[molpro-user] a new problem about ngt=411 ngto=433

[Kirk Peterson]

Dear Jules,

the origin of the problem is probably in trying to use cartesian polarization functions (int,,cart) with correlation consistent basis
sets - these are designed only for spherical harmonic functions only. Using cartesian d, f, etc. leads to
linear dependency issues. 

regards,

Kirk

On Jun 19, 2012, at 5:24 PM, Jules Yuan wrote:

> Hi, 
>    I found a problem about blocking error, in detail,the error message appears as--
> 
>  All pairs explicitly correlated. Number of r12-pairs:          528
>  
>  ngt=                   411   ngto=                   433
>  ? Error
>  ? Blocking error
>  ? The problem occurs in df_basis_blocks
> 
>  GLOBAL ERROR fehler on processor   0                             
>             
> 
> ,which sound pretty confusing, that it seems to be throwing away too many basis functions. Also, it can take up 52GB in one hour during the calculation, which is formidable.  The input file is
> 
>  ***,CTSdHc CCSD(T)
>   
>  Symmetry,nosym
>  geometry={
>   20
>  0  1
> ~~~~~~~~~~~ 
> }
>  wf,NELEC=86,SYMMETRY=1,SPIN=0,CHARGE=0
>  basis=vdz-f12
>  int,,cart
>  rhf
>  ccsd(t)-F12a,scale_trip=1
> 
> Later, I found that this problem can be solved by deleting the line 'int,,cart'. Do people have any idea what is the mechanism for this phenomenon?
> Hope this lucky trick might be useful for people encounter the same problem in future.
> 
> Jules
> 
> _______________________________________________
> 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/20120620/dfb7325a/attachment.html>