Parallel Dirichlet Process Slice Sampler

Dirichlet Process is a promising prior for learning non-parametric mixture models. It determines the number of parameters used in the model according to data. In this implementation slice sampler is used to have parallelization. The code is implemented in C++ with avoiding unnecessary memory allocations. Wrapper functions in Matlab and Python allow processing input/output easier.
Check details of the slice sampler : http://www.jmlr.org/proceedings/papers/v37/gea15.pdf