Hdp-bayes-vowels is a project mainly written in R and PYTHON, it's free.
Hierarchical Dirichlet Process vowel-acquisition device Gibbs sampler (e.g. Feldman et al., 2009)
HDP-BAYES-VOWELS
Ultimately, this will be an implementation of Feldman et al.'s (2009) hierarchical DP model of phonetic/lexical learning.
lex_hdp.r -- code for iterating the lexical-HDP model. fledman2009.r -- code to set-up and run the 1-D tests from the paper
Some warm-up stuff, if you're curious: finiteNormMixtureGibbs.r -- Gibbs sampler for a FMM Gaussian model of arbitrary dimensions dpNormMixtureGibbs.r -- Gibbs sampler for a DP Guassian Mixture model. This needs to be updated to sample from the conditional posteriors for the prior parameters on the component means/covariances, etc. rasmussen2000.r -- Functions/distributions from the Rasmussen (2000) paper on DP as the infinite limit of FMMs