Survival Analysis
Welcome to the Survival Analysis module of PyBH.
While our core expertise and the heart of this library lie in Bayesian Inference, we recognize the importance of established frequentist methods for benchmarking, validation, and speed.
Our SurvivalAnalysis module provides a bridge between these two worlds, offering a unified interface to both advanced Bayesian models and standard frequentist approaches.
Strategic Approach
Expert Bayesian Modeling: We leverage PyMC to provide sophisticated survival models that handle complex priors, hierarchical structures, and provide a complete picture of uncertainty via posterior distributions.
Frequentist Compatibility: To ensure your workflow is complete, we integrate the Lifelines library. This allows you to run classical statistical tests and models (like standard Kaplan-Meier or Cox PH) directly through our API.
Core Capabilities
Seamless Comparison: Easily compare Bayesian posterior estimates against Frequentist point estimates and confidence intervals.
Non-parametric & Regression: Support for Kaplan-Meier, Weibull, and Cox Proportional Hazards across both engines.
Unified API: Maintain a consistent coding style regardless of the mathematical engine running under the hood.