Bambi is an open-source Python package built on top of PyMC3 that makes it extremely easy to specify and fit Generalized Linear Multilevel Models (GLMMs) using a Bayesian approach with an intuitive and simple formula syntax very similar to the one popularized by the R package lme4. With Bambi, practitioners are benefited from all the power of PyMC3 without having to write any PyMC3 code directly.
This talk introduces Bambi, an open-source Python package built on top of PyMC3 that makes it extremely easy to specify and fit Generalized Linear Multilevel Models (GLMMs) using a Bayesian approach.
Albeit PyMC3 being a solid and flexible probabilistic programming language, writing models directly in PyMC3 may be time-consuming, or even hard in some cases. This may prevent some practitioners and researchers that primarily work with Python to benefit from the Bayesian approach to statistical modeling.
Bambi is the first Python library that allows specifying GLMMs with an intuitive and simple formula syntax very similar to the one popularized by the R package lme4. With Bambi, practitioners are benefited from all the power of PyMC3 without having to write any PyMC3 code directly.
This is an informative/hands-on talk aimed at researchers and practitioners in general who are interested in using a new simpler tool to specify and fit Bayesian models in Python. This talk requires some familiarity with regression modeling and basic concepts of Bayesian statistics.