DukeDarwin

A library for running genetic algorithms

Use

Call the functions run_algorithm or simplified_run in the class GeneticLibrary. This runs the algorithm

Examples

Problem name	File name	implemented	wikipedia link
Maximimize volume of a cylinder	MaximizeCylinderVolumeTest.java	✅️
Eight queens problem	EightQueensTest.java	✅️	https://en.wikipedia.org/wiki/Eight_queens_puzzle
Combine resistors	TuneResistanceTest.java	✅️

API

The api is made to look similar to my other libraries for genetic algorithms, for Rust and Python. The implementation of the libraries may however differ.

There are two function that you can call in the library, in order to run the algorithm, the simplified_run is supposed to be a bit easier to use.

public GenomeType simplified_run(Supplier<GenomeType> newGenomeSupplier, 
        UnaryOperator<GenomeType> mutator,
        ToDoubleFunction<GenomeType> scoreFunction, 
        Predicate<GenomeType> finishPredicate)

/*
 @param Supplier<GenomeType> newGenomeSupplier. makes new genomes.
 @param List<Function<GenomeType[], GenomeType>> mutators. list of functions mutating the genomes. 
 @param Consumer<GenomeType> printGenome. prints a genome/solution.
 @param ToDoubleFunction<GenomeType> scoreFunction. scores a genome, the scores are use in the selection of genomes, it tells how fit a genome is.
 @param Predicate<GenomeType> finishPredicate. Tells whether to halt the algorithm. Use a condition to determine if you want to halt.

 @return GenomeType the result of the calculation
*/
public GenomeType run_algorithm(Supplier<GenomeType> newGenomeSupplier, 
        List<Function<GenomeType[], GenomeType>> mutators,
        Consumer<GenomeType> printGenome,
        ToDoubleFunction<GenomeType> scoreFunction,
        Predicate<GenomeType> finishPredicate)

Due to a quirk in how java works, you also have to supply the class of the genome at instantiation of the genetic library

//instantiate the geneticlibrary with the class of the genome, double[].class
GeneticLibrary<double[]> geneticLibrary = new GeneticLibrary<double[]>(double[].class);
//run the algorithm and return the found solution for further manipulation 
double[] result = geneticLibrary.simplified_run(this::newGenome, this::mutate, this::scoreFunction, this::finished);