NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
> The current VisAD Field implementations attach values to > points rather than lines (or polygons or polyhedra). It > would be possible but a lot of work to define new extensions > of Function for values attached to line segments. > > However, you could easily generate an appropriate type > of Set for your line segments and attach values to the > end points of the line segments. You might use a UnionSet > of Gridded2DSets with manifold dimension = 1, as we do for > map outlines (see visad/examples/Rivers.java for a simple > example). For lines in 3-D you can use a UnionSet of > Gridded3DSets with manifold dimension = 1, and for triangles > in 3-D an Irregular3DSet with manifold dimension = 2. What hit me a bit later was that the regions are one or more closed curves. I just have to sort them before use them as samples for the grid. That would also reduce the number of samples to half of the original since every end point in a line is another lines start point. Something like: a->b b->c c->a Would result in: a->b->c The algorithm for that is rather slow, but it shouldn't be do many line segments so it won't be that much of a problem. Gunnar
visad
archives: