/**
* This method will return a naming String
*
* @return The name of the algorithm
*/
public String getName() {
String title = "PSOD-";
String cognitivePart = "";
if (algoTypeSimplified.isSelectedString(strNo)) {
cognitivePart = getPhi1() + "_";
} else {
title += "VG-";
}
String typePart = "";
if (algoTypePsod.isSelectedString(strAlgoTypeSixSigma)) {
typePart = "6o_";
}
if (algoTypePsod.isSelectedString(strAlgoTypeStagnationRiot)) {
typePart = "RS.";
if (stagnationMeasure.isSelectedString("Generations")) typePart += "Gn.";
else typePart += "Ft.";
if (convergenceCondition.isSelectedString("Relative")) typePart += "Rel.";
else typePart += "Abs.";
typePart += getStagnationTime() + "_";
typePart += getConvergenceThreshold() + "_";
}
return title + typePart + getTopology() + getTopologyRange() + "_" + cognitivePart + getPhi2();
}
public void setAlgoTypeSimplified(SelectedTag algoTypeSimplified) {
this.algoTypeSimplified = algoTypeSimplified;
if (algoTypeSimplified.isSelectedString(strYes)) {
algType.setSelectedTag(0); // inertness
m_Phi1 = 0.0;
m_Phi2 = 1.49445;
} else {
algType.setSelectedTag(1); // constriction
m_Phi1 = 2.05;
setPhi2(2.05);
}
GenericObjectEditor.setShowProperty(
ParticleSwarmOptimization.class, "phi1", algoTypeSimplified.isSelectedString(strNo));
GenericObjectEditor.setShowProperty(
ParticleSwarmOptimization.class, "algoType", algoTypeSimplified.isSelectedString(strNo));
}
public PSOD() {
super();
algoTypePsod = new SelectedTag(strAlgoTypeSixSigma, strAlgoTypeStagnationRiot);
algoTypePsod.setSelectedTag(strAlgoTypeSixSigma);
algoTypeSimplified = new SelectedTag(strNo, strYes);
algoTypeSimplified.setSelectedTag(0);
stagnationMeasure = new SelectedTag("Fitness calls", "Generations");
this.pMetric = new PhenotypeMetric();
this.stagnationMeasure.setSelectedTag(1);
convergenceCondition = new SelectedTag("Relative", "Absolute");
this.convergenceCondition.setSelectedTag(0);
sigmaFilter = 0.7;
lowLimit = 1;
upLimit = 2;
}
private boolean stagnationTimeHasPassed(PopulationInterface pop) {
if (stagnationMeasure.isSelectedString("Fitness calls")) { // by fitness calls
// System.out.println("stagnationTimeHasPassed returns " + ((pop.getFunctionCalls() -
// popFitCalls) >= m_stagTime) + " after " + (pop.getFunctionCalls() - popFitCalls));
return (pop.getFunctionCalls() - popFitCalls) >= m_stagTime;
} else { // by generation
// System.out.println("stagnationTimeHasPassed returns " + ((pop.getFunctionCalls() -
// popGens) >= m_stagTime) + " after " + (pop.getFunctionCalls() - popGens));
return (pop.getGeneration() - popGens) >= m_stagTime;
}
}
/**
* Return true if |oldFit - curFit| < |oldFit| * thresh (relative case) and if |oldFit - curFit| <
* thresh (absolute case).
