/**
* Returns the inverted generational distance of solution set
* @param solutionSet Solution set
* @return The value of the hypervolume indicator
*/
public double GetIgd(SolutionSet solutionSet)
{
return new InvertedGenerationalDistance().Compute(
solutionSet.WriteObjectivesToMatrix(),
_trueParetoFront.WriteObjectivesToMatrix(),
_problem.NumberOfObjectives);
}
public QualityIndicator(Problem problem, String paretoFrontFile)
{
_problem = problem;
_trueParetoFront = MetricsUtil.ReadNonDominatedSolutionSet(paretoFrontFile);
_trueParetoFrontHypervolume = new Hypervolume().HypervolumeValue(
_trueParetoFront.WriteObjectivesToMatrix(),
_trueParetoFront.WriteObjectivesToMatrix(),
_problem.NumberOfObjectives);
}
/**
* Computes the feasibility ratio
* Return the ratio of feasible solutions
*/
public double MeanOveralViolation(SolutionSet solutionSet)
{
double aux = 0.0;
for (int i = 0; i < solutionSet.Size(); i++)
{
aux += Math.Abs(solutionSet[i].NumberOfViolatedConstraints*
solutionSet[i].OverallConstraintViolation);
}
return aux/solutionSet.Size();
}
/** Assigns crowding distances to all solutions in a <code>SolutionSet</code>.
* @param solutionSet The <code>SolutionSet</code>.
* @param nObjs Number of objectives.
*/
public static void CrowdingDistanceAssignment(SolutionSet solutionSet, int nObjs)
{
int size = solutionSet.Size();
if (size == 0)
{
return;
}
if (size == 1)
{
solutionSet[0].CrowdingDistance = (double.PositiveInfinity);
return;
} // if
if (size == 2)
{
solutionSet[0].CrowdingDistance = (double.PositiveInfinity);
solutionSet[1].CrowdingDistance = (double.PositiveInfinity);
return;
} // if
//Use a new SolutionSet to evite alter original solutionSet
SolutionSet front = new SolutionSet(size);
for (int i = 0; i < size; i++)
{
front.Add(solutionSet[i]);
}
for (int i = 0; i < size; i++)
{
front[i].CrowdingDistance = (0.0);
}
for (int i = 0; i < nObjs; i++)
{
// Sort the population by Obj n
front.Sort(new ObjectiveComparator(i));
double objetiveMinn = front[0].Objective[i];
double objetiveMaxn = front[front.Size() - 1].Objective[i];
//Set de crowding distance
front[0].CrowdingDistance = (double.PositiveInfinity);
front[size - 1].CrowdingDistance = (double.PositiveInfinity);
for (int j = 1; j < size - 1; j++)
{
double distance = front[j + 1].Objective[i] - front[j - 1].Objective[i];
distance = distance/(objetiveMaxn - objetiveMinn);
distance += front[j].CrowdingDistance;
front[j].CrowdingDistance = (distance);
} // for
} // for
}
/**
* Computes the feasibility ratio
* Return the ratio of feasible solutions
*/
public double FeasibilityRatio(SolutionSet solutionSet)
{
double aux = 0.0;
for (int i = 0; i < solutionSet.Size(); i++)
{
if (solutionSet[i].OverallConstraintViolation < 0)
{
aux = aux + 1.0;
}
}
return aux/solutionSet.Size();
}
public override object Execute(object obj)
{
SolutionSet population = (SolutionSet) obj;
int populationSize = (int) Parameters["populationSize"];
SolutionSet result = new SolutionSet(populationSize);
//->Ranking the union
Ranking ranking = new Ranking(population);
int remain = populationSize;
int index = 0;
SolutionSet front = null;
population.Clear();
//-> Obtain the next front
front = ranking.GetSubfront(index);
while ((remain > 0) && (remain >= front.Size()))
{
//Asign crowding distance to individuals
Distance.CrowdingDistanceAssignment(front, _problem.NumberOfObjectives);
//Add the individuals of this front
for (int k = 0; k < front.Size(); k++)
{
result.Add(front[k]);
}
//Decrement remaint
remain = remain - front.Size();
//Obtain the next front
index++;
if (remain > 0)
{
front = ranking.GetSubfront(index);
}
}
//-> remain is less than front(index).size, insert only the best one
if (remain > 0)
{
// front containt individuals to insert
