org.generation5.demos
Class EvoDiophantine

java.lang.Object
  org.generation5.bio.EvolveStandard
      org.generation5.demos.EvoDiophantine

All Implemented Interfaces:

java.lang.Comparable, Evolvable

public final class EvoDiophantine

extends EvolveStandard

EvoDiophantine is an example of a genetic algorithm solving a diophantine equation. A diophantine equation is an integer-only equation of the form ax+by+cz...=t. The class solves for the variables given the coefficients a, b, c etc. and target t.

Field Summary

Fields inherited from class org.generation5.bio.EvolveStandard

fitnessValue, minimizeFitness, random

Constructor Summary

EvoDiophantine()
Creates a new instance of EvoDiophantine

Method Summary

void	calculateFitness() Diophantine fitness is calculated as the error between the target value and the calculated value.
double	getFitness() Return the fitness of this object.
static void	main(java.lang.String[] args) Run the genetic algorithm on the specified coefficients and target values.
Evolvable	mate(Evolvable partner) One-point crossover of the equation values.
void	mutate() Mutates one of the equation values by a random value between -2 and 2.
void	randomInitialize() Initialize the equation values to a number within the specified range.
static void	setCoefficients(int[] coeff, int target) Set the coefficients and target of the equation.
void	setRange(int min, int max) Set the range the genetic algorithm checks between.
java.lang.String	toString()

Methods inherited from class org.generation5.bio.EvolveStandard

compareTo

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

EvoDiophantine

public EvoDiophantine()

Creates a new instance of EvoDiophantine

Method Detail

calculateFitness

public void calculateFitness()

Diophantine fitness is calculated as the error between the target value and the calculated value.

Specified by:

calculateFitness in interface Evolvable

Specified by:

calculateFitness in class EvolveStandard

getFitness

public double getFitness()

Return the fitness of this object.

Specified by:

getFitness in interface Evolvable

Overrides:

getFitness in class EvolveStandard

Returns:

the fitness value of this instance.

mate

public Evolvable mate(Evolvable partner)

One-point crossover of the equation values.

Specified by:

mate in interface Evolvable

Specified by:

mate in class EvolveStandard

Parameters:

partner - the other parent equation.

Returns:

the new child equation.

mutate

public void mutate()

Mutates one of the equation values by a random value between -2 and 2.

Specified by:

mutate in interface Evolvable

Specified by:

mutate in class EvolveStandard

randomInitialize

public void randomInitialize()

Initialize the equation values to a number within the specified range.

Specified by:

randomInitialize in interface Evolvable

Specified by:

randomInitialize in class EvolveStandard

See Also:

setRange(int, int)

setRange

public void setRange(int min,
                     int max)

Set the range the genetic algorithm checks between.

Parameters:

min - minimum value.

max - maximum value.

toString

public java.lang.String toString()

Returns:

a string with the equation and current fitness.

setCoefficients

public static void setCoefficients(int[] coeff,
                                   int target)

Set the coefficients and target of the equation. As a static member function, this specifies the values for all members of the population.

Parameters:

coeff - the coefficients.

target - target value.

main

public static void main(java.lang.String[] args)

Run the genetic algorithm on the specified coefficients and target values. If nothing is specified, a+2b+3c+4d=50 is attempted. The GA runs for a maximum of 1000 iterations or until fitness reaches zero.

Parameters:

args - coefficients and target value to run the GA on.