Molcas Forum

ka3

Member

Registered: 2026-03-24

Posts: 7

Using RASSCF H5 File Data

Hi,

I'm trying to extract some information from the H5 files generated by the RASSCF program to calculate some MO overlaps between a closed-shell molecule and it's core-ionized equivalent. With the primitive basis from h5 I compute the MO overlap using the coefficients read from the field MO_VECTORS stored on the h5 files. I was wondering if constructing the MO's from MO_VECTORS corresponds to state averaged orbitals? Is there a way to get both state-averaged and state-specific MO's from the h5 file?

Thanks for any help!

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,204

Re: Using RASSCF H5 File Data

Yes, MO_VECTORS should be the averaged orbitals. To get the state-specific MOs, you need to diagonalize the state-specific density matrix (from DENSITY_MATRIX). The eigenvectors are the state-specific natural orbitals. Note that for RASSCF the DM is in the active space only: so each eigenvector corresponds to a linear combination of the averaged active MOs. (The non-active MOs are identical for all states, of course)

ka3

Member

Registered: 2026-03-24

Posts: 7

Re: Using RASSCF H5 File Data

Great, thanks Ignacio!