API Reference

The optimal-classification-cutoffs library provides a comprehensive API for threshold optimization across binary and multiclass classification problems.

Overview

The main entry points are:

• get_optimal_threshold() - Core optimization function for binary and multiclass problems

• ThresholdOptimizer - Scikit-learn compatible wrapper class

• cv_threshold_optimization() - Cross-validation utilities

Core Functions

The primary interface for threshold optimization:

• Core Functions

Metrics System

Built-in metrics and utilities for custom metric registration:

• Metrics System

Scikit-learn Interface

High-level wrapper for integration with scikit-learn workflows:

• Scikit-learn Interface

Cross-Validation

Utilities for robust threshold estimation:

• Cross-Validation Utilities

Multiclass Support

Specialized functionality for multiclass classification:

• Multiclass Support

Type Definitions

Type hints and protocols used throughout the library: