forcedotcom / distributions

Low-level primitives for collapsed Gibbs sampling in python and C++
BSD 3-Clause "New" or "Revised" License
33 stars 25 forks source link

Add example using pymc #8

Open fritzo opened 10 years ago

twiecki commented 10 years ago

Was gonna open an issue about that as well.

Note that with this Theano change, adding pure-python (wrapped) prob distributions to pymc3 is now much easier: https://github.com/pymc-devs/pymc/issues/507

twiecki commented 10 years ago

Does distributions have most common (log) probability distributions in C++ (e.g. normal)? Couldn't find them by quickly skimming the code base.

fritzo commented 10 years ago

Distributions implements a few probability distributions as score...(...) and sample...(...) in https://github.com/forcedotcom/distributions/blob/master/include/distributions/random.hpp But distributions is focused on collapsed Gibbs sampling of conjugate distributions, and only a few score and sample functions are needed, e.g. sample_normal(...) and score_student_t(...), but neither score_normal(...) nor sample_student_t(...).

On Sat, Aug 2, 2014 at 2:32 AM, Thomas Wiecki notifications@github.com wrote:

Does distributions have most common (log) probability distributions in C++ (e.g. normal)? Couldn't find them by quickly skimming the code base.

— Reply to this email directly or view it on GitHub https://github.com/forcedotcom/distributions/issues/8#issuecomment-50958557 .