Markov Chain Monte-Carlo

This plugin provides a notebook template to perform a simple Bayesian inference analysis on one numerical column of a dataset (observations).
The main goal is to perform inference on the variables that parametrize the distribution from which the observations were theoretically generated.

Plugin information

AuthorDataiku (Pierre P & Du P)
Last updated2018-08-01
LicenseApache Software License
Source codeGithub
Reporting issuesGithub

How To Use

The notebook template is available by selecting a dataset, go to "Lab", and in the notebooks section, choosing the appropriate pre-defined template.

The notebook content is mostly self-explanatory, but it should not be executed in one single run of all the cells, it was designed so that the user is guided to build his own probabilistic model that will vary depending on the use-case and the data itself.