Standardize Explainer on Bambi in PyMC Examples

I’ve been reviewing some of the GLM documents for pymc v4 and saw that .glm() has been deprecated and bambi seems to be the preferred route.

Some of the examples that were originally done without bambi read a little confusing to me (or an outsider), that an example from pymc is about bambi.

I was wondering if there is some standard info that can be added to the intro/preamble of each example that explains what bambi is, that it is based on pymc, and that it is the preferred method.

I think bambi should even be in some of the titles of the examples. Ex:

(Generalized) Linear and Hierarchical Linear Models in PyMC3 and Bambi

Even though bambi is built on top of pymc, I imagine people coming to the examples may be confused that they are technically learning to model with bambi and not pymc directly.

This example does a pretty good job about explaining bambi but its use could be explained in an intro that comes before the imports:

(Generalized) Linear and Hierarchical Linear Models in PyMC3

Here is an example that could be fixed:

GLM: Logistic Regression

For the record, I like this example that shows the manual way, and the simpler way with bambi:

GLM: Poisson Regression

3 Likes