dolfin team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

On Wed, Oct 07, 2009 at 12:11:57PM +0100, Garth N. Wells wrote:
>
>
> Anders Logg wrote:
> > On Wed, Oct 07, 2009 at 09:08:53AM +0100, Garth N. Wells wrote:
> >> We seem to have a problem with Python and interpolating functions. When
> >> I try
> >>
> >>    P1 = FunctionSpace(mesh_post, "CG", 1)
> >>    u_post = interpolate(u, P1);
> >>
> >> I get the below error.
> >>
> >> Garth
> >>
> >>
> >> Traceback (most recent call last):
> >>    File "demo.py", line 128, in <module>
> >>      u_post = interpolate(u, P1);
> >>    File
> >> "/home/garth/code/fenics/dolfin/dolfin-all/local/lib/python2.6/site-packages/dolfin/interpolate.py",
> >> line 28, in interpolate
> >>      Pv.interpolate(v)
> >> RuntimeError: *** Error: Treating Function as UFC function. Is that correct?
> >
> > Yes, it is correct.
> >
> > I assume you are interpolating u from another space, possibly DG?
> >
> > Then we should get to the else{} clause in the restrict() which
> > triggers the error. I added the error there temporarily so that we can
> > check that we only get to else{} when we should. Before the hash fix
> > yesterday, we got there all the time which works but is very
> > inefficient.
> >
> > I have not had time yet to run more than a couple of demos. Feel free
> > to remove the error if you have checked that all demos run except the
> > few (doing interpolation) where the error should be triggered.
> >
>
> I tried removing the error and running
>
>     demo/function/nonmatching-interpolation/cpp
>
> but and there is a very nasty bug which I can't track down. The code is
> failing inside the generated ufc code for tabulating the dof map.

I'll take a look later tonight.

--
Anders


> Garth
>
>
>
> _______________________________________________
> DOLFIN-dev mailing list
> DOLFIN-dev@xxxxxxxxxx
> http://www.fenics.org/mailman/listinfo/dolfin-dev

Attachment: signature.asc
Description: Digital signature