Probability mass functions (PMFs), probability density functions (PDFs), cumulative distribution functions (CDFs) and quantile functions, mainly via (optionally bounded/truncated) kernel smoothing. In the continuous case, there's support for univariate, multivariate and conditional distributions, including distributions that are both multivariate and conditional. Refer to the book "Kernel Smoothing" by Wand and Jones (1995), whose methods are generalized by the methods here. Also, supports categorical distributions, mixed conditional distributions (with mixed input types) and smooth empirical-like distributions, some of which, can be used for statistical classification. There are extensions for computing distance matrices (between distributions), multivariate probabilities, multivariate random numbers, moment-based statistics and mode estimates.