[11315] | 1 | #region License Information
|
---|
| 2 |
|
---|
| 3 | /* HeuristicLab
|
---|
[15583] | 4 | * Copyright (C) 2002-2018 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
[11315] | 5 | *
|
---|
| 6 | * This file is part of HeuristicLab.
|
---|
| 7 | *
|
---|
| 8 | * HeuristicLab is free software: you can redistribute it and/or modify
|
---|
| 9 | * it under the terms of the GNU General Public License as published by
|
---|
| 10 | * the Free Software Foundation, either version 3 of the License, or
|
---|
| 11 | * (at your option) any later version.
|
---|
| 12 | *
|
---|
| 13 | * HeuristicLab is distributed in the hope that it will be useful,
|
---|
| 14 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 15 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 16 | * GNU General Public License for more details.
|
---|
| 17 | *
|
---|
| 18 | * You should have received a copy of the GNU General Public License
|
---|
| 19 | * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
|
---|
| 20 | */
|
---|
| 21 |
|
---|
| 22 | #endregion
|
---|
| 23 |
|
---|
| 24 | using System;
|
---|
| 25 | using System.Collections.Generic;
|
---|
| 26 | using System.Linq;
|
---|
| 27 | using System.Linq.Expressions;
|
---|
| 28 | using System.Threading.Tasks;
|
---|
| 29 | using HeuristicLab.Common;
|
---|
[11362] | 30 | using HeuristicLab.Core;
|
---|
[11315] | 31 | using HeuristicLab.Data;
|
---|
[11443] | 32 | using HeuristicLab.Parameters;
|
---|
| 33 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
[11315] | 34 | using HeuristicLab.Problems.DataAnalysis;
|
---|
[11362] | 35 | using HeuristicLab.Random;
|
---|
[11315] | 36 |
|
---|
| 37 | namespace HeuristicLab.Algorithms.DataAnalysis {
|
---|
[11443] | 38 | [Item("RFParameter", "A random forest parameter collection")]
|
---|
| 39 | [StorableClass]
|
---|
| 40 | public class RFParameter : ParameterCollection {
|
---|
| 41 | public RFParameter() {
|
---|
| 42 | base.Add(new FixedValueParameter<IntValue>("N", "The number of random forest trees", new IntValue(50)));
|
---|
| 43 | base.Add(new FixedValueParameter<DoubleValue>("M", "The ratio of features that will be used in the construction of individual trees (0<m<=1)", new DoubleValue(0.1)));
|
---|
| 44 | base.Add(new FixedValueParameter<DoubleValue>("R", "The ratio of the training set that will be used in the construction of individual trees (0<r<=1)", new DoubleValue(0.1)));
|
---|
| 45 | }
|
---|
[11315] | 46 |
|
---|
[11443] | 47 | [StorableConstructor]
|
---|
[11445] | 48 | protected RFParameter(bool deserializing)
|
---|
[11443] | 49 | : base(deserializing) {
|
---|
| 50 | }
|
---|
| 51 |
|
---|
[11445] | 52 | protected RFParameter(RFParameter original, Cloner cloner)
|
---|
[11443] | 53 | : base(original, cloner) {
|
---|
| 54 | this.N = original.N;
|
---|
| 55 | this.R = original.R;
|
---|
| 56 | this.M = original.M;
|
---|
| 57 | }
|
---|
| 58 |
|
---|
| 59 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 60 | return new RFParameter(this, cloner);
|
---|
| 61 | }
|
---|
| 62 |
|
---|
| 63 | private IFixedValueParameter<IntValue> NParameter {
|
---|
| 64 | get { return (IFixedValueParameter<IntValue>)base["N"]; }
