HΛZΛRDOUS

Predictive survival and competing risks analysis in Python

The objective of this library is to provide a Python implementation of time-to-event prediction models in the presence of right-censored data.

The estimators of this library build on top of scikit-learn components and extend the scikit-learn API to offer dedicated prediction methods for survival and competing risks analysis.

They should be interoperable with scikit-learn tools such as pipelines, column transformers, cross-validation, hyper-parameter seach tools, etc.

This package will also offer neural network based estimators by leveraging PyTorch and skorch.

This library puts a focus on predictive accuracy rather than on inference. Quantifying the statistical association or causal effect of covariates with/on the cumulated event incidence or instantaneous hazard rate is not in the scope of this library at this time.

• License: MIT

• GitHub repository: soda-inria/hazardous

• Changelog: soda-inria/hazardous

• Status: under development, API is subject to change without notice.

Contents:

• API
  • Estimators
  • Metrics
  • Datasets
  • Inverse Probability Censoring Weight
• Example gallery
• How to get the SEER dataset