3 ms·
Stan (http://mc-stan.org/documentation/ http://mc-stan.org/documentation/) is arguably the most advanced language. It's especially pushing the bounds of doing a
by tansey 10y ago
Stan (http://mc-stan.org/documentation/ http://mc-stan.org/documentation/) is arguably the most advanced language. It's especially pushing the bounds of doing automatic variational inference, for the scenario where your model does not have a nice conjugate form that would be amenable to Gibbs sampling. It's not quite reached what I would say is production-quality, but some of the best people in the world of computational Bayesian methods (e.g., Michael Betancourt, most of David Blei's lab, etc.) are working on it.
- tbenst 10y agoYes Stan is awesome! The main difference between something like Stan and "next gen" languages like Anglican and Webppl, is there are basically no restrictions in where you use a distribution. Nested inference, probabilistic recursion, etc are all possible. For certain classes of problems this leads to greatly enhanced expressiveness. On the flip side, Stan is more production ready right now
- jmde 10y agoA major downside of Stan is its lack of support for discrete priors. This isn't really advertised very well, but is more of a problem than it might sound initially. Its type handling also can get a little frustrating at times. Overall, I highly recommend it but it does have its downsides, and there's some room for alternatives or improvement.