On Tue, Apr 14, 2020 at 5:49 AM Hill, Reuben <reuben.hil...@imperial.ac.uk>
wrote:
> Hi all,
>
> I might be missing something obvious, but I can't tell from the
> documentation if 1D coordinate DMSWARMs (immersed in interval DMPlex
> meshes) are supported. Does anyone know?
>
> I've successfully implemented 2D and 3D coordinate DMSwarms in 2D and 3D
> DMPlexes using DMSwarmSetPointCoordinates in Firedrake using petsc4py (via
> swarm.setPointCoordinates. The petsc4py function forces the input numpy
> array to have 2 dimensions with 1 column per dimension. In the 1D case,
> where each row of the coordinates array therefore has one column, I get the
> following error:
>
> E petsc4py.PETSc.Error: error code 63
> E [0] DMSwarmSetPointCoordinates() line 305 in
> /Users/rwh10/firedrake/src/petsc/src/dm/impls/swarm/swarmpic.c
> E [0] DMLocatePoints() line 6499 in
> /Users/rwh10/firedrake/src/petsc/src/dm/interface/dm.c
> E [0] DMLocatePoints_Plex() line 744 in
> /Users/rwh10/firedrake/src/petsc/src/dm/impls/plex/plexgeometry.c
> E [0] DMPlexLocatePoint_Internal() line 462 in
> /Users/rwh10/firedrake/src/petsc/src/dm/impls/plex/plexgeometry.c
> E [0] Argument out of range
> E [0] No point location for cell 0 with type segment
>
Point location in 1D was not implemented because no one ever asked for it.
Do you need this? or is this just for completeness?
Thanks,
Matt
> Thanks
>
> Reuben Hill
>
--
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/>