Distance.CrowdingDistanceAssignment(front, _problem.NumberOfObjectives);
front.Sort(CrowdingComparator);
for (int k = 0; k < remain; k++)
{
result.Add(front[k]);
}
}
return result;
}
public Ranking(SolutionSet solutionSet)
{
_solutionSet = solutionSet;
// dominateMe[i] contains the number of solutions dominating i
int[] dominateMe = new int[_solutionSet.Size()];
// iDominate[k] contains the list of solutions dominated by k
List<int>[] iDominate = new List<int>[_solutionSet.Size()];
// front[i] contains the list of individuals belonging to the front i
List<int>[] front = new List<int>[_solutionSet.Size() + 1];
// flagDominate is an auxiliar encodings.variable
int flagDominate;
// Initialize the fronts
for (int k = 0; k < front.Length; k++)
{
front[k] = new List<int>();
}
//-> Fast non dominated sorting algorithm
// Contribution of Guillaume Jacquenot
for (int p = 0; p < _solutionSet.Size(); p++)
{
// Initialize the list of individuals that i dominate and the number
// of individuals that dominate me
iDominate[p] = new List<int>();
dominateMe[p] = 0;
}
for (int p = 0; p < (_solutionSet.Size() - 1); p++)
{
// For all q individuals , calculate if p dominates q or vice versa
for (int q = p + 1; q < _solutionSet.Size(); q++)
{
flagDominate = Constraint.Compare(solutionSet[p], solutionSet[q]);
if (flagDominate == 0)
{
flagDominate = Dominance.Compare(solutionSet[p], solutionSet[q]);
}
if (flagDominate == -1)
{
iDominate[p].Add(q);
dominateMe[q]++;
}
else if (flagDominate == 1)
{
iDominate[q].Add(p);
dominateMe[p]++;
}
}
// If nobody dominates p, p belongs to the first front
}
for (int p = 0; p < _solutionSet.Size(); p++)
{
if (dominateMe[p] == 0)
{
front[0].Add(p);
solutionSet[p].Rank = 0;
}
}
//Obtain the rest of fronts
int i = 0;
IEnumerator<int> it1; // Iterators
while (front[i].Count != 0)
{
i++;
it1 = front[i - 1].GetEnumerator();
while (it1.MoveNext())
{
IEnumerator<int> it2 = iDominate[it1.Current].GetEnumerator(); // Iterators
while (it2.MoveNext())
{
int index = it2.Current;
dominateMe[index]--;
if (dominateMe[index] == 0)
{
front[i].Add(index);
_solutionSet[index].Rank = i;
}
}
}
}
//<-
_ranking = new SolutionSet[i];
//0,1,2,....,i-1 are front, then i fronts
for (int j = 0; j < i; j++)
{
_ranking[j] = new SolutionSet(front[j].Count);
it1 = front[j].GetEnumerator();
while (it1.MoveNext())
{
_ranking[j].Add(solutionSet[it1.Current]);
}
}
}
/**
* Updates the threshold value using the population
*/
public void UpdateThreshold(SolutionSet set)
{
_threshold = FeasibilityRatio(set)*MeanOveralViolation(set);
}
/**
* Updates the grid limits and the grid content adding the solutions contained
* in a specific <code>SolutionSet</code>.
* @param solutionSet The <code>SolutionSet</code>.
*/
public void updateGrid(SolutionSet solutionSet)
{
//Update lower and upper limits
UpdateLimits(solutionSet);
//Calculate the division Size
for (int obj = 0; obj < _objectives; obj++)
{
_divisionSize[obj] = _upperLimits[obj] - _lowerLimits[obj];
}
//Clean the hypercubes
for (int i = 0; i < Hypercubes.Length; i++)
{
Hypercubes[i] = 0;
}
//Add the population
AddSolutionSet(solutionSet);
}
/**
* Return the index of the nearest solution in the solution set to a given solution
* @param solution
* @param solutionSet
* @return The index of the nearest solution; -1 if the solutionSet is empty
*/
public static int IndexToNearestSolutionInSolutionSpace(Solution solution, SolutionSet solutionSet)
{
int index = -1;
double minimumDistance = double.MaxValue;
for (int i = 0; i < solutionSet.Size(); i++)
{
double distance = DistanceBetweenSolutions(solution, solutionSet[i]);
if (distance < minimumDistance)
{
minimumDistance = distance;
index = i;
}
}
return index;
}
/**
* Updates the grid limits and the grid content adding a new
* <code>Solution</code>.
* If the solution falls out of the grid bounds, the limits and content of the
* grid must be re-calculated.
* @param solution <code>Solution</code> considered to update the grid.
* @param solutionSet <code>SolutionSet</code> used to update the grid.
*/
public void updateGrid(Solution solution, SolutionSet solutionSet)
{
int location = Location(solution);
if (location == -1)
{
//Re-build the Adaptative-Grid
//Update lower and upper limits
UpdateLimits(solutionSet);
//Actualize the lower and upper limits whit the individual
for (int obj = 0; obj < _objectives; obj++)
{
double solutionObjective = solution.Objective[obj];
if (solutionObjective < _lowerLimits[obj])
{
_lowerLimits[obj] = solutionObjective;
}
if (solutionObjective > _upperLimits[obj])
{
_upperLimits[obj] = solutionObjective;
}
}
//Calculate the division Size
for (int obj = 0; obj < _objectives; obj++)
{
_divisionSize[obj] = _upperLimits[obj] - _lowerLimits[obj];
}
//Clean the hypercube
for (int i = 0; i < Hypercubes.Length; i++)
{
Hypercubes[i] = 0;
}
//add the population
AddSolutionSet(solutionSet);
}
}
// NSGAII
public override SolutionSet Execute()
{
int populationSize;
int maxEvaluations;
QualityIndicator indicators;
Operator mutationOperator;
Operator crossoverOperator;
Operator selectionOperator;
object parameter;
if (InputParameters.TryGetValue("populationSize", out parameter))
{
populationSize = (int) parameter;
}
else
{
throw new Exception("populationSize does not exist");
}
if (InputParameters.TryGetValue("maxEvaluations", out parameter))
{
maxEvaluations = (int) parameter;
}
else
{
throw new Exception("maxEvaluations does not exist");
}
if (InputParameters.TryGetValue("indicators", out parameter))
{
indicators = (QualityIndicator) parameter;
}
else
{
throw new Exception("maxEvaluations does not exist");
}
// Initializing variables
var population = new SolutionSet(populationSize);
var evaluations = 0;
int requiredEvaluations = 0;
Operator unknownIOperator;
//Read the operators
if (UsedOperators.TryGetValue("mutation", out unknownIOperator))
{
mutationOperator = unknownIOperator;
}
else
{
throw new Exception("mutation does not exist");
}
if (UsedOperators.TryGetValue("crossover", out unknownIOperator))
{
crossoverOperator = unknownIOperator;
}
else
{
throw new Exception("crossover does not exist");
}
if (UsedOperators.TryGetValue("selection", out unknownIOperator))
{
selectionOperator = unknownIOperator;
}
else
{
throw new Exception("selection does not exist");
}
// Create the initial solutionSet
for (int i = 0; i < populationSize; i++)
{
var newSolution = new Solution(Problema);
Problema.Evaluate(newSolution);
Problema.EvaluateConstraints(newSolution);
evaluations++;
population.Add(newSolution);
} //for
// Generations
while (evaluations < maxEvaluations)
{
// Create the offSpring solutionSet
var offspringPopulation = new SolutionSet(populationSize);
Solution[] parents = new Solution[2];
for (int i = 0; i < (populationSize/2); i++)
{
if (evaluations < maxEvaluations)
{
//obtain parents
parents[0] = (Solution) selectionOperator.Execute(population);
parents[1] = (Solution) selectionOperator.Execute(population);
Solution[] offSpring = (Solution[]) crossoverOperator.Execute(parents);
mutationOperator.Execute(offSpring[0]);
mutationOperator.Execute(offSpring[1]);
Problema.Evaluate(offSpring[0]);
Problema.EvaluateConstraints(offSpring[0]);
Problema.Evaluate(offSpring[1]);
Problema.EvaluateConstraints(offSpring[1]);
offspringPopulation.Add(offSpring[0]);
offspringPopulation.Add(offSpring[1]);
evaluations += 2;
} // if
} // for
System.Console.WriteLine("evaluation #: " + evaluations);
// Create the solutionSet union of solutionSet and offSpring
SolutionSet union = population.Union(offspringPopulation);
// Ranking the union
Ranking localRanking = new Ranking(union);
int remain = populationSize;
int index = 0;
SolutionSet front = null;
population.Clear();
// Obtain the next front
front = localRanking.GetSubfront(index);
while ((remain > 0) && (remain >= front.Size()))
{
//Assign crowding distance to individuals
Distance.CrowdingDistanceAssignment(front, Problema.NumberOfObjectives);
//Add the individuals of this front
for (int k = 0; k < front.Size(); k++)
{
population.Add(front[k]);
} // for
//Decrement remain
remain = remain - front.Size();
//Obtain the next front
index++;
if (remain > 0)
{
front = localRanking.GetSubfront(index);
} // if
} // while
// Remain is less than front(index).size, insert only the best one
if (remain > 0)
{
// front contains individuals to insert
Distance.CrowdingDistanceAssignment(front, Problema.NumberOfObjectives);
front.Sort(new CrowdingComparator());
for (int k = 0; k < remain; k++)
{
population.Add(front[k]);
} // for
remain = 0;
} // if
// This piece of code shows how to use the indicator object into the code
// of NSGA-II. In particular, it finds the number of evaluations required
// by the algorithm to obtain a Pareto front with a hypervolume higher
// than the hypervolume of the true Pareto front.