|
---|
| 65 | }
|
---|
| 66 |
|
---|
| 67 | private IFixedValueParameter<DoubleValue> RParameter {
|
---|
| 68 | get { return (IFixedValueParameter<DoubleValue>)base["R"]; }
|
---|
| 69 | }
|
---|
| 70 |
|
---|
| 71 | private IFixedValueParameter<DoubleValue> MParameter {
|
---|
| 72 | get { return (IFixedValueParameter<DoubleValue>)base["M"]; }
|
---|
| 73 | }
|
---|
| 74 |
|
---|
| 75 | public int N {
|
---|
| 76 | get { return NParameter.Value.Value; }
|
---|
| 77 | set { NParameter.Value.Value = value; }
|
---|
| 78 | }
|
---|
| 79 |
|
---|
| 80 | public double R {
|
---|
| 81 | get { return RParameter.Value.Value; }
|
---|
| 82 | set { RParameter.Value.Value = value; }
|
---|
| 83 | }
|
---|
| 84 |
|
---|
| 85 | public double M {
|
---|
| 86 | get { return MParameter.Value.Value; }
|
---|
| 87 | set { MParameter.Value.Value = value; }
|
---|
| 88 | }
|
---|
[11315] | 89 | }
|
---|
| 90 |
|
---|
| 91 | public static class RandomForestUtil {
|
---|
[11338] | 92 | private static void CrossValidate(IRegressionProblemData problemData, Tuple<IEnumerable<int>, IEnumerable<int>>[] partitions, int nTrees, double r, double m, int seed, out double avgTestMse) {
|
---|
| 93 | avgTestMse = 0;
|
---|
| 94 | var ds = problemData.Dataset;
|
---|
| 95 | var targetVariable = GetTargetVariableName(problemData);
|
---|
| 96 | foreach (var tuple in partitions) {
|
---|
[11315] | 97 | double rmsError, avgRelError, outOfBagAvgRelError, outOfBagRmsError;
|
---|
[11338] | 98 | var trainingRandomForestPartition = tuple.Item1;
|
---|
| 99 | var testRandomForestPartition = tuple.Item2;
|
---|
[11343] | 100 | var model = RandomForestModel.CreateRegressionModel(problemData, trainingRandomForestPartition, nTrees, r, m, seed, out rmsError, out avgRelError, out outOfBagRmsError, out outOfBagAvgRelError);
|
---|
[11338] | 101 | var estimatedValues = model.GetEstimatedValues(ds, testRandomForestPartition);
|
---|
| 102 | var targetValues = ds.GetDoubleValues(targetVariable, testRandomForestPartition);
|
---|
[11315] | 103 | OnlineCalculatorError calculatorError;
|
---|
[11338] | 104 | double mse = OnlineMeanSquaredErrorCalculator.Calculate(estimatedValues, targetValues, out calculatorError);
|
---|
[11315] | 105 | if (calculatorError != OnlineCalculatorError.None)
|
---|
| 106 | mse = double.NaN;
|
---|
[11338] | 107 | avgTestMse += mse;
|
---|
[11315] | 108 | }
|
---|
[11338] | 109 | avgTestMse /= partitions.Length;
|
---|
| 110 | }
|
---|
[11443] | 111 |
|
---|
[11338] | 112 | private static void CrossValidate(IClassificationProblemData problemData, Tuple<IEnumerable<int>, IEnumerable<int>>[] partitions, int nTrees, double r, double m, int seed, out double avgTestAccuracy) {
|
---|
| 113 | avgTestAccuracy = 0;
|
---|
| 114 | var ds = problemData.Dataset;
|
---|
| 115 | var targetVariable = GetTargetVariableName(problemData);
|
---|
| 116 | foreach (var tuple in partitions) {
|
---|
| 117 | double rmsError, avgRelError, outOfBagAvgRelError, outOfBagRmsError;
|
---|
| 118 | var trainingRandomForestPartition = tuple.Item1;
|
---|
| 119 | var testRandomForestPartition = tuple.Item2;
|
---|
