[petsc-users] Getting unsupported type message from MatCholeskyFactor

Hong hzhang at mcs.anl.gov
Mon May 4 17:23:33 CDT 2015


Alp :

> I have tested seqaij and mpiaij types but still getting unspoorted type
> error from MatCholeskyFactor function. Should I activate an external
> packege like SuperLU ?
>

Petsc only supports sequential  Cholesky. You must use MUMPS for parallel
Cholesky.
SuperLU is a sequential; SuperLU_DIST supports parallel LU, not Cholesky.

Hong
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20150504/7f644bfd/attachment.html>


More information about the petsc-users mailing list