jupyter-bayes

Jupyter environment made for Bayesian inference and graphical modeling.

What it Gives You

It is based on Jupyter Notebook Data Science Stack and cdeck3r’s R docker image, provides the ability to perform analyses using R, Python and Julia. Pre-loaded with dozens of common data science packages. ✨

Additionally, it provides the following “standard” packages used for Bayesian inference:

• R
  • jags
  • rjags
  • runjags
  • HydeNet
  • gRain
  • gRim
  • bnlearn
  • rStan
  • parallel
  • compute.es
• Python
  • PyMC3
  • PyStan
  • BamBI
• Julia
  • pull requests welcome!

Note: Plots of bayesian nets require you to use Google Chrome. It will not work within Firefox.

Basic Use

Spin up the container using the command

docker run -it --rm -p 8888:8888 leblancfg/jupyter-bayes:latest

For other startup options check out Jupyter Notebook Data Science Stack.

Contributions Welcome!

Bayesian inference is still a rapidly-moving field, and the popularity of its package ecosystem is evolving with time — what is cutting-edge now will almost certainly not be within a few years. Contributions are very welcome!

TODO

• Package Kruschke’s essentials for Doing Bayesian Data Analysis 2nd edition into the notebook workspace
• Set up semantic versinoing for docker image tags