[molpro-user] problem in CI

[Toru Shiozaki]

Dear Glen,
with the latest Molpro of mine, it fails with a different error message, which implies that
the current set-up (the choice of active space etc) is somewhat unrealistic.

In my opinion, it is a good idea to perform a full-valence MULTI (quite feasible in this case) to see
what kind of active space is needed. Of course they are lower than the ones computed
by the smaller CAS, but it is important to make sure that you are not missing states in the same irrep.

MULTI for 3 states in irrep 1:
with full valence active space
>  TOTAL ENERGIES                      -620.63129643  -620.51906188  -620.50128320

with your input, occ,12,3
>  TOTAL ENERGIES                      -620.58599756  -620.46529423  -620.23047903

with occ,13,3, for instance,
>  TOTAL ENERGIES                      -620.59191546  -620.48158728  -620.46216682


Best wishes,
Toru

On Apr 13, 2011, at 9:36 PM, Glen Jenness wrote:

> Dear Molpro users,
> I have a colleague trying to do a CI calculation, but he keeps getting the following error:
> 
>  FILE 7 RECORD    3300 OFFSET=          0. NOT FOUND
> 
>  Records on file 7
> 
>  IREC   NAME  TYPE        OFFSET    LENGTH   IMPLEMENTATION   EXT   PREV   PARENT  MPP_STATE
>    1    3000                  0.       11.         sf          0      0      0      0
>    2    3100                 11.        6.         sf          0      0      0      0
>    3    4000                 17.       33.         sf          0      0      0      0
>    4    4001                 50.       33.         sf          0      0      0      0
>    5    4002                 83.       33.         sf          0      0      0      0
>    6    3800                116.       33.         sf          0      0      0      0
>    7    3802                149.       33.         sf          0      0      0      0
>    8    3700                182.        9.         sf          0      0      0      0
>    9    3200                191.     1452.         sf          0      0      0      0
> 
>  ? Error
>  ? Record not found
>  ? The problem occurs in readm
> 
> Since I've never done any CI calculations with MOLPRO, I'm not quite sure how to overcome this.  Does anyone have any suggestions?
> 
> Thanks!
> Glen Jenness
> 
> P.S.  Here is his input.
> 
> 
> memory,300,m
> 
> gprint,orbitals,civector
> symmetry,x
> orient,noorient
> geometry={
>          Na;
>          Cl1,Na,r2}
> 
> basis=avtz
> 
> r2=2.4
> 
> 
> 
> reforb=2140.2
> refci=6000.2
> savci=6100.2
> 
> text,compute wavefunction at reference geometry
> 
> {hf;occ,11,3;wf,28,1;
> orbital,2100.2}
> 
> {multi;occ,12,3;closed,10,3;
> wf,28,1;state,3;
> natorb,reforb
> noextra}
> 
> {ci;occ,12,3;closed,10,3;
> wf,28,1;state,3;
> thresh,thrdls=1.d-8,thrdlp=8
> orbital,reforb
> save,refci}
> 
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user