dolfin team mailing list archive

[Johan Hake <johan.hake@xxxxxxxxx>]

On Friday 19 March 2010 02:30:40 Anders Logg wrote:
> Question #104855 on DOLFIN changed:
> https://answers.launchpad.net/dolfin/+question/104855
> 
>     Status: Open => Answered
> 
> Anders Logg proposed the following answer:
> When using eval() you need to subclass Expression (not Function).
> 
> That gives you an Expression that can be used in a variational
> problem. But an Expression does not have a vector so if you need the
> norm of the vector (I guess just for debugging), you need to project
> or interpolate the Expression to a Function.
> 
> class Alpha(Expression):
>     def eval(self, v, x):
>         v[0] = alpha1
>         if x[0] > 0.5: v[0] = alpha2
> 
> alpha = Alpha()
> d = interpolate(alpha, V)
> 
> print " |d| = %g" %(d.vector().norm("l2"))
> 
> --
> Anders
> 
> On Fri, Mar 19, 2010 at 08:49:35AM -0000, Achim Schroll wrote:
> > New question #104855 on DOLFIN:
> > https://answers.launchpad.net/dolfin/+question/104855
> >
> > Hi!
> > My piecewise defined function alpha (see below) just returns zero.
> > What is wrong? How to define it instead?
> >
> > Best, Achim
> >
> > # set parameters
> > alpha1 = 0.2; alpha2 = 0.8
> >
> > # create mesh and finite element
> > mesh = UnitSquare(20, 20)
> > V = FunctionSpace(mesh, "CG", 1)
> >
> > # define trial & testfunctions
> > u_trial = TrialFunction(V); phi = TestFunction(V)
> >
> > # define diffusion parameter
> > class alpha(Function):
> >     def __init__(self, V):
> >         Function.__init__(self, V)
> >     def eval(self, v, x):
> >         v[0] = alpha1
> >         if x[0] > 0.5: v[0] = alpha2
> > d = alpha(V)
> > print " |d| = %g" %(d.vector().norm("l2"))

What version of DOLFIN are you using. It has not been possible to overload 
Function.eval in PyDOLFIN since 0.9.5. 

You might want to upgrade ;)

Johan