Changeset 16386 for branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis

Timestamp:

12/15/18 12:07:16 (6 years ago)

Author:

gkronber

Message:

#2925 merged changes r15972:16382 from trunk to branch

Location:

branches/2925_AutoDiffForDynamicalModels

Files:

• . (modified) (1 prop)
• HeuristicLab.Algorithms.DataAnalysis (modified) (1 prop)
• HeuristicLab.Algorithms.DataAnalysis/3.4 (modified) (1 prop)
• HeuristicLab.Algorithms.DataAnalysis/3.4/GBM/GradientBoostingRegressionAlgorithm.cs (modified) (1 diff)
• HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/GradientBoostedTreesAlgorithm.cs (modified) (1 diff)
• HeuristicLab.Algorithms.DataAnalysis/3.4/NearestNeighbour/NearestNeighbourModel.cs (modified) (2 diffs)
• HeuristicLab.Algorithms.DataAnalysis/3.4/NeuralNetwork/NeuralNetworkEnsembleModel.cs (modified) (5 diffs)
• HeuristicLab.Algorithms.DataAnalysis/3.4/NeuralNetwork/NeuralNetworkModel.cs (modified) (5 diffs)
• HeuristicLab.Algorithms.DataAnalysis/3.4/NonlinearRegression/NonlinearRegression.cs (modified) (1 diff)
• HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestClassification.cs (modified) (1 diff)
• HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestModel.cs (modified) (1 diff)
• HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestRegression.cs (modified) (1 diff)
• HeuristicLab.Algorithms.DataAnalysis/3.4/SupportVectorMachine/SupportVectorMachineModel.cs (modified) (2 diffs)
• HeuristicLab.Algorithms.DataAnalysis/3.4/TSNE/TSNEAlgorithm.cs (modified) (1 diff)
• HeuristicLab.Algorithms.DataAnalysis/3.4/kMeans/KMeansClusteringModel.cs (modified) (3 diffs)

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/GBM/GradientBoostingRegressionAlgorithm.cs

@@ -210 +210 @@
     protected override void Run(CancellationToken cancellationToken) {
       // Set up the algorithm
-      if (SetSeedRandomly) Seed = new System.Random().Next();
+      if (SetSeedRandomly) Seed = RandomSeedGenerator.GetSeed();
       var rand = new MersenneTwister((uint)Seed);
 

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/GradientBoostedTreesAlgorithm.cs

@@ -186 +186 @@
     protected override void Run(CancellationToken cancellationToken) {
       // Set up the algorithm
-      if (SetSeedRandomly) Seed = new System.Random().Next();
+      if (SetSeedRandomly) Seed = Random.RandomSeedGenerator.GetSeed();
 
       // Set up the results display

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/NearestNeighbour/NearestNeighbourModel.cs

@@ -130 +130 @@
           // automatic determination of weights (all features should have variance = 1)
           this.weights = this.allowedInputVariables
-            .Select(name => 1.0 / dataset.GetDoubleValues(name, rows).StandardDeviationPop())
+            .Select(name => {
+              var pop = dataset.GetDoubleValues(name, rows).StandardDeviationPop();
+              return  pop.IsAlmost(0) ? 1.0 : 1.0/pop;
+            })
             .Concat(new double[] { 1.0 }) // no scaling for target variable
             .ToArray();
@@ -259 +262 @@
 
 
+    public bool IsProblemDataCompatible(IRegressionProblemData problemData, out string errorMessage) {
+      return RegressionModel.IsProblemDataCompatible(this, problemData, out errorMessage);
+    }
+
+    public override bool IsProblemDataCompatible(IDataAnalysisProblemData problemData, out string errorMessage) {
+      if (problemData == null) throw new ArgumentNullException("problemData", "The provided problemData is null.");
+
+      var regressionProblemData = problemData as IRegressionProblemData;
+      if (regressionProblemData != null)
+        return IsProblemDataCompatible(regressionProblemData, out errorMessage);
+
+      var classificationProblemData = problemData as IClassificationProblemData;
+      if (classificationProblemData != null)
+        return IsProblemDataCompatible(classificationProblemData, out errorMessage);
+
+      throw new ArgumentException("The problem data is not a regression nor a classification problem data. Instead a " + problemData.GetType().GetPrettyName() + " was provided.", "problemData");
+    }
+
     IRegressionSolution IRegressionModel.CreateRegressionSolution(IRegressionProblemData problemData) {
       return new NearestNeighbourRegressionSolution(this, new RegressionProblemData(problemData));

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/NeuralNetwork/NeuralNetworkEnsembleModel.cs

@@ -36 +36 @@
   public sealed class NeuralNetworkEnsembleModel : ClassificationModel, INeuralNetworkEnsembleModel {
 
