Development

This section provides guidance for developers contributing to glmpynet. It outlines the strategic roadmap, the tactical phase-by-phase development plans, and the reference guides for building and testing the code.

Our development philosophy is phased and incremental. We begin with a high-level strategy, create a detailed action plan for the current phase, and then execute that plan with focused, bite-sized tasks.

Strategy & Action Plans:

• glmpynet: Development Roadmap
  • Project Vision
  • Development Strategy
  • Phased Roadmap
  • Known Risks & Open Questions
  • Future Enhancements
• Phase 1: glmnetpp Foundation Analysis
  • Objective
  • Background: Initial Build Failures
  • Methodology: Static Analysis
  • Requirements Analysis
• Phase 1: glmnetpp Foundation Action Plan
  • Objective
  • Action Plan
• Phase 2: Python-C++ Binding
  • Technical Strategy
  • The C++ Wrapper Function: API Contract
  • Key Challenges
• Phase 3: Scikit-learn Compatible API
  • Model Name: LogisticRegression
  • API Design Philosophy: A Hybrid Approach
  • API Contract
  • Future Expansion

Developer Guides:

• Architecture and Design
  • 1. Introduction
  • 2. Core Philosophy: The Wrapper Pattern
  • 3. Component Layers
  • 4. The LogisticRegression Class: Bridging the Gap
  • 5. Data Flow
  • 6. Key Design Decision: Adopting the Scikit-learn API
• Bazel Setup and Rationale
  • Design Decision: Why Bazel?
  • Installation using the Official Apt Repository
• Environment Setup
  • Prerequisites
  • Step 1: Create the Conda Environment
  • Step 2: Activate the Environment
  • Step 3: Verify the Installation
• Building and Testing
  • Building the Library
  • Running Unit Tests
• Benchmarking the C++ Core
  • Listing Available Benchmarks
  • Running a Benchmark
  • Interpreting the Results
• Building glmpynet
  • Steps to Build glmpynet
• Contributing to glmpynet
  • How Can I Contribute?
  • Submitting Code Changes (Pull Requests)