dolfin team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

On Mon, Dec 04, 2006 at 02:35:21PM +0100, Garth N. Wells wrote:
> Anders Logg wrote:
> > On Mon, Dec 04, 2006 at 02:14:51PM +0100, Garth N. Wells wrote:
> >> Could we add something to FFC to describe where the various degrees of 
> >> freedom live (on vertices, edges, internal)?
> >>
> >> Garth
> > 
> > Yes we could, but I'd rather not. Why do we need it? I'd prefer if
> > DOLFIN did not know anything about dofs, other than how to reorder
> > them.
> >
> 
> Not sure that it's this simple if you want to assemble some terms 
> block-wise. Also, for parallel assembly we might need to know where dofs 
> lie. I'm still thinking about this so I can't be concrete in what's 
> needed just yet.

ok. I'm certainly open to generating everything we need to do
efficient assembly, but my feeling is that it won't be needed.

> > I see two options:
> > 
> >  1. Reorder the dofs computed by FFC, not knowing why and how they
> >     are numbered.
> >
> 
> This is one option I'm considering. Just take the FFC numbering, and run 
> a renumbering algorithm over it to minimise bandwidth.

ok.

> >  2. Compute a different mapping in FFC.
> > 
> 
> This could be difficult as it depends on properties of the mesh, 
> partition, etc. I think that this should remain simple.

Agree.

/Anders

> > I can be convinced otherwise of course, but I don't understand why
> > this is needed.
> > 
> > /Anders
> > _______________________________________________
> > FFC-dev mailing list
> > FFC-dev@xxxxxxxxxx
> > http://www.fenics.org/mailman/listinfo/ffc-dev
> > 
> 
> 
> _______________________________________________
> FFC-dev mailing list
> FFC-dev@xxxxxxxxxx
> http://www.fenics.org/mailman/listinfo/ffc-dev