On Fri, Dec 2, 2016 at 8:35 AM Roy Stogner <royst...@ices.utexas.edu> wrote:
We maintain node_to_element maps in MOOSE. We're already paying for the
time to construct that data structure. If you have that (or as you suggest,
something similar) determining whether the node is evaluable is easy.
Perhaps we should entertain moving the optional construction and
maintenance of that data structure to libMesh?
Cody
> On Thu, 1 Dec 2016, Cody Permann wrote:
>
> > I'm starting to use the new Elem::isEvaluable() method available in
> libMesh. It seems like it would be handy to promote this method up to
> > DofObject though so it could be used with Node. If I understand
> correctly, a node "should" be evaluable if it's attached to any element
> that's
> > evaluable so this method could be implemented rather easily. Does this
> sound reasonable?
>
> The bad news is that getting a list of all the elements attached to a
> Node is surprisingly hard - there's basically no way to do it without
> a PointLocator, and no way to do it in O(1) time.
> The good news is that we don't need a list of all the elements sharing
> the node, just a list of all the DoF indices hosted on the node. My
> inclination would be to add an extra overload to DofMap::dof_indices()
> to grab those, and then the new DofMap::is_evaluable() implementation
> would be the exact same as the old DofMap::is_evaluable(); we could
> actually just make them two instantiations of the same templated
> function.
> ---
> Roy
>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Libmesh-devel mailing list
Libmesh-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/libmesh-devel
