Weibull

Hypothesis

The Weibull model is a parametric model that makes strong assumptions about the distribution of survival times. Unlike semi-parametric models, it assumes that the hazard rate follows a specific shape (monotonic increase or decrease). Validating these assumptions is crucial to ensure your predictions are not biased.

You will find here an extensive list of these hypotheses along with existing solutions to test them.

Example Notebook

You’ll find here an example notebook displaying how you could use the Weibull model using the SurvivalAnalysis class.

Model

class PyBH.SurvivalAnalysis.pymc_models.Weibull[source]

Bayesian Weibull Survival Model implementation. Parameters: alpha (shape k), beta (scale eta).

predict_survival_function(times, X_new, credible_interval=0.95)[source]: Predict S(t) = exp(-(t/beta)^alpha). Returns a DataFrame with mean survival and uncertainty bounds.