schnellerhase / dolfinx

Next generation FEniCS problem solving environment
https://fenicsproject.org
GNU Lesser General Public License v3.0
0 stars 0 forks source link

Read up on PETSc MG implementation (DM data structure) #7

Closed schnellerhase closed 5 days ago

schnellerhase commented 3 weeks ago

We are interested in understanding how the DM data objects look like and are constructed.

This is probably a blueprint for a necessary data structure in dolfinx to allow for more efficient restriction/prolongation operations as well!

schnellerhase commented 6 days ago

The data structure of choice should be a DMPlex object.

schnellerhase commented 6 days ago

Especially we will need a transfer operation of a dolfinx topology/mesh representation to a DMPlex object. Need to figure out how refinement is implemented in the petsc world.

schnellerhase commented 5 days ago

We will not want to adapt to another mesh data structure, but only work with the transfer operator view of things, i.e. work with PETSc by only supplying the restriction and prolongation operator