source: branches/DataAnalysis.PopulationDiversityAnalysis/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Symbolic/Analyzers/FineGrainedStructuralPopulationDiversityAnalyzer.cs @ 13806

#region License Information
/* HeuristicLab
 * Copyright (C) 2002-2010 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 HeuristicLab.Analysis;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Data;
using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
using HeuristicLab.Operators;
using HeuristicLab.Optimization;
using HeuristicLab.Optimization.Operators;
using HeuristicLab.Parameters;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
using HeuristicLab.Problems.DataAnalysis.Symbolic;
using System.Collections.Generic;
using HeuristicLab.Problems.DataAnalysis.Symbolic.Symbols;
using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding.Symbols;

namespace HeuristicLab.Problems.DataAnalysis.Regression.Symbolic.Analyzers {
  /// <summary>
  /// An operator that analyzes the population diversity using fine grained structural tree similarity estimation.
  /// </summary>
  [Item("FineGrainedStructuralPopulationDiversityAnalyzer", "An operator that analyzes the population diversity using fine grained structural tree similarity estimation.")]
  [StorableClass]
  public sealed class FineGrainedStructuralPopulationDiversityAnalyzer : SymbolicRegressionPopulationDiversityAnalyzer {

    #region Properties and Parameters

    private const string FunctionTreeGrammarParameterName = "FunctionTreeGrammar";
    private const string MinimumLevelDeltaParameterName = "MinimumLevelDelta";
    private const string MaximumLevelDeltaParameterName = "MaximumLevelDelta";
    private const string PreventMultipleComparisonContributionParameterName = "PreventMultipleComparisonContribution";
    private const string MaximumExpressionDepthParameterName = "MaxExpressionDepth";
    private const string LevelDifferenceCoefficientParameterName = "LevelDifferenceCoefficient";
    private const string AncestorIndexCoefficientParameterName = "AncestorIndexCoefficient";
    private const string ConstantValueCoefficientParameterName = "ConstantValueCoefficient";
    private const string VariableWeightCoefficientParameterName = "VariableWeightCoefficient";
    private const string TimeOffsetCoefficientParameterName = "TimeOffsetCoefficient";
    private const string VariableIndexCoefficientParameterName = "VariableIndexCoefficient";
    private const string AdditiveSimilarityCalculationParameterName = "AdditiveSimilarityCalculation";

    public IValueLookupParameter<GlobalSymbolicExpressionGrammar> FunctionTreeGrammarParameter {
      get { return (IValueLookupParameter<GlobalSymbolicExpressionGrammar>)Parameters[FunctionTreeGrammarParameterName]; }
    }
    public GlobalSymbolicExpressionGrammar FunctionTreeGrammar {
      get { return FunctionTreeGrammarParameter.ActualValue; }
    }
    public IValueLookupParameter<IntValue> MaximumExpressionDepthParameter {
      get { return (IValueLookupParameter<IntValue>)Parameters[MaximumExpressionDepthParameterName]; }
    }
    public int MaximumExpressionDepth {
      get { return MaximumExpressionDepthParameter.ActualValue.Value; }
    }

    public IValueParameter<IntValue> MinimumLevelDeltaParameter {
      get { return (IValueParameter<IntValue>)Parameters[MinimumLevelDeltaParameterName]; }
    }
    public int MinimumLevelDelta {
      get { return MinimumLevelDeltaParameter.Value.Value; }
    }
    public IValueParameter<IntValue> MaximumLevelDeltaParameter {
      get { return (IValueParameter<IntValue>)Parameters[MaximumLevelDeltaParameterName]; }
    }
    public int MaximumLevelDelta {
      get { return MaximumLevelDeltaParameter.Value.Value; }
    }
    public IValueParameter<BoolValue> PreventMultipleComparisonContributionParameter {
      get { return (IValueParameter<BoolValue>)Parameters[PreventMultipleComparisonContributionParameterName]; }
    }
    public bool PreventMultipleComparisonContribution {
      get { return PreventMultipleComparisonContributionParameter.Value.Value; }
    }