if ((indicators != null) &&
(requiredEvaluations == 0))
{
double hv = indicators.GetHypervolume(population);
if (hv >= (0.98*indicators.GetTrueParetoFrontHypervolume()))
{
requiredEvaluations = evaluations;
} // if
} // if
} // while
// Return as output parameter the required evaluations
OutputParameters["evaluations"] = requiredEvaluations;
// Return the first non-dominated front
Ranking ranking = new Ranking(population);
ranking.GetSubfront(0).PrintFeasibleFUN("FUN_NSGAII");
return ranking.GetSubfront(0);
}
/** Returns the minimum distance from a <code>Solution</code> to a
* <code>SolutionSet according to the encodings.variable values</code>.
* @param solution The <code>Solution</code>.
* @param solutionSet The <code>SolutionSet</code>.
* @return The minimum distance between solution and the set.
* @throws JMException
*/
public static double DistanceToSolutionSetInSolutionSpace(Solution solution,
SolutionSet solutionSet)
{
//At start point the distance is the max
double distance = double.MaxValue;
// found the min distance respect to population
for (int i = 0; i < solutionSet.Size(); i++)
{
double aux = DistanceBetweenSolutions(solution, solutionSet[i]);
if (aux < distance)
{
distance = aux;
}
} // for
//->Return the best distance
return distance;
}
private void NormalizePopulation(SolutionSet pop)
{
Solution z = new Solution(Problema);
for (var i = 0; i < Problema.NumberOfObjectives; i++)
{
z.Objective[i] = double.NegativeInfinity;
}
foreach (var x in pop.SolutionList)
{
for (var i = 0; i < Problema.NumberOfObjectives; i++)
{
if (x.Objective[i] > z.Objective[i])
{
z.Objective[i] = x.Objective[i];
}
}
}
foreach (var x in pop.SolutionList)
{
for (var i = 0; i < Problema.NumberOfObjectives; i++)
{
x.Objective[i] = (x.Objective[i] - IdealPoint.Objective[i])/
(z.Objective[i] - IdealPoint.Objective[i]);
}
}
}
public override SolutionSet Execute()
{
int populationSize;
int maxEvaluations;
QualityIndicator indicators;
Operator mutationOperator;
Operator crossoverOperator;
Operator selectionOperator;
object parameter;
if (InputParameters.TryGetValue("populationSize", out parameter))
{
populationSize = (int) parameter;
}
else
{
throw new Exception("populationSize does not exist");
}
if (InputParameters.TryGetValue("maxEvaluations", out parameter))
{
maxEvaluations = (int) parameter;
}
else
{
throw new Exception("maxEvaluations does not exist");
}
if (InputParameters.TryGetValue("indicators", out parameter))
{
indicators = (QualityIndicator) parameter;
}
else
{
throw new Exception("maxEvaluations does not exist");
}
if (InputParameters.TryGetValue("theta", out parameter))
{
Theta = (int) parameter;
}
else
{
throw new Exception("Theta does not exist");
}
if (InputParameters.TryGetValue("H", out parameter))
{
K = MetalMath.BinomialCoeff(Problema.NumberOfObjectives - 1,
((int) parameter) + Problema.NumberOfObjectives - 1);
}
else
{
throw new Exception("H does not exist");
}
// Initializing variables
var population = new SolutionSet(populationSize);
var evaluations = 0;
int requiredEvaluations = 0;
Operator unknownIOperator;
//Read the operators
if (UsedOperators.TryGetValue("mutation", out unknownIOperator))
{
mutationOperator = unknownIOperator;
}
else
{
throw new Exception("mutation does not exist");
}
if (UsedOperators.TryGetValue("crossover", out unknownIOperator))
{
crossoverOperator = unknownIOperator;
}
else
{
throw new Exception("crossover does not exist");
}
if (UsedOperators.TryGetValue("selection", out unknownIOperator))
{
selectionOperator = unknownIOperator;
}
else
{
throw new Exception("selection does not exist");
}
K = populationSize;
ReferencePoints = new Solution[K];
//Create references points
GenerateReferencePoint();