*
* @param
* @return
*/
protected boolean isStillConverged(PopulationInterface pop) {
double[] curFit = pop.getBestFitness();
double dist = EuclideanMetric.euclideanDistance(oldFit, curFit);
boolean ret;
if (convergenceCondition.isSelectedString("Relative")) {
ret = (dist < (oldNorm * convThresh));
} else {
ret = (dist < convThresh);
}
return ret;
}
public void doHiding() {
GenericObjectEditor.setShowProperty(
ParticleSwarmOptimization.class, "phi1", algoTypeSimplified.isSelectedString(strNo));
GenericObjectEditor.setShowProperty(
getClass(),
"convergenceCondition",
algoTypePsod.isSelectedString(strAlgoTypeStagnationRiot));
GenericObjectEditor.setShowProperty(
getClass(), "stagnationMeasure", algoTypePsod.isSelectedString(strAlgoTypeStagnationRiot));
GenericObjectEditor.setShowProperty(
getClass(), "stagnationTime", algoTypePsod.isSelectedString(strAlgoTypeStagnationRiot));
GenericObjectEditor.setShowProperty(
getClass(),
"convergenceThreshold",
algoTypePsod.isSelectedString(strAlgoTypeStagnationRiot));
GenericObjectEditor.setShowProperty(
getClass(), "sigmaFilter", algoTypePsod.isSelectedString(strAlgoTypeSixSigma));
}
protected double[] getAcceleration(
double[] personalBestPos, double[] neighbourBestPos, double[] curPosition, double[][] range) {
double[] accel = new double[curPosition.length];
double chi;
RNG.setRandomSeed();
boolean bIsAlgoSixSigma = algoTypePsod.isSelectedString(strAlgoTypeSixSigma);
boolean bIsStagOnRiot = algoTypePsod.isSelectedString(strAlgoTypeStagnationRiot);
double gen = 0;
boolean is_extreme = false;
boolean is_partial = false;
if (bIsAlgoSixSigma) {
gen = RNG.gaussianDouble(sigmaFilter);
// double chance_of_disagreements = RNG.randomDouble(0, 1);
// if (execCursor < chance_of_disagreements) {
if ((-2 < gen && gen <= -1) || (1 < gen && gen < 2)) // +- 2 sigma
{
partial_disagreements++;
is_partial = true;
} else {
if (gen <= -2 || gen >= 2) {
extreme_disagreements++;
is_extreme = true;
}
}
// }
} else {
if (bIsStagOnRiot && riotReady) {
extreme_disagreements++;
is_extreme = true;
}
}
for (int i = 0; i < personalBestPos.length; i++) {
// the component from the old velocity
accel[i] = 0;
if (algType.getSelectedTag().getID() == 1) chi = m_InertnessOrChi;
else chi = 1.;
// the component from the social model
double agreedAccel = neighbourBestPos[i] - curPosition[i]; // common PSO case
if (bIsAlgoSixSigma) {
if (is_partial) // +- 2 sigma
{
// make partial disagreements
double partialDisagreement = RNG.randomDouble(-1.0 * lowLimit, lowLimit);
agreedAccel *= partialDisagreement;
} else {
if (is_extreme) {
double extremeDisagreement = RNG.randomDouble(-1.0 * lowLimit, lowLimit);
// agreedAccel *= extremeDisagreement;
if (extremeDisagreement < 0)
agreedAccel *= (extremeDisagreement - (upLimit - lowLimit)); // -2 to -1
else agreedAccel *= (extremeDisagreement + (upLimit - lowLimit)); // 1 to 2*/
}
}
} else {
if (bIsStagOnRiot && is_extreme) {
double extremeDisagreement = RNG.randomDouble(-1.0 * lowLimit, lowLimit);
if (extremeDisagreement < 0)
agreedAccel *= (extremeDisagreement - (upLimit - lowLimit)); // -2 to -1
else agreedAccel *= (extremeDisagreement + (upLimit - lowLimit)); // 1 to 2
}
}
// the component from the cognition model (invalid in Pederson case)
if (algoTypeSimplified.isSelectedString(strNo))
accel[i] =
this.m_Phi1 * chi * RNG.randomDouble(0, 1) * (personalBestPos[i] - curPosition[i]);
// the component from social model
accel[i] += this.m_Phi2 * chi * RNG.randomDouble(0, 1) * agreedAccel;
}
return accel;
}
