If the data set is large enough to warrant it, Sparse Matrix theory should allow you to set up an ~N-dimensional mechanism with semi-reasonable amounts of coding pain. I've been meaning to write something along that line for a long time and never quite gotten around to more than nibbling at the edges of the problem. Since I only have two or three substantial data sets to put into it, I've been ~lazy and kept the indexes in my head. It definitely limits sharing and expanding them.
If you or madbodger find anything that even is a start or an approximation of the full deal, please let me know. And iff it's open source, maybe I/we can help?
(no subject)
If you or