Cholesky decompose failed: the matrix is not positive definite or ill-conditioned.

699 views
Skip to first unread message

Andy Wang

unread,
Oct 14, 2022, 6:28:28 PM10/14/22
to cp2k
Dear all,

I have a structure which has been optimized by Gaussian. Then I want to use CP2K to scan a pair of dihedrals. I am using M06 with all electron potential. The basis set is aug-cc-pvtz. Is the error related to this large basis set?

Thanks in advance for any suggestions.

Bests,
Andy


geo.in
tmp.pdb

Krack Matthias (PSI)

unread,
Oct 15, 2022, 11:19:36 AM10/15/22
to cp...@googlegroups.com

Hi Andy

 

Such an error can be caused by a bad input structure or by a basis set with relatively small exponents. Maybe the latter is true in your case. You can try to switch CHOLESKY off.

 

HTH

 

Matthias

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/6aeb5cbf-82ad-4172-9340-64dd2ab13c49n%40googlegroups.com.

Reply all
Reply to author
Forward
0 new messages