Hi there. I'm investigating using matplotlib for plotting of Adaptive
Mesh Refinement (
http://en.wikipedia.org/wiki/Adaptive_mesh_refinement ) data -- the
primary characteristic of which is that it is of non-equal resolution.
I've used the scipy/delaunay method, as mentioned in the cookbook,
but unfortunately that provides a level of interpolation that is not
always desirable; very often when plotting data, we want to be able to
see clear cell boundaries, as well as boundaries between resolution
levels.
Essentially what I have are the following pieces of data: x, y, dx,
dy, z. The simplest way is for me to sample this using a loop over
points in the module where I handle the data; however, what I'd like
to be able to do is hand it off to matplotlib, and the on-the-fly
change the x,y (and z) bounds. (This seems as though it would be the
more efficient manner of handling the data, anyway.)
Is there a way to do this? If not, would it be terribly difficult for
me to implement? I've browsed the code, and it seems that the best
starting place would by pcolor in lib/matplotlib/pylab.py or
src/_image.cpp. I very much would like to leverage the abilities of
matplotlib -- specifically, I'm very excited about being able to plot
this data, and then overplot contour or quiver plots (which I have
done with my data using the delaunay method.)
Any ideas? Thanks!
-Matt