// Create the initial solutionSet
for (int i = 0; i < populationSize; i++)
{
var newSolution = new Solution(Problema);
Problema.Evaluate(newSolution);
Problema.EvaluateConstraints(newSolution);
evaluations++;
population.Add(newSolution);
}
IdealPoint = new Solution(Problema);
for (var i = 0; i < Problema.NumberOfObjectives; i++)
{
IdealPoint.Objective[i] = double.PositiveInfinity;
}
ComputeIdealPoint(population);
ThetaNonDominatedSort ranking = null;
// Generations
SolutionSet union = null;
while (evaluations < maxEvaluations)
{
// Create the offSpring solutionSet
var offspringPopulation = new SolutionSet(populationSize);
Solution[] parents = new Solution[2];
for (int i = 0; i < (populationSize/2); i++)
{
if (evaluations < maxEvaluations)
{
//obtain parents
parents[0] = (Solution) selectionOperator.Execute(population);
parents[1] = (Solution) selectionOperator.Execute(population);
Solution[] offSpring = (Solution[]) crossoverOperator.Execute(parents);
mutationOperator.Execute(offSpring[0]);
mutationOperator.Execute(offSpring[1]);
Problema.Evaluate(offSpring[0]);
Problema.EvaluateConstraints(offSpring[0]);
Problema.Evaluate(offSpring[1]);
Problema.EvaluateConstraints(offSpring[1]);
offspringPopulation.Add(offSpring[0]);
offspringPopulation.Add(offSpring[1]);
evaluations += 2;
}
}
//Update ideal point
// Create the solutionSet union of solutionSet and offSpring
union = population.Union(offspringPopulation);
ComputeIdealPoint(offspringPopulation);
//Normalize
NormalizePopulation(union);
//Clustering and non dominated sort
ranking = new ThetaNonDominatedSort(union, Clustering(union), Theta);
//Generate next pop
population = new SolutionSet(populationSize);
int front = 0;
while (population.Size() + ranking.GetSubfront(front).Size() <= populationSize)
{
var frontResult = ranking.GetSubfront(front);
foreach (var r in frontResult.SolutionList)
{
population.Add(r);
}
front++;
}
//Generate new pop
var localList =
ranking.GetSubfront(front).SolutionList.OrderBy(item => PseudoRandom.Instance().Next()).ToList();
int localSizePt1 = population.Size();
for (int i = 0; i < populationSize - localSizePt1; i++)
{
population.Add(localList[i]);
}
if ((indicators != null) && (requiredEvaluations == 0))
{
/*
double hv = indicators.GetHypervolume(population);
if (hv >= (0.98 * indicators.GetTrueParetoFrontHypervolume()))
{
requiredEvaluations = evaluations;
}*/
double gd = indicators.GetGd(ranking.GetSubfront(0));
Console.WriteLine(gd);
}
}
OutputParameters["evaluations"] = requiredEvaluations;
ranking.GetSubfront(0).PrintFeasibleFUN("FUN_THETA_NSGAIII");
Console.WriteLine("before quitting");
Console.WriteLine(indicators.GetGd(ranking.GetSubfront(0)));
Console.WriteLine("before quitting");
return ranking.GetSubfront(0);
}
private List<Cluster> Clustering(SolutionSet normalizedPop)
{
List<Cluster> clusters = new List<Cluster>(K);
for (var c = 0; c < K; c++)
{
clusters.Add(new Cluster(ReferencePoints[c], normalizedPop.Size()));
}
for (var x = 0; x < normalizedPop.Size(); x++)
{
var k = 0;
var min = Distance(normalizedPop.SolutionList[x], ReferencePoints[k]);
for (var j = 1; j < K; j++)
{
var d2 = Distance(normalizedPop.SolutionList[x], ReferencePoints[j]);
if (d2.Item2 < min.Item2)
{
min = d2;
k = j;
}
}
clusters[k].Add(normalizedPop.SolutionList[x], min);
}
return clusters;
}
/**
* Returns the spread of solution set
* @param solutionSet Solution set
* @return The value of the hypervolume indicator
*/
public double GetSpread(SolutionSet solutionSet)
{
return new Spread().Compute(solutionSet.WriteObjectivesToMatrix(),
_trueParetoFront.WriteObjectivesToMatrix(),
_problem.NumberOfObjectives);
}
/**
* Updates the grid adding solutions contained in a specific
* <code>SolutionSet</code>.
