printing MO coefficients?

Mihály,
Possibly a bug or oversight in the Molpro interface:

From the Molpro manual:
EOM_NSTATES NDOUB -1 Number of states in open shell case.

This is different than calling "nstates" in the MRCC standalone input file. The MRCC setup within molpro is doing things correctly:
Spin quantum number: 1.5

When I run the subsequent MRCC energy, it seems that MRCC is forcing the electronic state to actually be a doublet (when I am trying to specify either a quartet or sextet):

Number of alpha electrons: 6
Number of beta electrons: 4
Number of active particles: 0
Number of active holes: 4
z-component of spin: 1.0

Dear Nate,
You should set the directory where mrcc is executed in the molpro input file using the dir option, and then you can place a MINP file including iface=molpro in that directory. In the MINP file you can set the multiplicity and occupation you need.

Dear Mihály,
I save the fort.10/14/24/55/56 files from the Molpro output. Create a new MINP file. I should not assume that MRCC is going to correctly read the molecule information, right? I need to specify the precise geometry, basis set, and occupation in the MINP?

Will I still be able to pass a Douglas-Kroll reference energy into MRCC using the MINP?

When I try to run MRCC in this manner, it skips the SCF step, thank goodness. However, it seems to skip the first CC calculation and heads straight into the LRCCSDT part:

Executing mrcc...

**********************************************************************
LR-CCSDT transition moment calculation

...
Number of alpha electrons: 6
Number of beta electrons: 3
z-component of spin: 1.5
Spatial symmetry: 1
...
that's good! It's getting the right occupation.

---
Energy of reference determinant [au]: -1269.096180538512
Property integrals are missing!
Unable to calculate transition moments!

Fatal error in mrcc.
Program will stop.

---
that's bad. Any suggestions?

Dear Nate,
If mrcc is used in this way, only the transformed MO integrals are passed over to mrcc. You do not need to specify the geometry, basis set, occupation, and any molecular information. You should only set the memory, the type of calculation, the number of states, and similar things.
The reference HF energy with the DK integrals will be calculated by mrcc.
The ground-state CC calculation is not skipped. The program says that an LR-CC calculation is performed, but, of course, the ground-state CC equations will also be solved.
The problem is that the dipole integrals are not printed by molpro. As far as I remember this has not been done. So you will only be able to run excitation energy calculations.