[Dune] YaspGrid refinement question

[Martin Nolte]

The method preAdapt returns true if there is an entity that might be caorsened.
This is also clearly stated in the documentation of Dune::Grid.

So, if you only mark entities for refinement, the method preAdapt() should
return false. You can find out whether entities have been refined by the return
value of adapt().

This means the implementation in YaspGrid does exactly what it should do, since
YaspGrid does not support coarsening.

Cheers,

Martin


Andreas Dedner wrote:
> I far as I can see it, this is a bug in YASP.
> I understand that preAdapt only returns false if
> a call to adapt will lead to no modification of the
> grid. The default implementation returns false and that
> is used in YASP. The default implementation for adapt
> is of course to do nothing - so this is consistent.
> Yhe implementation of adapt on YASP is to call a 
> globalRefine - independent of any marker or anything
> of that sort.
> I think that this is not intuitive behaviour and can
> be quite problematic in adaptive computations since
> the grid is globally refined each time adapt is called.
> I had assumed that a grid with no local
> adaptation (as given by the Capabilities) does nothing
> if adapt is called.
> But prehaps I misundersood the code...
> Andreas
> On Tue, 3 Mar 2009 17:15:49 +0100 (CET)
>  Christian May <cmay at phys.ethz.ch> wrote:
>> Dear developers,
>>
>> as I understand it, preAdapt() should return true if an
>> entity of the grid has been marked for refinement by the
>> mark() method.
>> In the attached program, all entities are marked.
>> However, if I run it, it always returns false.
>> Can you please give me a hint?
>>
>> Thanks
>> Christian
>>
>>
>>
> ---------------------------------------------------------------------------
>> Christian May
>>
>> ETH Zurich
>> Institute for Theoretical Physics, HIT K 31.5
>> CH-8093 Zurich, Switzerland
>>
>> Tel: + 41 44 633 79 61
>> Fax: + 41 44 633 11 15
>>
>> Email: cmay at itp.phys.ethz.ch
>>
> 
> 
> _______________________________________________
> Dune mailing list
> Dune at dune-project.org
> http://lists.dune-project.org/mailman/listinfo/dune

-- 
Martin Nolte <nolte at mathematik.uni-freiburg.de>

Universität Freiburg                                   phone: +49-761-203-5642
Abteilung für angewandte Mathematik                    fax:   +49-761-203-5632
Hermann-Herder-Straße 10
79104 Freiburg, Germany