A Python package for the statistical analysis of A/B tests
-
Updated
Jun 12, 2026 - Python
A Python package for the statistical analysis of A/B tests
Causalis - State-of-the-art robust causal inference for experiments and observational data in python
Воспроизводимые эксперименты по снижению дисперсии оценки эффекта: plain diff, CUPED, VWE и их комбинация.
896 tests / 42 scripts validating Zetyra calculators (GSD, CUPED, Bayesian, SSR incl. Single-Arm, RAR, Master Protocol). Replicates 8 published trials (Salk, DAPA-HF, PACIFIC, MONALEESA-7, I-SPY 2, STAMPEDE, REMAP-CAP, NCT03377023) + Leyrat 2024.
A Python package for the full A/B test lifecycle, from power analysis to sequential monitoring and variance reduction
End-to-end A/B testing framework — power analysis, SRM checks, CUPED variance reduction, BH-corrected segment breakouts, and an auto-generated ship/no-ship decision doc. Built on a 60k-user synthetic checkout experiment.
A/B test lift simulation & analysis in Python: CUPED variance reduction, power analysis, and guardrail metrics with reproducible pytest CI.
Local-first experimentation and causal decision platform for A/B testing, CUPED, trust checks, segment-aware rollout policy, and simulation-driven validation.
Modern A/B experimentation utilities: CUPED/CUPAC hooks, triggered analysis, SRM, switchback helpers, and power sims.
Production A/B testing framework with Budget A/B, CUPED variance reduction, and sequential testing - DoorDash methodology
Production-simulated experimentation, metrics intelligence, CUPED/guardrail decisioning, and streaming observability platform.
Trustworthy A/B test readout as a Claude Code skill — SRM, CUPED, BH-corrected, ship/no-ship verdict
Causal failure attribution + Cladder-style benchmark + A/B testing harness with power analysis, CUPED, propensity matching, and DiD
Bayesian A/B testing pipeline on GCP - MovieLens 25M with closed-form Bayesian + multivariate CUPED + Cloud Run
End-to-end product experimentation case study. A +5% A/B test win hides returning-user harm, refund rate degradation, and a treatment effect that decays after 48 hours. Includes CUPED, BH correction, Bayesian analysis, and guardrail metrics.
Experimentation decision lab with A/B simulation, CUPED, guardrails, segment readouts, power analysis, and sequential evidence snapshots.
Small, well-tested Python package for online controlled experiments: sample size, CUPED, mSPRT sequential tests, delta-method ratio metrics.
Add a description, image, and links to the cuped topic page so that developers can more easily learn about it.
To associate your repository with the cuped topic, visit your repo's landing page and select "manage topics."