ricc2 / rimp2 and rohf

[alex_p]

Hi,

I'm trying to do single point mp2 energy calculations on an open shell system. I tried a rohf calculation with rimp2 and ricc2. Surprisingly the two programs calculated different energies (different by 0.0125 Hartree).
With uhf the energy is exactly the same.
Since the manual states that ricc2 includes the rimp2 functionalities as a subset, I would have expected that both programs produce the same result. Is any of this programs calculating the "correct" energy?
(I know that the manual doesn't mention rohf as a valid wave function for rimp2 but was hoping that there might be an undocumented feature.).

Turbomole version used is 6.1 and 6.2

Best regards,
Alexander Patzer

[christof.haettig]

The rimp2 calculation is wrong. (Just check the orbital statistics in the output!)
Christof