[petsc-dev] PetscSection

[Matthew Knepley]

On Sat, Nov 10, 2012 at 4:13 PM, Jed Brown <jedbrown at mcs.anl.gov> wrote:
> On Sat, Nov 10, 2012 at 3:02 PM, Matthew Knepley <knepley at gmail.com> wrote:
>>
>>
>> > How does one get an IS containing _only_ the labeled points?
>>
>> DMComplexGetStratumIS()
>
> Great. Under what circumstances do you call this routine and get back an IS
> with mixed dimension?

Everything is just a point, so I get back mixed dimension things all the time. I
usually separate by stratum.

>>
>>
>> >>
>> >> > Can you answer how you can distinguish a quad from a tet in a
>> >> > non-interpolated mesh?
>> >>
>> >> You can't, but that is the point. You are not supposed to distinguish
>> >> them.
>> >
>> >
>> > Hmm, so these are the same at the category-theory level, but definitely
>> > not
>> > at the mesh topology level. It doesn't seem right to not distinguish
>> > them at
>> > all. How do you write DMComplexInterpolate for mixed-dim strata?
>>
>> Interpolation is analysis. You need to know what dimension you are in,
>> etc.
>
>
> But it's still topological rather than geometric. You don't need
> coordinates, for example.

You don't need coordinates for interpolation?

   Matt

--
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