跳转至

🗺️ PyStatPower Roadmap

Welcome to the PyStatPower project roadmap! This document outlines the project's core vision, currently supported functionality, and planned future features.

Info

This roadmap is a living document and will evolve as new features are implemented and community feedback is integrated.

🎯 Project Vision

PyStatPower aims to provide a streamlined, accurate, and easy-to-use Python toolkit for statistical power analysis, sample size estimation, power calculation, and effect size solving.

📚 Current Feature Status

Here is the current implementation status of statistical models in PyStatPower:

📊 Mean Models

  • Single Group
  • Confidence Interval
  • Inequality Test
  • Non-Inferiority Test
  • Superiority Test
  • Equivalence Test

  • Two Independent Groups

  • Confidence Interval
  • Inequality Test
  • Non-Inferiority Test
  • Superiority Test
  • Equivalence Test

  • Two Correlated Groups

  • Confidence Interval
  • Inequality Test
  • Non-Inferiority Test
  • Superiority Test
  • Equivalence Test

🍰 Proportion Models

  • Single Group
  • Confidence Interval
  • Inequality Test
  • Non-Inferiority Test
  • Superiority Test
  • Equivalence Test

  • Two Independent Groups

  • Confidence Interval
  • Inequality Test
  • Non-Inferiority Test
  • Superiority Test
  • Equivalence Test

  • Two Correlated Groups

  • Confidence Interval
  • Inequality Test
  • Non-Inferiority Test
  • Superiority Test
  • Equivalence Test

📈 Correlation Models

  • Confidence Interval
  • Inequality Test

🧩 Miscellaneous Models

  • Observe At Least One Event
  • Multi-Reader Multi-Case
  • Bayesian Single-Arm Phase II Trial Designs with Time-to-Event Endpoints

🚀 Future Development Plans

🎯 Short-term Priorities

  • Complete Equivalence Tests (TOST) for remaining Mean and Proportion models.
  • Add support for Paired Samples in Mean and Proportion models.
  • Improve exception handling and helpful error messages for edge parameter cases.

🔮 Mid-term Expansion

  • Survival Analysis: Log-rank tests and Hazard Ratio power calculations.
  • ANOVA Models: One-way and two-way ANOVA sample size estimations.

🌌 Long-term Vision

  • Support advanced designs (e.g., Crossover trials, Repeated measures).
  • Optional interactive Web UI for quick estimates without coding.

🤝 Contributing

Contributions are always welcome! If you would like to help build any of the features listed above, feel free to open a Pull Request or start a discussion in our Issues tracker.