indmftpr helper script

[harrisonlabollita]

Hi all,

This is a small python script (init_dmftpr) that has helped me prepare the case.indmftpr file for the dmftproj program. I tried to generalize it to handle most cases, but I'm sure there are some holes. I wanted to share this as it might be helpful to enhance the user experience especially for new users that are learning the TRIQS and dft_tools software.

Best, Harry

[the-hampel]

Hi @harrisonlabollita,

thank you for you PR! I myself do not have to much experience with the dmftproj program, but imagine it to be a nice addition. Could someone who has some more experience using dmftproj maybe take a look? Specifically I we should make sure, that this script does not lead to some pitfalls in special cases. Are there cases with this does not work, and the script should exit?

Best, Alex

[hschnait]

Hi @ all

Thanks for this contribution, I'm sure this will come in handy. I've tried it out for my cases and it seems to work, but there are some (minor) things I might want to add:

• In the script it is possible to chose an orbital both as correlated and non-correlated, thus leading to a "3" for lsort in the .indmftpr file. This should not be allowed.
• For SOC-calculations it should not be allowed to use irreps in the cubic basis. While this is in principle allowed in the .indmftpr file, dmftproj will use the whole d-shell anyways when SOC=1 (as eg/t2g are no irreps in that case). This might lead to confusion.
• This is more of a convenience feature, but imho it would be nice to have a possibility to give the projection window in eV as well as Ry.
• basis=fromfile is not possible with this script, but as this is more of an advanced feature, I'm not sure if this is too important here.

Best, Hermann

[harrisonlabollita]

Hi @the-hampel and @hschnait,

Thanks for the feedback! This is something I've just used for my own applications, so thanks @hschnait for pointing out some edge cases that should be accounted for. If this seems useful, I'm happy to fix them in a new commit!

Best, Harry

[the-hampel]

Thanks for the updates. @hschnait can you quickly confirm if all the things have been fixed? For me this looks good now. Thanks!

[hschnait]

Looks good to me

Alexander Hampel @.***> schrieb am Mo., 8. Nov. 2021, 17:25:

Thanks for the updates. @hschnait https://github.com/hschnait can you quickly confirm if all the things have been fixed? For me this looks good now. Thanks!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/TRIQS/dft_tools/pull/186#issuecomment-963325654, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKHTVMQ3NN3PWENG2AIYGZLUK72VZANCNFSM5GHDVILA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[the-hampel]

Okay thanks @hschnait . Than we can go ahead and merge it. I would like to merge it into unstable instead of 3.0.x. I can rebase on top of unstable and change the PR, but this will change quite a bit your 3.0.x branch @harrisonlabollita. Is that fine with you? Otherwise you could do it by yourself.

• Alex

[harrisonlabollita]

Yes, that's fine with me!

[the-hampel]

Hi, I updated your branch to unstable and I also updated the cmake list and documentation for the new file. Now the file should be correctly copied to the bin dir. Could you maybe add a docstring in the file to briefly state what the script does? Thank you.

[harrisonlabollita]

Thanks! Done!

[the-hampel]

Perfect - thanks. The tests are only failing because I changed the base branch. I will merge this now.