[11343] | 120 | var model = RandomForestModel.CreateClassificationModel(problemData, trainingRandomForestPartition, nTrees, r, m, seed, out rmsError, out avgRelError, out outOfBagRmsError, out outOfBagAvgRelError);
|
---|
[11338] | 121 | var estimatedValues = model.GetEstimatedClassValues(ds, testRandomForestPartition);
|
---|
| 122 | var targetValues = ds.GetDoubleValues(targetVariable, testRandomForestPartition);
|
---|
| 123 | OnlineCalculatorError calculatorError;
|
---|
| 124 | double accuracy = OnlineAccuracyCalculator.Calculate(estimatedValues, targetValues, out calculatorError);
|
---|
| 125 | if (calculatorError != OnlineCalculatorError.None)
|
---|
| 126 | accuracy = double.NaN;
|
---|
| 127 | avgTestAccuracy += accuracy;
|
---|
| 128 | }
|
---|
| 129 | avgTestAccuracy /= partitions.Length;
|
---|
[11315] | 130 | }
|
---|
| 131 |
|
---|
[12509] | 132 | /// <summary>
|
---|
| 133 | /// Grid search without crossvalidation (since for random forests the out-of-bag estimate is unbiased)
|
---|
| 134 | /// </summary>
|
---|
| 135 | /// <param name="problemData">The regression problem data</param>
|
---|
| 136 | /// <param name="parameterRanges">The ranges for each parameter in the grid search</param>
|
---|
| 137 | /// <param name="seed">The random seed (required by the random forest model)</param>
|
---|
| 138 | /// <param name="maxDegreeOfParallelism">The maximum allowed number of threads (to parallelize the grid search)</param>
|
---|
[11362] | 139 | public static RFParameter GridSearch(IRegressionProblemData problemData, Dictionary<string, IEnumerable<double>> parameterRanges, int seed = 12345, int maxDegreeOfParallelism = 1) {
|
---|
| 140 | var setters = parameterRanges.Keys.Select(GenerateSetter).ToList();
|
---|
| 141 | var crossProduct = parameterRanges.Values.CartesianProduct();
|
---|
| 142 | double bestOutOfBagRmsError = double.MaxValue;
|
---|
| 143 | RFParameter bestParameters = new RFParameter();
|
---|
| 144 |
|
---|
[12509] | 145 | var locker = new object();
|
---|
[11362] | 146 | Parallel.ForEach(crossProduct, new ParallelOptions { MaxDegreeOfParallelism = maxDegreeOfParallelism }, parameterCombination => {
|
---|
| 147 | var parameterValues = parameterCombination.ToList();
|
---|
| 148 | var parameters = new RFParameter();
|
---|
| 149 | for (int i = 0; i < setters.Count; ++i) { setters[i](parameters, parameterValues[i]); }
|
---|
| 150 | double rmsError, outOfBagRmsError, avgRelError, outOfBagAvgRelError;
|
---|
[11443] | 151 | RandomForestModel.CreateRegressionModel(problemData, problemData.TrainingIndices, parameters.N, parameters.R, parameters.M, seed, out rmsError, out outOfBagRmsError, out avgRelError, out outOfBagAvgRelError);
|
---|
[11426] | 152 |
|
---|
| 153 | lock (locker) {
|
---|
| 154 | if (bestOutOfBagRmsError > outOfBagRmsError) {
|
---|
[11362] | 155 | bestOutOfBagRmsError = outOfBagRmsError;
|
---|
| 156 | bestParameters = (RFParameter)parameters.Clone();
|
---|
| 157 | }
|
---|
| 158 | }
|
---|
| 159 | });
|
---|
| 160 | return bestParameters;
|
---|
| 161 | }
|
---|
| 162 |
|
---|
[12509] | 163 | /// <summary>
|
---|
| 164 | /// Grid search without crossvalidation (since for random forests the out-of-bag estimate is unbiased)
