source: branches/GeneralizedQAP/HeuristicLab.Problems.GeneralizedQuadraticAssignment/3.3/Operators/ApproximateLocalSearch.cs @ 7407

#region License Information
/* HeuristicLab
 * Copyright (C) 2002-2012 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
 *
 * This file is part of HeuristicLab.
 *
 * HeuristicLab is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * HeuristicLab 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 General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
 */
#endregion

using System;
using System.Collections.Generic;
using System.Linq;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Data;
using HeuristicLab.Encodings.IntegerVectorEncoding;
using HeuristicLab.Operators;
using HeuristicLab.Optimization;
using HeuristicLab.Parameters;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;

namespace HeuristicLab.Problems.GeneralizedQuadraticAssignment.Operators {
  [Item("ApproximateLocalSearch", @"The approximate local search is described in Mateus, G., Resende, M., and Silva, R. 2011. GRASP with path-relinking for the generalized quadratic assignment problem. Journal of Heuristics 17, Springer Netherlands, pp. 527-565.

The implementation differs slightly from Mateus et al. in that the maximumIterations parameter defines a cap on the number of steps that the local search can perform. While the maxSampleSize parameter corresponds to the maxItr parameter defined by Mateus et al.")]
  [StorableClass]
  public class ApproximateLocalSearch : SingleSuccessorOperator, IGQAPLocalImprovementOperator, IStochasticOperator {
    public IProblem Problem { get; set; }
    public Type ProblemType {
      get { return typeof(GeneralizedQuadraticAssignmentProblem); }
    }

    public ILookupParameter<IRandom> RandomParameter {
      get { return (ILookupParameter<IRandom>)Parameters["Random"]; }
    }
    public IValueLookupParameter<IntValue> MaximumIterationsParameter {
      get { return (IValueLookupParameter<IntValue>)Parameters["MaximumIterations"]; }
    }
    public ILookupParameter<IntValue> EvaluatedSolutionsParameter {
      get { return (ILookupParameter<IntValue>)Parameters["EvaluatedSolutions"]; }
    }
    public ILookupParameter<DoubleValue> QualityParameter {
      get { return (ILookupParameter<DoubleValue>)Parameters["Quality"]; }
    }
    public ILookupParameter<DoubleValue> InfeasibilityParameter {
      get { return (ILookupParameter<DoubleValue>)Parameters["Infeasibility"]; }
    }
    public ILookupParameter<ResultCollection> ResultsParameter {
      get { return (ILookupParameter<ResultCollection>)Parameters["Results"]; }
    }
    public IValueLookupParameter<IntValue> MaximumCandidateListSizeParameter {
      get { return (IValueLookupParameter<IntValue>)Parameters["MaximumCandidateListSize"]; }
    }
    public IValueLookupParameter<IntValue> MaximumSampleSizeParameter {
      get { return (IValueLookupParameter<IntValue>)Parameters["MaximumSampleSize"]; }
    }
    public ILookupParameter<IntegerVector> AssignmentParameter {
      get { return (ILookupParameter<IntegerVector>)Parameters["Assignment"]; }
    }
    public ILookupParameter<DoubleMatrix> WeightsParameter {
      get { return (ILookupParameter<DoubleMatrix>)Parameters["Weights"]; }
    }
    public ILookupParameter<DoubleMatrix> DistancesParameter {
      get { return (ILookupParameter<DoubleMatrix>)Parameters["Distances"]; }
    }
    public ILookupParameter<DoubleMatrix> InstallationCostsParameter {
      get { return (ILookupParameter<DoubleMatrix>)Parameters["InstallationCosts"]; }
    }
    public ILookupParameter<DoubleValue> TransportationCostsParameter {
      get { return (ILookupParameter<DoubleValue>)Parameters["TransportationCosts"]; }
    }
    public ILookupParameter<DoubleArray> DemandsParameter {
      get { return (ILookupParameter<DoubleArray>)Parameters["Demands"]; }
    }
    public ILookupParameter<DoubleArray> CapacitiesParameter {
      get { return (ILookupParameter<DoubleArray>)Parameters["Capacities"]; }
    }


