On Thu, Aug 6, 2015 at 10:09 AM, Gautam Bisht <[email protected]> wrote:
> Hi Matt,
>
> During the last week discussion with Mark, it became obvious that to add
> PETSc's AMR capability in PFLOTRAN would require incorporation of DMPlex.
>
> Currently, PFLOTRAN has multiple `modes` which include:
> - RICHARDS: Single phase mass conservation,
> -TH: Single phase mass + energy conservation,
> -GENERAL: Two phase mass + energy conservation, etc.
>
> All the different modes have portions of duplicate code for
> Residual/Jacobian evaluation. I believe incorporation of DMComposite would
> possibly shrink the current PFLOTRAN code.
>
> Thus, I was curious to check if an example for DMPlex+DMComposite exists.
>
We have only used DMComposite to deal with things that have really regular
discretizations and topologies, namely
that work for DMDA. DMComposite will not calculate the Jacobian layout (you
can tell it), and its not trivial to calculate
the blocks yourself (in my opinion).
I tend to use Plex by itself for these kind of situations. I guess it
depends on what you want to be in charge of. If you
want to run the whole show (cell iteration, local numbering, etc.) then
DMComposite is a good choice. If you want to
give up some control, and only manage element-wise residuals/Jacobians then
Plex may be a good choice.
Maybe I am missing something in how you want to set this up.
Thanks,
Matt
> -Gautam.
>
>
> On Thu, Aug 6, 2015 at 8:08 AM, Gautam Bisht <[email protected]> wrote:
>
>> I'm going to move this thread over to dev mailing list.
>>
>> -Gautam.
>>
>> On Thu, Aug 6, 2015 at 4:44 AM, Matthew Knepley <[email protected]>
>> wrote:
>>
>>> On Wed, Aug 5, 2015 at 10:21 PM, Gautam Bisht <[email protected]> wrote:
>>>
>>>> Hi Matt,
>>>>
>>>> Instead of using gcc4.9, I reinstalled PETSc using clang on mac os x
>>>> 10.10 and the example runs fine.
>>>>
>>>> Btw, are there any examples that use DMPlex+DMComposite?
>>>>
>>>
>>> I don't think so. What would you anticipate using it for?
>>>
>>> Thanks,
>>>
>>> Matt
>>>
>>>
>>>> Thanks,
>>>> -Gautam.
>>>>
>>>
>>>
>>>
>>> --
>>> 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
>>>
>>
>>
>
--
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