|
---|
| 165 | /// </summary>
|
---|
| 166 | /// <param name="problemData">The classification problem data</param>
|
---|
| 167 | /// <param name="parameterRanges">The ranges for each parameter in the grid search</param>
|
---|
| 168 | /// <param name="seed">The random seed (required by the random forest model)</param>
|
---|
| 169 | /// <param name="maxDegreeOfParallelism">The maximum allowed number of threads (to parallelize the grid search)</param>
|
---|
[11362] | 170 | public static RFParameter GridSearch(IClassificationProblemData problemData, Dictionary<string, IEnumerable<double>> parameterRanges, int seed = 12345, int maxDegreeOfParallelism = 1) {
|
---|
| 171 | var setters = parameterRanges.Keys.Select(GenerateSetter).ToList();
|
---|
| 172 | var crossProduct = parameterRanges.Values.CartesianProduct();
|
---|
| 173 |
|
---|
| 174 | double bestOutOfBagRmsError = double.MaxValue;
|
---|
| 175 | RFParameter bestParameters = new RFParameter();
|
---|
| 176 |
|
---|
[12509] | 177 | var locker = new object();
|
---|
[11362] | 178 | Parallel.ForEach(crossProduct, new ParallelOptions { MaxDegreeOfParallelism = maxDegreeOfParallelism }, parameterCombination => {
|
---|
| 179 | var parameterValues = parameterCombination.ToList();
|
---|
| 180 | var parameters = new RFParameter();
|
---|
| 181 | for (int i = 0; i < setters.Count; ++i) { setters[i](parameters, parameterValues[i]); }
|
---|
| 182 | double rmsError, outOfBagRmsError, avgRelError, outOfBagAvgRelError;
|
---|
[11443] | 183 | RandomForestModel.CreateClassificationModel(problemData, problemData.TrainingIndices, parameters.N, parameters.R, parameters.M, seed,
|
---|
[11362] | 184 | out rmsError, out outOfBagRmsError, out avgRelError, out outOfBagAvgRelError);
|
---|
[11426] | 185 |
|
---|
| 186 | lock (locker) {
|
---|
| 187 | if (bestOutOfBagRmsError > outOfBagRmsError) {
|
---|
[11362] | 188 | bestOutOfBagRmsError = outOfBagRmsError;
|
---|
| 189 | bestParameters = (RFParameter)parameters.Clone();
|
---|
| 190 | }
|
---|
| 191 | }
|
---|
| 192 | });
|
---|
| 193 | return bestParameters;
|
---|
| 194 | }
|
---|
| 195 |
|
---|
[12509] | 196 | /// <summary>
|
---|
| 197 | /// Grid search with crossvalidation
|
---|
| 198 | /// </summary>
|
---|
| 199 | /// <param name="problemData">The regression problem data</param>
|
---|
| 200 | /// <param name="numberOfFolds">The number of folds for crossvalidation</param>
|
---|
| 201 | /// <param name="shuffleFolds">Specifies whether the folds should be shuffled</param>
|
---|
| 202 | /// <param name="parameterRanges">The ranges for each parameter in the grid search</param>
|
---|
| 203 | /// <param name="seed">The random seed (required by the random forest model)</param>
|
---|
| 204 | /// <param name="maxDegreeOfParallelism">The maximum allowed number of threads (to parallelize the grid search)</param>
|
---|
| 205 | /// <returns>The best parameter values found by the grid search</returns>
|
---|
[11362] | 206 | public static RFParameter GridSearch(IRegressionProblemData problemData, int numberOfFolds, bool shuffleFolds, Dictionary<string, IEnumerable<double>> parameterRanges, int seed = 12345, int maxDegreeOfParallelism = 1) {
|
---|
[11315] | 207 | DoubleValue mse = new DoubleValue(Double.MaxValue);
|
---|
[11443] | 208 | RFParameter bestParameter = new RFParameter();
|
---|
[11315] | 209 |
|
---|
[11343] | 210 | var setters = parameterRanges.Keys.Select(GenerateSetter).ToList();
|
---|
[11338] | 211 | var partitions = GenerateRandomForestPartitions(problemData, numberOfFolds);
|
---|
[11343] | 212 | var crossProduct = parameterRanges.Values.CartesianProduct();
|
---|
[11315] | 213 |
|
---|
[12509] | 214 | var locker = new object();
|
---|
[11343] | 215 | Parallel.ForEach(crossProduct, new ParallelOptions { MaxDegreeOfParallelism = maxDegreeOfParallelism }, parameterCombination => {
|
---|
| 216 | var parameterValues = parameterCombination.ToList();
|
---|
[11315] | 217 | double testMSE;
|
---|
| 218 | var parameters = new RFParameter();
|
---|
[11343] | 219 | for (int i = 0; i < setters.Count; ++i) {
|
---|
| 220 | setters[i](parameters, parameterValues[i]);
|
---|
[11315] | 221 | }
|
---|
[11443] | 222 | CrossValidate(problemData, partitions, parameters.N, parameters.R, parameters.M, seed, out testMSE);
|
---|
[11426] | 223 |
|
---|
| 224 | lock (locker) {
|
---|
| 225 | if (testMSE < mse.Value) {
|
---|
[11343] | 226 | mse.Value = testMSE;
|
---|
| 227 | bestParameter = (RFParameter)parameters.Clone();
|
---|
| 228 | }
|
---|
[11315] | 229 | }
|
---|
| 230 | });
|
---|
| 231 | return bestParameter;
|
---|
| 232 | }
|
---|
[11338] | 233 |
|
---|
[12509] | 234 | /// <summary>
|
---|
| 235 | /// Grid search with crossvalidation
|
---|
| 236 | /// </summary>
|
---|
| 237 | /// <param name="problemData">The classification problem data</param>
|
---|
| 238 | /// <param name="numberOfFolds">The number of folds for crossvalidation</param>
|
---|
| 239 | /// <param name="shuffleFolds">Specifies whether the folds should be shuffled</param>
|
---|
| 240 | /// <param name="parameterRanges">The ranges for each parameter in the grid search</param>
|
---|
| 241 | /// <param name="seed">The random seed (for shuffling)</param>
|
---|
| 242 | /// <param name="maxDegreeOfParallelism">The maximum allowed number of threads (to parallelize the grid search)</param>
|
---|
[11362] | 243 | public static RFParameter GridSearch(IClassificationProblemData problemData, int numberOfFolds, bool shuffleFolds, Dictionary<string, IEnumerable<double>> parameterRanges, int seed = 12345, int maxDegreeOfParallelism = 1) {
|
---|
[11338] | 244 | DoubleValue accuracy = new DoubleValue(0);
|
---|
[11443] | 245 | RFParameter bestParameter = new RFParameter();
|
---|
[11338] | 246 |
|
---|
[11343] | 247 | var setters = parameterRanges.Keys.Select(GenerateSetter).ToList();
|
---|
| 248 | var crossProduct = parameterRanges.Values.CartesianProduct();
|
---|
[11362] | 249 | var partitions = GenerateRandomForestPartitions(problemData, numberOfFolds, shuffleFolds);
|
---|
[11338] | 250 |
|
---|
[12509] | 251 | var locker = new object();
|
---|
[11343] | 252 | Parallel.ForEach(crossProduct, new ParallelOptions { MaxDegreeOfParallelism = maxDegreeOfParallelism }, parameterCombination => {
|
---|
| 253 | var parameterValues = parameterCombination.ToList();
|
---|
[11338] | 254 | double testAccuracy;
|
---|
| 255 | var parameters = new RFParameter();
|
---|
[11343] | 256 | for (int i = 0; i < setters.Count; ++i) {
|
---|
| 257 | setters[i](parameters, parameterValues[i]);
|
---|
[11338] | 258 | }
|
---|
[11443] | 259 | CrossValidate(problemData, partitions, parameters.N, parameters.R, parameters.M, seed, out testAccuracy);
|
---|
[11426] | 260 |
|
---|
| 261 | lock (locker) {
|
---|
| 262 | if (testAccuracy > accuracy.Value) {
|
---|
[11343] | 263 | accuracy.Value = testAccuracy;
|
---|
| 264 | bestParameter = (RFParameter)parameters.Clone();
|
---|
| 265 | }
|
---|
[11338] | 266 | }
|
---|
| 267 | });
|
---|
| 268 | return bestParameter;
|
---|
| 269 | }
|
---|
| 270 |
|
---|
[11362] | 271 | private static Tuple<IEnumerable<int>, IEnumerable<int>>[] GenerateRandomForestPartitions(IDataAnalysisProblemData problemData, int numberOfFolds, bool shuffleFolds = false) {
|
---|
| 272 | var folds = GenerateFolds(problemData, numberOfFolds, shuffleFolds).ToList();
|
---|
[11343] | 273 | var partitions = new Tuple<IEnumerable<int>, IEnumerable<int>>[numberOfFolds];
|
---|
| 274 |
|
---|
| 275 | for (int i = 0; i < numberOfFolds; ++i) {
|
---|
| 276 | int p = i; // avoid "access to modified closure" warning
|
---|
| 277 | var trainingRows = folds.SelectMany((par, j) => j != p ? par : Enumerable.Empty<int>());
|
---|
| 278 | var testRows = folds[i];
|
---|
| 279 | partitions[i] = new Tuple<IEnumerable<int>, IEnumerable<int>>(trainingRows, testRows);
|
---|
| 280 | }
|
---|
| 281 | return partitions;
|
---|
| 282 | }
|
---|
| 283 |
|
---|
[11362] | 284 | public static IEnumerable<IEnumerable<int>> GenerateFolds(IDataAnalysisProblemData problemData, int numberOfFolds, bool shuffleFolds = false) {
|
---|
| 285 | var random = new MersenneTwister((uint)Environment.TickCount);
|
---|
| 286 | if (problemData is IRegressionProblemData) {
|
---|
| 287 | var trainingIndices = shuffleFolds ? problemData.TrainingIndices.OrderBy(x => random.Next()) : problemData.TrainingIndices;
|
---|
| 288 | return GenerateFolds(trainingIndices, problemData.TrainingPartition.Size, numberOfFolds);
|
---|
| 289 | }
|
---|
| 290 | if (problemData is IClassificationProblemData) {
|
---|
| 291 | // when shuffle is enabled do stratified folds generation, some folds may have zero elements
|
---|
| 292 | // otherwise, generate folds normally
|
---|
| 293 | return shuffleFolds ? GenerateFoldsStratified(problemData as IClassificationProblemData, numberOfFolds, random) : GenerateFolds(problemData.TrainingIndices, problemData.TrainingPartition.Size, numberOfFolds);
|
---|
| 294 | }
|
---|
| 295 | throw new ArgumentException("Problem data is neither regression or classification problem data.");
|
---|
| 296 | }
|
---|
[11343] | 297 |
|
---|
[11362] | 298 | /// <summary>
|
---|
| 299 | /// Stratified fold generation from classification data. Stratification means that we ensure the same distribution of class labels for each fold.
|
---|
| 300 | /// The samples are grouped by class label and each group is split into @numberOfFolds parts. The final folds are formed from the joining of
|
---|
| 301 | /// the corresponding parts from each class label.
|
---|
| 302 | /// </summary>
|
---|
| 303 | /// <param name="problemData">The classification problem data.</param>
|
---|
| 304 | /// <param name="numberOfFolds">The number of folds in which to split the data.</param>
|
---|
| 305 | /// <param name="random">The random generator used to shuffle the folds.</param>
|
---|
| 306 | /// <returns>An enumerable sequece of folds, where a fold is represented by a sequence of row indices.</returns>
|
---|
| 307 | private static IEnumerable<IEnumerable<int>> GenerateFoldsStratified(IClassificationProblemData problemData, int numberOfFolds, IRandom random) {
|
---|
| 308 | var values = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TrainingIndices);
|
---|
| 309 | var valuesIndices = problemData.TrainingIndices.Zip(values, (i, v) => new { Index = i, Value = v }).ToList();
|
---|
| 310 | IEnumerable<IEnumerable<IEnumerable<int>>> foldsByClass = valuesIndices.GroupBy(x => x.Value, x => x.Index).Select(g => GenerateFolds(g, g.Count(), numberOfFolds));
|
---|
| 311 | var enumerators = foldsByClass.Select(f => f.GetEnumerator()).ToList();
|
---|
| 312 | while (enumerators.All(e => e.MoveNext())) {
|
---|
| 313 | yield return enumerators.SelectMany(e => e.Current).OrderBy(x => random.Next()).ToList();
|
---|
| 314 | }
|
---|
| 315 | }
|
---|
| 316 |
|
---|
| 317 | private static IEnumerable<IEnumerable<T>> GenerateFolds<T>(IEnumerable<T> values, int valuesCount, int numberOfFolds) {
|
---|
| 318 | // if number of folds is greater than the number of values, some empty folds will be returned
|
---|
| 319 | if (valuesCount < numberOfFolds) {
|
---|
| 320 | for (int i = 0; i < numberOfFolds; ++i)
|
---|
| 321 | yield return i < valuesCount ? values.Skip(i).Take(1) : Enumerable.Empty<T>();
|
---|
| 322 | } else {
|
---|
| 323 | int f = valuesCount / numberOfFolds, r = valuesCount % numberOfFolds; // number of folds rounded to integer and remainder
|
---|
| 324 | int start = 0, end = f;
|
---|
| 325 | for (int i = 0; i < numberOfFolds; ++i) {
|
---|
| 326 | if (r > 0) {
|
---|
| 327 | ++end;
|
---|
| 328 | --r;
|
---|
| 329 | }
|
---|
| 330 | yield return values.Skip(start).Take(end - start);
|
---|
| 331 | start = end;
|
---|
| 332 | end += f;
|
---|
| 333 | }
|
---|
| 334 | }
|
---|
| 335 | }
|
---|
| 336 |
|
---|
[11343] | 337 | private static Action<RFParameter, double> GenerateSetter(string field) {
|
---|
| 338 | var targetExp = Expression.Parameter(typeof(RFParameter));
|
---|
| 339 | var valueExp = Expression.Parameter(typeof(double));
|
---|
[11443] | 340 | var fieldExp = Expression.Property(targetExp, field);
|
---|
[11343] | 341 | var assignExp = Expression.Assign(fieldExp, Expression.Convert(valueExp, fieldExp.Type));
|
---|
| 342 | var setter = Expression.Lambda<Action<RFParameter, double>>(assignExp, targetExp, valueExp).Compile();
|
---|
| 343 | return setter;
|
---|
| 344 | }
|
---|
| 345 |
|
---|
[11338] | 346 | private static string GetTargetVariableName(IDataAnalysisProblemData problemData) {
|
---|
| 347 | var regressionProblemData = problemData as IRegressionProblemData;
|
---|
| 348 | var classificationProblemData = problemData as IClassificationProblemData;
|
---|
| 349 |
|
---|
| 350 | if (regressionProblemData != null)
|
---|
| 351 | return regressionProblemData.TargetVariable;
|
---|
| 352 | if (classificationProblemData != null)
|
---|
| 353 | return classificationProblemData.TargetVariable;
|
---|
| 354 |
|
---|
| 355 | throw new ArgumentException("Problem data is neither regression or classification problem data.");
|
---|
| 356 | }
|
---|
[11315] | 357 | }
|
---|
| 358 | }
|
---|