graphevo 0.1.0

Genetic Algorithm for Graph Optimization

GraphEvo: Genetic Algorithm for Graph Optimization

GraphEvo is a powerful C++ library with Python bindings that implements genetic algorithms for graph optimization problems. It provides efficient tools for evolving and optimizing graph structures using genetic algorithms.

Features

• Genetic Algorithm Implementation: Core genetic algorithm operations including selection, crossover, and mutation
• Graph Generation: Tools for generating and evolving graph structures
• Fitness Evaluation: Customizable fitness functions for graph optimization
• High Performance: C++ implementation with Python bindings for optimal performance

Installation

From PyPI

pip install graphevo

From Source

# Clone the repository
git clone https://github.com/Js-Hwang1/GraphEvo.git
cd GraphEvo/python

# Install in development mode
pip install -e .

Requirements

• Python 3.7 or higher
• C++17 compatible compiler
• CMake 3.10 or higher
• Eigen3
• pybind11

Quick Start

import graphevo as ge
import networkx as nx

# Create a graph generator
generator = ge.GraphGenerator()

# Generate an initial population
population = generator.generate_population(
    population_size=100,
    num_nodes=50,
    edge_probability=0.3
)

# Create a genetic algorithm instance
ga = ge.GeneticAlgorithm(
    population=population,
    mutation_rate=0.1,
    crossover_rate=0.8,
    elite_size=5
)

# Run the genetic algorithm
best_graph = ga.evolve(
    generations=100,
    fitness_function=lambda g: nx.density(g)  # Example fitness function
)

# Access the best graph
print(f"Best graph density: {nx.density(best_graph)}")

Core Components

GraphGenerator

• Generates random graphs
• Creates initial populations
• Supports various graph generation strategies

GeneticAlgorithm

• Implements the core genetic algorithm
• Handles selection, crossover, and mutation
• Supports customizable fitness functions

GeneticOperators

• Provides mutation and crossover operations
• Customizable for different graph types
• Optimized for performance

FitnessEvaluator

• Evaluates graph fitness
• Supports custom fitness functions
• Handles multi-objective optimization

Advanced Usage

Custom Fitness Functions

def custom_fitness(graph):
    # Calculate graph properties
    density = nx.density(graph)
    clustering = nx.average_clustering(graph)
    
    # Combine into a single fitness score
    return density * clustering

# Use in genetic algorithm
ga = ge.GeneticAlgorithm(
    population=population,
    fitness_function=custom_fitness
)

Custom Genetic Operators

def custom_mutation(graph):
    # Implement custom mutation logic
    return mutated_graph

def custom_crossover(parent1, parent2):
    # Implement custom crossover logic
    return child_graph

# Use custom operators
ga.set_mutation_operator(custom_mutation)
ga.set_crossover_operator(custom_crossover)

Citation

If you use GraphEvo in your research, please cite:

@software{GraphEvo,
  author = {Junsung Hwang},
  title = {GraphEvo: Genetic Algorithm for Graph Optimization},
  year = {2025},
  publisher = {GitHub},
  url = {https://github.com/Js-Hwang1/GraphEvo}
}

Contact

• Author: Junsung Hwang
• Email: [e-mail]
• GitHub: Js-Hwang1