How do I run a mixed effects regression using Python?

I normally use lme4 package in R to run mixed effects regression models. What are some recommended packages in Python to do this?

There are several Python packages that can run mixed effects models.

  1. The statsmodels package has a linear mixed effects module. However, the implementation is not as fast or as fully featured as lme4.

  2. The Bayesian Model-Building Interface (BAMBI) package was developed by Jake Westfall and Tal Yarkoni and provides a high-level bayesian model building interface to pymc3. The syntax is designed to be similar to running models in lme4.

  3. Pymer4 was developed by Eshin Jolly and provides a python wrapper to lme4 in R. It is designed to provide the same functionality and interface used in lme4 in R, but using pandas dataframes.

A comparison between a few of these (and some other) implementations: https://github.com/junpenglao/GLMM-in-Python/blob/master/GLMM_in_python.ipynb