Additional inputs to phonopy

[raynol-dsouza]

1. In order to calculate the projected DOS (either internally with phonopy or externally using a custom script), one needs the eigenvectors of the dynamical matrix at each (ideally irreducible) qpoint. This is by default hard-coded to be False in pyiron. I add an additional input to allow to save these eigenvectors if necessary.

2. Phonopy has 2(?) methods for the integration over the Brillouin zone: the tetrahedron method (phonopy default, hard coded in pyiron) and the Gaussian smearing method (sigma calculated internally by phonopy). I also add an additional input to allow to change the default.

[coveralls]

Pull Request Test Coverage Report for Build 9094382462

Details

• 2 of 4 (50.0%) changed or added relevant lines in 1 file are covered.
• 1 unchanged line in 1 file lost coverage.
• Overall coverage increased (+0.004%) to 71.059%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
pyiron_atomistics/atomistics/master/phonopy.py	2	4	50.0%

Files with Coverage Reduction	New Missed Lines	%
pyiron_atomistics/atomistics/master/phonopy.py	1	45.5%

Totals
Change from base Build 9047055407:	0.004%
Covered Lines:	10629
Relevant Lines:	14958

---

💛 - Coveralls

[raynol-dsouza]

@jan-janssen Thanks for the format black commit!

[pmrv]

If I remember correctly one can get the eigenvector also via some post processing from phonopy even from a loaded job, so it might not be necessary to save them and rather just compute them on demand (not sure how long this takes anymore). But that would be a bigger change and this is self contained and works, so looks good to me.

[raynol-dsouza]

@pmrv yes, you are right. But this involves executing job.phonopy.run_mesh() again, which is basically repeating what could just be done once. It is also quite memory intensive to do it on demand depending on how dense the qpoint mesh is.

[jan-janssen]

Looks good to me - please go ahead and merge the pull request so we include the changes in the next release.