Thanks, but that one I know.
I need the nearest neighbor to each point, for all points. The method
you used takes of order n^2 operations: 100 for each point and 100 to
search for the minimum distance to that point.
I thought that there were some algorithms that were linear in the
number of points (n^1).
Here's a reference to an N^1.5 algorithm:

http://cgm.cs.mcgill.ca/~soss/cs644/projects/perrier/

Thanks,
Jim