#region License Information
/* HeuristicLab
 * Copyright (C) 2002-2013 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.Linq;
using HeuristicLab.Analysis;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Data;
using HeuristicLab.Optimization;
using HeuristicLab.Parameters;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;

namespace HeuristicLab.Problems.DataAnalysis.Symbolic.SlidingWindow {
  [StorableClass]
  [Item("Sliding Window Qualities Analyzer", "Analyzer that computes the qualities of the best solution on past, current, and future regions of the sliding window training data.")]
  public sealed class SlidingWindowQualitiesAnalyzer : SymbolicDataAnalysisAnalyzer {
    private const string SlidingWindowQualitiesResultName = "Sliding Window Qualities";
    private const string ProblemDataParameterName = "ProblemData";
    private const string EvaluatorParameterName = "Evaluator";
    private const string FitnessCalculationPartitionParameterName = "FitnessCalculationPartition";
    private const string ValidationPartitionParameterName = "ValidationPartition";
    private const string SymbolicExpressionTreeInterpreterParameterName = "SymbolicExpressionTreeInterpreter";

    public IValueLookupParameter<IDataAnalysisProblemData> ProblemDataParameter {
      get { return (IValueLookupParameter<IDataAnalysisProblemData>)Parameters[ProblemDataParameterName]; }
    }
    public ILookupParameter<IntRange> FitnessCalculationPartitionParameter {
      get { return (ILookupParameter<IntRange>)Parameters[FitnessCalculationPartitionParameterName]; }
    }
    public ILookupParameter<IntRange> ValidationPartitionParameter {
      get { return (ILookupParameter<IntRange>)Parameters[ValidationPartitionParameterName]; }
    }
    public ILookupParameter<IEvaluator> EvaluatorParameter {
      get { return (ILookupParameter<IEvaluator>)Parameters[EvaluatorParameterName]; }
    }

    public ILookupParameter<ISymbolicDataAnalysisExpressionTreeInterpreter> SymbolicExpressionTreeInterpreter {
      get { return (ILookupParameter<ISymbolicDataAnalysisExpressionTreeInterpreter>)Parameters[SymbolicExpressionTreeInterpreterParameterName]; }
    }

    public override IDeepCloneable Clone(Cloner cloner) {
      return new SlidingWindowQualitiesAnalyzer(this, cloner);
    }
    private SlidingWindowQualitiesAnalyzer(SlidingWindowQualitiesAnalyzer original, Cloner cloner)
      : base(original, cloner) {
    }



    [StorableConstructor]
    private SlidingWindowQualitiesAnalyzer(bool deserializing) : base(deserializing) { }

    public SlidingWindowQualitiesAnalyzer() {
      Parameters.Add(new ValueLookupParameter<IDataAnalysisProblemData>(ProblemDataParameterName, "The problem data on which the symbolic data analysis solution should be evaluated."));
      Parameters.Add(new LookupParameter<IEvaluator>(EvaluatorParameterName, ""));
      Parameters.Add(new LookupParameter<IntRange>(FitnessCalculationPartitionParameterName, ""));
      Parameters.Add(new LookupParameter<ISymbolicDataAnalysisExpressionTreeInterpreter>(SymbolicExpressionTreeInterpreterParameterName, ""));

      ProblemDataParameter.Hidden = true;
    }

    [StorableHook(HookType.AfterDeserialization)]
    private void AfterDeserialization() {
    }

    public override IOperation Apply() {
      if (FitnessCalculationPartitionParameter.ActualValue == null)
        // do nothing because the sliding window hasn't been initialized yet
        return base.Apply();


      var results = ResultCollectionParameter.ActualValue;
      if (!results.ContainsKey("Best training solution")) return base.Apply();

      var problemData = (IRegressionProblemData)ProblemDataParameter.ActualValue;
      var evaluator = (SymbolicDataAnalysisSingleObjectiveEvaluator<IRegressionProblemData>)EvaluatorParameter.ActualValue;
      var context = new Core.ExecutionContext(ExecutionContext, evaluator, new Scope());
      var fitnessCalculationPartition = FitnessCalculationPartitionParameter.ActualValue;

      var bestSolution = (ISymbolicDataAnalysisSolution)results["Best training solution"].Value;
      var bestModel = bestSolution.Model;
      var bestTree = bestModel.SymbolicExpressionTree;
      // add result
      if (!results.ContainsKey(SlidingWindowQualitiesResultName)) {
        results.Add(new Result(SlidingWindowQualitiesResultName, new DataTable(SlidingWindowQualitiesResultName)));
      }
      var swQualitiesTable = (DataTable)results[SlidingWindowQualitiesResultName].Value;

      // compute before quality
      var beforeQuality = 0.0;
      if (!swQualitiesTable.Rows.ContainsKey("Before Quality"))
        swQualitiesTable.Rows.Add(new DataRow("Before Quality") { VisualProperties = { StartIndexZero = true } });
      if (fitnessCalculationPartition.Start > problemData.TrainingPartition.Start) {
        var beforeRange = new IntRange(problemData.TrainingPartition.Start, fitnessCalculationPartition.Start);
        beforeQuality = evaluator.Evaluate(context, bestTree, problemData, Enumerable.Range(beforeRange.Start, beforeRange.Size));
      }
      swQualitiesTable.Rows["Before Quality"].Values.Add(beforeQuality);

      // compute current quality
      var currentQuality = ((DoubleValue)results["CurrentBestQuality"].Value).Value;
      if (!swQualitiesTable.Rows.ContainsKey("Current Quality"))
        swQualitiesTable.Rows.Add(new DataRow("Current Quality") { VisualProperties = { StartIndexZero = true } });
      swQualitiesTable.Rows["Current Quality"].Values.Add(currentQuality);

      // compute after quality
      if (fitnessCalculationPartition.End < problemData.TrainingPartition.End) {
        var afterRange = new IntRange(fitnessCalculationPartition.End, problemData.TrainingPartition.End);
        var afterQuality = evaluator.Evaluate(context, bestTree, problemData,
          Enumerable.Range(afterRange.Start, afterRange.Size));
        if (!swQualitiesTable.Rows.ContainsKey("After Quality"))
          swQualitiesTable.Rows.Add(new DataRow("After Quality") { VisualProperties = { StartIndexZero = true } });
        swQualitiesTable.Rows["After Quality"].Values.Add(afterQuality);
      }
      // compute test quality
      if (!swQualitiesTable.Rows.ContainsKey("Test Quality"))
        swQualitiesTable.Rows.Add(new DataRow("Test Quality") { VisualProperties = { StartIndexZero = true } });
      var regressionSolution = (IRegressionSolution)bestSolution;
      swQualitiesTable.Rows["Test Quality"].Values.Add(regressionSolution.TestRSquared);
      return base.Apply();
    }

    public override bool EnabledByDefault { get { return false; } }
  }
}