Dear Psi4 team,
I’ve been using your program (python interface) to build an SCF method with corrections for dispersion added to the core Hamiltonian, with good results. We’d like to move to performing geometry optimisations, but this is notably limited to the built in methods.
I note that the optimise function works by repeatedly calling gradients and OPTKING. Assuming I can generate the gradients for the additional terms independently (analytically/numerically/whatever), is there a straightforward way to add them to the optimisation step?