[molpro-user] getting state average MO in a multi computation

[Peter Knowles]

The following input does what I think you want. You can choose the standard Molden format, or Molpro's xml-based format, according to your taste.

Or you may be asking which orbitals are actually used in the printed CI wavefunction. In the example given, these are the state-averaged orbitals, but you can control this using the NATORB subcommand of MCSCF (see http://www.molpro.net/info/current/doc/manual/node236.html)

Peter

geometry={b;h,b,2.3}
rhf
{mcscf;wf,,1;state,3;wf,,4;natorb,5000.2,ci}
put,molden,modump.molden
put,xml,modump.xml

On 27 Dec 2009, at 13:45, Francoise Remacle wrote:

> Dear Prof. Knowles,
> We are using molpro and the multi keyword to compute excited states.
> We then need to build a one electron rho matrix for a given superposition of these excited states.
> for this, we need the LCAO coefficients of the average MO's that are used to build the CSF's , whose integer occupation numbers are listed below the subtitle 'CI vectors' in the molpro output.
> We would like to know which keyword we need to use to dump the matrix for these LCAO coefficients.
> As far as we understand, these are not the LCAO of the state averaged natural orbitals that diagonalize the state averaged rho matrix.
> Could you kindly tell us how to proceed?
> Many thanks in advance ,
> Best wishes for the New Year,
> Francoise Remacle
> 

--
Prof. Peter J. Knowles             
School of Chemistry, Cardiff University, Main Building, Park Place, Cardiff CF10 3AT, UK
Telephone +44 29 208 74805 Email KnowlesPJ at Cardiff.ac.uk 
WWW http://www.cardiff.ac.uk/chemy/contactsandpeople/academicstaff/knowles.html