Updated the protocols to accomodate the suggestions proposed on issue #959 in aiidalab_qe.

[Minotakm]

Hey all,

There is an ongoing discussion regarding why the DOS sometimes misses peaks at low energies and why PDOS and DOS results can differ. The discussion concluded with the following changes:

• Decreasing the DeltaE value to 0.01 in the DOS protocol.
• Changing the occupations setting from tetrahedra to tetrahedra_opt.

Following that I did some tests, where here we can see that the low peak now is tracked.

Additionally, the DOS calculated with dos.x, projwfc.x, and the sum of the projections from pDOS are now consistent.

Furthermore, I updated the PdosWorkChain to check for the tetrahedra parameters.

Closes aiidalab/aiidalab-qe#959

[AndresOrtegaGuerrero]

Could you test with a magnetic system as well ?

[Minotakm]

Sure, I can. Will do, and update thanks for the suggestion.

[AndresOrtegaGuerrero]

Could you test with a magnetic system as well ?

Thank you , I suggested this since i think the projwfc.x calculation doest parse properly total PDOS for magnetic systems

[Minotakm]

Just a small update. You are right (@AndresOrtegaGuerrero ) projwfc.x calculation parses DOS for spin-up as total dos. I will look into that. For spin up it seems that is fine.

[AndresOrtegaGuerrero]

thank you @Minotakm could you maybe update the PR to merged it

[AndresOrtegaGuerrero]

@Minotakm could you update the tests with the changes ,

E       -      DeltaE: 0.02
E       +      DeltaE: 0.01

[Minotakm]

@AndresOrtegaGuerrero should be fine now.

[AndresOrtegaGuerrero]

This ones too

E       -        occupations: tetrahedra
E       +        occupations: tetrahedra_opt

[AndresOrtegaGuerrero]

Here as well ,
[in workflows/protocols
/test_pdos.py](

aiida-quantumespresso/tests/workflows/protocols/test_pdos.py

Line 45 in bcd5508

def test_electronic_type(get_pdos_generator_inputs):

)

    for namespace, occupations in zip((builder.scf, builder.nscf), ('fixed', 'tetrahedra')):
        parameters = namespace['pw']['parameters'].get_dict()
        assert parameters['SYSTEM']['occupations'] == occupations
        assert 'degauss' not in parameters['SYSTEM']
        assert 'smearing' not in parameters['SYSTEM']

[Minotakm]

Should be there.

[AndresOrtegaGuerrero]

LGTM! thank you @Minotakm

[mbercx]

@edan-bainglass if you needed this change urgently, please let me know and/or merge the commit with a more sensible commit message next time. This is not explaining anything, and the "issue" that is referred to is actually an unrelated PR. We try hard to keep a good commit history around here.

@Minotakm can you write a commit message on what problem the changes in this PR try to solve, and what was changed and why? Then I can amend c5617c0.

[edan-bainglass]

Apologies @mbercx 🙏 Great if we can document these commit rules. The "unrelated" PR is referenced by @Minotakm in the description of this PR, which I used for the commit message. Anyhow, as you suggest, I'll disregard your vacation status next time and bother you with work 👍 Let me practice now by asking you to make a release, so we can include the change in the app. That'd be bonza 💪 Cheers mate 🦘