paradiseo/trunk/paradiseo-moeo/src/diversity/moeoSharingDiversityAssignment.h

// -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-

//-----------------------------------------------------------------------------
// moeoSharingDiversityAssignment.h
// (c) OPAC Team (LIFL), Dolphin Project (INRIA), 2007
/*
    This library...

    Contact: paradiseo-help@lists.gforge.inria.fr, http://paradiseo.gforge.inria.fr
 */
//-----------------------------------------------------------------------------

#ifndef MOEOSHARINGDIVERSITYASSIGNMENT_H_
#define MOEOSHARINGDIVERSITYASSIGNMENT_H_

#include <eoPop.h>
#include <comparator/moeoDiversityThenFitnessComparator.h>
#include <distance/moeoDistance.h>
#include <distance/moeoDistanceMatrix.h>
#include <distance/moeoEuclideanDistance.h>
#include <diversity/moeoDiversityAssignment.h>

/**
 * Sharing assignment scheme originally porposed by:
 * D. E. Goldberg, "Genetic Algorithms in Search, Optimization and Machine Learning", Addision-Wesley, MA, USA (1989).
 */
template < class MOEOT >
class moeoSharingDiversityAssignment : public moeoDiversityAssignment < MOEOT >
{
public:

    /** the objective vector type of the solutions */
    typedef typename MOEOT::ObjectiveVector ObjectiveVector;


    /**
     * Ctor
     * @param _distance the distance used to compute the neighborhood of solutions (can be related to the decision space or the objective space)
     * @param _nicheSize neighborhood size in terms of radius distance (closely related to the way the distances are computed)
     * @param _alpha parameter used to regulate the shape of the sharing function
     */
    moeoSharingDiversityAssignment(moeoDistance<MOEOT,double> & _distance, double _nicheSize = 0.5, double _alpha = 1.0) : distance(_distance), nicheSize(_nicheSize), alpha(_alpha)
    {}


    /**
     * Ctor with an euclidean distance (with normalized objective values) in the objective space is used as default
     * @param _nicheSize neighborhood size in terms of radius distance (closely related to the way the distances are computed)
     * @param _alpha parameter used to regulate the shape of the sharing function
     */
    moeoSharingDiversityAssignment(double _nicheSize = 0.5, double _alpha = 1.0) : distance(defaultDistance), nicheSize(_nicheSize), alpha(_alpha)
    {}


    /**
     * Sets diversity values for every solution contained in the population _pop
     * @param _pop the population
     */
    void operator()(eoPop < MOEOT > & _pop)
    {
        // 1 - set simuilarities
        setSimilarities(_pop);
        // 2 - a higher diversity is better, so the values need to be inverted
        moeoDiversityThenFitnessComparator < MOEOT > divComparator;
        double max = std::max_element(_pop.begin(), _pop.end(), divComparator)->diversity();
        for (unsigned int i=0 ; i<_pop.size() ; i++)
        {
            _pop[i].diversity(max - _pop[i].diversity());
        }
    }


    /**
     * @warning NOT IMPLEMENTED, DO NOTHING !
     * Updates the diversity values of the whole population _pop by taking the deletion of the objective vector _objVec into account.
     * @param _pop the population
     * @param _objVec the objective vector
     * @warning NOT IMPLEMENTED, DO NOTHING !
     */
    void updateByDeleting(eoPop < MOEOT > & _pop, ObjectiveVector & _objVec)
    {
        std::cout << "WARNING : updateByDeleting not implemented in moeoSharingDiversityAssignment" << std::endl;
    }


protected:

    /** the distance used to compute the neighborhood of solutions */
    moeoDistance < MOEOT , double > & distance;
    /** euclidean distancein the objective space (can be used as default) */
    moeoEuclideanDistance < MOEOT > defaultDistance;
    /** neighborhood size in terms of radius distance */
    double nicheSize;
    /** parameter used to regulate the shape of the sharing function */
    double alpha;


    /**
     * Sets similarities for every solution contained in the population _pop
     * @param _pop the population
     */
    virtual void setSimilarities(eoPop < MOEOT > & _pop)
    {
        // compute distances between every individuals
        moeoDistanceMatrix < MOEOT , double > dMatrix (_pop.size(), distance);
        dMatrix(_pop);
        // compute similarities
        double sum;
        for (unsigned int i=0; i<_pop.size(); i++)
        {
            sum = 0.0;
            for (unsigned int j=0; j<_pop.size(); j++)
            {
                sum += sh(dMatrix[i][j]);
            }
            _pop[i].diversity(sum);
        }
    }


    /**
     * Sharing function
     * @param _dist the distance value
     */
    double sh(double _dist)
    {
        double result;
        if (_dist < nicheSize)
        {
            result = 1.0 - pow(_dist / nicheSize, alpha);
        }
        else
        {
            result = 0.0;
        }
        return result;
    }

};


#endif /*MOEOSHARINGDIVERSITYASSIGNMENT_H_*/