source: branches/DataAnalysis/HeuristicLab.Problems.DataAnalysis.MultiVariate.TimeSeriesPrognosis/3.3/Symbolic/SymbolicTimeSeriesPrognosisSolution.cs @ 11194

#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.Collections.Generic;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding.Symbols;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
using HeuristicLab.Problems.DataAnalysis.MultiVariate.TimeSeriesPrognosis.Symbolic.Symbols;
using HeuristicLab.Problems.DataAnalysis.Symbolic.Symbols;
using System.Linq;
using System.Drawing;
using System;
using HeuristicLab.Data;

namespace HeuristicLab.Problems.DataAnalysis.MultiVariate.TimeSeriesPrognosis.Symbolic {
  [StorableClass]
  [Item("SymbolicTimeSeriesPrognosisSolution", "Represents a solution for time series prognosis.")]
  public class SymbolicTimeSeriesPrognosisSolution : NamedItem, IMultiVariateDataAnalysisSolution, IStorableContent {
    [Storable]
    private MultiVariateDataAnalysisProblemData problemData;
    [Storable]
    private SymbolicTimeSeriesPrognosisModel model;
    [Storable]
    private int horizon;
    [Storable]
    private string conditionalEvaluationVariable;
    [Storable]
    private double[] lowerEstimationLimit;
    [Storable]
    private double[] upperEstimationLimit;

    public string Filename { get; set; }

    [StorableConstructor]
    protected SymbolicTimeSeriesPrognosisSolution(bool deserializing) : base(deserializing) { }
    protected SymbolicTimeSeriesPrognosisSolution(SymbolicTimeSeriesPrognosisSolution original, Cloner cloner)
      : base(original, cloner) {
      problemData = (MultiVariateDataAnalysisProblemData)cloner.Clone(original.problemData);
      model = (SymbolicTimeSeriesPrognosisModel)cloner.Clone(original.model);
      horizon = original.horizon;
      conditionalEvaluationVariable = original.conditionalEvaluationVariable;
      lowerEstimationLimit = (double[])original.lowerEstimationLimit.Clone();
      upperEstimationLimit = (double[])original.upperEstimationLimit.Clone();
    }
    public SymbolicTimeSeriesPrognosisSolution() {
      horizon = 1;
    }

    public SymbolicTimeSeriesPrognosisSolution(MultiVariateDataAnalysisProblemData problemData, SymbolicTimeSeriesPrognosisModel model, int horizon, string conditionalEvaluationVariable, double[] lowerEstimationLimit, double[] upperEstimationLimit)
      : this() {
      this.problemData = problemData;
      this.model = model;
      this.horizon = horizon;
      this.conditionalEvaluationVariable = conditionalEvaluationVariable;
      this.lowerEstimationLimit = (double[])lowerEstimationLimit.Clone();
      this.upperEstimationLimit = (double[])upperEstimationLimit.Clone();
    }
    public override IDeepCloneable Clone(Cloner cloner) {
      return new SymbolicTimeSeriesPrognosisSolution(this, cloner);
    }
    [StorableHook(HookType.AfterDeserialization)]
    private void AfterDeserialization() {
      if (problemData != null)
        RegisterProblemDataEvents();
    }

    public override Image ItemImage {
      get { return HeuristicLab.Common.Resources.VS2008ImageLibrary.Function; }
    }

    public int Horizon {
      get { return horizon; }
      set {
        if (value <= 0) throw new ArgumentException();
        horizon = value;
      }
    }

    public SymbolicTimeSeriesPrognosisModel Model {
      get { return model; }
      set {
        if (model != value) {
          if (value == null) throw new ArgumentNullException();
          model = value;
          RaiseModelChanged();
        }
      }
    }

    public string ConditionalEvaluationVariable {
      get { return conditionalEvaluationVariable; }
      set {
        if (conditionalEvaluationVariable != value) {
          conditionalEvaluationVariable = value;
          RaiseEstimatedValuesChanged();
        }
      }
    }

    public double GetLowerEstimationLimit(int i) {
      return lowerEstimationLimit[i];
    }
    public double GetUpperEstimationLimit(int i) {
      return upperEstimationLimit[i];
    }


    public IEnumerable<double[]> GetPrognosis(int t) {
      return model.GetEstimatedValues(problemData, t, t + 1, horizon);
    }

    #region IMultiVariateDataAnalysisSolution Members

    public MultiVariateDataAnalysisProblemData ProblemData {
      get { return problemData; }
      set {
        if (problemData != value) {
          if (value == null) throw new ArgumentNullException();
          if (model != null && problemData != null &&
            !(problemData.InputVariables
              .Select(c => c.Value)
              .SequenceEqual(value.InputVariables
                             .Select(c => c.Value)) &&
             problemData.TargetVariables
             .Select(c => c.Value)
             .SequenceEqual(value.TargetVariables
                            .Select(c => c.Value)))) {
            throw new ArgumentException("Could not set new problem data with different structure");
          }

          if (problemData != null) DeregisterProblemDataEvents();
          problemData = value;
          RaiseProblemDataChanged();
          RegisterProblemDataEvents();
        }
      }
    }

    IMultiVariateDataAnalysisModel IMultiVariateDataAnalysisSolution.Model {
      get { return model; }
    }

    public IEnumerable<double[]> EstimatedValues {
      get {
        return ApplyEstimationLimits(model.GetEstimatedValues(problemData, 0, problemData.Dataset.Rows));
      }
    }

    public IEnumerable<double[]> EstimatedTrainingValues {
      get {
        return ApplyEstimationLimits(model.GetEstimatedValues(problemData, problemData.TrainingSamplesStart.Value, problemData.TrainingSamplesEnd.Value));
      }
    }

    public IEnumerable<double[]> EstimatedTestValues {
      get {
        return ApplyEstimationLimits(model.GetEstimatedValues(problemData, problemData.TestSamplesStart.Value, problemData.TestSamplesEnd.Value));
      }
    }

    #endregion


    #region Events
    protected virtual void RegisterProblemDataEvents() {
      ProblemData.ProblemDataChanged += new EventHandler(ProblemData_Changed);
    }
    protected virtual void DeregisterProblemDataEvents() {
      ProblemData.ProblemDataChanged += new EventHandler(ProblemData_Changed);
    }
    private void ProblemData_Changed(object sender, EventArgs e) {
      RaiseProblemDataChanged();
    }

    public event EventHandler ProblemDataChanged;
    protected virtual void RaiseProblemDataChanged() {
      var listeners = ProblemDataChanged;
      if (listeners != null)
        listeners(this, EventArgs.Empty);
    }

    public event EventHandler ModelChanged;
    protected virtual void RaiseModelChanged() {
      EventHandler handler = ModelChanged;
      if (handler != null)
        handler(this, EventArgs.Empty);
    }

    public event EventHandler EstimatedValuesChanged;
    protected virtual void RaiseEstimatedValuesChanged() {
      var listeners = EstimatedValuesChanged;
      if (listeners != null)
        listeners(this, EventArgs.Empty);
    }
    #endregion



    private IEnumerable<double[]> ApplyEstimationLimits(IEnumerable<double[]> values) {
      foreach (var xs in values) {
        for (int i = 0; i < xs.Length; i++) {
          if (double.IsNaN(xs[i])) {
            xs[i] = (upperEstimationLimit[i] - lowerEstimationLimit[i]) / 2.0 + lowerEstimationLimit[i];
          } else {
            xs[i] = Math.Max(lowerEstimationLimit[i], Math.Min(upperEstimationLimit[i], xs[i]));
          }
        }
        yield return xs;
      }
    }
  }
}