[molpro-user] SA-CASSCF CI Vector is missing electrons - what is going on?

Seth Olsen seth.olsen at uq.edu.au
Thu Sep 21 05:07:20 CEST 2017

Hi Molpro-user,

I’m running a SA-CASSCF (4 electrons over 6 orbitals) calculation averaged over singlet, triplet, reduced and oxidized states of a molecule (1 state each, even weighting).  The calculation runs very well, but the CI vectors have the wrong number of electrons!  This seems to be a bug, because it seems the program _wants_ to report the right labels, but is printing zeros instead of a,b,2 resulting in apparent redundancies, for example, the singlet state has 9 different coefficients all labelled ‘200000’:

 200000      0.2297617
 200000      0.2297617
 200000     -0.2173852
 200000     -0.2173852
 200000     -0.2173849
 200000     -0.2173849
 200000      0.1932838
 200000      0.1932822
 200000      0.1606696

Moreover, in the MULTI output the correct electron number and active space dimension are reported — for example, again for the singlet state:

Number of closed-shell orbitals:  43 (  43 )
 Number of active  orbitals:        6 (   6 )
 Number of external orbitals:      95 (  95 )

 State symmetry 1

 Number of electrons:     4    Spin symmetry=Singlet   Space symmetry=1
 Number of states:        1
 Number of CSFs:        105   (225 determinants, 225 intermediate states)
…indicating that the program does recognise the electron number specifications given in the input.

So what’s going on?  Can these results be trusted?  Is the program meant to behave this way?  Have I found a bug in the output code?  Any clarification is appreciated.  I’m running 2015.1.

Many Thanks,

Seth Olsen, PhD.
Honorary Fellow
School of Mathematics & Physics
The University of Queensland
QLD 4072  Australia
Ph: +61 7 3365 2816
A PGP public key for this address has been uploaded to the key servers.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20170921/f9a69e61/attachment.html>

More information about the Molpro-user mailing list