org.eclipse.viatra.dse.objectives

Class ObjectiveComparatorHelper

java.lang.Object

org.eclipse.viatra.dse.objectives.ObjectiveComparatorHelper

public class ObjectiveComparatorHelper
extends java.lang.Object

This class is responsible to compare and sort fitness values. TrajectoryFitness instances can be added to an instance of this class, that it can sort them.

Constructor Summary

Constructors

Constructor and Description
ObjectiveComparatorHelper(IObjective[][] leveledObjectives)

Method Summary

Methods

Modifier and Type	Method and Description
void	addTrajectoryFitness(TrajectoryFitness trajectoryFitness) Adds a TrajectoryFitness to an inner list to compare later.
void	clearTrajectoryFitnesses() Clears the inner TrajectoryFitness list.
int	compare(Fitness o1, Fitness o2) Compares two fitnesses based on hierarchical dominance.
static void	crowdingDistanceAssignment(java.util.List<TrajectoryFitness> front, IObjective[][] leveledObjectives) Executes the crowding distance assignment for the specified front.
java.util.List<? extends java.util.List<TrajectoryFitness>>	getFronts() Returns the previously added TrajectoryFitness instances in fronts.
java.util.List<TrajectoryFitness>	getParetoFront() Returns the pareto front of the previously added TrajectoryFitness.
TrajectoryFitness	getRandomBest() Returns a random TrajectoryFitness from the pareto front.
java.util.List<TrajectoryFitness>	getTrajectoryFitnesses() Returns the inner TrajectoryFitness list.
void	setComputeCrowdingDistance(boolean computeCrowdingDistance)

Methods inherited from class java.lang.Object

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

Constructor Detail

ObjectiveComparatorHelper

public ObjectiveComparatorHelper(IObjective[][] leveledObjectives)

Method Detail

setComputeCrowdingDistance

public void setComputeCrowdingDistance(boolean computeCrowdingDistance)

compare

public int compare(Fitness o1,
          Fitness o2)

Compares two fitnesses based on hierarchical dominance. Returns -1 if the second parameter o2 is a better solution (o2 dominates o1), 1 if the first parameter o1 is better (o1 dominates o2) and returns 0 if they are non-dominating each other.

Parameters:

o1 -

o2 -

Returns:

addTrajectoryFitness

public void addTrajectoryFitness(TrajectoryFitness trajectoryFitness)

Adds a TrajectoryFitness to an inner list to compare later.

Parameters:

trajectoryFitness -

clearTrajectoryFitnesses

public void clearTrajectoryFitnesses()

Clears the inner TrajectoryFitness list.

getTrajectoryFitnesses

public java.util.List<TrajectoryFitness> getTrajectoryFitnesses()

Returns the inner TrajectoryFitness list.

Returns:

getRandomBest

public TrajectoryFitness getRandomBest()

Returns a random TrajectoryFitness from the pareto front.

Returns:

getParetoFront

public java.util.List<TrajectoryFitness> getParetoFront()

Returns the pareto front of the previously added TrajectoryFitness.

Returns:

getFronts

public java.util.List<? extends java.util.List<TrajectoryFitness>> getFronts()

Returns the previously added TrajectoryFitness instances in fronts.

Returns:

crowdingDistanceAssignment

public static void crowdingDistanceAssignment(java.util.List<TrajectoryFitness> front,
                              IObjective[][] leveledObjectives)

Executes the crowding distance assignment for the specified front.

Parameters:

front -