From adde4c48987db687b6272087a511668f2d2dbb32 Mon Sep 17 00:00:00 2001 From: verel Date: Tue, 4 May 2010 16:37:22 +0000 Subject: [PATCH] Add the autocorrelation sampling git-svn-id: svn://scm.gforge.inria.fr/svnroot/paradiseo@1777 331e1502-861f-0410-8da2-ba01fb791d7f --- .../src/continuator/moFitnessStat.h | 52 ++--- trunk/paradiseo-mo/src/mo.h | 1 + .../src/sampling/moAutocorrelationSampling.h | 88 ++++++++ trunk/paradiseo-mo/src/sampling/moSampling.h | 6 + .../tutorial/Lesson6/CMakeLists.txt | 2 + .../tutorial/Lesson6/autocorrelation.cpp | 209 ++++++++++++++++++ .../tutorial/Lesson6/sampling.cpp | 3 +- .../Lesson6/testRandomNeutralWalk.cpp | 2 +- .../tutorial/Lesson6/testRandomWalk.cpp | 2 +- 9 files changed, 335 insertions(+), 30 deletions(-) create mode 100644 trunk/paradiseo-mo/src/sampling/moAutocorrelationSampling.h create mode 100644 trunk/paradiseo-mo/tutorial/Lesson6/autocorrelation.cpp diff --git a/trunk/paradiseo-mo/src/continuator/moFitnessStat.h b/trunk/paradiseo-mo/src/continuator/moFitnessStat.h index 718a4e1c4..d3919ffab 100644 --- a/trunk/paradiseo-mo/src/continuator/moFitnessStat.h +++ b/trunk/paradiseo-mo/src/continuator/moFitnessStat.h @@ -42,35 +42,35 @@ * that need to be calculated over the solution * It is a moStatBase AND an eoValueParam so it can be used in Monitors. */ -template -class moFitnessStat : public moStat +template +class moFitnessStat : public moStat { public : - typedef T Fitness; - using moStat< EOT, Fitness >::value; + typedef typename EOT::Fitness Fitness; + using moStat< EOT, Fitness >::value; - /** - * Default Constructor - * @param _description a description of the stat - */ - moFitnessStat(std::string _description = "fitness"): - moStat(Fitness(), _description) {} - - /** - * store fitness value - * @param _sol the corresponding solution - */ - virtual void operator()(EOT & _sol) - { - value() = _sol.fitness(); - } - - /** - * @return the name of the class - */ - virtual std::string className(void) const { - return "moFitnessStat"; - } + /** + * Default Constructor + * @param _description a description of the stat + */ + moFitnessStat(std::string _description = "fitness"): + moStat(Fitness(), _description) {} + + /** + * store fitness value + * @param _sol the corresponding solution + */ + virtual void operator()(EOT & _sol) + { + value() = _sol.fitness(); + } + + /** + * @return the name of the class + */ + virtual std::string className(void) const { + return "moFitnessStat"; + } }; #endif diff --git a/trunk/paradiseo-mo/src/mo.h b/trunk/paradiseo-mo/src/mo.h index fb51540f1..9f1553444 100755 --- a/trunk/paradiseo-mo/src/mo.h +++ b/trunk/paradiseo-mo/src/mo.h @@ -132,6 +132,7 @@ #include #include +#include #include #include diff --git a/trunk/paradiseo-mo/src/sampling/moAutocorrelationSampling.h b/trunk/paradiseo-mo/src/sampling/moAutocorrelationSampling.h new file mode 100644 index 000000000..e667a0cd5 --- /dev/null +++ b/trunk/paradiseo-mo/src/sampling/moAutocorrelationSampling.h @@ -0,0 +1,88 @@ +/* + + Copyright (C) DOLPHIN Project-Team, INRIA Lille - Nord Europe, 2006-2010 + + Sebastien Verel, Arnaud Liefooghe, Jeremie Humeau + + This software is governed by the CeCILL license under French law and + abiding by the rules of distribution of free software. You can use, + modify and/ or redistribute the software under the terms of the CeCILL + license as circulated by CEA, CNRS and INRIA at the following URL + "http://www.cecill.info". + + As a counterpart to the access to the source code and rights to copy, + modify and redistribute granted by the license, users are provided only + with a limited warranty and the software's author, the holder of the + economic rights, and the successive licensors have only limited liability. + + In this respect, the user's attention is drawn to the risks associated + with loading, using, modifying and/or developing or reproducing the + software by the user in light of its specific status of free software, + that may mean that it is complicated to manipulate, and that also + therefore means that it is reserved for developers and experienced + professionals having in-depth computer knowledge. Users are therefore + encouraged to load and test the software's suitability as regards their + requirements in conditions enabling the security of their systems and/or + data to be ensured and, more generally, to use and operate it in the + same conditions as regards security. + The fact that you are presently reading this means that you have had + knowledge of the CeCILL license and that you accept its terms. + + ParadisEO WebSite : http://paradiseo.gforge.inria.fr + Contact: paradiseo-help@lists.gforge.inria.fr +*/ + +#ifndef moAutocorrelationSampling_h +#define moAutocorrelationSampling_h + +#include +#include +#include +#include +#include +#include + +/** + * To compute the autocorrelation function: + * Perform a random walk based on the neighborhood, + * The fitness values of solutions are collected during the random walk + * The autocorrelation can be computed from the serie of fitness values + * + */ +template +class moAutocorrelationSampling : public moSampling +{ +public: + typedef typename Neighbor::EOT EOT ; + + using moSampling::localSearch; + + /** + * Default Constructor + * @param _init initialisation method of the solution + * @param _neighborhood neighborhood giving neighbor in random order + * @param _nbStep Number of steps of the random walk + */ + moAutocorrelationSampling(eoInit & _init, + moNeighborhood & _neighborhood, + eoEvalFunc& _fullEval, moEval& _eval, + unsigned int _nbStep) : + moSampling(_init, * new moRandomWalk(_neighborhood, _fullEval, _eval, _nbStep), fitnessStat) + { + } + + /** + * default destructor + */ + ~moAutocorrelationSampling() { + // delete the pointer on the local search which has been constructed in the constructor + delete &localSearch; + } + +protected: + moFitnessStat fitnessStat; + +}; + + +#endif diff --git a/trunk/paradiseo-mo/src/sampling/moSampling.h b/trunk/paradiseo-mo/src/sampling/moSampling.h index 2e9c70174..3c99d82ac 100644 --- a/trunk/paradiseo-mo/src/sampling/moSampling.h +++ b/trunk/paradiseo-mo/src/sampling/moSampling.h @@ -41,6 +41,7 @@ #include #include #include +#include #include /** @@ -69,10 +70,15 @@ public: add(_stat); } + /** + * default destructor + */ ~moSampling() { + // delete all monitors for(unsigned i = 0; i < monitorVec.size(); i++) delete monitorVec[i]; + // delete the checkpoint delete checkpoint ; } diff --git a/trunk/paradiseo-mo/tutorial/Lesson6/CMakeLists.txt b/trunk/paradiseo-mo/tutorial/Lesson6/CMakeLists.txt index 83a8b073e..f92dd8712 100644 --- a/trunk/paradiseo-mo/tutorial/Lesson6/CMakeLists.txt +++ b/trunk/paradiseo-mo/tutorial/Lesson6/CMakeLists.txt @@ -10,8 +10,10 @@ ADD_EXECUTABLE(testRandomWalk testRandomWalk.cpp) ADD_EXECUTABLE(testMetropolisHasting testMetropolisHasting.cpp) ADD_EXECUTABLE(testRandomNeutralWalk testRandomNeutralWalk.cpp) ADD_EXECUTABLE(sampling sampling.cpp) +ADD_EXECUTABLE(autocorrelation autocorrelation.cpp) TARGET_LINK_LIBRARIES(testRandomWalk eoutils ga eo) TARGET_LINK_LIBRARIES(testMetropolisHasting eoutils ga eo) TARGET_LINK_LIBRARIES(testRandomNeutralWalk eoutils ga eo) +TARGET_LINK_LIBRARIES(autocorrelation eoutils ga eo) TARGET_LINK_LIBRARIES(sampling eoutils ga eo) diff --git a/trunk/paradiseo-mo/tutorial/Lesson6/autocorrelation.cpp b/trunk/paradiseo-mo/tutorial/Lesson6/autocorrelation.cpp new file mode 100644 index 000000000..dfb608381 --- /dev/null +++ b/trunk/paradiseo-mo/tutorial/Lesson6/autocorrelation.cpp @@ -0,0 +1,209 @@ +//----------------------------------------------------------------------------- +/** sampling.cpp + * + * SV - 03/05/10 + * + */ +//----------------------------------------------------------------------------- + +// standard includes +#define HAVE_SSTREAM + +#include // runtime_error +#include // cout +#include // ostrstream, istrstream +#include +#include + +// the general include for eo +#include + +// declaration of the namespace +using namespace std; + +//----------------------------------------------------------------------------- +// representation of solutions, and neighbors +#include // bit string : see