Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/HeuristicLab.Algorithms.DataAnalysis/3.4/TimeSeries/AutoregressiveModeling.cs @ 17180

Last change on this file since 17180 was 17180, checked in by swagner, 5 years ago

#2875: Removed years in copyrights

File size: 6.5 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) Heuristic and Evolutionary Algorithms Laboratory (HEAL)
4 *
5 * This file is part of HeuristicLab.
6 *
7 * HeuristicLab is free software: you can redistribute it and/or modify
8 * it under the terms of the GNU General Public License as published by
9 * the Free Software Foundation, either version 3 of the License, or
10 * (at your option) any later version.
11 *
12 * HeuristicLab is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
15 * GNU General Public License for more details.
16 *
17 * You should have received a copy of the GNU General Public License
18 * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
19 */
20#endregion
21
22using System;
23using System.Linq;
24using System.Threading;
25using HeuristicLab.Common;
26using HeuristicLab.Core;
27using HeuristicLab.Data;
28using HeuristicLab.Optimization;
29using HeuristicLab.Parameters;
30using HEAL.Attic;
31using HeuristicLab.Problems.DataAnalysis;
32using HeuristicLab.Problems.DataAnalysis.Symbolic;
33using HeuristicLab.Problems.DataAnalysis.Symbolic.TimeSeriesPrognosis;
34
35namespace HeuristicLab.Algorithms.DataAnalysis.TimeSeries {
36  [Item("Autoregressive Modeling (AR)", "Timeseries modeling algorithm that creates AR-N models.")]
37  [Creatable(CreatableAttribute.Categories.DataAnalysis, Priority = 130)]
38  [StorableType("3BB9D723-944A-420F-A44A-A86F6FB4DABE")]
39  public class AutoregressiveModeling : FixedDataAnalysisAlgorithm<ITimeSeriesPrognosisProblem> {
40    private const string TimeOffesetParameterName = "Maximum Time Offset";
41
42    public IFixedValueParameter<IntValue> TimeOffsetParameter {
43      get { return (IFixedValueParameter<IntValue>)Parameters[TimeOffesetParameterName]; }
44    }
45
46
47    public int TimeOffset {
48      get { return TimeOffsetParameter.Value.Value; }
49      set { TimeOffsetParameter.Value.Value = value; }
50    }
51
52    [StorableConstructor]
53    protected AutoregressiveModeling(StorableConstructorFlag _) : base(_) { }
54    protected AutoregressiveModeling(AutoregressiveModeling original, Cloner cloner) : base(original, cloner) { }
55    public override IDeepCloneable Clone(Cloner cloner) {
56      return new AutoregressiveModeling(this, cloner);
57    }
58
59    public AutoregressiveModeling()
60      : base() {
61      Parameters.Add(new FixedValueParameter<IntValue>(TimeOffesetParameterName, "The maximum time offset for the model ranging from 1 to infinity.", new IntValue(1)));
62      Problem = new TimeSeriesPrognosisProblem();
63    }
64
65    protected override void Run(CancellationToken cancellationToken) {
66      double rmsError, cvRmsError;
67      var solution = CreateAutoRegressiveSolution(Problem.ProblemData, TimeOffset, out rmsError, out cvRmsError);
68      Results.Add(new Result("Autoregressive solution", "The autoregressive time series prognosis solution.", solution));
69      Results.Add(new Result("Root mean square error", "The root of the mean of squared errors of the autoregressive time series prognosis solution on the training set.", new DoubleValue(rmsError)));
70      Results.Add(new Result("Estimated root mean square error (cross-validation)", "The estimated root of the mean of squared errors of the autoregressive time series prognosis solution via cross validation.", new DoubleValue(cvRmsError)));
71    }
72
73    /// <summary>
74    /// Calculates an AR(p) model. For further information see http://en.wikipedia.org/wiki/Autoregressive_model
75    /// </summary>
76    /// <param name="problemData">The problem data which should be used for training</param>
77    /// <param name="timeOffset">The parameter p of the AR(p) specifying the maximum time offset [1,infinity] </param>
78    /// <returns>The times series autoregressive solution </returns>
79    public static ITimeSeriesPrognosisSolution CreateAutoRegressiveSolution(ITimeSeriesPrognosisProblemData problemData, int timeOffset) {
80      double rmsError, cvRmsError;
81      return CreateAutoRegressiveSolution(problemData, timeOffset, out rmsError, out cvRmsError);
82    }
83
84    private static ITimeSeriesPrognosisSolution CreateAutoRegressiveSolution(ITimeSeriesPrognosisProblemData problemData, int timeOffset, out double rmsError, out double cvRmsError) {
85      string targetVariable = problemData.TargetVariable;
86
87      double[,] inputMatrix = new double[problemData.TrainingPartition.Size, timeOffset + 1];
88      var targetValues = problemData.Dataset.GetDoubleValues(targetVariable).ToList();
89      for (int i = 0, row = problemData.TrainingPartition.Start; i < problemData.TrainingPartition.Size; i++, row++) {
90        for (int col = 0; col < timeOffset; col++) {
91          inputMatrix[i, col] = targetValues[row - col - 1];
92        }
93      }
94      // set target values in last column
95      for (int i = 0; i < inputMatrix.GetLength(0); i++)
96        inputMatrix[i, timeOffset] = targetValues[i + problemData.TrainingPartition.Start];
97
98      if (inputMatrix.ContainsNanOrInfinity())
99        throw new NotSupportedException("Linear regression does not support NaN or infinity values in the input dataset.");
100
101
102      alglib.linearmodel lm = new alglib.linearmodel();
103      alglib.lrreport ar = new alglib.lrreport();
104      int nRows = inputMatrix.GetLength(0);
105      int nFeatures = inputMatrix.GetLength(1) - 1;
106      double[] coefficients = new double[nFeatures + 1]; // last coefficient is for the constant
107
108      int retVal = 1;
109      alglib.lrbuild(inputMatrix, nRows, nFeatures, out retVal, out lm, out ar);
110      if (retVal != 1) throw new ArgumentException("Error in calculation of linear regression solution");
111      rmsError = ar.rmserror;
112      cvRmsError = ar.cvrmserror;
113
114      alglib.lrunpack(lm, out coefficients, out nFeatures);
115
116      var tree = LinearModelToTreeConverter.CreateTree(
117        variableNames: Enumerable.Repeat(problemData.TargetVariable, nFeatures).ToArray(),
118        lags: Enumerable.Range(0, timeOffset).Select(i => (i + 1) * -1).ToArray(),
119        coefficients: coefficients.Take(nFeatures).ToArray(),
120        @const: coefficients[nFeatures]
121        );
122
123      var interpreter = new SymbolicTimeSeriesPrognosisExpressionTreeInterpreter(problemData.TargetVariable);
124      var model = new SymbolicTimeSeriesPrognosisModel(problemData.TargetVariable, tree, interpreter);
125      var solution = model.CreateTimeSeriesPrognosisSolution((ITimeSeriesPrognosisProblemData)problemData.Clone());
126      return solution;
127    }
128  }
129}
Note: See TracBrowser for help on using the repository browser.