Fractional occupancies.
hongyi-zhao opened this issue · 6 comments
For some elements, fractional occupancies of the atomic reference configuration are used when generating the pseudopotentials, as shown below:
But this is not reflected in the lobster_basis used by this package. Is this reasonable?
See here and here for the related discussions.
Regards,
Zhao
The min file does not have them, the max file does. Thus, these orbitals are tested as a part of thr basis.
I think we had an updated version of the scripts at some point or we manually removed these orbitals from min for Be. They led to poor projections for semiconductors but are needed for metals.
Ah, one more thing: there is an issue for Ba. You might want to check out our database paper. We explain it there. An orbital that is needed is not yet existing as a part of the Lobster basis.
The min file does not have them, the max file does.
However, the form of fractional occupancies used on the VASP wiki does not appear in the following document:
You might want to check out our database paper.
Do you mean this one: https://doi.org/10.1002/cplu.202200123?
The min file does not have them, the max file does.
However, the form of fractional occupancies used on the VASP wiki does not appear in the following document:
You might want to check out our database paper.
It depends which orbitals arw available in Lobster and Lobster can only project into the whole orbtial. There is no way to reflect this in the files other than adding the whole orbital to the max file.
Will close this as we cannot directly enforce the occupancy in Lobster