also EO tutorial lesson 1: FirstBitGA.cpp +#include // neighbor of bit string + +//----------------------------------------------------------------------------- +// fitness function, and evaluation of neighbors +#include +#include +#include + +//----------------------------------------------------------------------------- +// neighborhood description +#include // visit one random neighbor possibly the same one several times + +//----------------------------------------------------------------------------- +// the sampling class +#include + +// Declaration of types +//----------------------------------------------------------------------------- +// Indi is the typedef of the solution type like in paradisEO-eo +typedef eoBit Indi; // bit string with unsigned fitness type +// Neighbor is the typedef of the neighbor type, +// Neighbor = How to compute the neighbor from the solution + information on it (i.e. fitness) +// all classes from paradisEO-mo use this template type +typedef moBitNeighbor Neighbor ; // bit string neighbor with unsigned fitness type + + +void main_function(int argc, char **argv) +{ + /* ========================================================= + * + * Parameters + * + * ========================================================= */ + // more information on the input parameters: see EO tutorial lesson 3 + // but don't care at first it just read the parameters of the bit string size and the random seed. + + // First define a parser from the command-line arguments + eoParser parser(argc, argv); + + // For each parameter, define Parameter, read it through the parser, + // and assign the value to the variable + + // random seed parameter + eoValueParam seedParam(time(0), "seed", "Random number seed", 'S'); + parser.processParam( seedParam ); + unsigned seed = seedParam.value(); + + // length of the bit string + eoValueParam vecSizeParam(20, "vecSize", "Genotype size", 'V'); + parser.processParam( vecSizeParam, "Representation" ); + unsigned vecSize = vecSizeParam.value(); + + // the number of steps of the random walk + eoValueParam stepParam(100, "nbStep", "Number of steps of the random walk", 'n'); + parser.processParam( stepParam, "Representation" ); + unsigned nbStep = stepParam.value(); + + // the name of the output file + string str_out = "out.dat"; // default value + eoValueParam outParam(str_out.c_str(), "out", "Output file of the sampling", 'o'); + parser.processParam(outParam, "Persistence" ); + + // the name of the "status" file where all actual parameter values will be saved + string str_status = parser.ProgramName() + ".status"; // default value + eoValueParam statusParam(str_status.c_str(), "status", "Status file"); + parser.processParam( statusParam, "Persistence" ); + + // do the following AFTER ALL PARAMETERS HAVE BEEN PROCESSED + // i.e. in case you need parameters somewhere else, postpone these + if (parser.userNeedsHelp()) { + parser.printHelp(cout); + exit(1); + } + if (statusParam.value() != "") { + ofstream os(statusParam.value().c_str()); + os << parser;// and you can use that file as parameter file + } + + /* ========================================================= + * + * Random seed + * + * ========================================================= */ + + // reproducible random seed: if you don't change SEED above, + // you'll aways get the same result, NOT a random run + // more information: see EO tutorial lesson 1 (FirstBitGA.cpp) + rng.reseed(seed); + + /* ========================================================= + * + * Initialization of the solution + * + * ========================================================= */ + + // a Indi random initializer: each bit is random + // more information: see EO tutorial lesson 1 (FirstBitGA.cpp) + eoUniformGenerator uGen; + eoInitFixedLength random(vecSize, uGen); + + /* ========================================================= + * + * Eval fitness function (full evaluation) + * + * ========================================================= */ + + // the fitness function is just the number of 1 in the bit string + oneMaxEval