Back to browse
GitHub Repository

Bayesian CausalImpact for Python — R-compatible Gibbs sampler in Rust (PyO3), 30x faster, no TensorFlow

19 starsPython

New Causal Impact Library

by djwjjtw·Mar 27, 2026·9 points·2 comments

AI Analysis

●●SolidNiche GemBig Brain

Rust-powered CausalImpact port that's 10-30x faster than the R original.

Strengths
  • Rust Gibbs sampler via PyO3 delivers genuine 10-30x speedup over R
  • Numerical equivalence tests ensure results match Google's CausalImpact exactly
  • No TensorFlow dependency — lighter weight than modern ML alternatives
Weaknesses
  • Niche audience — causal inference is already a specialized field
  • Python ecosystem already has CausalPy and DoWhy for similar problems
Category
Target Audience

Data scientists and analysts doing causal inference

Similar To

CausalImpact (R) · CausalPy · DoWhy

Similar Projects