Updated vasp input sets with MPSurfaceSet from atomate1 #4287
Conversation
|
I would rather not have two surface sets with only slgihtly different parameters and confusing names. The changes to the inputs are small enough that you can go ahead and modify MVLSlabSet directly and atomate should use MVLSlabSet. |
|
@abhardwaj73 it looks like the adsorption workflows in |
There was a problem hiding this comment.
Looks good to me @abhardwaj73 , thanks. See a few small comments.
Also, please update the tests here if needed so they pass with the new parameters.
src/pymatgen/io/vasp/sets.py
Outdated
| class MPSurfaceSet(MVLSlabSet): | ||
| """ | ||
| Input class for MP slab calcs, mostly to change parameters | ||
| and defaults slightly | ||
| """ | ||
|
|
||
| def __init__(self, structure, bulk=False, auto_dipole=None, **kwargs): | ||
| # If not a bulk calc, turn get_locpot/auto_dipole on by default | ||
| auto_dipole = auto_dipole or not bulk | ||
| super().__init__(structure, bulk=bulk, auto_dipole=False, **kwargs) | ||
| # This is a hack, but should be fixed when this is ported over to | ||
| # pymatgen to account for vasp native dipole fix | ||
| if auto_dipole: | ||
| self._config_dict["INCAR"].update({"LDIPOL": True, "IDIPOL": 3}) | ||
| self.auto_dipole = True | ||
|
|
||
| @property | ||
| def incar(self): | ||
| incar = super().incar | ||
|
|
||
| # Determine LDAU based on slab chemistry without adsorbates | ||
| ldau_elts = {"O", "F"} | ||
| if self.structure.site_properties.get("surface_properties"): | ||
| non_adsorbate_elts = { | ||
| s.specie.symbol for s in self.structure if s.properties["surface_properties"] != "adsorbate" | ||
| } | ||
| else: | ||
| non_adsorbate_elts = {s.specie.symbol for s in self.structure} | ||
| ldau = bool(non_adsorbate_elts & ldau_elts) | ||
|
|
||
| # Should give better forces for optimization | ||
| incar_config = { | ||
| "EDIFFG": -0.05, | ||
| "ENAUG": 4000, | ||
| "IBRION": 1, | ||
| "POTIM": 1.0, | ||
| "LDAU": ldau, | ||
| "EDIFF": 1e-5, | ||
| "ISYM": 0, | ||
| } | ||
| incar.update(incar_config) | ||
| incar.update(self.user_incar_settings) | ||
| return incar | ||
|
|
||
|
|
There was a problem hiding this comment.
Thanks @abhardwaj73 . Since we were asked only to update MVLSlabSet to match the MPSurfaceSet parameters, this class can now be removed.
|
|
||
| """Determine LDAU based on slab chemistry without adsorbates""" | ||
|
|
There was a problem hiding this comment.
For historical provenance purposes, please add some comment lines either here or at the top of the MVLSlabSet class explaining that the parameters were updated in March 2025 to match the MPSurfaceSet in atomate1.
There was a problem hiding this comment.
Thanks @abhardwaj73 ! @shyuep I think this should be ready to merge, pending tests all passing.
Summary
Major changes:
Todos
Mentioned in this issue #4247 @rkingsbury