#region License Information
/* HeuristicLab
* Copyright (C) 2002-2016 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 .
*/
#endregion
using System.Collections.Generic;
using System.Linq;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Data;
using HeuristicLab.Encodings.RealVectorEncoding;
using HeuristicLab.Operators;
using HeuristicLab.Optimization;
using HeuristicLab.Parameters;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
namespace HeuristicLab.Problems.ParameterOptimization {
[Item("BestSolutionsAnalyzer", "Tracks the best parameter vector solutions of the current algorithm run.")]
[StorableClass]
public class BestSolutionsAnalyzer : SingleSuccessorOperator, IAnalyzer {
private const string MaximizationParameterName = "Maximization";
private const string ParameterVectorParameterName = "RealVector";
private const string ParameterNamesParameterName = "ParameterNames";
private const string QualityParameterName = "Quality";
private const string PreviousBestQualityParameterName = "PreviousBestQuality";
private const string BestQualityParameterName = "BestQuality";
private const string BestKnownQualityParameterName = "BestKnownQuality";
private const string ResultsParameterName = "Results";
private const string BestSolutionsResultName = "Best Solutions Store";
public virtual bool EnabledByDefault {
get { return false; }
}
public ILookupParameter MaximizationParameter {
get { return (ILookupParameter)Parameters[MaximizationParameterName]; }
}
public IScopeTreeLookupParameter ParameterVectorParameter {
get { return (IScopeTreeLookupParameter)Parameters[ParameterVectorParameterName]; }
}
public ILookupParameter ParameterNamesParameter {
get { return (ILookupParameter)Parameters[ParameterNamesParameterName]; }
}
public IScopeTreeLookupParameter QualityParameter {
get { return (IScopeTreeLookupParameter)Parameters[QualityParameterName]; }
}
public ILookupParameter PreviousBestQualityParameter {
get { return (ILookupParameter)Parameters[PreviousBestQualityParameterName]; }
}
public ILookupParameter BestQualityParameter {
get { return (ILookupParameter)Parameters[BestQualityParameterName]; }
}
public ILookupParameter BestKnownQualityParameter {
get { return (ILookupParameter)Parameters[BestKnownQualityParameterName]; }
}
public IValueLookupParameter ResultsParameter {
get { return (IValueLookupParameter)Parameters[ResultsParameterName]; }
}
[StorableConstructor]
protected BestSolutionsAnalyzer(bool deserializing) : base(deserializing) { }
protected BestSolutionsAnalyzer(BestSolutionsAnalyzer original, Cloner cloner)
: base(original, cloner) { }
public override IDeepCloneable Clone(Cloner cloner) {
return new BestSolutionsAnalyzer(this, cloner);
}
public BestSolutionsAnalyzer()
: base() {
Parameters.Add(new LookupParameter(MaximizationParameterName, "True if the problem is a maximization problem."));
Parameters.Add(new ScopeTreeLookupParameter(ParameterVectorParameterName, "The parameter vector which should be evaluated."));
Parameters.Add(new LookupParameter(ParameterNamesParameterName, "The names of the elements in the parameter vector."));
Parameters.Add(new ScopeTreeLookupParameter(QualityParameterName, "The quality name for the parameter vectors."));
Parameters.Add(new LookupParameter(PreviousBestQualityParameterName, "The best quality of the previous iteration."));
Parameters.Add(new LookupParameter(BestQualityParameterName, "The best quality found so far."));
Parameters.Add(new LookupParameter(BestKnownQualityParameterName, "The quality of the best known solution."));
Parameters.Add(new ValueLookupParameter(ResultsParameterName, "The result collection where the results should be stored."));
}
public override IOperation Apply() {
ItemArray parameterVectors = ParameterVectorParameter.ActualValue;
ItemArray qualities = QualityParameter.ActualValue;
bool max = MaximizationParameter.ActualValue.Value;
DoubleValue bestKnownQuality = BestKnownQualityParameter.ActualValue;
var solutions = parameterVectors.Zip(qualities, (ParameterVector, Quality) => new { ParameterVector, Quality });
if (max) solutions = solutions.MaxItems(s => s.Quality.Value);
else solutions = solutions.MinItems(s => s.Quality.Value);
if (BestQualityParameter.ActualValue == null) {
if (max) BestQualityParameter.ActualValue = new DoubleValue(double.MinValue);
else BestQualityParameter.ActualValue = new DoubleValue(double.MaxValue);
}
if (PreviousBestQualityParameter.ActualValue == null)
PreviousBestQualityParameter.ActualValue = (DoubleValue)BestQualityParameter.ActualValue.Clone();
//add result for best solutions
ResultCollection results = ResultsParameter.ActualValue;
if (!results.ContainsKey(BestSolutionsResultName))
results.Add(new Result(BestSolutionsResultName, new ItemSet(new DoubleArrayEqualityComparer())));
var previousBestQuality = PreviousBestQualityParameter.ActualValue.Value;
var bestQuality = solutions.First().Quality.Value;
var bestSolutions = (ItemSet)results[BestSolutionsResultName].Value;
//clear best solutions if new found quality is better than the existing one
if (max && bestQuality > previousBestQuality || !max && bestQuality < previousBestQuality)
bestSolutions.Clear();
//add new found solutions
if (max && bestQuality >= BestQualityParameter.ActualValue.Value
|| !max && bestQuality <= BestQualityParameter.ActualValue.Value) {
foreach (var solution in solutions) {
var newSolution = (DoubleArray)solution.ParameterVector.Clone();
newSolution.ElementNames = ParameterNamesParameter.ActualValue;
bestSolutions.Add(newSolution);
}
}
//update best quality
if (max && bestQuality >= BestQualityParameter.ActualValue.Value
|| !max && bestQuality <= BestQualityParameter.ActualValue.Value) {
BestQualityParameter.ActualValue.Value = bestQuality;
}
//update best known quality
if (bestKnownQuality == null || max && bestQuality > bestKnownQuality.Value
|| !max && bestQuality < bestKnownQuality.Value) {
BestKnownQualityParameter.ActualValue = new DoubleValue(bestQuality);
}
PreviousBestQualityParameter.ActualValue = (DoubleValue)BestQualityParameter.ActualValue.Clone();
return base.Apply();
}
}
public class DoubleArrayEqualityComparer : IEqualityComparer {
public bool Equals(DoubleArray x, DoubleArray y) {
if (x == null && y == null) return true;
if (x == null) return false;
if (y == null) return false;
return x.SequenceEqual(y);
}
public int GetHashCode(DoubleArray obj) {
if (obj == null) return 0;
return (int)obj.Aggregate(23L, (current, item) => current ^ System.BitConverter.DoubleToInt64Bits(item));
}
}
}