    [StorableConstructor]
    protected ApproximateLocalSearch(bool deserializing) : base(deserializing) { }
    protected ApproximateLocalSearch(ApproximateLocalSearch original, Cloner cloner) : base(original, cloner) { }
    public ApproximateLocalSearch()
      : base() {
      Parameters.Add(new LookupParameter<IRandom>("Random", "The random number generator to use."));
      Parameters.Add(new ValueLookupParameter<IntValue>("MaximumIterations", "The maximum number of iterations that should be performed."));
      Parameters.Add(new LookupParameter<IntValue>("EvaluatedSolutions", "The number of evaluated solution equivalents."));
      Parameters.Add(new LookupParameter<DoubleValue>("Quality", "The solution quality."));
      Parameters.Add(new LookupParameter<DoubleValue>("Infeasibility", "The infeasibility describes the sum of the overbooked capacities."));
      Parameters.Add(new LookupParameter<ResultCollection>("Results", "The result collection that stores the results."));
      Parameters.Add(new ValueLookupParameter<IntValue>("MaximumCandidateListSize", "The maximum number of candidates that should be found in each step.", new IntValue(10)));
      Parameters.Add(new ValueLookupParameter<IntValue>("MaximumSampleSize", "The maximum number of candidates that should be sampled in each step.", new IntValue(100)));
      Parameters.Add(new LookupParameter<IntegerVector>("Assignment", "The equipment-location assignment vector."));
      Parameters.Add(new LookupParameter<DoubleMatrix>("Weights", "The weights matrix describes the flows between the equipments."));
      Parameters.Add(new LookupParameter<DoubleMatrix>("Distances", "The distances matrix describes the distances between the locations at which the equipment can be installed."));
      Parameters.Add(new LookupParameter<DoubleMatrix>("InstallationCosts", "The installation costs matrix describes the installation costs of installing equipment i at location j."));
      Parameters.Add(new LookupParameter<DoubleValue>("TransportationCosts", "The transportation cost represents the flow-unit per distance-unit cost factor. This value can also be set to 1 if these costs are factored into the weights matrix already."));
      Parameters.Add(new LookupParameter<DoubleArray>("Demands", "The demands vector describes the space requirements of the equipments."));
      Parameters.Add(new LookupParameter<DoubleArray>("Capacities", "The capacities vector describes the available space at the locations."));
    }

    public override IDeepCloneable Clone(Cloner cloner) {
      return new ApproximateLocalSearch(this, cloner);
    }

    /// <summary>
    /// The implementation differs slightly from Mateus et al. in that the maximumIterations parameter defines a cap
    /// on the number of steps that the local search can perform. While the maxSampleSize parameter corresponds to
    /// the maxItr parameter defined by Mateus et al.
    /// </summary>
    /// <param name="random">The random number generator to use.</param>
    /// <param name="assignment">The equipment-location assignment vector.</param>
    /// <param name="quality">The solution quality.</param>
    /// <param name="infeasibility">The infeasibility describes the sum of the overbooked capacities.</param>
    /// <param name="maxCLS">The maximum number of candidates that should be found in each step.</param>
    /// <param name="maxSampleSize">The maximum number of candidates that should be sampled in each step.</param>
    /// <param name="maximumIterations">The maximum number of iterations that should be performed.</param>
    /// <param name="demands">The demands vector describes the space requirements of the equipments.</param>
    /// <param name="capacities">The capacities vector describes the available space at the locations.</param>
    /// <param name="weights">The weights matrix describes the flows between the equipments.</param>
    /// <param name="distances">The distances matrix describes the distances between the locations at which the equipment can be installed.</param>
    /// <param name="installationCosts">The installation costs matrix describes the installation costs of installing equipment i at location j</param>
    /// <param name="transportationCosts">The transportation cost represents the flow-unit per distance-unit cost factor. This value can also be set to 1 if these costs are factored into the weights matrix already.</param>
    public static void Apply(IRandom random, IntegerVector assignment, DoubleValue quality, DoubleValue infeasibility,
                             IntValue maxCLS, IntValue maxSampleSize, IntValue maximumIterations,
                             DoubleArray demands, DoubleArray capacities,
                             DoubleMatrix weights, DoubleMatrix distances,
                             DoubleMatrix installationCosts,
                             DoubleValue transportationCosts) {
      for (int i = 0; i < maximumIterations.Value; i++) {
        int count = 0;
        var CLS = new List<Tuple<NMove, DoubleValue, DoubleValue>>();
        double sum = 0.0;
        do {
          var move = StochasticNMoveSingleMoveGenerator.Generate(random, assignment, 2, capacities);
          DoubleValue moveInfeasibility;
          var moveQuality = GQAPNMoveEvaluator.Evaluate(move, assignment, quality,
            demands, capacities, weights, distances,
            installationCosts, transportationCosts, out moveInfeasibility);
          if (moveInfeasibility.Value <= 0.0 && moveQuality.Value < quality.Value) {
            CLS.Add(Tuple.Create(move, moveQuality, moveInfeasibility));
            sum += 1.0 / moveQuality.Value;
          }
          count++;
        } while (CLS.Count < maxCLS.Value && count < maxSampleSize.Value);
        if (CLS.Count == 0) return; // END
        else {
          var ball = random.NextDouble() * sum;
          var selected = CLS.Last();
          foreach (var candidate in CLS) {
            ball -= 1.0 / candidate.Item2.Value;
            if (ball <= 0.0) {
              selected = candidate;
              break;
            }
          }
          NMoveMaker.Apply(assignment, selected.Item1);
          quality.Value = selected.Item2.Value;
          infeasibility.Value = selected.Item3.Value;
        }
      }
    }

    public override IOperation Apply() {
      Apply(RandomParameter.ActualValue,
        AssignmentParameter.ActualValue,
        QualityParameter.ActualValue,
        InfeasibilityParameter.ActualValue,
        MaximumCandidateListSizeParameter.ActualValue,
        MaximumSampleSizeParameter.ActualValue,
        MaximumIterationsParameter.ActualValue,
        DemandsParameter.ActualValue, CapacitiesParameter.ActualValue,
        WeightsParameter.ActualValue, DistancesParameter.ActualValue,
        InstallationCostsParameter.ActualValue,
        TransportationCostsParameter.ActualValue);
      return base.Apply();
    }
  }
}