    public IValueParameter<DoubleValue> LevelDifferenceCoefficientParameter {
      get { return (IValueParameter<DoubleValue>)Parameters[LevelDifferenceCoefficientParameterName]; }
    }
    public double LevelDifferenceCoefficient {
      get { return LevelDifferenceCoefficientParameter.Value.Value; }
    }
    public IValueParameter<DoubleValue> AncestorIndexCoefficientParameter {
      get { return (IValueParameter<DoubleValue>)Parameters[AncestorIndexCoefficientParameterName]; }
    }
    public double AncestorIndexCoefficient {
      get { return AncestorIndexCoefficientParameter.Value.Value; }
    }
    public IValueParameter<DoubleValue> ConstantValueCoefficientParameter {
      get { return (IValueParameter<DoubleValue>)Parameters[ConstantValueCoefficientParameterName]; }
    }
    public double ConstantValueCoefficient {
      get { return ConstantValueCoefficientParameter.Value.Value; }
    }
    public IValueParameter<DoubleValue> VariableWeightCoefficientParameter {
      get { return (IValueParameter<DoubleValue>)Parameters[VariableWeightCoefficientParameterName]; }
    }
    public double VariableWeightCoefficient {
      get { return VariableWeightCoefficientParameter.Value.Value; }
    }
    public IValueParameter<DoubleValue> TimeOffsetCoefficientParameter {
      get { return (IValueParameter<DoubleValue>)Parameters[TimeOffsetCoefficientParameterName]; }
    }
    public double TimeOffsetCoefficientCoefficient {
      get { return TimeOffsetCoefficientParameter.Value.Value; }
    }
    public IValueParameter<DoubleValue> VariableIndexCoefficientParameter {
      get { return (IValueParameter<DoubleValue>)Parameters[VariableIndexCoefficientParameterName]; }
    }
    public double VariableIndexCoefficient {
      get { return VariableIndexCoefficientParameter.Value.Value; }
    }
    public IValueParameter<BoolValue> AdditiveSimilarityCalculationParameter {
      get { return (IValueParameter<BoolValue>)Parameters[AdditiveSimilarityCalculationParameterName]; }
    }
    public bool AdditiveSimilarityCalculation {
      get { return AdditiveSimilarityCalculationParameter.Value.Value; }
    }

    #endregion

    [StorableConstructor]
    private FineGrainedStructuralPopulationDiversityAnalyzer(bool deserializing) : base(deserializing) { }
    private FineGrainedStructuralPopulationDiversityAnalyzer(FineGrainedStructuralPopulationDiversityAnalyzer original, Cloner cloner) : base(original, cloner) { }
    public FineGrainedStructuralPopulationDiversityAnalyzer() : base() {
      Parameters.Add(new ValueLookupParameter<GlobalSymbolicExpressionGrammar>(FunctionTreeGrammarParameterName, "The grammar that is used for symbolic regression models."));
      Parameters.Add(new ValueLookupParameter<IntValue>(MaximumExpressionDepthParameterName, "Maximal depth of the analyzed symbolic expressions."));
      Parameters.Add(new ValueParameter<IntValue>(MinimumLevelDeltaParameterName, "Minimum value for the level delta of the analyzed genetic information items.", new IntValue(0)));
      Parameters.Add(new ValueParameter<IntValue>(MaximumLevelDeltaParameterName, "Maximum value for the level delta of the analyzed genetic information items.", new IntValue(int.MaxValue)));
      Parameters.Add(new ValueParameter<BoolValue>(PreventMultipleComparisonContributionParameterName, "Flag that denotes whether genetic information items are hindered from contributing to the similarity function multiple times.", new BoolValue(false)));
      Parameters.Add(new ValueParameter<DoubleValue>(LevelDifferenceCoefficientParameterName, "Weighting coefficient for level differences.", new DoubleValue(0.2)));
      Parameters.Add(new ValueParameter<DoubleValue>(AncestorIndexCoefficientParameterName, "Weighting coefficient for ancestor index differences.", new DoubleValue(0.2)));
      Parameters.Add(new ValueParameter<DoubleValue>(ConstantValueCoefficientParameterName, "Weighting coefficient for constant value differences.", new DoubleValue(0.2)));
      Parameters.Add(new ValueParameter<DoubleValue>(VariableWeightCoefficientParameterName, "Weighting coefficient for variable weight differences.", new DoubleValue(0.2)));
      Parameters.Add(new ValueParameter<DoubleValue>(TimeOffsetCoefficientParameterName, "Weighting coefficient for time lag differences.", new DoubleValue(0.2)));
      Parameters.Add(new ValueParameter<DoubleValue>(VariableIndexCoefficientParameterName, "Weighting coefficient for variable index differences.", new DoubleValue(0.2)));
      Parameters.Add(new ValueParameter<BoolValue>(AdditiveSimilarityCalculationParameterName, "Flag that denotes whether the similarity of genetic information items shall be calculated using additive calculation.", new BoolValue(true)));
    }

