source: branches/GeneralizedQAP/HeuristicLab.Problems.GeneralizedQuadraticAssignment/3.3/SolutionCreators/GreedyRandomizedSolutionCreator.cs @ 12214

#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 System.Threading;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Data;
using HeuristicLab.Encodings.IntegerVectorEncoding;
using HeuristicLab.Parameters;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
using HeuristicLab.Random;

namespace HeuristicLab.Problems.GeneralizedQuadraticAssignment {
  [Item("GreedyRandomizedSolutionCreator", "Creates a solution according to the procedure 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.")]
  [StorableClass]
  public class GreedyRandomizedSolutionCreator : GQAPStochasticSolutionCreator,
    IEvaluatorAwareGQAPOperator {

    public IValueLookupParameter<IntValue> MaximumTriesParameter {
      get { return (IValueLookupParameter<IntValue>)Parameters["MaximumTries"]; }
    }
    public IValueLookupParameter<BoolValue> CreateMostFeasibleSolutionParameter {
      get { return (IValueLookupParameter<BoolValue>)Parameters["CreateMostFeasibleSolution"]; }
    }
    public IValueLookupParameter<IGQAPEvaluator> EvaluatorParameter {
      get { return (IValueLookupParameter<IGQAPEvaluator>)Parameters["Evaluator"]; }
    }

    [StorableConstructor]
    protected GreedyRandomizedSolutionCreator(bool deserializing) : base(deserializing) { }
    protected GreedyRandomizedSolutionCreator(GreedyRandomizedSolutionCreator original, Cloner cloner)
      : base(original, cloner) { }
    public GreedyRandomizedSolutionCreator()
      : base() {
      Parameters.Add(new ValueLookupParameter<IntValue>("MaximumTries", "The maximum number of tries to create a feasible solution after which an exception is thrown. If it is set to 0 or a negative value there will be an infinite number of attempts.", new IntValue(100000)));
      Parameters.Add(new ValueLookupParameter<BoolValue>("CreateMostFeasibleSolution", "If this is set to true the operator will always succeed, and outputs the solution with the least violation instead of throwing an exception.", new BoolValue(false)));
      Parameters.Add(new ValueLookupParameter<IGQAPEvaluator>("Evaluator", "The evaluator that is used to evaluate GQAP solutions."));
    }

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

    public static IntegerVector CreateSolution(IRandom random, DoubleArray demands, DoubleArray capacities,
      IGQAPEvaluator evaluator,
      int maximumTries, bool createMostFeasibleSolution, CancellationToken cancelToken) {
      int tries = 0;
      var assignment = new Dictionary<int, int>(demands.Length);
      DoubleArray slack = new DoubleArray(capacities.Length);
      double minViolation = double.MaxValue;
      Dictionary<int, int> bestAssignment = null;
      HashSet<int> CF = new HashSet<int>(), // set of chosen facilities / equipments
          T = new HashSet<int>(), // set of facilities / equpiments that can be assigned to the set of chosen locations (CL)
          CL = new HashSet<int>(), // set of chosen locations
          F = new HashSet<int>(Enumerable.Range(0, demands.Length)), // set of (initially) all facilities / equipments
          L = new HashSet<int>(Enumerable.Range(0, capacities.Length)); // set of (initially) all locations

      while (maximumTries <= 0 || tries < maximumTries) {
        cancelToken.ThrowIfCancellationRequested();

        assignment.Clear();
        for (int i = 0; i < capacities.Length; i++) slack[i] = capacities[i];
        CF.Clear();
        T.Clear();
        CL.Clear();
        F.Clear(); F.UnionWith(Enumerable.Range(0, demands.Length));
        L.Clear(); L.UnionWith(Enumerable.Range(0, capacities.Length));

        double threshold = 1.0;
        do {
          if (L.Any() && random.NextDouble() < threshold) {
            int l = L.SampleRandom(random);
            L.Remove(l);
            CL.Add(l);
            T = new HashSet<int>(WithDemandEqualOrLess(F, GetMaximumSlack(slack, CL), demands));
          }
          if (T.Any()) {
            int f = T.SampleRandom(random);
            T.Remove(f);
            F.Remove(f);
            CF.Add(f);
            int l = WithSlackGreaterOrEqual(CL, demands[f], slack).SampleRandom(random);
            assignment.Add(f, l);
            slack[l] -= demands[f];
            T = new HashSet<int>(WithDemandEqualOrLess(F, GetMaximumSlack(slack, CL), demands));
            threshold = 1.0 - (double)T.Count / Math.Max(F.Count, 1.0);
          }
        } while (T.Any() || L.Any());
        if (maximumTries > 0) tries++;
        if (!F.Any()) {
          bestAssignment = assignment;
          break;
        } else if (createMostFeasibleSolution) {
          // complete the solution and remember the one with least violation
          foreach (var l in L.ToArray()) {
            CL.Add(l);
            L.Remove(l);
          }
          while (F.Any()) {
            var f = F.MaxItems(x => demands[x]).SampleRandom(random);
            var l = CL.MaxItems(x => slack[x]).SampleRandom(random);
            F.Remove(f);
            assignment.Add(f, l);
            slack[l] -= demands[f];
          }
          double violation = evaluator.EvaluateOverbooking(slack, capacities);
          if (violation < minViolation) {
            bestAssignment = assignment;
            assignment = new Dictionary<int, int>(demands.Length);
            minViolation = violation;
          }
        }
      }

      if (bestAssignment == null || bestAssignment.Count != demands.Length) throw new InvalidOperationException(String.Format("No solution could be found in {0} tries.", maximumTries));

      return new IntegerVector(bestAssignment.OrderBy(x => x.Key).Select(x => x.Value).ToArray());
    }

    protected override IntegerVector CreateRandomSolution(IRandom random, DoubleArray demands, DoubleArray capacities) {
      return CreateSolution(random, demands, capacities,
        EvaluatorParameter.ActualValue,
        MaximumTriesParameter.ActualValue.Value,
        CreateMostFeasibleSolutionParameter.ActualValue.Value,
        CancellationToken);
    }

    private static IEnumerable<int> WithDemandEqualOrLess(IEnumerable<int> facilities, double maximum, DoubleArray demands) {
      foreach (int f in facilities) {
        if (demands[f] <= maximum) yield return f;
      }
    }

    private static double GetMaximumSlack(DoubleArray slack, HashSet<int> CL) {
      return slack.Select((val, idx) => new { idx, val }).Where(x => CL.Contains(x.idx)).Select(x => x.val).Max();
    }

    private static IEnumerable<int> WithSlackGreaterOrEqual(HashSet<int> locations, double minimum, DoubleArray slack) {
      foreach (int l in locations) {
        if (slack[l] >= minimum) yield return l;
      }
    }
  }
}