AlgElimination.cpp

#include "ECF_base.h"
#include "AlgElimination.h"
#include "ECF_macro.h"
 
 
Elimination :: Elimination()
{
    name_ = "Elimination";
 
    selFitPropOp = static_cast<SelFitnessProportionalOpP> (new SelFitnessProportionalOp);
    selRandomOp = static_cast<SelRandomOpP> (new SelRandomOp);
    selBestOp = static_cast<SelBestOpP> (new SelBestOp);
}
 
 
void Elimination :: registerParameters(StateP state)
{
    registerParameter(state, "gengap", (voidP) new double(0.6), ECF::DOUBLE,
        "generation gap (percentage of population to be eliminated)");
    registerParameter(state, "selpressure", (voidP) new double(10), ECF::DOUBLE,
        "selection pressure: how much is the worst individual 'worse' than the best");
}
 
 
bool Elimination :: initialize(StateP state)
{
    selFitPropOp->initialize(state);
    selRandomOp->initialize(state);
    selBestOp->initialize(state);
 
    voidP genGapP = getParameterValue(state, "gengap");
    genGap_ = *((double*) genGapP.get());
 
    if(genGap_ <= 0 || genGap_ > 1) {
        ECF_LOG_ERROR(state, "Error: generation gap parameter in Elimination algorithm must be in <0, 1]!");
        throw "";
    }
 
    voidP selPressP = getParameterValue(state, "selpressure");
    selPressure_ = *((double*) selPressP.get());
 
    // selection chooses worse individuals
    selFitPropOp->setSelPressure(1./selPressure_);
 
    return true;
}
 
 
bool Elimination :: advanceGeneration(StateP state, DemeP deme)
{
    // elitism: copy current best individual
    IndividualP best = selBestOp->select(*deme);
    best = copy(best);
 
    // copy pointers
    std::vector<IndividualP> newGen;
    for(uint i = 0; i < deme->size(); i++)
        newGen.push_back(deme->at(i));
 
    // eliminate genGap_ worse individuals
    uint generationGap = (uint) (genGap_ * deme->size());
    for(uint i = 0; i < generationGap; i++) {
        IndividualP victim = selFitPropOp->select(newGen);
        removeFrom(victim, newGen);
    }
 
    // make genGap_ new ones
    for(uint i = 0; i < generationGap; i++) {
        IndividualP parent1 = selRandomOp->select(*deme);
        IndividualP parent2 = selRandomOp->select(*deme);
        IndividualP child = copy(parent1);
        mate(parent1, parent2, child);
        newGen.push_back(child);
    }
 
    // replace
    for(uint i = 0; i < deme->size(); i++)
        replaceWith(deme->at(i), newGen[i]);
 
    // mutate all
    mutate(*deme);
 
    // evaluate new individuals
    for(uint i = 0; i < deme->size(); i++)
        if(!deme->at(i)->fitness->isValid()) {
            evaluate(deme->at(i));
        }
 
    // elitism: preserve best individual
    IndividualP random = selFitPropOp->select(*deme);
    if(best->fitness->isBetterThan(random->fitness))
        replaceWith(random, best);
 
    return true;
}