Fix phonon + Lobster flow by removing magmoms before passing to phonopy

[naik-aakash]

Changes

1. Update pymatgen, mp-api and Lobsterpy version
2. Add checks to remove magmom from structure to avoid phonopy failures

Closes #739

[naik-aakash]

Hi @JaGeo and @utf , this changes should in principle fix the phonon workflow once mp-api is compatible with the latest pymatgen version.

Tests fail due to dependencies conflicts

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 76.22%. Comparing base (62ae659) to head (031521d).
Report is 18 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #751      +/-   ##
==========================================
+ Coverage   76.19%   76.22%   +0.02%     
==========================================
  Files          87       88       +1     
  Lines        7178     7191      +13     
  Branches     1057     1051       -6     
==========================================
+ Hits         5469     5481      +12     
- Misses       1388     1393       +5     
+ Partials      321      317       -4     

Files	Coverage Δ
src/atomate2/common/jobs/phonons.py	86.20% <100.00%> (+1.50%)	⬆️

[naik-aakash]

Hi @utf and @JaGeo , this PR could be merged now if there are no comments on it 😄

[JaGeo]

@naik-aakash I am not sure removing the initial magnetic moments in all cases is a good solution. We should rather disenable the seekpath option implemented in Phonopy in that case.

[JaGeo]

Here is a discussion about it: materialsproject/pymatgen#3555

[naik-aakash]

Here is a discussion about it: materialsproject/pymatgen#3555

Ohh, ok, I see. I will check it. Thanks! 😃

[JaGeo]

Easy solution for now: removing all magnetic moments and adding a warning that we currently do not consider the initial magnetic structure for determining the symmetry and adding this also clearly in the documentation.

@utf what do you think? In this way, we could get a new version out very fast.

[JaGeo]

As I have not run magnetic structures with this workflow so far, I would like to do some tests if we implement this, add some additional features (manually setting the primitive cell, skipping band structure or options to set the kpath manually)

[utf]

Easy solution for now: removing all magnetic moments and adding a warning that we currently do not consider the initial magnetic structure for determining the symmetry and adding this also clearly in the documentation.

Yes, this is a good solution.

[naik-aakash]

Thanks @utf and @JaGeo , I will now add a warning and also update this in documentation of phonon workflow.

[JaGeo]

Tagging @JonathanSchmidt1 in case he might want to contribute to atomate2 with his implementation of a magnetic phonon workflow in a follow-up PR.

[JonathanSchmidt1]

It might take a few weeks till I find the time but I would be happy to add it.
It will just not have the full functionality of the normal phonon workflow as
PRIMITIVE_AXES = auto and BAND = auto are not allowed with MAGMOM in phonopy.

[JaGeo]

@JonathanSchmidt1 Great. Yes, sure. I have mentioned some suggestions above as well

[utf]

Thanks @naik-aakash and @JaGeo