Multireference input

Dear MRCC developers,

is it possible to provide any simplest example of multireference CC input?
Manual seems to be rather clear, but i would like to avoid misunderstanding.
All test jobs seems to be with all inactive orbitals.

Thanks in advance.

P.S. Sorry, accidentally posted in wrong sub-forum.

This is a sample MINP file for a water MRCCSD calculation with a 4x4 CAS including the HOMO-1, HOMO, LUMO, and LUMO+1 orbitals:

basis=cc-pVDZ
calc=CCSD
mem=500MB
nacto=4
nactv=4
ccprog=mrcc

geom
H
O 1 R1
H 2 R1 1 A

R1=0.9575
A=104.51

#########################################################
The same using the active keyword (please note the frozen 1s orbital):

basis=cc-pVDZ
calc=CCSD
mem=500MB
active=vector
0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ccprog=mrcc

geom
H
O 1 R1
H 2 R1 1 A

R1=0.9575
A=104.51

#########################################################
Please note that currently only HF orbitals are available if mrcc is used in standalone mode. If you need MCSCF orbitals you should use Molpro. A sample input file:

memory,100,M

R=0.9575 angs
A=104.51

geometry={
H
O 1 R
H 2 R 1 A
}

basis,cc-pVDZ
wf,10,1,0
HF
{multi
occ,4,1,2,0;
closed,2,1,0,0;
natorb;
}
mrcc,method=ccsd,nacto=4,nactv=4

Dear Kallay,
I tried the same input, but with mult=3.
After mrcc finished mult changes to 2.
Inputs with nacto/nactv and vector=active (as suggested above) were run and the energy value changed.
Could you suggest changes and new options for (this) open-shell calculations?
Is it normal mult changed?

Thanks in advance.

mult is only changed if the number of electrons or the occupation specified by the user is not compatible with the multiplicity. You should check if the charge of the system or the occupation is correct.

Dear Kallay,

I tested your last input, and the results are
HF-SCF=-76.02678480
CASSCF=-76.07784353
MRCCSD=-76.23934209
However, MRCC prints a strange reference energy before the first CC iteration, which is neither the CASSCF nor the HF-SCF one.
Energy of reference determinant [au]: -76.026194030599
Is this correct?

Another question. Can we apply some occupation restrictions in MULTI (i.e., RASSCF) to speed up the MRCCSD calculation? My tests showed this is not true, but maybe the system is too small.

Best regards,
Wenli

Dear Wenli,
It prints out the energy of the determinant which is composed of MCSCF natural orbitals. This is used as reference in the CC calculation.
Unfortunately, occupation restrictions are not available.