Turing.Jl

Turing.jl is a probabilistic programming language and Bayesian modeling framework for the Julia programming language. Users write models in syntax that is close to the mathematical specification, declaring priors and a likelihood, and then draw posterior samples with inference algorithms such as the No U Turn Sampler. It is written entirely in Julia and works with the wider Julia ecosystem. It is aimed at researchers, statisticians and data scientists who do Bayesian inference, whether they are new to the method or experienced. The framework supports models with discrete parameters and stochastic control flow, so it is not limited to simple continuous models. The Turing ecosystem is built from composable packages, including DynamicPPL.jl, a domain specific language and backend used by Turing, JuliaBUGS.jl, a modern implementation of the BUGS language, TuringGLM.jl for Bayesian generalized linear models written with formula syntax, and AdvancedHMC.jl, a modular implementation of Hamiltonian Monte Carlo algorithms. The site offers step by step tutorials, a beginner's walkthrough, a list of research papers that use Turing.jl, guidance on citing it, a newsletter, and Google Summer of Code project ideas. Community support is through GitHub and a Slack channel. The project is developed by the TuringLang community on GitHub. The site publishes no pricing and lists no paid plans.