/**
* Mutates the classifier. It mutates the action and the condition.
*
* @param envState is the current environtment state. It is necessary for the niched mutation.
* @return a boolean indicating if the action has been mutated.
*/
public boolean mutate(double[] envState) {
int i;
// First, the condition is mutated
for (i = 0; i < rep.length; i++) {
rep[i].mutate(envState[i]);
}
// Second, the action is mutated
int act = 0;
if (Config.rand() < Config.pM) {
do {
act = (int) (Config.rand() * (double) Config.numberOfActions);
} while (action == act);
action = act;
return true;
}
return false;
} // end mutate
/**
* Applies crossover. It generates two children.
*
* @param parent1 is the first parent.
* @param parent2 is the second parent
* @param child1 is the first child
* @param child2 is the second child.
*/
public void makeCrossover(
Classifier parent1, Classifier parent2, Classifier child1, Classifier child2) {
int i = 0;
// cross1 is a number between [0.. clLength-1]
int cross1 = (int) (Config.rand() * (double) Config.clLength);
// cross2 is a number between [1..clLenght].
int cross2 = (int) (Config.rand() * (double) Config.clLength) + 1;
if (cross1 > cross2) {
int aux = cross2;
cross2 = cross1;
cross1 = aux;
// In the else-if condition is not necessary to check if (cross2<clLength)
// to increment the point, because cross1 [0..length-1]
} else if (cross1 == cross2) cross2++;
// All the intervals (real representation) or genes (ternary representation) that
// are not in the cross point are crossed.
if (!Config.ternaryRep) {
for (i = cross1 + 1; i < cross2 - 1; i++) {
child2.setAllele(i, parent1);
child1.setAllele(i, parent2);
}
// Now we have to cross the border allele
child1.crossAllele(cross1, parent1, parent2);
child1.crossAllele(cross2 - 1, parent2, parent1);
child2.crossAllele(cross1, parent2, parent1);
child2.crossAllele(cross2 - 1, parent1, parent2);
} else {
for (i = cross1; i < cross2 - 1; i++) {
child2.setAllele(i, parent1);
child1.setAllele(i, parent2);
}
}
} // end makeCrossover
/**
* It crosses a real allele within two parents. If the representation is a ternary representation,
* a crossover within intervals is not possible because there is only one gene in each position.
* So, in this case, the gene of the second parent will be copied. In case of being a real
* representation, a random number is generated to decide where to cross the interval. It it's
* crossed within the inteval, the crossAllele method will do it.
*
* @param parent1 is the first parent classifier.
* @param parent2 is the second parent classifier.
*/
public void crossAllele(int i, Classifier parent1, Classifier parent2) {
if (Config.ternaryRep) {
rep[i].setAllele(parent2.rep[i]);
} else {
if (Config.typeOfAttributes[i].equals("ternary")) {
rep[i].setAllele(parent2.rep[i]);
} else {
if (Config.rand() < 0.5) rep[i].setAllele(parent2.rep[i]);
else { // The alleles has to be crossed
rep[i].setAllele(parent1.rep[i].getLowerAllele(), parent2.rep[i].getUpperAllele());
rep[i].verifyInterval();
}
}
}
} // end crossAllele
/**
* It constructs a classifier with the condition and the action specified. It's used by the
* covering operator.
*
* <p>
*
* @param condition is the environtment state
* @param action is the action chosen
* @param size is the size of the set.
* @param tStamp is the time
*/
public Classifier(double[] envState, int classOfExample, int size, int tStamp) {
int i = 0;
rep = new Attribute[Config.clLength];
if (Config.ternaryRep) {
for (i = 0; i < rep.length; i++) rep[i] = new TernaryRep(envState[i]);
} else {
for (i = 0; i < rep.length; i++) {
if (Config.typeOfAttributes[i].equals("ternary")) rep[i] = new TernaryRep(envState[i]);
else {
if (envState[i] == Config.unknownValue) rep[i] = new RealRep(Config.rand());
else rep[i] = new RealRep(envState[i]);
}
}
}
action = classOfExample;
parameters = new Parameters(tStamp, size, getGenerality());
} // end Classifier
