I am working on joint and conditional density trees for approximating clique potentials in Bayesian Belief Networks. A brief introduction to topic is available from this paper in case you'd like to get a better description what I'm talking about.
I am looking for an implementation that supports both discrete and continuous variables in a joint probability distribution. The leaf nodes in the learned tree would provide probability distributions for both discrete and continuous variables.
I'd be very glad to find an open source implementation, but any implementation in any form would work for me, at least as a test for my own implementation (which I'd have to do anyway).
I've seen this approach described in various papers in different contexts, but I could not find any implementations.