[molpro-user] CASSCF and MCSCF in Molpro

[Peter Knowles]

Dear Pavel,

Yes, CASSCF is a special case of MCSCF. Yes, Molpro uses determinants for
CASSCF purely for reasons of computational efficiency.

Peter
 
>>> Pawel Wielgus <wielgus at mml.ch.pwr.wroc.pl> 08/16/06 4:18 PM >>> 
Dear Molpro Users,

I have a very confusing problem with understanding the basic concept of
general MCSCF and CASSCF in Molpro. Both of them optimize MOs and CSFs
coefficients. General MCSCF must use CSFs as a basis, and the CAS uses
Slater determinants by default (by the way, what is the reason for that
except the efficiency?).
Now, for the following piece of input for N2 we perform CAS:

MULTI
 OCC,3,1,1,0,3,1,1,0
 CORE,1,0,0,0,1,0,0,0
 CLOSED,1,0,0,0,1,0,0,0
 WF,14,1,0
 STATE,1

This job leaves 1sigma(g) and 1sigma(u) frozen (non- optimized) and
2sigma(g) with 2sigma(u) always doubly occupied and optimized. Except
those all of the combinations are included within the valence space
consisting of 10 orbitals.

Now, I can't imagine how to perform MCSCF instead of CASSCF. Is it
enough to restrict the number of configurations available, i.e. instead
of complete active space use SELECT and CON for instance (and MULTI
switches to CSFs by default) ? And then we have MCSCF but not CAS, right?

Or in other words: each CAS is of MCSCF type, but the opposite is not
true. Is that statement correct ?

Regards

Pawel Wielgus