[molpro-user] Problem with state-averaged RASSCF

Dr Seth OLSEN s.olsen1 at uq.edu.au
Tue Jun 14 08:16:26 BST 2005


Hello Molpro-Users,

Thank you Professor Werner for your reply earlier to my question.  I do
have a related question for the list, however.  I understand that
deletion of single excitations from a RASSCF calculation will improve
convergence due to the redundancy of the single excitations and orbital
transformations.  However, this might complicate state-averaged
calculations, for example on both an A and a B state of a planar
congugated or aromatic hydrocarbon.  In these cases, wouldn't single
excitations be necessary in order to capture 'orbital relaxation'
effects for each state?  Or would double excitations take care of this?
 Would one just see a change in the second excitation CI coefficients
compensating for the missing singles, as these would normally compensate
to first order for the restriction of casting both states in the same
orbital basis?

Cheers,

Seth


ccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccms

Dr Seth Olsen, PhD
Postdoctoral Fellow, Computational Systems Biology Group
Centre for Computational Molecular Science
Chemistry Building,
The University of Queensland
Qld 4072, Brisbane, Australia

tel (617) 33653732
fax (617) 33654623
email: s.olsen1 at uq.edu.au
Web: www.ccms.uq.edu.au 

ccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccms




----- Original Message -----
From: "H. -J. Werner" <werner at theochem.uni-stuttgart.de>
Date: Tuesday, May 24, 2005 5:29 pm
Subject: Re: [molpro-user] Problem with state-averaged RASSCF
> Dear Seth,
> It is exactly what is prints: if you want to do this calculation,
> you have to increase the parameter mxshel and recompile. 
> But it will probably be expensive and very hard to converge
> unless you remove single excitations (restrict,-1,0 see manual).
> An easier way to see which orbitals are important would
> be to do an MRCI with a small active space and compute/print
> the natural orbitals and their occupation numbers.
> Best regards
> Joachim Werner
> Likely this will be too big in any case. 
> On Di, 24 Mai 2005, Dr Seth OLSEN wrote:
> 
> >
> >
> >Hi Molpro-users,
> >
> >I'm trying to do a 2-state averaged RASSCF calculation which 
> behaves like a CAS(2,2)*MRCISD calculation within a larger active 
> space (18,18).  My system has 126 electrons, and I am interested in 
> the singlet state, in spatial point group C1. I have tried to do 
> this using the RESTRICT card in multi.  My MULTI deck looks like this:
> >
> >mcscf
> >start,2140.2
> >occ,72
> >closed,54
> >config
> >wf,126,1,0
> >rotate,53.1,57.1,0
> >rotate,50.1,56.1,0
> >rotate,45.1,55.1,0
> >restrict,0,2,65.1,66.1,67.1,68.1,69.1,70.1,72.1
> >print,ref1
> >state,2
> >weight,1,1
> >natorb,2141.2
> >
> >That is, restricting configurations to those with up to two 
> electrons in the orbitals above 'HOMO+1', up to the top of the 
> active space.
> >
> >The job is failing with the following error message:
> >Number of closed-shell orbitals: 54 ( 54 )
> > Number of active  orbitals:      18 ( 18 )
> > Number of external orbitals:    196 (196 )
> > 
> > State symmetry 1
> > ********************
> > *  in CIREF        *
> > * nshell is     10 *
> > *  Hardwired !!!   *
> > ********************
> > 
> > Number of electrons: 18    Spin symmetry=Singlet     Space 
> symmetry=1> Number of states:     1
> > Number of CSFs: *******
> > Number of states:     2
> > ********************
> >    in CICON
> >   nshell,mxshel     12     11
> >  nshell.ge.mxshel !!
> >  Set mxshel=13 in commons cconf,cconf1
> >       AND
> >  Set mxshlr=13 in common cref
> > ********************
> > 
> > ERROR EXIT
> > CURRENT STACK:      MAIN
> > 
> >
> >Do I have my RESTRICT syntax wrong?  Is there something else that 
> I've missed?
> >
> >Cheers,
> >
> >Seth
> >
> >
> >
> >ccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccms
> >
> >Dr Seth Olsen, PhD
> >Postdoctoral Fellow, Computational Systems Biology Group
> >Centre for Computational Molecular Science
> >Chemistry Building,
> >The University of Queensland
> >Qld 4072, Brisbane, Australia
> >
> >tel (617) 33653732
> >fax (617) 33654623
> >email: s.olsen1 at uq.edu.au
> >Web: www.ccms.uq.edu.au 
> >
> >ccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccms
> >
> >
> >
> 
> -- 
> Prof. Hans-Joachim Werner
> Institute for Theoretical Chemistry
> University of Stuttgart
> Pfaffenwaldring 55
> D-70569 Stuttgart, Germany
> Tel.: (0049) 711 / 685 4400
> Fax.: (0049) 711 / 685 4442
> e-mail: werner at theochem.uni-stuttgart.de
> 



More information about the Molpro-user mailing list