A Ruby gem for genetic algorithms.
Ruby
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
examples
lib
spec
.gitignore
Gemfile
README.md
Rakefile
darwinning.gemspec

README.md

Darwinning

Gem Version

A Ruby gem to aid in the use of genetic algorithms.

Installation

gem install darwinning

Usage

There are two ways of using Darwinning. You can either start with a class that is a subclass of Darwinning::Organism or you can include Darwinning in an existing class that you would like to make evolveable.

Include Style:

This style is useful if you want to take existing objects and use Darwinning on them when needed.

Here's an dumb example of how you might use Darwinning to solve a pointless problem:

Let's say for some reason you need a set of 3 numbers (within a range of 0-100) that add up to 100. This is a strange problem to have, but let's solve it anyway.

require 'darwinning'

class Triple
  include Darwinning

  GENE_RANGES = {
    first_number: (0..100),
    second_number: (0..100),
    third_number: (0..100)
  }

  attr_accessor :first_number, :second_number, :third_number

  def fitness
    # Try to get the sum of the 3 digits to add up to 100
    (first_number + second_number + third_number - 100).abs
  end
end

In order to use Darwinning this way you must:

  • include Darwinning in your class
  • define a GENE_RANGES constant with your relivant value names as keys
  • define attr_accessors for all your values
  • define a fitness method

Once you have your organism class that includes Darwinning, you can create a population and evolve it:

if Triple.is_evolveable?
  triple_pop = Triple.build_population(0, 10, 100)
  triple_pop.evolve! # evolve until fitness goal is or generations limit is met

  pp "Best member: #{triple_pop.best_member}"
end

Inheritance Style:

This style is good when you just want to set up some quick objects to only use in Darwinning tasks.

Let's solve the same dumb problem we looked at before...

require 'darwinning'

class Triple < Darwinning::Organism
  @name = "Triple"
  @genes = [
    Darwinning::Gene.new(name: "first digit", value_range: (0..100)),
    Darwinning::Gene.new(name: "second digit", value_range: (0..100)),
    Darwinning::Gene.new(name: "third digit", value_range: (0..100))
  ]

  def fitness
  # Try to get the sum of the 3 digits to add up to 100
    (genotypes.values.inject { |sum, x| sum + x } - 100).abs
  end
end

With this Darwinning::Organism class, you can now build a population and evolve it:

triple_pop = Darwinning::Population.new(
  organism: Triple, population_size: 10,
  fitness_goal: 0, generations_limit: 100
)
triple_pop.evolve!

pp "Best member: #{triple_pop.best_member}"

Stepping Through Generations Manually

# pop is a Darwinning::Population of organisms
puts pop.generation
# 0
pop.make_next_generation!
puts pop.generation
# 1

# you can view a history of population's evolion
puts pop.history
# [[gen0_member1, gen0_member2,...], [gen1_member1, gen1_member2,...]]

More Examples

Check out the /examples folder for more examples. That seems like a good place to put examples, right?

Built by:

With help from:

lots of great contributers and work based on forks from: Nanosim-LIG