groupedpaneldatamodels documentation
groupedpaneldatamodels is the first package to bring Grouped Fixed Effects (GFE) and Grouped Interactive Fixed Effects (GIFE) estimators
(Bonhomme & Manresa 2015; Su, Shi & Phillips 2016; Ando & Bai 2016; Su & Ju, 2018) to Python.
Compared with classic fixed effects, grouping boosts efficiency, uncovers hidden groupings, and scales to large N panels.
This package implements some of the most popular GFE and GIFE estimators, analytical or bootstrap standard errors, and automatic selection for hyperparameters such as the number of groupings using Information Criterion.
import groupedpaneldatamodels as gpdm
model = gpdm.GroupedFixedEffects(dependent=Y, exog=X, G=3)
model.fit().summary() # prints coefficients, group sizes and IC
New here? Start with Usage Guide for a quick start guide, then you can dive into the API docs.