2D kpoint grid being interpreted as 3D

Concerns any physical issues arising during the setup step (-i option). This includes problems with symmetries, k/q-point sets, and so on. For technical problems (running in parallel, etc), refer to the Technical forum.

Moderators: Davide Sangalli, andrea.ferretti, myrta gruning, Daniele Varsano

Post Reply
arb83@cam.ac.uk
Posts: 98
Joined: Thu Jul 02, 2020 3:56 pm

Re: 2D kpoint grid being interpreted as 3D

Post by arb83@cam.ac.uk » Tue Oct 27, 2020 7:10 pm

Dear Daniele

I tried this but received the same error message. I assume there is nothing else to try?

Just so I understand what problems might occur if I increase the tolerance?

Best wishes,

Alan
Alan Bowman
University of Cambridge

User avatar
Daniele Varsano
Posts: 3773
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: 2D kpoint grid being interpreted as 3D

Post by Daniele Varsano » Wed Oct 28, 2020 7:42 am

Dear Alan,
it is possible that such large tolerance mess up the indexes of Gvectors and it is not safe.
Anyway, it sounds very strange to me you solved the problem for the GW and not for BSE, as the G search should be the same.
I will try to reproduce your error in the next days and let you know. Can you provide the BSE input you are using? Please provide also the report file.

Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

arb83@cam.ac.uk
Posts: 98
Joined: Thu Jul 02, 2020 3:56 pm

Re: 2D kpoint grid being interpreted as 3D

Post by arb83@cam.ac.uk » Wed Oct 28, 2020 11:31 am

Dear Daniele

Please find the files attached. Many thanks for looking into this.

With best wishes,

Alan
You do not have the required permissions to view the files attached to this post.
Alan Bowman
University of Cambridge

arb83@cam.ac.uk
Posts: 98
Joined: Thu Jul 02, 2020 3:56 pm

Re: 2D kpoint grid being interpreted as 3D

Post by arb83@cam.ac.uk » Wed Oct 28, 2020 11:31 am

Final file now attached.
You do not have the required permissions to view the files attached to this post.
Alan Bowman
University of Cambridge

arb83@cam.ac.uk
Posts: 98
Joined: Thu Jul 02, 2020 3:56 pm

Re: 2D kpoint grid being interpreted as 3D

Post by arb83@cam.ac.uk » Thu Oct 29, 2020 11:34 am

Dear Daniele

I have had one additional thought. I tried rotating the unit cell so the first lattice vector was aligned with the x axis. The minimum value of GmG-(G-G') on the three axes is now

0.0 0.02285403 0.00090557337

Is there a way to quickly calculate GmG-(G-G') (i.e. without doing a full yambo initialisation and calculating the BSE dielectric). If yes then I could just try many different unit cells slightly rotated with respect to each other and see which give the minimum set of GmG-(G-G').

With best wishes,

Alan
Alan Bowman
University of Cambridge

User avatar
Daniele Varsano
Posts: 3773
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: 2D kpoint grid being interpreted as 3D

Post by Daniele Varsano » Thu Oct 29, 2020 3:53 pm

Dear Alan,

I had a deep look at your problem.
Forget all I told you before regarding the tolerance in the search of G-G1.
Your problem is related to the kpoint grid which is not seen at uniform using standard algorithm when searching k points index for the BSE.
You should solve the problem using the new algorithm, so adding the keyword NoDiagSC also in the BSE input.

Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

arb83@cam.ac.uk
Posts: 98
Joined: Thu Jul 02, 2020 3:56 pm

Re: 2D kpoint grid being interpreted as 3D

Post by arb83@cam.ac.uk » Thu Oct 29, 2020 5:36 pm

Dear Daniele

Many thanks for this. This partly solved the problem but the code still does not run fully. Instead of getting stuck in the Transferred momenta grid calculation it now has an error in the BSE calculation. Specifically I received the following error messages:

"
<48s> P132-cpu-e-766: Response_T_space parallel ENVIRONMENT is incomplete. Switching to defaults
P132-cpu-e-766: [ERROR] STOP signal received while in :[07] Bethe Salpeter Equation @q1


[ERROR] STOP signal received while in :[07] Bethe Salpeter Equation @q1

[ERROR]Impossible to define an appropriate parallel structure
"

I have not defined a parallel environment as I was letting Yambo choose its defaults (which has been fine for everything else). I've been running the code on 160 cores. Is this to do with the newer algorithm? Either way could you suggest a way to progress please? Everything is the same apart from the additional line you asked me to put in.

With best wishes and thank you for all your help thus far,

Alan
Alan Bowman
University of Cambridge

User avatar
Daniele Varsano
Posts: 3773
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: 2D kpoint grid being interpreted as 3D

Post by Daniele Varsano » Thu Oct 29, 2020 6:08 pm

Dear Alan,
it seems you need to guide the code in defining a parallel structure.
If you send your input/report of the job reporting the error?
I will have a look.

Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

arb83@cam.ac.uk
Posts: 98
Joined: Thu Jul 02, 2020 3:56 pm

Re: 2D kpoint grid being interpreted as 3D

Post by arb83@cam.ac.uk » Thu Oct 29, 2020 7:16 pm

Dear Daniele

Please find the files attached. Many thanks in advance.

With best wishes,

Alan
You do not have the required permissions to view the files attached to this post.
Alan Bowman
University of Cambridge

arb83@cam.ac.uk
Posts: 98
Joined: Thu Jul 02, 2020 3:56 pm

Re: 2D kpoint grid being interpreted as 3D

Post by arb83@cam.ac.uk » Thu Oct 29, 2020 7:17 pm

Final file now attached.
You do not have the required permissions to view the files attached to this post.
Alan Bowman
University of Cambridge

Post Reply