[molpro-user] A bug of CASPT2-F12 numerical gradient?

[Hans-Joachim Werner]

Please use 

forces,numerical 
or 
optg,numerical

then it will hopefully work.
We will try to fix this problem as soon as possible.
Best regards
Joachim Werner
-- 
Prof. Dr. Hans-Joachim Werner
Institute for Theoretical Chemistry
University of Stuttgart
Pfaffenwaldring 55
D-70569 Stuttgart, Germany
Tel.: (0049) 711 / 685 64400
Fax.: (0049) 711 / 685 64442
e-mail: werner at theochem.uni-stuttgart.de




Am 17.12.2012 um 16:46 schrieb SCAr Chung <s4c6ar at gmail.com>:

> Dear Molpro developers,
> 
>   I am sorry that my numerical gradient failed (ALASKA was called), when using CASPT2-F12. The same problem was also found for MRCI-F12. No problem for non-F12 cases.
> 
> Input:
>  ***,
>  memory,200,m;
>  basis=vdz-f12;
>  explicit,ri_basis=vdz,df_basis=vdz,df_basis_exch=vdz
>  geomtyp=xyz
>  geometry=test.xyz
>  {multi;occ,9;closed,7;
>  wf,charge=0,symmetry=1,spin=0;
>  STATE,1;
>  DM;}
>  {RS2-f12,shift=0.2,IGNORE
>  }
>  force
> 
> End of my output:
> 1PROGRAM * ALASKA (Gradient of the energy)     Author: R. Lindh
> 
> 
>  RECORD SPECIFICATION MISSING ON MCSCF CARD
> 
>  GLOBAL ERROR fehler on processor   0
> 
> Best Wishes,
> 
> Oscar
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user