Welcome to glmpynet’s documentation!

glmpynet is a Python package providing a scikit-learn-compatible LogisticRegression API powered by the high-performance glmnetpp C++ library.

This project bridges the computational speed of glmnetpp with the ease-of-use of the Python data science ecosystem. It provides a user-friendly hybrid API that accepts both standard Scikit-learn parameters (e.g., C, penalty) for seamless integration and glmnet-native parameters (e.g., alpha) for advanced control.

This documentation guides you through installing the package, using it in data science workflows, and understanding its design.

Project Status

The Python API for glmpynet.LogisticRegression is now complete and fully tested against a mock backend. The next major phase of development is to implement the real C++ binding that connects this API to the glmnetpp engine.

See the glmpynet: Development Roadmap for the full development plan.

User Documentation:

• Getting Started
  • Installation
  • Quick Start
• Usage Guide
  • API Philosophy: A Hybrid Approach
  • Basic Usage
  • Advanced Usage: Direct glmnet Parameters
  • Integration with Scikit-learn
• API Reference
  • LogisticRegression Class
• Examples
  • Example 1: Feature Selection with Lasso
• Example Notebooks
  • Basic Binary Classification with LogisticNet
  • Using LogisticRegression in a Pipeline
  • Hyperparameter Tuning with GridSearchCV
  • Handling Sparse Data with LogisticRegression

Development:

• Development
  • glmpynet: Development Roadmap
  • Phase 1: glmnetpp Foundation Analysis
  • Phase 1: glmnetpp Foundation Action Plan
  • Phase 2: Python-C++ Binding
  • Phase 3: Scikit-learn Compatible API
  • Architecture and Design
  • Bazel Setup and Rationale
  • Environment Setup
  • Building and Testing
  • Benchmarking the C++ Core
  • Building glmpynet
  • Contributing to glmpynet

Project Information:

• Project Information
  • Dependencies
  • Citation
  • Contributing & License

Indices and tables

• Index

• Module Index

• Search Page