melmi / EvolveSharp

.NET Genetic Algorithm Framework

A Genetic Algorithm Framework for .NET

Visit our website

Getting Started

1. Setting Up

Create a Console Application and install EvoleSharp using Nuget:

PM> Install-Package EvolveSharp

2. Create a fitness function

In this simple example we are creating a fitness function to sum all genes of an individual

public class ExampleFitnessFunction : IFitnessFunction<double>
{
    public double Evaluate(IIndividual<double> individual)
    {
        var sum = 0.0;
        for (var i = 0; i < individual.Length; i++)
        {
            sum += individual[i];
        }
        return sum;
    }
}

3. Instantiate the GeneticAlgorithm class and call the Evolve method

You can set 3 parameters:

1. Population count: How many individuals each population will have
2. Gene count: How many genes each individual will have
3. Generation count: How many generations will occour

class Program
{
    static void Main(string[] args)
    {
        const int populationCount = 100;
        const int generationCount = 10000;
        const int geneCount = 10;
       
        var ga = new GeneticAlgorithm(populationCount, geneCount, new ExampleFitnessFunction());
        ga.Evolve(generationCount);
    }
}

Other information

1. You can create your own mutation, selection, crossover and initializer class. You just need to inherit the respective interface.
2. You can set the Mutator, CrossoverMethod, Selector and Initializer after instantiate the GeneticAlgorithm class.
3. You can create and set your own reporter in GeneticAlgorithm class.

Sample

Clone EvolveSharp and try the Travelling salesman problem sample

Contributing

Please use the issue tracker and pull requests.

License

Copyright (c) 2014 Afonso França Licensed under the MIT license.