[petsc-dev] Adjacency relations in Sieve and stencils

Matthew Knepley knepley at gmail.com
Thu Aug 23 13:22:11 CDT 2012


On Thu, Aug 23, 2012 at 12:19 PM, Chris Eldred <chris.eldred at gmail.com>wrote:

> I am working with a 2D unstructured mesh using sieve and I wanted to
> get some advice on how to determine adjacency relations using the
> topological operators provided by Sieve (cone, support, closure,
> star). I have figured out the following:
>
> Vertices for a given edge: cone(edge)
> Cells for a given edge: support(edge)
>
> Cells for a given vertex: star(vertex) \ support(vertex)
> Edges for a given vertex: support(vertex)
>
> Edges for a given cell: cone(cell)
> Vertices for a given cell: closure(cell) \ cone(cell)
>
> The set of edges associated with the cells for a given edge: U
> cone(support(edge)) - ie the union of the cones of the support of the
> edge
>
> Are there other topological operators that more naturally express
> these relations (especially the ones involving unions or complements)?


The only other builtin operations are meet and join. You are correct that
you
want to use union there. We could add routines like

  DMComplexGetConeUnion(dm, [p], &unionSize, &unionArray);

but it was not clear that would do anything other than bloat the interface.
I
think the best way to procede is to write these convenience routines using
the lower level primitives, and move them into PETSc if they turn out to be
useful. I can help you write them.

   Matt


>
> --
> Chris Eldred
> DOE Computational Science Graduate Fellow
> Graduate Student, Atmospheric Science, Colorado State University
> B.S. Applied Computational Physics, Carnegie Mellon University, 2009
> chris.eldred at gmail.com
>



-- 
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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120823/c4391994/attachment.html>


More information about the petsc-dev mailing list