On Mon, Nov 22, 2021 at 7:24 PM Matthew Knepley <[email protected]> wrote:

On Mon, Nov 22, 2021 at 7:21 PM Alexis Marboeuf < [email protected]> wrote:

...

Dear PETSc Team,

I am writing a DMPlexCreateNaturalVector routine in src/dm/impls/plex/plexnatural.c. What is the way to properly interface the call such as for DMCreateGlobalVector? In particular, do you want to unify the call with its DMDA counterpart DMDACreateNaturalVector of src/dm/impls/da/dadist.c?

This exists currently. However, we are debating how it should behave. Currently the natural vector has a global order that matches the original global order, but has the current partition. It could be that we want a vector that matches the original partition, so we are discussing this online.

I forgot to point you to the discussion: https://gitlab.com/petsc/petsc/-/merge_requests/4547 Thanks, Matt

Thanks,

Matt

...

Thanks, Alexis

-- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener

https://www.cse.buffalo.edu/~knepley/ <http://www.cse.buffalo.edu/~knepley/>

-- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener https://www.cse.buffalo.edu/~knepley/ <http://www.cse.buffalo.edu/~knepley/>