[molpro-user] Numerical Freq calculations using F12-CCSD(T)

[John Travers]

Dear Werner
Thank you for the suggestions! I guess it is related to the choice of the basis. Since I used a aug-cc-pcwVQZ basis, there is no corresponding JK fit ones. I used an aug-cc-pV5Z JKFIT (see attached input). Do you have other suggestions for the choice of ri, df, and jk basis for the aug-cc-pcwVQZ orbital basis? BTW, the molpro version I used is 2012.1. Thanks again!
Best wishes
John 

    On Tuesday, January 17, 2017 8:13 AM, Werner Győrffy <gyorffy at theochem.uni-stuttgart.de> wrote:
 

 Dear John,

There might be some problem with the choice of the basis sets you use. 
It is recommended to use the optri basis sets, for example, for 
basis=avtz use ccsd(t)-f12b,ri_basis=avtz/optri. Could you please 
provide an input for your calculation failed?

Regards,

Werner.

On 01/16/2017 05:41 PM, John Travers wrote:
> Hi
>
> I was trying to do a numerical frequency calculation using f12-ccsd(t)
> in c1 symmetry for a small molecule. The program always produces some
> warning messages like : "Poor overlap with frozen basis set. May have
> deleted the wrong vectors!
>  ? The problem occurs in SvdConstructOrthoBasis"
>
> When checked the detailed output,  I also found some possible numerical
> errors, something looks like :
>
>  Construction of CABS:
>  Pseudo-inverse stability              NaN
>  Smallest eigenvalue of S        -3.71E-13  (threshold= 1.00E-08)
>  Ratio eigmin/eigmax              -3.71E-13  (threshold= 1.00E-09)
>  Smallest eigenvalue of S kept    -3.71E-13  (threshold=-3.71E-13, 0
> functions deleted, 658 kept)
>  Maximum overlap with frozen deleted vectors: (nDel=  44)
>    Vector:        1      2      3      4      5      6      7
>    8      9      10      11      12      13      14      15      16
>  17      18      51      25      67      44      36      35      43
>  32      40      38      33      39      55      52      34      22
>  37      21      42      50      20      41      74      19      53
>  54      26      23    231      28      24      95      93      27
> 158      75      61    123      68    108      31      63    163
> 121    203    152      62      29    157      45      94      72
> 97      77      65    204    103      30    115    126      66
> 250      73    131    101    162    116      91    200    122
>    Overlap:      NaN    NaN    NaN    NaN    NaN    NaN    NaN
>  NaN    NaN    NaN    NaN    NaN    NaN    NaN    NaN    NaN
>  NaN    NaN  1.436  1.248  1.102  1.000  0.999  0.997  0.996
> 0.994  0.985  0.983  0.973  0.964  0.962  0.912  0.896  0.871
> 0.858  0.853  0.827  0.825  0.742  0.596  0.594  0.537  0.523
> 0.485  0.465  0.386  0.375  0.373  0.346  0.327  0.321  0.310
> 0.307  0.268  0.259  0.257  0.249  0.247  0.216  0.204  0.194
> 0.191  0.183  0.166  0.164  0.159  0.135  0.134  0.130  0.126
> 0.118  0.114  0.110  0.100  0.095  0.093  0.092  0.092  0.091
> 0.089  0.080  0.079  0.079  0.078  0.076  0.075  0.072  0.070
>
>
> My questions are related to these "NaN" and warnings. Do these matter
> and how to avoid this? Thanks!
>
> Best wishes
>
>
> John
>
>
>
>
>
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user
>
_______________________________________________
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/20170118/d6f21a00/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.inp
Type: application/octet-stream
Size: 3161 bytes
Desc: not available
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20170118/d6f21a00/attachment.obj>