paradiseo/edo/src/edoEDASA.h

/*
The Evolving Distribution Objects framework (EDO) is a template-based,
ANSI-C++ evolutionary computation library which helps you to write your
own estimation of distribution algorithms.

This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.

This library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA

Copyright (C) 2010 Thales group
*/
/*
Authors:
    Johann Dréo <johann.dreo@thalesgroup.com>
    Caner Candan <caner.candan@thalesgroup.com>
*/

#ifndef _edoEDASA_h
#define _edoEDASA_h

#include <eo>
//#include <mo>

#include <utils/eoRNG.h>

#include "edoAlgo.h"
#include "edoEstimator.h"
#include "edoModifierMass.h"
#include "edoSampler.h"
#include "edoContinue.h"

//! edoEDASA< D >

template < typename D >
class edoEDASA : public edoAlgo< D >
{
public:
    //! Alias for the type EOT
    typedef typename D::EOType EOType;

    //! Alias for the atom type
    typedef typename EOType::AtomType AtomType;

    //! Alias for the fitness
    typedef typename EOType::Fitness Fitness;

public:

    //! edoEDASA constructor
    /*!
      All the boxes used by a EDASA need to be given.

      \param selector Population Selector
      \param estimator Distribution Estimator
      \param selectone SelectOne
      \param modifier Distribution Modifier
      \param sampler Distribution Sampler
      \param pop_continue Population Continuator
      \param distribution_continue Distribution Continuator
      \param evaluation Evaluation function.
      \param sa_continue Stopping criterion.
      \param cooling_schedule Cooling schedule, describes how the temperature is modified.
      \param initial_temperature The initial temperature.
      \param replacor Population replacor
    */
    edoEDASA (eoSelect< EOType > & selector,
	      edoEstimator< D > & estimator,
	      eoSelectOne< EOType > & selectone,
	      edoModifierMass< D > & modifier,
	      edoSampler< D > & sampler,
	      eoContinue< EOType > & pop_continue,
	      edoContinue< D > & distribution_continue,
	      eoEvalFunc < EOType > & evaluation,
	      moContinuator< moDummyNeighbor<EOType> > & sa_continue,
	      moCoolingSchedule<EOType> & cooling_schedule,
	      double initial_temperature,
	      eoReplacement< EOType > & replacor
	      )
	: _selector(selector),
	  _estimator(estimator),
	  _selectone(selectone),
	  _modifier(modifier),
	  _sampler(sampler),
	  _pop_continue(pop_continue),
	  _distribution_continue(distribution_continue),
	  _evaluation(evaluation),
	  _sa_continue(sa_continue),
	  _cooling_schedule(cooling_schedule),
	  _initial_temperature(initial_temperature),
	  _replacor(replacor)

    {}

    //! function that launches the EDASA algorithm.
    /*!
      As a moTS or a moHC, the EDASA can be used for HYBRIDATION in an evolutionary algorithm.

      \param pop A population to improve.
      \return TRUE.
    */
    void operator ()(eoPop< EOType > & pop)
    {
	assert(pop.size() > 0);

	double temperature = _initial_temperature;

	eoPop< EOType > current_pop;

	eoPop< EOType > selected_pop;


	//-------------------------------------------------------------
	// Estimating a first time the distribution parameter thanks
	// to population.
	//-------------------------------------------------------------

	D distrib = _estimator(pop);

	double size = distrib.size();
	assert(size > 0);

	//-------------------------------------------------------------


	do
	    {
		//-------------------------------------------------------------
		// (3) Selection of the best points in the population
		//-------------------------------------------------------------

		selected_pop.clear();

		_selector(pop, selected_pop);

		assert( selected_pop.size() > 0 );

		//-------------------------------------------------------------


		//-------------------------------------------------------------
		// (4) Estimation of the distribution parameters
		//-------------------------------------------------------------

		distrib = _estimator(selected_pop);

		//-------------------------------------------------------------


		// TODO: utiliser selected_pop ou pop ???

		assert(selected_pop.size() > 0);


		//-------------------------------------------------------------
		// Init of a variable contening a point with the bestest fitnesses
		//-------------------------------------------------------------

		EOType current_solution = _selectone(selected_pop);

		//-------------------------------------------------------------


		//-------------------------------------------------------------
		// Fit the current solution with the distribution parameters (bounds)
		//-------------------------------------------------------------

		// FIXME: si besoin de modifier la dispersion de la distribution
		// _modifier_dispersion(distribution, selected_pop);
		_modifier(distrib, current_solution);

		//-------------------------------------------------------------


		//-------------------------------------------------------------
		// Evaluating a first time the current solution
		//-------------------------------------------------------------

		_evaluation( current_solution );

		//-------------------------------------------------------------


		//-------------------------------------------------------------
		// Building of the sampler in current_pop
		//-------------------------------------------------------------

		_sa_continue.init( current_solution );

		current_pop.clear();

		do
		    {
			EOType candidate_solution = _sampler(distrib);
			_evaluation( candidate_solution );

			// TODO: verifier le critere d'acceptation
			if ( candidate_solution.fitness() < current_solution.fitness() ||
			     rng.uniform() < exp( ::fabs(candidate_solution.fitness() - current_solution.fitness()) / temperature ) )
			    {
				current_pop.push_back(candidate_solution);
				current_solution = candidate_solution;
			    }
		    }
 		while ( _sa_continue( current_solution ) );

		//-------------------------------------------------------------


		_replacor(pop, current_pop); // copy current_pop in pop

		pop.sort();

		if ( ! _cooling_schedule( temperature ) ){ eo::log << eo::debug << "_cooling_schedule" << std::endl; break; }

		if ( ! _distribution_continue( distrib ) ){ eo::log << eo::debug << "_distribution_continue" << std::endl; break; }

		if ( ! _pop_continue( pop ) ){ eo::log << eo::debug << "_pop_continue" << std::endl; break; }

	    }
	while ( 1 );
    }

private:

    //! A EOType selector
    eoSelect < EOType > & _selector;

    //! A EOType estimator. It is going to estimate distribution parameters.
    edoEstimator< D > & _estimator;

    //! SelectOne
    eoSelectOne< EOType > & _selectone;

    //! A D modifier
    edoModifierMass< D > & _modifier;

    //! A D sampler
    edoSampler< D > & _sampler;

    //! A EOType population continuator
    eoContinue < EOType > & _pop_continue;

    //! A D continuator
    edoContinue < D > & _distribution_continue;

    //! A full evaluation function.
    eoEvalFunc < EOType > & _evaluation;

    //! Stopping criterion before temperature update
    moContinuator< moDummyNeighbor<EOType> > & _sa_continue;

    //! The cooling schedule
    moCoolingSchedule<EOType> & _cooling_schedule;

    //! Initial temperature
    double  _initial_temperature;

    //! A EOType replacor
    eoReplacement < EOType > & _replacor;
};

#endif // !_edoEDASA_h