On Nov 7, 2008, at 11:21 AM, Angus Hendrick wrote:
The pyamg test code also does not build a very complicated matrix,
using a pyamg "poisson" function that takes only a domain size as an
argument to deliver the sparse matrix.
I think what Daniel is suggesting is that it's not surprising that
there's not much difference in speed since neither is spending very
much time building the matrix and both are solving the matrix with the
same (or similar) fast code. We're still working on improving FiPy's
speed at building matrices, but we're glad you're happy with the
results you're getting so far.
> I can extract all the problem
> related arrays without much difficulty, but the boundary
conditions are more
> vexing since in the forms I can readily access, the locations
where each
> boundary condition is applied live as unresolved union and
intersection of
> boolean arrays buried inside the FixedValue or FixedFlux
instances. Is
> there an approved method for extracting and resolving this location
> information? If not, is there an easy hack I can use?
This is not particularly "approved", but not particularly
"disapproved", either:
As you've found, bc.faces contains a binOp Variable representing
whatever conditions defined the boundary, so, e.g.,
>>> bc = FixedValue(faces=mesh.getFacesLeft() & (Y > 2.), value=1.)
>>> bc.faces
will dump a mess like
(FaceVariable(value=array([False, False, False, False, False,
False, False, False, False,
False, False, False, True, False, False, False, True,
False,
False, False, True, False, False, False], dtype=bool),
mesh=UniformGrid2D(dx=1.0, dy=1.0, nx=3, ny=3)) & [False False False
False False False False False False True True True
False False False False False False False False True True
True True])
but
>>> bc.faces.getValue()
will return a mask array.
That's hopefully what you needed?
