source: branches/HeuristicLab.EvolutionaryTracking/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/SymbolicDataAnalysisExpressionTreeSimilarity.cs @ 9317

#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.Linq;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Data;
using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
using HeuristicLab.Operators;
using HeuristicLab.Parameters;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;

namespace HeuristicLab.Problems.DataAnalysis.Symbolic {
  [StorableClass]
  public class SymbolicDataAnalysisExpressionTreeSimilarityCalculator : SingleSuccessorOperator {
    private const string SymbolicExpressionTreeParameterName = "SymbolicExpressionTree";
    private const string CurrentSymbolicExpressionTreeParameterName = "CurrentSymbolicExpressionTree";
    private const string SimilarityValuesParmeterName = "Similarity";
    // comparer parameters
    private const string MatchVariablesParameterName = "MatchVariableNames";
    private const string MatchVariableWeightsParameterName = "MatchVariableWeights";
    private const string MatchConstantValuesParameterName = "MatchConstantValues";


    public IScopeTreeLookupParameter<ISymbolicExpressionTree> SymbolicExpressionTreeParameter {
      get { return (IScopeTreeLookupParameter<ISymbolicExpressionTree>)Parameters[SymbolicExpressionTreeParameterName]; }
    }
    public IValueParameter<ISymbolicExpressionTree> CurrentSymbolicExpressionTreeParameter {
      get { return (IValueParameter<ISymbolicExpressionTree>)Parameters[CurrentSymbolicExpressionTreeParameterName]; }
    }
    public ILookupParameter<BoolValue> MatchVariableNamesParameter {
      get { return (ILookupParameter<BoolValue>)Parameters[MatchVariablesParameterName]; }
    }
    public ILookupParameter<BoolValue> MatchVariableWeightsParameter {
      get { return (ILookupParameter<BoolValue>)Parameters[MatchVariableWeightsParameterName]; }
    }
    public ILookupParameter<BoolValue> MatchConstantValuesParameter {
      get { return (ILookupParameter<BoolValue>)Parameters[MatchConstantValuesParameterName]; }
    }
    public ILookupParameter<DoubleValue> SimilarityParameter {
      get { return (ILookupParameter<DoubleValue>)Parameters[SimilarityValuesParmeterName]; }
    }

    public ISymbolicExpressionTree CurrentSymbolicExpressionTree {
      get { return CurrentSymbolicExpressionTreeParameter.Value; }
      set { CurrentSymbolicExpressionTreeParameter.Value = value; }
    }

    public SymbolicExpressionTreeNodeSimilarityComparer SimilarityComparer { get; set; }

    protected SymbolicDataAnalysisExpressionTreeSimilarityCalculator(SymbolicDataAnalysisExpressionTreeSimilarityCalculator original, Cloner cloner) : base(original, cloner) { }
    public override IDeepCloneable Clone(Cloner cloner) { return new SymbolicDataAnalysisExpressionTreeSimilarityCalculator(this, cloner); }
    [StorableConstructor]
    protected SymbolicDataAnalysisExpressionTreeSimilarityCalculator(bool deserializing) : base(deserializing) { }

    public SymbolicDataAnalysisExpressionTreeSimilarityCalculator()
      : base() {
      Parameters.Add(new ScopeTreeLookupParameter<ISymbolicExpressionTree>(SymbolicExpressionTreeParameterName, "The symbolic expression trees to analyze."));
      Parameters.Add(new ValueParameter<ISymbolicExpressionTree>(CurrentSymbolicExpressionTreeParameterName, ""));
      Parameters.Add(new LookupParameter<BoolValue>(MatchVariablesParameterName, "Specify if the symbolic expression tree comparer should match variable names."));
      Parameters.Add(new LookupParameter<BoolValue>(MatchVariableWeightsParameterName, "Specify if the symbolic expression tree comparer should match variable weights."));
      Parameters.Add(new LookupParameter<BoolValue>(MatchConstantValuesParameterName, "Specify if the symbolic expression tree comparer should match constant values."));
      Parameters.Add(new LookupParameter<DoubleValue>(SimilarityValuesParmeterName, ""));
    }

    public override IOperation Apply() {
      var trees = SymbolicExpressionTreeParameter.ActualValue;

      bool found = false;
      double similarity = 0.0;
      var current = CurrentSymbolicExpressionTree;

      foreach (var tree in trees) {
        if (tree == current) { found = true; }
        if (!found) continue;

        similarity += SymbolicDataAnalysisExpressionTreeSimilarity.MaxCommonSubtreeSimilarity(current, tree, SimilarityComparer);
      }
      lock (SimilarityParameter.ActualValue) {
        SimilarityParameter.ActualValue.Value += similarity;
      }
      return base.Apply();
    }
  }



  public static class SymbolicDataAnalysisExpressionTreeSimilarity {
    /// <summary>
    /// Returns a similarity value based on the size of the maximum common subtree according to the given equality comparison.
    /// </summary>
    /// <param name="a"></param>
    /// <param name="b"></param>
    /// <param name="comparer"></param>
    /// <returns>Similarity degree between the two trees, scaled between [0,1], where 1 = similar, 0 = non-similar</returns>
    public static double MaxCommonSubtreeSimilarity(ISymbolicExpressionTree a, ISymbolicExpressionTree b, SymbolicExpressionTreeNodeSimilarityComparer comparer) {
      double max = 0;
      foreach (var aa in a.Root.GetSubtree(0).GetSubtree(0).IterateNodesBreadth()) {
        int lenA = aa.GetLength();
        if (lenA <= max) continue;
        foreach (var bb in b.Root.GetSubtree(0).GetSubtree(0).IterateNodesBreadth()) {
          int lenB = bb.GetLength();
          if (lenB <= max) continue;
          int matches = SymbolicExpressionTreeMatching.Match(aa, bb, comparer);
          if (max < matches) max = matches;
        }
      }
      return max / Math.Max(a.Length, b.Length);
    }

    private static double CalculateSimilarity(ISymbolicExpressionTreeNode a, ISymbolicExpressionTreeNode b, SymbolicExpressionTreeNodeSimilarityComparer comp) {
      return (double)SymbolicExpressionTreeMatching.Match(a, b, comp) / Math.Max(a.GetLength(), b.GetLength());
    }

    public static double CalculateCompoundSimilarity(ISymbolicExpressionTree a, ISymbolicExpressionTree b, SymbolicExpressionTreeNodeSimilarityComparer comparer) {
      var nA = a.Root.GetSubtree(0).GetSubtree(0);
      var nB = b.Root.GetSubtree(0).GetSubtree(0);

      var itemsA = nA.IterateNodesBreadth().Where(n => n.SubtreeCount > 0).Select(n => new MatchItem { Node = n, Matched = false }).ToArray();
      var itemsB = nB.IterateNodesBreadth().Where(n => n.SubtreeCount > 0).Select(n => new MatchItem { Node = n, Matched = false }).ToArray();


      double similaritySum = 0;

      foreach (var ia in itemsA) {
        foreach (var ib in itemsB) {
          similaritySum += CalculateSimilarity(ia.Node, ib.Node, comparer);
        }
      }
      return similaritySum / (itemsA.Length * itemsB.Length);
    }
  }

  class MatchItem {
    public ISymbolicExpressionTreeNode Node;
    public bool Matched;
  }
}