On 02/03/10 11:19, Richard Norton wrote:
New question #102925 on DOLFIN:
https://answers.launchpad.net/dolfin/+question/102925
I cannot figure out how to fix the following problem... In the following code
I would like to modify sigma so that it defines an operator from u to a vector
in R^2 in the following way...
def sigma(u):
If abs(u.dx(1))< 1:
return as_vector( [u.dx(0), 0])
else:
return as_vector( [u.dx(0), 2*(u.dx(1)**2 -1)*u.dx(1)] )
so that sigma returns something different if the gradient of u in the y direction has
magnitude less than 1 (at each point in the domain). I suspect there is a problem with
"abs(u.dx(1))< 1"
Conditional statement are not (yet) supported in the generated code.
What you can do is create a new Function and make it equal to
w = 2*(u.dx(1)**2 -1)*u.dx(1)
and then zero it when abs(u.dx(1)) < 1.
There is a blueprint to add support for conditional statements.
https://blueprints.launchpad.net/ffc/+spec/ufl-conditionals
Garth
---------------------------------------
mesh = UnitSquare(n, n)
V = FunctionSpace(mesh, "CG", 1)
...
# Test and trial functions
v = TestFunction(V)
u = TrialFunction(V)
uold = Function(V)
def sigma(u):
return as_vector( [u.dx(0), 2*(u.dx(1)**2 -1)*u.dx(1)] )
# Variational problem
a = dot(grad(u), grad(v))*dx
L = dot(sigma(uold),grad(v))*dx
_______________________________________________
Mailing list: https://launchpad.net/~dolfin
Post to : dolfin@lists.launchpad.net
Unsubscribe : https://launchpad.net/~dolfin
More help : https://help.launchpad.net/ListHelp
