Different CC energy with localized orbitals

Dear Developers,

as far as I am concerned, the CC energy should be invariant with respect to orbital localization.
Indeed, if I run CCS or CCSD calculations with localized orbitals (orbloco=pm, orblocv=pm), the energy equals to that from the same calculation with canonical (HF) orbitals.

However, if I run
-higher-order CC (e.g. CCSDT, CCSDTQ), or
- CCSD with ccprog=mrcc
I obtain different energies with and without orbloco/orblocv (which should not be the case in theory).

What is the reason for this, and can it be fixed somehow?

I attach four output files to illustrate the above points. (CCSD with/without orbloc with same energy; CCSDT with/without orbloc with different energy.)

Best,
Zsolt


 







 

Dear Zsolt,

This behavior is due to the symmetry of your molecule, which is not handled at all in the orbital localization module.
Please, set symm=off for the calculations using localized orbitals and ccprog=mrcc.

Best wishes,
Peter