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

    protected override double[,] CalculateSimilarities(SymbolicExpressionTree[] solutions) {
      // collect information stored int the problem's parameters
      double variableWeightSigma = 0;
      double constantMinimumValue = 0;
      double constantMaximumValue = 0;
      int minimumTimeOffset = 0;
      int maximumTimeOffset = 0;
      foreach (Symbol symbol in FunctionTreeGrammar.Symbols) {
        Constant constant = symbol as Constant;
        if (constant !=null) {
          constantMinimumValue = constant.MinValue;
          constantMaximumValue = constant.MaxValue;
        }
        DataAnalysis.Symbolic.Symbols.Variable variable = symbol as DataAnalysis.Symbolic.Symbols.Variable;
        if (variable != null)
          variableWeightSigma = variable.WeightSigma;
        LaggedVariable laggedVariable = symbol as LaggedVariable;
        if (laggedVariable !=null) {
          minimumTimeOffset = laggedVariable.MinLag;
          maximumTimeOffset = laggedVariable.MaxLag;
        }
      }
      int n = solutions.Length;
      List<string> variableNames = new List<string>();
      foreach (StringValue variableName in ProblemData.InputVariables) {
        variableNames.Add(variableName.Value);
      }
      variableNames.Add(ProblemData.TargetVariable.Value);
      // collect genetic information item lists and store them also in dictionaries
      IList<GeneticInformationItem>[] geneticInformationItemsLists = new List<GeneticInformationItem>[n];
      IDictionary<string, IList<GeneticInformationItem>>[] geneticInformationItemsListsDictionaries = new IDictionary<string, IList<GeneticInformationItem>>[n];
      for (int i = 0; i < n; i++) {
        geneticInformationItemsLists[i] = GeneticInformationItem.GetGeneticInformationItems(solutions[i].Root, variableNames, MinimumLevelDelta, MaximumLevelDelta);
        geneticInformationItemsListsDictionaries[i] = GeneticInformationItem.GetDictionary(geneticInformationItemsLists[i]);
      }
      // calculate solution similarities
      double[,] similarities = new double[n, n];
      for (int i = 0; i < n; i++) {
        for (int j = 0; j < n; j++) {
          if (i == j)
            similarities[i, j] = 1;
          else {
            IList<GeneticInformationItem> solution1GeneticItems = geneticInformationItemsLists[i];
            IDictionary<string, IList<GeneticInformationItem>> solution2GeneticItemsDictionary = GeneticInformationItem.CopyDictionary(geneticInformationItemsListsDictionaries[j]);
            double similarity = 0;
            for (int k = 0; k < solution1GeneticItems.Count; k++) {
              double bestPendantSimilarity = 0;
              GeneticInformationItem item = solution1GeneticItems[k];
              GeneticInformationItem bestPendant = null;
              IList<GeneticInformationItem> geneticInformationItemsList = null;
              string key = GeneticInformationItem.GetKey(item);
              if (solution2GeneticItemsDictionary.ContainsKey(key)) {
                geneticInformationItemsList = solution2GeneticItemsDictionary[GeneticInformationItem.GetKey(item)];
                bestPendant = GeneticInformationItem.FindBestPendant(item, geneticInformationItemsList,
                  constantMinimumValue, constantMaximumValue, variableWeightSigma,
                  MaximumExpressionDepth, minimumTimeOffset, maximumTimeOffset,
                  LevelDifferenceCoefficient, AncestorIndexCoefficient, ConstantValueCoefficient, VariableWeightCoefficient,
                  TimeOffsetCoefficientCoefficient, VariableIndexCoefficient, AdditiveSimilarityCalculation,
                  out bestPendantSimilarity);
              }
              if (bestPendant != null) {
                similarity += bestPendantSimilarity;
                if (PreventMultipleComparisonContribution)
                  geneticInformationItemsList.Remove(bestPendant);
              }
            }
            similarities[i, j] = similarity / solution1GeneticItems.Count;
          }
        }
      }
      return similarities;
    }

  }

}