BUG: Potential incompatibility with "old" MP GGA workflow w.r.t. k-points in GGA static calculations

[Andrew-S-Rosen]

In Pymatgen, the MPStaticSet has a reciprocal_density of 100, as shown here. This overrides the default value of 64 in the MPRelaxSet.yaml. However, in atomate2, the MPGGAStaticSetGenerator does not seem to change the default of 64 to 100.

I'm not sure if I'm missing this somewhere, but that's a potential incompatibility. One would have to dig through atomate1 to be 100% certain (or compare against structures on MP). Note: This bug report was pulled from a larger comment in #724 since I felt it was important to distinguish it.

Tagging @esoteric-ephemera.

from pymatgen.core import Structure
from atomate2.vasp.jobs.mp import MPGGAStaticMaker

s = Structure(
    lattice=[[0, 2.13, 2.13], [2.13, 0, 2.13], [2.13, 2.13, 0]],
    species=["Ni", "O"],
    coords=[[0, 0, 0], [0.5, 0.5, 0.5]],
)
kpts = MPGGAStaticMaker().input_set_generator.get_input_set(s, potcar_spec=True).kpoints.kpts[0]
print(kpts) # (9, 9, 9)

from pymatgen.core import Structure
from pymatgen.io.vasp.sets import MPStaticSet

s = Structure(
    lattice=[[0, 2.13, 2.13], [2.13, 0, 2.13], [2.13, 2.13, 0]],
    species=["Ni", "O"],
    coords=[[0, 0, 0], [0.5, 0.5, 0.5]],
)
kpts = MPStaticSet(s).kpoints.kpts[0]
print(kpts) # (10, 10, 10)

[JaGeo]

Thanks for pointing this out. We in our group mostly adapt the sets for our runs and therefore don't notice such problems. Especially for future MP production runs, however, this might be highly important.

[esoteric-ephemera]

Thanks for pointing this out! While we'll eventually use KSPACING in lieu of KPOINTS for MP production runs, this has to be fixed for legacy MP workflows. I'm currently working on getting the sets ported to PMG and should probably put the draft PR for that up soon