* <b>REQUIRE</b> The grid limits must have been previously calculated.
* @param solutionSet The <code>SolutionSet</code> considered.
*/
private void AddSolutionSet(SolutionSet solutionSet)
{
//Calculate the location of all individuals and update the grid
MostPopulated = 0;
for (int ind = 0; ind < solutionSet.Size(); ind++)
{
int location = Location(solutionSet[ind]);
Hypercubes[location]++;
if (Hypercubes[location] > Hypercubes[MostPopulated])
{
MostPopulated = location;
}
}
//The grid has been updated, so also update ocuppied's hypercubes
CalculateOccupied();
}
/**
* Returns the hypervolume of solution set
* @param solutionSet Solution set
* @return The value of the hypervolume indicator
*/
public double GetHypervolume(SolutionSet solutionSet)
{
return new Hypervolume().HypervolumeValue(solutionSet.WriteObjectivesToMatrix(),
_trueParetoFront.WriteObjectivesToMatrix(),
_problem.NumberOfObjectives);
}
/**
* Updates the grid limits considering the solutions contained in a
* <code>SolutionSet</code>.
* @param solutionSet The <code>SolutionSet</code> considered.
*/
private void UpdateLimits(SolutionSet solutionSet)
{
//Init the lower and upper limits
for (int obj = 0; obj < _objectives; obj++)
{
//Set the lower limits to the max real
_lowerLimits[obj] = double.MaxValue;
//Set the upper limits to the min real
_upperLimits[obj] = double.MinValue;
}
//Find the max and min limits of objetives into the population
for (int ind = 0; ind < solutionSet.Size(); ind++)
{
Solution tmpIndividual = solutionSet[ind];
for (int obj = 0; obj < _objectives; obj++)
{
double tmpIndividualObjective = tmpIndividual.Objective[obj];
if (tmpIndividualObjective < _lowerLimits[obj])
{
_lowerLimits[obj] = tmpIndividualObjective;
}
if (tmpIndividualObjective > _upperLimits[obj])
{
_upperLimits[obj] = tmpIndividualObjective;
}
}
}
}
private void ComputeIdealPoint(SolutionSet pop)
{
if (pop.Size() <= 0)
{
throw new Exception("Empty pop");
}
pop.SolutionList.ForEach(sol =>
{
for (var i = 0; i < sol.NumberOfObjectives; i++)
{
if (sol.Objective[i] < IdealPoint.Objective[i])
{
IdealPoint.Objective[i] = sol.Objective[i];
}
}
});
}
public Cluster(Solution referencePoint, int populationSize)
{
Solutions = new SolutionSet(populationSize);
Distances = new List<Tuple<double, double>>();
ReferencePoint = referencePoint.Clone();
}
private void printObjectives(string filename, SolutionSet pop)
{
Thread.CurrentThread.CurrentCulture = new CultureInfo("en-US");
TextWriter tw = new StreamWriter(filename);
pop.SolutionList.ForEach(sol => { tw.WriteLine(string.Join(",", sol.Objective)); });
tw.Close();
}
/**
* Returns a matrix with distances between solutions in a
* <code>SolutionSet</code>.
* @param solutionSet The <code>SolutionSet</code>.
* @return a matrix with distances.
