FAIF - Evolutionary Algorithm

Template based evolutionary algorithm

Fast start

    //the class represent the problem
    class EASpace  : public EvolutionaryAlgorithmSpace {
    public:
        //custom mutation
        static Individual& mutation(Individual& ind) {
            RandomDouble r; //returns the double from 0 to 1 with uniform distribution
            ind = 4.0*(r() - 0.5)*ind;
            return ind;
        }
        //the fitness function
        static double fitness(const Individual& ind) { return -ind*ind; }
    };

  #include <faif/search/EvolutionaryAlgorithm.hpp>

  EASpace::Population pop; //initial population
  pop.push_back(1.0);
  pop.push_back(-2.0);
  pop.push_back(-4.0);
  pop.push_back(8.0);

  EvolutionaryAlgorithm< EASpace, MutationCustom, CrossoverNone, SelectionRanking, StopAfterNSteps<30> > ea;
  EASpace::Individual best = ea.solve(pop); //run the ea