Mike Rosulek > Algorithm-Evolve-0.03 > ArrayEvolver

Download:
Algorithm-Evolve-0.03.tar.gz

Dependencies

Annotate this POD

View/Report Bugs
Module Version: 0.03   Source  

NAME ^

ArrayEvolver - A generic base critter class for use with Algorithm::Evolve

SYNOPSIS ^

  package ArrayCritters;
  use ArrayEvolver gene_length => 50,
                   alphabet => [qw(foo bar baz boo)],
                   ...;
  our @ISA = ('ArrayEvolver');
  ## ArrayCritters is now a valid critter class
  
  sub foo_method {
      my $self = shift;
      $self->{foo}++;   ## You can add object attributes
  }
  
  sub fitness {
      my $self = shift;

      ## You can override the default inherited methods to suit the
      ## task at hand
  }

You can use this class as a base class any time your representation is an array gene.

USE ARGUMENTS ^

gene_length

The length of arrays to evolve. Defaults to 20.

alphabet

A reference to an array of valid tokens for the genes. Defaults to [0,1]. Unlike in StringEvolver, the tokens can be any length.

reference_gene

By default, fitness is measured as the number of positions in which a critter's gene agrees with a reference array. However, if you are implementing a non-trivial evolver, you will probably override the fitness method and this argument won't make a difference. It defaults to ('1') x 20.

mutation_rate

If this number is less than one, then it is the probablistic mutation rate for each position in the array. If it is greater than or equal to one, then exactly that many mutations will be performed per child (so it must be an integer). Defaults to 0.05.

crossover_pts

The number of crossover points when performing crossover. See Algorithm::Evolve::Util for more information on crossover.

INHERITED METHODS ^

When used as a base class, the subclass inherits the following methods:

Class->new([ \@init ])

When used with an argument, the new critter is created with the given array reference as the initial value for its gene. Otherwise, this method creates a random array gene over the alphabet.

$obj->mutate()

Mutates the critter's gene according to the given mutation rate.

Class->crossover($obj1, $obj2)

Takes two critters and returns a random crossover of the two, according to the given number of crossover points.

$obj->fitness()

Returns the fitness of the critter, measured as the number of positions which agree with a reference array. You will probably override this method.

$obj->gene([ \@new_gene ])

Returns or sets the value of the critter's array gene. Returns an array reference.

SEE ALSO ^

Algorithm::Evolve, Algorithm::Evolve::Util, the rest of the examples/ directory.

AUTHOR ^

Algorithm::Evolve is written by Mike Rosulek <mike@mikero.com>. Feel free to contact me with comments, questions, patches, or whatever.

COPYRIGHT ^

Copyright (c) 2003 Mike Rosulek. All rights reserved. This module is free software; you can redistribute it and/or modify it under the same terms as Perl itself.

syntax highlighting: