medevalkit 0.1.3.1

A toolkit for evaluating machine learning models for healthcare applications.

MedEvalKit

MedEvalKit is a modular and extensible Python toolkit for evaluating machine learning models, especially in healthcare applications. It provides unified APIs for computing metrics, calibration curves, bootstrap confidence intervals, and plotting diagnostic curves.

Features

• Binary and multiclass classification support
• Threshold optimization
• Calibration error estimation
• Bootstrap confidence intervals
• ROC, PR, and calibration plotting
• Simulate results at different incidence rates

Installation

pip install medevalkit

Usage Example

from medevalkit import Evaluate

clf.fit(X_train, y_train)
y_prob = clf.predict_proba(X_test)

evaluator = Evaluate(y_true=y_test, y_prob=y_prob, classification=True, threshold=threshold)
report = evaluator.generate_report(bootstrap=True)
print(report["text_report"])

GitHub Repo

Visit https://github.com/wesleyyeung/medevalkit for more examples.

License

This project is licensed under the MIT License.