On Oct 9, 2008, at 7:34 PM, Angus Hendrick wrote:
I have a 3x3 domain. I want the center of the left face to have a
fixed value boundary condition, and the remainder to have a fixed
flux condition.
In fipy 1.2.1, I could just say bcs = FixedValue(mesh.getFacesLeft()
[1], value=0) and it worked. More comprehensive implementation of
lazy evaluation to thwarts that approach in the trunk version, and
instead I get the complaint: AttributeError: 'unOp' object has no
attribute 'getMesh', presumably because the index operator is not
resolving? Of course even if it did resolve, it doesn't look like
that would work anymore since getFacesLeft returns boolean
FaceVariable, instead of a list of faceIDs.
I think this is covered in
http://article.gmane.org/gmane.comp.python.fipy/945
Basically, we've never promised that asking for a particular index of
a CellVariable is reproducible or meaningful (I'm occasionally tempted
to scramble them up out of pure spite). The right way to do it is to
specify the coordinates, like in the examples in that message.
Just so you don't think we're ignoring you: I'm on leave for the next
week or so. I think Daniel may be out for awhile, too.
