Dynamic Exploration Graph Documentation

The Dynamic Exploration Graph (DEG) algorithm is a fast graph-based approximate nearest neighbor search library. It offers performance comparable to HNSW while providing dynamic graph modification and robust exploration capabilities.

Getting Started

• Installation
  • Requirements
  • Standard Installation
  • Building from Source
  • Verifying Installation
• Quickstart
  • Core Workflow
  • Complete Example
  • Batch Querying
  • Understanding the eps Parameter
  • Next Steps

User Guides

• Building Graphs
  • Graph Types
  • Supported Data Types & Metrics
  • Optimization Target
  • Builder Parameters
  • Quick Build
  • Dynamic Modification
  • Saving & Loading
• Searching
  • Vector Search (k-NN)
  • Search Parameter: eps
  • Filtering
  • Exploration (Graph Walk)
  • Re-Ranking

API Reference

• API Reference
  • Graph
  • Builder
  • Distances & Spaces
  • Search & Re-Ranking
  • Analysis
  • Optimization
  • CPU & Hardware Acceleration