KernelCondDensity

Multidimensional kernel density estimation with optimal and asymptotically optimal bandwidth choices

Implements the Rosenblatt-Parzen density estimator for multidimensional domains. It can implement different bandwidth choices (Silverman plugin is asymptotically optimal, a leave one out estimator leads to a result which might be more appriate in smaller samples). There are several features implemented (bootstrapping, log likelihood calculations, etc). Please let me know if any mistakes show up.