ffc team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

On Mon, Dec 04, 2006 at 08:00:28AM -0600, Kirby, Robert C wrote:
> Note that all of this information is conveniently stored
> in FIAT in the DualBasis.entity_ids dictionary.

Yes, this is what is used by FFC to generate the code that computes
the mapping. If needed, FFC could generate code that propagates this
info to DOLFIN.

/Anders


> As a reference point on this discussion, only the evaluation
> engine in Sundance seems to need to know about locations
> (to get mappings and to figure out who should be nonzero on
> boundaries).  On the other hand, as a more integrated environment
> than DOLFIN/FFC/FIAT, it's much easier to get the information
> whenever you need it -- same runtime library used in defining
> elements,forms, operations, etc.
> 
> Robert C. Kirby
> Associate Professor
> Department of Mathematics and Statistics
> Texas Tech University
> 
> "Mathematical software should be mathematical"
> 
> 
> 
> -----Original Message-----
> From: ffc-dev-bounces@xxxxxxxxxx on behalf of Anders Logg
> Sent: Mon 12/4/2006 7:21 AM
> To: ffc-dev@xxxxxxxxxx
> Subject: Re: [FFC-dev] dof locations
>  
> 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.
> 
> I see two options:
> 
>  1. Reorder the dofs computed by FFC, not knowing why and how they
>     are numbered.
> 
>  2. Compute a different mapping in FFC.
> 
> 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
>