Genetic

Introduction

Genetic is a swift package that makes it incredibly simple to include the Genetic Algorithm within a project.

Usage

Genetic incredibly simple to use. You just create a Genetic object, set it fitness and mutation functions, set its starting population, call apply() and it will continue through to the next generation.

  private var gene: Genetic = Genetic<Int>(mutationFactor: 10, numberOfChildren: 10, startingPopulation: [])

Defined:

  public init(mutationFactor: Int = 100,
              numberOfChildren: Int = 10,
              startingPopulation: [[T]] = [])

• Genetic uses generic typing to build its population. T must conform to Equatable
• mutationFactor - A number greater than 0 that determines the chances of a single mutation. The probability is 1 / {this_number}
• numberOfChildren - Number of children per generation.
• startingPopulation - An array of arrays that defines the first population to pass through the algorithm.

Properties

Setting Starting population

You can set the starting population of the Genetic object.

  gene.startingPopulation = newPop

Fitness Function

Genetic has a property for defining the fitness function of the algorithm. This function should return a Double that ranks the child that is passed to this function.

  gene.fitnessFunction = { (child: [Int], index: Int) -> Double in
    //return some rank as Doubleh
  }

Mutation Function

Genetic has a property for defining the mutation function of the algorithm. This function should return T. This will be used to replace a portion of the child array when selected to mutate.

  gene.mutationFunction = { () -> Int in
    return Int.random(in: 0...10)
  }

Metrics

You can receive certain metrics from the Genetic object such as current generation and current highest rank.

  gene.generations
  gene.highestRanking

Applying the algorithm

You can nudge forward the algorithm by one generation by calling apply(). This will return the next generation. Genetic will store the most recent generation and calling this again will apply the algorithm to that generation, forwarding the alorithm along.

  public func apply() -> [[T]]

Clearing the algorithm

You can clear the whole algorithm and reset the class by calling clear()

public func clear()

• This will set the generations back to 0 and reset the working population (you will not have to reset the starting population)