How expensive/accurate is OMP2?

For (1), you can see an example here. The trick is that you create two molecules with the same name. The first is the cation. Converge that. Then set guess to read and perform the next SCF. When guess is set to read, Psi realizes we have orbitals on hand for another molecule with the same name and reads them in. I haven’t done this before myself, so let me know if you encounter anything suspicious.

(2) should be simple, but I’m not sure how you’d do it just yet. The difficulty is that the only orbital passing mechanism is to send in a wavefunction, but I don’t know of a way to force a new set of orbitals onto a Wavefunction Python-side.