[petsc-users] Building a rectangular MPI matrix relating to PetscSections

[Jed Brown]

Alejandro D Otero <aotero at fi.uba.ar> writes:

> Hi all, is it posible to build a rectangular matrix that relates two fields
> with (posible) different number of DoF per point defining the nonzero
> structure from 2 petscsections representing one field each?

There isn't a "convenience function", but of course you can create this
rectangular matrix.

What you plan to do with it and your discretization will affect whether
or not representing it as an assembled matrix is a good choice.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 818 bytes
Desc: not available
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20150714/dc6b2f32/attachment.pgp>