public void postGeneration() {
// Create the solutionSet union of solutionSet and offSpring
union_ = ((SolutionSet) population_).union(offspringPopulation_);
// Ranking the union
Ranking ranking = new Ranking(union_);
if (ranking.getNumberOfSubfronts() == 0) System.out.println("No hay subfrentes!!");
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())) {
// Assign crowding distance to individuals
distance_.crowdingDistanceAssignment(front, problem_.getNumberOfObjectives());
// Add the individuals of this front
for (int k = 0; k < front.size(); k++) {
population_.add(front.get(k));
} // for
// Decrement remain
remain = remain - front.size();
// Obtain the next front
index++;
if (remain > 0) {
front = ranking.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, problem_.getNumberOfObjectives());
front.sort(new jmetal.coevolutionary.base.operator.comparator.CrowdingComparator());
for (int k = 0; k < remain; k++) {
population_.add(front.get(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
prepareBestSolutions();
} // postGeneration
private void prepareBestSolutions() {
bestDecisionVariables_ = new DecisionVariables[numberOfSolutions_];
Ranking ranking = new Ranking(population_);
int remain;
SolutionSet subfront = ranking.getSubfront(0);
int sz = subfront.size();
int rest = population_.size() - sz;
int i;
if ((sz >= bestSolutionsFirstLevel_)
&& (ranking.getNumberOfSubfronts() > 1)
&& (rest >= (numberOfSolutions_ - bestSolutionsFirstLevel_))) {
// Subfront is enought big and the remain is assured
int[] indices = RandomVector.getRandomVector_Int(bestSolutionsFirstLevel_, sz);
for (i = 0; i < bestSolutionsFirstLevel_; ++i)
bestDecisionVariables_[i] = subfront.get(indices[i]).getDecisionVariables();
remain = numberOfSolutions_ - bestSolutionsFirstLevel_;
} // if
else if (rest < (numberOfSolutions_ - bestSolutionsFirstLevel_)) {
// The rest of subfronts are so little
int[] indices = RandomVector.getRandomVector_Int(numberOfSolutions_, sz);
for (i = 0; i < numberOfSolutions_; ++i)
bestDecisionVariables_[i] = subfront.get(indices[i]).getDecisionVariables();
remain = 0;
} // else if
else if (numberOfSolutions_ == 1) {
int[] indices = RandomVector.getRandomVector_Int(bestSolutionsFirstLevel_, sz);
bestDecisionVariables_[0] = subfront.get(indices[0]).getDecisionVariables();
remain = 0;
i = 1;
} // else if
else {
int[] indices = RandomVector.getRandomVector_Int(sz, sz);
for (i = 0; i < sz; ++i)
bestDecisionVariables_[i] = subfront.get(indices[i]).getDecisionVariables();
remain = numberOfSolutions_ - sz;
} // else
int index = 1;
while ((index < ranking.getNumberOfSubfronts()) && (remain > 0)) {
subfront = ranking.getSubfront(index);
sz = subfront.size();
if (sz >= remain) {
int[] indices = RandomVector.getRandomVector_Int(remain, sz);
for (int j = 0; j < remain; ++j, ++i)
bestDecisionVariables_[i] = subfront.get(indices[j]).getDecisionVariables();
remain = 0;
} // if
else {
int[] indices = RandomVector.getRandomVector_Int(sz, sz);
for (int j = 0; j < sz; ++j, ++i, --remain)
bestDecisionVariables_[i] = subfront.get(indices[j]).getDecisionVariables();
} // else
++index;
} // while
} // prepareBestSolutions