*/
public static double[][] DistanceMatrix(SolutionSet solutionSet)
{
//The matrix of distances
double[][] distance = new double[solutionSet.Size()][];
for (int i = 0; i > solutionSet.Size(); i++)
{
distance[i] = new double[solutionSet.Size()];
}
//-> Calculate the distances
for (int i = 0; i < solutionSet.Size(); i++)
{
distance[i][i] = 0.0;
Solution solutionI = solutionSet[i];
for (int j = i + 1; j < solutionSet.Size(); j++)
{
Solution solutionJ = solutionSet[j];
distance[i][j] = DistanceBetweenObjectives(solutionI, solutionJ);
distance[j][i] = distance[i][j];
} // for
} // for
//->Return the matrix of distances
return distance;
}
public ThetaNonDominatedSort(SolutionSet population, List<Cluster> clusters, double theta)
{
Dominance = new ThetaComparator(theta);
_solutionSet = population;
_ranking = new List<SolutionSet>();
// dominateMe[i] contains the number of solutions dominating i
int[] dominateMe = new int[_solutionSet.Size()];
// iDominate[k] contains the list of solutions dominated by k
List<int>[] iDominate = new List<int>[_solutionSet.Size()];
// front[i] contains the list of individuals belonging to the front i
List<int>[] front = new List<int>[_solutionSet.Size() + 1];
// flagDominate is an auxiliar encodings.variable
int flagDominate;
for (int p = 0; p < (clusters.Count - 1); p++)
{
// Initialize the fronts
for (int k = 0; k < front.Length; k++)
{
front[k] = new List<int>();
}
for (int q = 0; q < clusters[p].Count; q++)
{
// Initialize the list of individuals that i dominate and the number
// of individuals that dominate me
iDominate[q] = new List<int>();
dominateMe[q] = 0;
}
for (int q = 0; q < (clusters[p].Count - 1); q++)
{
for (int r = q + 1; r < (clusters[p].Count - 1); r++)
{
flagDominate = Dominance.Compare(clusters[p][q].Item2, clusters[p][r].Item2);
//flagDominate = Constraint.Compare(clusters[p][q].Item1, clusters[p][r].Item1);
if (flagDominate == 0)
{
flagDominate = Dominance.Compare(clusters[p][q].Item2, clusters[p][r].Item2);
}
if (flagDominate == -1)
{
iDominate[q].Add(r);
dominateMe[r]++;
}
else if (flagDominate == 1)
{
iDominate[r].Add(q);
dominateMe[q]++;
}
}
}
//Fill front
if (_ranking.Count == 0)
{
_ranking.Add(new SolutionSet(population.Size()));
}
for (int q = 0; q < clusters[p].Count; q++)
{
if (dominateMe[q] == 0)
{
front[0].Add(q);
_ranking[0].Add(clusters[p][q].Item1);
clusters[p][q].Item1.Rank = 0;
}
}
//Obtain the rest of fronts
int i = 0;
IEnumerator<int> it1; // Iterators
while (front[i].Count != 0)
{
i++;
if (i >= _ranking.Count)
{
_ranking.Add(new SolutionSet(population.Size()));
}
it1 = front[i - 1].GetEnumerator();
while (it1.MoveNext())
{
IEnumerator<int> it2 = iDominate[it1.Current].GetEnumerator(); // Iterators
while (it2.MoveNext())
{
int index = it2.Current;
dominateMe[index]--;
if (dominateMe[index] == 0)
{
front[i].Add(index);
_ranking[i].Add(clusters[p][index].Item1);
clusters[p][index].Item1.Rank = i;
}
}
}
}
}
}
/// <summary>
/// Joins two solution sets: the current one and the one passed as argument
/// </summary>
/// <param name="solutionSet">
/// A <see cref="SolutionSet" />
/// </param>
/// <returns>
/// A new solution set
/// A <see cref="SolutionSet" />
/// </returns>
public SolutionSet Union(SolutionSet solutionSet)
{
if (solutionSet == null)
{
throw new ArgumentNullException("solutionSet");
}
//Check the correct size
int newSize = Size() + solutionSet.Size();
if (newSize < Capacity)
{
newSize = Capacity;
}
//Create a new population
var union = new SolutionSet(newSize);
foreach (Solution solution in SolutionList)
{
union.Add(solution);
}
for (int i = Size(); i < (Size() + solutionSet.Size()); i++)
{
union.Add(solutionSet.SolutionList[i - Size()]);
}
return union;
}
