Introduction

Overview

AnalysisG is a high-performance Graph Neural Network (GNN) framework designed specifically for High Energy Physics (HEP) analysis. It combines the power of C++ for computational efficiency with the flexibility of Python for ease of use.

What is AnalysisG?

AnalysisG provides a complete ecosystem for:

Data Processing: Read and process physics events from ROOT files
Graph Construction: Transform physics events into graph representations
Neural Networks: Train and deploy graph neural networks
Analysis: Perform physics analysis with built-in tools
Visualization: Create plots and visualizations of results

Framework Architecture

The framework is organized into several key components:

Core Components

Core Module: Base templates and fundamental functionality
Events: Event processing for various physics analyses
Graphs: Graph construction from physics events
Models: Neural network model implementations
Metrics: Performance evaluation and metrics

Infrastructure

Modules: Low-level infrastructure components
PyC: Python-C++ interface layer via Cython
Templates: Code templates for custom implementations

Design Philosophy

AnalysisG follows these design principles:

Performance: C++ implementation for computational efficiency
Usability: Python interface for ease of use
Extensibility: Template-based architecture for customization
Modularity: Independent, reusable components
Flexibility: Support for various analysis workflows

Use Cases

AnalysisG is suitable for:

Top quark pair production analysis
Beyond Standard Model (BSM) searches
Multi-lepton final states
Jet reconstruction and classification
Graph-based event classification
Neural network model development for HEP

Technology Stack

The framework is built on:

C++17: Core implementation language
Cython: Python-C++ interface
PyTorch/LibTorch: Neural network framework
ROOT: High-energy physics data format
CMake: Build system
Doxygen: C++ documentation
Sphinx: Python and overall documentation

Getting Help

For questions and support:

Check the API documentation in API Reference
Review the module documentation in Module Documentation
Look at examples in Examples
Visit the GitHub repository for issues and discussions

Next Steps

Continue to:

API Reference - Detailed API documentation
Module Documentation - Module-by-module documentation
Examples - Usage examples and tutorials