Dirac interface to MRCC - working for anyone?

I guess we have that part sorted out (note the manual entry is not consistent with this). Even if we run in C1, I get the invalid symmetry error for open shell systems.

Mihally - do you have any time to look into this? With Dirac18, closed shell cases work fine. So the MDCINT and MRCONEE files seem to be ordered ok. But going to an open shell system, e.g., Be to Be+, always fail with the invalid spatial symmetry error. I've been doing all sort of debugging today. In both xmrcc.f and mrcc.f this error trap kicks in whenever isym is not equal to 1 for ccsd. For a relativistic open shell system I can't see how this will not always fail. Turning symmetry off gets me further but then it fails saying to check the number of g/u orbitals for MDCINT and MRCONEE.

Could you post the MINP and fort.56 files as well as the header of the fort.55 file?

Dear Mihaly,

attached are the MINP and fort.55 for Be+ (all electrons correlated). MINP is only slightly different from neutral Be (just removed one electron from refdet and changed 2 to 3 in the top line). A fort.56 was not generated by the dirac_mointegral_export program.

Dear Kirk,
If your state is not total symmetric, you need to specify the symmetry of the state in the input file. In this case, you should set symm=5 in the MINP file. This is because the unpaired electron is located on the E_1/2g Fermion irrep, and the serial number of this irrep is 5. The latter can most simply inferred from file fort.55. The integers in the second and subsequent lines are the irreps of the orbitals. In your case:
5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 8 8 8 8
If you compare it with the occupation vector available in MINP:
1 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
you can see that the open-shell orbital, that is, the second one, belongs to irrep 5.

 

Thank you! I'll try that right now.

-Kirk