org.evosuite.ga

Class FitnessFunction<T extends Chromosome>

java.lang.Object

org.evosuite.ga.FitnessFunction<T>

All Implemented Interfaces:

Serializable

Direct Known Subclasses:

RegressionTestFitnessFunction, TestFitnessFunction, TestSuiteFitnessFunction

public abstract class FitnessFunction<T extends Chromosome>
extends Object
implements Serializable

Abstract base class of fitness functions

Author:

Gordon Fraser

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected static org.slf4j.Logger	logger Constant logger

Constructor Summary

Constructors

Constructor and Description
FitnessFunction()

Method Summary

Modifier and Type	Method and Description
T	getBestStoredIndividual() If the fitness function as an archive, returns the best individual in the archive.
abstract double	getFitness(T individual) Calculate and set fitness function #TODO the 'set fitness' part should be done by some abstract super class of all FitnessFunctions
abstract boolean	isMaximizationFunction() Do we need to maximize, or minimize this function?
static double	normalize(double value) Normalize a value using Andrea's normalization function
boolean	updateCoveredGoals() if the fitness function contains an archive, updates the archive and the fitness values of the population, and returns true.
protected void	updateIndividual(FitnessFunction<?> ff, T individual, double fitness) Make sure that the individual gets to know about its fitness
protected void	updateIndividual(FitnessFunction<?> ff, T individual, double fitness, int toCoverTargets)

Methods inherited from class java.lang.Object

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

Field Detail

logger

protected static final org.slf4j.Logger logger

Constant logger

Constructor Detail

FitnessFunction

public FitnessFunction()

Method Detail

updateIndividual

protected void updateIndividual(FitnessFunction<?> ff,
                                T individual,
                                double fitness)

Make sure that the individual gets to know about its fitness

Parameters:

individual - a Chromosome object.

fitness - a double.

updateIndividual

protected void updateIndividual(FitnessFunction<?> ff,
                                T individual,
                                double fitness,
                                int toCoverTargets)

getBestStoredIndividual

public T getBestStoredIndividual()

If the fitness function as an archive, returns the best individual in the archive. returns null otherwise

Returns:

getFitness

public abstract double getFitness(T individual)

Calculate and set fitness function #TODO the 'set fitness' part should be done by some abstract super class of all FitnessFunctions

Parameters:

individual - a Chromosome object.

Returns:

new fitness

normalize

public static double normalize(double value)
                        throws IllegalArgumentException

Normalize a value using Andrea's normalization function

Parameters:

value - a double.

Returns:

a double.

Throws:

IllegalArgumentException - if any.

isMaximizationFunction

public abstract boolean isMaximizationFunction()

Do we need to maximize, or minimize this function?

Returns:

a boolean.

updateCoveredGoals

public boolean updateCoveredGoals()

if the fitness function contains an archive, updates the archive and the fitness values of the population, and returns true. if the fitness function doesn't contain an archive, return false. This method has to be invoked after each generation.

Returns: