[8430] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
[14185] | 3 | * Copyright (C) 2002-2016 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
[8430] | 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 |
|
---|
| 22 | using System;
|
---|
| 23 | using System.Linq;
|
---|
[14523] | 24 | using System.Threading;
|
---|
[8430] | 25 | using HeuristicLab.Common;
|
---|
| 26 | using HeuristicLab.Core;
|
---|
| 27 | using HeuristicLab.Data;
|
---|
| 28 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
| 29 | using HeuristicLab.Optimization;
|
---|
| 30 | using HeuristicLab.Parameters;
|
---|
| 31 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
| 32 | using HeuristicLab.Problems.DataAnalysis;
|
---|
| 33 | using HeuristicLab.Problems.DataAnalysis.Symbolic;
|
---|
| 34 | using HeuristicLab.Problems.DataAnalysis.Symbolic.TimeSeriesPrognosis;
|
---|
| 35 |
|
---|
| 36 | namespace HeuristicLab.Algorithms.DataAnalysis.TimeSeries {
|
---|
[13238] | 37 | [Item("Autoregressive Modeling (AR)", "Timeseries modeling algorithm that creates AR-N models.")]
|
---|
[12504] | 38 | [Creatable(CreatableAttribute.Categories.DataAnalysis, Priority = 130)]
|
---|
[8430] | 39 | [StorableClass]
|
---|
| 40 | public class AutoregressiveModeling : FixedDataAnalysisAlgorithm<ITimeSeriesPrognosisProblem> {
|
---|
| 41 | private const string TimeOffesetParameterName = "Maximum Time Offset";
|
---|
| 42 |
|
---|
| 43 | public IFixedValueParameter<IntValue> TimeOffsetParameter {
|
---|
| 44 | get { return (IFixedValueParameter<IntValue>)Parameters[TimeOffesetParameterName]; }
|
---|
| 45 | }
|
---|
| 46 |
|
---|
| 47 |
|
---|
| 48 | public int TimeOffset {
|
---|
| 49 | get { return TimeOffsetParameter.Value.Value; }
|
---|
| 50 | set { TimeOffsetParameter.Value.Value = value; }
|
---|
| 51 | }
|
---|
| 52 |
|
---|
| 53 | [StorableConstructor]
|
---|
| 54 | protected AutoregressiveModeling(bool deserializing) : base(deserializing) { }
|
---|
| 55 | protected AutoregressiveModeling(AutoregressiveModeling original, Cloner cloner) : base(original, cloner) { }
|
---|
| 56 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 57 | return new AutoregressiveModeling(this, cloner);
|
---|
| 58 | }
|
---|
| 59 |
|
---|
| 60 | public AutoregressiveModeling()
|
---|
| 61 | : base() {
|
---|
| 62 | Parameters.Add(new FixedValueParameter<IntValue>(TimeOffesetParameterName, "The maximum time offset for the model ranging from 1 to infinity.", new IntValue(1)));
|
---|
| 63 | Problem = new TimeSeriesPrognosisProblem();
|
---|
| 64 | }
|
---|
| 65 |
|
---|
[14523] | 66 | protected override void Run(CancellationToken cancellationToken) {
|
---|
[8430] | 67 | double rmsError, cvRmsError;
|
---|
| 68 | var solution = CreateAutoRegressiveSolution(Problem.ProblemData, TimeOffset, out rmsError, out cvRmsError);
|
---|
| 69 | Results.Add(new Result("Autoregressive solution", "The autoregressive time series prognosis solution.", solution));
|
---|
| 70 | 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)));
|
---|
| 71 | 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)));
|
---|
| 72 | }
|
---|
| 73 |
|
---|
| 74 | /// <summary>
|
---|
| 75 | /// Calculates an AR(p) model. For further information see http://en.wikipedia.org/wiki/Autoregressive_model
|
---|
| 76 | /// </summary>
|
---|
| 77 | /// <param name="problemData">The problem data which should be used for training</param>
|
---|
| 78 | /// <param name="timeOffset">The parameter p of the AR(p) specifying the maximum time offset [1,infinity] </param>
|
---|
| 79 | /// <returns>The times series autoregressive solution </returns>
|
---|
[8762] | 80 | public static ITimeSeriesPrognosisSolution CreateAutoRegressiveSolution(ITimeSeriesPrognosisProblemData problemData, int timeOffset) {
|
---|
[8430] | 81 | double rmsError, cvRmsError;
|
---|
| 82 | return CreateAutoRegressiveSolution(problemData, timeOffset, out rmsError, out cvRmsError);
|
---|
| 83 | }
|
---|
| 84 |
|
---|
[8762] | 85 | private static ITimeSeriesPrognosisSolution CreateAutoRegressiveSolution(ITimeSeriesPrognosisProblemData problemData, int timeOffset, out double rmsError, out double cvRmsError) {
|
---|
[8430] | 86 | string targetVariable = problemData.TargetVariable;
|
---|
| 87 |
|
---|
| 88 | double[,] inputMatrix = new double[problemData.TrainingPartition.Size, timeOffset + 1];
|
---|
[8762] | 89 | var targetValues = problemData.Dataset.GetDoubleValues(targetVariable).ToList();
|
---|
[8430] | 90 | for (int i = 0, row = problemData.TrainingPartition.Start; i < problemData.TrainingPartition.Size; i++, row++) {
|
---|
[8762] | 91 | for (int col = 0; col < timeOffset; col++) {
|
---|
| 92 | inputMatrix[i, col] = targetValues[row - col - 1];
|
---|
[8430] | 93 | }
|
---|
| 94 | }
|
---|
[8762] | 95 | // set target values in last column
|
---|
| 96 | for (int i = 0; i < inputMatrix.GetLength(0); i++)
|
---|
| 97 | inputMatrix[i, timeOffset] = targetValues[i + problemData.TrainingPartition.Start];
|
---|
[8430] | 98 |
|
---|
| 99 | if (inputMatrix.Cast<double>().Any(x => double.IsNaN(x) || double.IsInfinity(x)))
|
---|
| 100 | throw new NotSupportedException("Linear regression does not support NaN or infinity values in the input dataset.");
|
---|
| 101 |
|
---|
| 102 |
|
---|
[12817] | 103 | alglib.linearmodel lm = new alglib.linearmodel();
|
---|
| 104 | alglib.lrreport ar = new alglib.lrreport();
|
---|
[8430] | 105 | int nRows = inputMatrix.GetLength(0);
|
---|
| 106 | int nFeatures = inputMatrix.GetLength(1) - 1;
|
---|
| 107 | double[] coefficients = new double[nFeatures + 1]; // last coefficient is for the constant
|
---|
| 108 |
|
---|
| 109 | int retVal = 1;
|
---|
| 110 | alglib.lrbuild(inputMatrix, nRows, nFeatures, out retVal, out lm, out ar);
|
---|
| 111 | if (retVal != 1) throw new ArgumentException("Error in calculation of linear regression solution");
|
---|
| 112 | rmsError = ar.rmserror;
|
---|
| 113 | cvRmsError = ar.cvrmserror;
|
---|
| 114 |
|
---|
| 115 | alglib.lrunpack(lm, out coefficients, out nFeatures);
|
---|
| 116 |
|
---|
| 117 |
|
---|
[14400] | 118 | ISymbolicExpressionTree tree = new SymbolicExpressionTree(new ProgramRootSymbol().CreateTreeNode());
|
---|
| 119 | ISymbolicExpressionTreeNode startNode = new StartSymbol().CreateTreeNode();
|
---|
| 120 | tree.Root.AddSubtree(startNode);
|
---|
| 121 | ISymbolicExpressionTreeNode addition = new Addition().CreateTreeNode();
|
---|
| 122 | startNode.AddSubtree(addition);
|
---|
| 123 |
|
---|
| 124 | for (int i = 0; i < timeOffset; i++) {
|
---|
| 125 | LaggedVariableTreeNode node = (LaggedVariableTreeNode)new LaggedVariable().CreateTreeNode();
|
---|
| 126 | node.VariableName = targetVariable;
|
---|
| 127 | node.Weight = coefficients[i];
|
---|
| 128 | node.Lag = (i + 1) * -1;
|
---|
| 129 | addition.AddSubtree(node);
|
---|
| 130 | }
|
---|
| 131 |
|
---|
| 132 | ConstantTreeNode cNode = (ConstantTreeNode)new Constant().CreateTreeNode();
|
---|
| 133 | cNode.Value = coefficients[coefficients.Length - 1];
|
---|
| 134 | addition.AddSubtree(cNode);
|
---|
| 135 |
|
---|
[8430] | 136 | var interpreter = new SymbolicTimeSeriesPrognosisExpressionTreeInterpreter(problemData.TargetVariable);
|
---|
[13941] | 137 | var model = new SymbolicTimeSeriesPrognosisModel(problemData.TargetVariable, tree, interpreter);
|
---|
[8430] | 138 | var solution = model.CreateTimeSeriesPrognosisSolution((ITimeSeriesPrognosisProblemData)problemData.Clone());
|
---|
| 139 | return solution;
|
---|
| 140 | }
|
---|
| 141 | }
|
---|
| 142 | }
|
---|