Interpolation onto function spaces with periodic boundary conditions appears to be broken

[Patrick Farrell]

Consider the following code:

from dolfin import *

mesh = RectangleMesh(0, 0, 1, 2, 50, 50)

class PeriodicBoundary(SubDomain):
  def inside(self, x, on_boundary):
    return abs(x[0]) < DOLFIN_EPS and on_boundary

  def map(self, x, y):
    y[0] = x[0] - 1

pbc = PeriodicBoundary()

V1 = FunctionSpace(mesh, "CG", 1)
V2 = FunctionSpace(mesh, "CG", 1, constrained_domain=pbc)

expr = Expression("x[1]")

func1 = interpolate(expr, V1); plot(func1, title="Without PBC")
func2 = interpolate(expr, V2); plot(func2, title="With PBC")

interactive()

I expected both of the plots to be the same: the function space is periodic in the x-boundary, and I've interpolated y. But they are not: there's an odd spike in the bottom-left corner, and the right-hand boundary has the wrong values.

[Patrick Farrell]

This happens both with 1.2 and trunk.

[Mikael Mortensen]

Try

def map(self, x, y):
    y[0] = x[0] -1
    y[1] = x[1] 

That works for me

Mikael

[Patrick Farrell]

Ah, what a silly mistake. Thanks, Mikael. I'll add a check in PeriodicBoundaryComputation.cpp to make sure the user has set all of the necessary dimensions.