Exchange and Coulomb components in HF energy


#1

Does anyone know of a simple, existing way to get the Coulomb and Exchange energy components to the HF energy?


#2

It looks like the Coulomb and Exchange energies are added together almost instantly upon computation. You’d either have to modify the compute_E functions in libscf_solver to store those contributions to variables or get the electron repulsion integral tensor and the orbital coefficients and manually compute the Coulomb and exchange components, Psi4Numpy style.


#3

See here for an example.


#4

Thanks to both of you. It’s peculiar that our output file contains the “exchange gradient” but no one seems to have ever wanted to know the exchange energy. Of course I was wondering if this data was already printed or stored somewhere. I will implement something like you suggest. Thanks again.


#5

I don’t see an “exchange gradient”, you might be referring to the exchange contributed to the nuclear gradient?


#6

Yes, that’s what I saw.


#7

But you are asking about energy? The two are not really comparable.


#8

My mistake then. I haven’t thought through the gradient contributions in awhile. But, yes, it’s the energy I’m interested in.


#9

For organizational purposes, I marked this question as solved. If it isn’t solved, let me know, and I can undo it.