GLM Granger

Python implementation and generalization of Granger causality for spike train data using Generalized Linear Models (GLMs), adapted from:

Kim et al. (2011).
A Granger Causality Measure for Point Process Models of Ensemble Neural Spiking Activity
PLOS Comput Biol, 7(3):e1001110

Features

• Supports any statsmodels GLM family (e.g. Poisson, Negative Binomial, Binomial)
• Optimal lag selection via K-fold cross-validation
• Optional filtering of indirect connections via causal pathway analysis
• Permutation-based significance testing with FDR correction
• Parallelized computation with joblib

Installation

This package depends on:

• Python 3.8+
• scikit-learn
• joblib
• numpy
• statsmodels >= 0.14.2

You can install the required packages using:

pip install scikit-learn joblib numpy statsmodels