Peter Kraus
Peter Kraus
FWIW, the `VV10` energy contribution is added to the `DFT FUNCTIONAL TOTAL ENERGY` here: https://github.com/psi4/psi4/blob/9a1853f004977d79b333a89b5fda237732aafe83/psi4/driver/procrouting/scf_proc/scf_iterator.py#L705 so implementing the change would be trivial.
This is now ready for review. I have had to move the dispersion correction bit in `proc.py` around to a spot after the ECP contribution is in, to keep the...
Yes, please go ahead.
This one should be ready to review again.
I'd appreciate some feedback, especially on the docs. There's one or two items where I'd like @loriab to weigh in further: function aliases and zero-if-not-present for VV10 and empirical dispersion....
> > > I am still completely lost when I try to read `_validate_cbs_inputs`. > > What I understand is that the user provides the energy call with metadata information...
I agree about the docs, see the point above. However, in lieu of DDD, I don't want to spend the time writing a rst file that will be immediately obsolete...
Ah yes, this is an issue. Neither of the three works as one would expect them to work: the first two don't do anything as the fctl stage is not...
So I was thinking about getting rid of `"component"` too. The difference between something like "MP2" and "B2PLYP" is that while there's a name for the SCF component of MP2...
This would be a wonderful stop-gap for our use case, in lieu of "adding" people to samples & collections.