>
> I don't understand this.  An element (and therefore all quadrature
> points on that element) is either entirely owned by a processor or is
> entirely not.  Do you have a coarser grid that you're integrating on?


Each element is entirely on one process, but the quantity that I'm
integrating isn't. This is why I ghosted the element and I'm performing
part of the integration one multiple processes.

What elements are you calling id() on?


For now I'm testing the code on one process, so all elements should be
local. I could also be calling id() on ghosted elements.

On Tue, Sep 22, 2015 at 4:55 PM, Roy Stogner <royst...@ices.utexas.edu>
wrote:

>
> On Tue, 22 Sep 2015, Harshad Sahasrabudhe wrote:
>
> I'm using ParallelMesh to discretize the Poisson equation. I want to
>> integrate a certain quantity over each element. That quantity is not
>> available for all quadrature points of some elements on one process (such
>> elements are ghosted), so I sum the parts which are available in
>> respective
>> processes for that particular element and MPI reduce the sum to get the
>> complete integral.
>>
>
> I don't understand this.  An element (and therefore all quadrature
> points on that element) is either entirely owned by a processor or is
> entirely not.  Do you have a coarser grid that you're integrating on?
>
> Now to be able to do this, I have to identify the elements for which the
>> integral has to be MPI reduced. Can anyone help me with how the element
>> can
>> be identified? I tried using elem->id() but I get an exception in
>> DofObject::id(), since the _id == invalid_id.
>>
>
> I also don't understand how this can happen.  Local and ghosted
> elements have their own ids.  Remote element shims return
> remote_elem_id (!= invalid_id).  What elements are you calling id()
> on?
> ---
> Roy
>
------------------------------------------------------------------------------
_______________________________________________
Libmesh-users mailing list
Libmesh-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/libmesh-users

Reply via email to