fullEval; + + /* ========================================================= + * + * evaluation of a neighbor solution + * + * ========================================================= */ + + // Use it if there is no incremental evaluation: a neighbor is evaluated by the full evaluation of a solution + // moFullEvalByModif neighborEval(fullEval); + + // Incremental evaluation of the neighbor: fitness is modified by +/- 1 + moOneMaxIncrEval neighborEval; + + /* ========================================================= + * + * the neighborhood of a solution + * + * ========================================================= */ + + // Exploration of the neighborhood in random order + // at each step one bit is randomly generated + moRndWithReplNeighborhood neighborhood(vecSize); + + /* ========================================================= + * + * The sampling of the search space + * + * ========================================================= */ + + // sampling object : + // - random initialization + // - local search to sample the search space + // - one statistic to compute + moAutocorrelationSampling sampling(random, neighborhood, fullEval, neighborEval, nbStep); + + /* ========================================================= + * + * execute the sampling + * + * ========================================================= */ + + sampling(); + + /* ========================================================= + * + * export the sampling + * + * ========================================================= */ + + // to export the statistics into file + sampling.fileExport(str_out); + + // to get the values of statistics + // so, you can compute some statistics in c++ from the data + const std::vector & fitnessValues = sampling.getVector(0); + + std::cout << "First values:" << std::endl; + std::cout << "Fitness " << fitnessValues[0] << std::endl; + + std::cout << "Last values:" << std::endl; + std::cout << "Fitness " << fitnessValues[fitnessValues.size() - 1] << std::endl; +} + +// A main that catches the exceptions + +int main(int argc, char **argv) +{ + try { + main_function(argc, argv); + } + catch (exception& e) { + cout << "Exception: " << e.what() << '\n'; + } + return 1; +} diff --git a/trunk/paradiseo-mo/tutorial/Lesson6/sampling.cpp b/trunk/paradiseo-mo/tutorial/Lesson6/sampling.cpp index 3c581a8dc..e501a88bd 100644 --- a/trunk/paradiseo-mo/tutorial/Lesson6/sampling.cpp +++ b/trunk/paradiseo-mo/tutorial/Lesson6/sampling.cpp @@ -50,7 +50,6 @@ using namespace std; // the sampling class #include - // Declaration of types //----------------------------------------------------------------------------- // Indi is the typedef of the solution type like in paradisEO-eo @@ -181,7 +180,7 @@ void main_function(int argc, char **argv) * ========================================================= */ // fitness of the solution at each step - moFitnessStat fStat; + moFitnessStat fStat; // Hamming distance to the global optimum eoHammingDistance distance; // Hamming distance diff --git a/trunk/paradiseo-mo/tutorial/Lesson6/testRandomNeutralWalk.cpp b/trunk/paradiseo-mo/tutorial/Lesson6/testRandomNeutralWalk.cpp index 7d76495d4..6c4aac11f 100644 --- a/trunk/paradiseo-mo/tutorial/Lesson6/testRandomNeutralWalk.cpp +++ b/trunk/paradiseo-mo/tutorial/Lesson6/testRandomNeutralWalk.cpp @@ -197,7 +197,7 @@ void main_function(int argc, char **argv) moCheckpoint checkpoint(continuator); - moFitnessStat fStat; + moFitnessStat fStat; eoHammingDistance distance; moDistanceStat distStat(distance, solution); // distance from the intial solution diff --git a/trunk/paradiseo-mo/tutorial/Lesson6/testRandomWalk.cpp b/trunk/paradiseo-mo/tutorial/Lesson6/testRandomWalk.cpp index 5eab2c1d1..546a29f99 100644 --- a/trunk/paradiseo-mo/tutorial/Lesson6/testRandomWalk.cpp +++ b/trunk/paradiseo-mo/tutorial/Lesson6/testRandomWalk.cpp @@ -162,7 +162,7 @@ void main_function(int argc, char **argv) moCheckpoint checkpoint(continuator); - moFitnessStat fStat; + moFitnessStat fStat; eoHammingDistance distance; Indi bestSolution(vecSize, true); moDistanceStat distStat(distance, bestSolution);