[molpro-user] WARNING: exit with wrong ci vector!!

[Kirk Peterson]

Dear Jun,

this usually happens when the initial guess orbitals are poor.  Since aVTZ works, I would recommend first doing an aVTZ CAS and then immediately follow that with an aVQZ calculation.

best wishes,

Kirk

On Feb 29, 2012, at 10:04 AM, Li Jun wrote:

> Dear all,
> When I used aVQZ in my SA-CAS(15e, 10o) calculations, the following warning occurred :
> WARNING: exit with wrong ci vector!!
> And then some calculations were successfully done, some failed with errors as:
> NON-ZERO P-SPACE GRADIENT
> While, these messages were never sent if aVTZ basis set was used. Then my question is:
> 1) What is necessary to avoid this?
> 2) For successful calculations with the "WARNING: exit with wrong ci vector!!", are the results reliable, or what is the criterion to verify its reliablity?
> I just used the following input:
> {casscf;natorb;occ,12;closed,2;pspace,2.0d0;wf,19,1,1;state,6;}
> Thanks in advance!
> Best
> Jun Li
> 
> 
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20120229/f016b1ac/attachment.html>