan 13, 2016 7:25 pm
Affiliation: EPFL

Re: Some questions about Lead example in 2018 worksop

Post by sponce »

Dear Inoka,

Thank you for attending the school.

Indeed everything we ran on the school is the development version of QE and EPW.

This is version 6.2.2 of QE (not yet released).

You can download the development version from https://gitlab.com/QEF/q-e

Keep in mind that this might be unstable. If you want a stable version you will have to wait the next release (a few months from now).

The calculation of the resistivity can be done with the python script or is now produced by EPW (this will be v5 in the next release).

Regarding the "n". It is indeed the number of electron that contributes to the mobility and vary with your system. You can read the book from Ziman called "Electrons and Phonons" that talks about that.

Best wishes,
Samuel
Prof. Samuel Poncé
Chercheur qualifié F.R.S.-FNRS / Professeur UCLouvain
Institute of Condensed Matter and Nanosciences
UCLouvain, Belgium
Web: https://www.samuelponce.com

inoka
Posts: 37
Joined: Sun Oct 16, 2016 5:41 pm
Affiliation:

Re: Some questions about Lead example in 2018 worksop

Post by inoka »

Dear Dr. Samuel

I actually didn't get that chance. Thus I was waiting for information on the workshop and now following them for better understanding of calculation of electron relaxation time and resistivity which needs for my PhD project. But still I am not yet very sure how to do this.

I have already calculated linewidth.elself file and analysed to get Im(self). Also resistivity have been calculated with the python script and a2f_tr.01. I apreciate it very much if i can send those files to you or any other developer personaly and get comments on it.

I have no way to proceed without clarifying about this.

Thank you
Inoka

sponce
Site Admin
Posts: 616
Joined: Wed Jan 13, 2016 7:25 pm
Affiliation: EPFL

Re: Some questions about Lead example in 2018 worksop

Post by sponce »

Hello Inoka,

Sorry to hear that.

If you have trouble computing the resistivity, I suggest your try with the latest development version since it is inside the code and does not require you to change a python script by hand (except the thing about the n).

You can download and compile the version by doing

git clone https://gitlab.com/QEF/q-e.git
cd q-e
git checkout develop
./configure
make epw

That should work. You can then follow the same step as for the workshop.

Best,
Samuel
Prof. Samuel Poncé
Chercheur qualifié F.R.S.-FNRS / Professeur UCLouvain
Institute of Condensed Matter and Nanosciences
UCLouvain, Belgium
Web: https://www.samuelponce.com

inoka
Posts: 37
Joined: Sun Oct 16, 2016 5:41 pm
Affiliation:

Re: Some questions about Lead example in 2018 worksop

Post by inoka »

Dear Dr. Samuel

Thank you for providing me the link. I downloaded and compiled it and now trying the lead example.
I am having a problem to get the output from epw.in file.
I have finished scf.in, ph.in (pp.py to create and save required files) and nscf.in. But epw.in crashes and gives the following error.



%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
task # 0
from phq_readin : error # 1
reading inputph namelist
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%


Could you please tell me why this comes? Do I need to have separate folders (pp, phonons and epw as the same way as old module?)


Thank you
Inoka

inoka
Posts: 37
Joined: Sun Oct 16, 2016 5:41 pm
Affiliation:

Re: Some questions about Lead example in 2018 worksop

Post by inoka »

Dear Dr. Samuel

I am really sorry for the previous post. I have done a mistake in the job script for epw.in.

Now it is working.

Thank you

Inoka

sponce
Site Admin
Posts: 616
Joined: Wed Jan 13, 2016 7:25 pm
Affiliation: EPFL

Re: Some questions about Lead example in 2018 worksop

Post by sponce »

Glad to hear it !
Prof. Samuel Poncé
Chercheur qualifié F.R.S.-FNRS / Professeur UCLouvain
Institute of Condensed Matter and Nanosciences
UCLouvain, Belgium
Web: https://www.samuelponce.com

inoka
Posts: 37
Joined: Sun Oct 16, 2016 5:41 pm
Affiliation:

Re: Some questions about Lead example in 2018 worksop

Post by inoka »

Dear Dr. Samuel

I am getting another error now.


I have completed scf.in, ph.in, pp.py and nscf.in.

Epw.in has completed wannierization part, after that I am getting the following error.

PWscf started at: Mon May 7 20:25:52 CST 2018
forrtl: severe (24): end-of-file during read, unit 5, file /proc/8616/fd/0
Image PC Routine Line Source
epw.x 00000000023C155E Unknown Unknown Unknown
epw.x 00000000023F7ECA Unknown Unknown Unknown
epw.x 0000000000441E1D Unknown Unknown Unknown
epw.x 000000000040E242 Unknown Unknown Unknown
epw.x 000000000040D52E Unknown Unknown Unknown
libc.so.6 00002AAAAC0862E0 Unknown Unknown Unknown
epw.x 000000000040D42A Unknown Unknown Unknown
srun: error: plato205: task 0: Exited with exit code 24



I am running my job with QE 6.2.2. I appreciate it very much if you can give me some guidance for this

Thank you

Inoka

sponce
Site Admin
Posts: 616
Joined: Wed Jan 13, 2016 7:25 pm
Affiliation: EPFL

Re: Some questions about Lead example in 2018 worksop

Post by sponce »

Hello,

It seems you are trying to use pw.x to run an epw.in file.

Best wishes,
Samuel
Prof. Samuel Poncé
Chercheur qualifié F.R.S.-FNRS / Professeur UCLouvain
Institute of Condensed Matter and Nanosciences
UCLouvain, Belgium
Web: https://www.samuelponce.com

inoka
Posts: 37
Joined: Sun Oct 16, 2016 5:41 pm
Affiliation:

Re: Some questions about Lead example in 2018 worksop

Post by inoka »

Dear Dr. Samuel

Actually I ran the job correctly with epw.x. However I couldn't avoid this error.

But when I remove parallel_k and parallel_q it works and I could reproduce what you have given in the tutorial. I am not sure why parallel_k and q does not work. I am trying this with QE 6.2.2(downloaded from github).

I got another problem with Wed5/exercise2/si.
epw2.in gave me the error below. (I copied only the error part of the epw2.out file). Can you please let me know how to solve this. I think I should mention that I copied fc.si file manually to the save folder (renamed as ifc.q2r) since the pp.py did not do that. Hope the error doesn't have any effect with that.

Scattering rate
===================================================================


Fermi Surface thickness = 1.200000 eV
This is computed with respect to the fine Fermi level 8.544315 eV
Only states between 7.344315 eV and 9.744315 eV will be included

Progression iq (fine) = 50/ 8000
Progression iq (fine) = 100/ 8000
Progression iq (fine) = 150/ 8000
Progression iq (fine) = 200/ 8000
Progression iq (fine) = 250/ 8000
Progression iq (fine) = 300/ 8000
Progression iq (fine) = 350/ 8000
Progression iq (fine) = 400/ 8000
Progression iq (fine) = 450/ 8000
Progression iq (fine) = 500/ 8000
Creation of a restart point

===================================================================
Temp [K] Fermi [eV] Hole density [cm^-3] Hole mobility [cm^2/Vs]
===================================================================


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine rdiagh (2):
diagonalization (DSYEV) failed
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

stopping ...

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine rdiagh (2):
diagonalization (DSYEV) failed
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Thank you very much for your advises

Inoka

Post Reply