+    private object mlpEnsembleLocker = new object();
     private alglib.mlpensemble mlpEnsemble;
-    public alglib.mlpensemble MultiLayerPerceptronEnsemble {
-      get { return mlpEnsemble; }
-      set {
-        if (value != mlpEnsemble) {
-          if (value == null) throw new ArgumentNullException();
-          mlpEnsemble = value;
-          OnChanged(EventArgs.Empty);
-        }
-      }
-    }
 
     public override IEnumerable<string> VariablesUsedForPrediction {
@@ -103 +94 @@
         }
         // mlpeprocess writes data in mlpEnsemble and is therefore not thread-safe
-        lock (mlpEnsemble) {
+        lock (mlpEnsembleLocker) {
           alglib.mlpeprocess(mlpEnsemble, x, ref y);
         }
@@ -123 +114 @@
         }
         // mlpeprocess writes data in mlpEnsemble and is therefore not thread-safe
-        lock (mlpEnsemble) {
+        lock (mlpEnsembleLocker) {
           alglib.mlpeprocess(mlpEnsemble, x, ref y);
         }
@@ -139 +130 @@
     }
 
+
+    public bool IsProblemDataCompatible(IRegressionProblemData problemData, out string errorMessage) {
+      return RegressionModel.IsProblemDataCompatible(this, problemData, out errorMessage);
+    }
+
+    public override bool IsProblemDataCompatible(IDataAnalysisProblemData problemData, out string errorMessage) {
+      if (problemData == null) throw new ArgumentNullException("problemData", "The provided problemData is null.");
+
+      var regressionProblemData = problemData as IRegressionProblemData;
+      if (regressionProblemData != null)
+        return IsProblemDataCompatible(regressionProblemData, out errorMessage);
+
+      var classificationProblemData = problemData as IClassificationProblemData;
+      if (classificationProblemData != null)
+        return IsProblemDataCompatible(classificationProblemData, out errorMessage);
+
+      throw new ArgumentException("The problem data is not a regression nor a classification problem data. Instead a " + problemData.GetType().GetPrettyName() + " was provided.", "problemData");
+    }
+
     public IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
       return new NeuralNetworkEnsembleRegressionSolution(this, new RegressionEnsembleProblemData(problemData));
@@ -145 +155 @@
       return new NeuralNetworkEnsembleClassificationSolution(this, new ClassificationEnsembleProblemData(problemData));
     }
-
-    #region events
-    public event EventHandler Changed;
-    private void OnChanged(EventArgs e) {
-      var handlers = Changed;
-      if (handlers != null)
-        handlers(this, e);
-    }
-    #endregion
-
+   
     #region persistence
     [Storable]

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/NeuralNetwork/NeuralNetworkModel.cs

@@ -36 +36 @@
   public sealed class NeuralNetworkModel : ClassificationModel, INeuralNetworkModel {
 
+    private object mlpLocker = new object();
     private alglib.multilayerperceptron multiLayerPerceptron;
-    public alglib.multilayerperceptron MultiLayerPerceptron {
-      get { return multiLayerPerceptron; }
-      set {
-        if (value != multiLayerPerceptron) {
-          if (value == null) throw new ArgumentNullException();
-          multiLayerPerceptron = value;
-          OnChanged(EventArgs.Empty);
-        }
-      }
-    }
 
     public override IEnumerable<string> VariablesUsedForPrediction {
@@ -107 +98 @@
         }
         // NOTE: mlpprocess changes data in multiLayerPerceptron and is therefore not thread-save!
-        lock (multiLayerPerceptron) {
+        lock (mlpLocker) {
           alglib.mlpprocess(multiLayerPerceptron, x, ref y);
         }
@@ -127 +118 @@
         }
         // NOTE: mlpprocess changes data in multiLayerPerceptron and is therefore not thread-save!
-        lock (multiLayerPerceptron) {
+        lock (mlpLocker) {
           alglib.mlpprocess(multiLayerPerceptron, x, ref y);
         }
@@ -143 +134 @@
     }
 
+    public bool IsProblemDataCompatible(IRegressionProblemData problemData, out string errorMessage) {
+      return RegressionModel.IsProblemDataCompatible(this, problemData, out errorMessage);
+    }
+
+    public override bool IsProblemDataCompatible(IDataAnalysisProblemData problemData, out string errorMessage) {
+      if (problemData == null) throw new ArgumentNullException("problemData", "The provided problemData is null.");
+
+      var regressionProblemData = problemData as IRegressionProblemData;
+      if (regressionProblemData != null)
+        return IsProblemDataCompatible(regressionProblemData, out errorMessage);
+
+      var classificationProblemData = problemData as IClassificationProblemData;
+      if (classificationProblemData != null)
+        return IsProblemDataCompatible(classificationProblemData, out errorMessage);
+
+      throw new ArgumentException("The problem data is not a regression nor a classification problem data. Instead a " + problemData.GetType().GetPrettyName() + " was provided.", "problemData");
+    }
+
     public IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
       return new NeuralNetworkRegressionSolution(this, new RegressionProblemData(problemData));
@@ -149 +158 @@
       return new NeuralNetworkClassificationSolution(this, new ClassificationProblemData(problemData));
     }
-
-    #region events
-    public event EventHandler Changed;
-    private void OnChanged(EventArgs e) {
-      var handlers = Changed;
-      if (handlers != null)
-        handlers(this, e);
-    }
-    #endregion
 
     #region persistence

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/NonlinearRegression/NonlinearRegression.cs

@@ -186 +186 @@
         qualityTable.Rows.Add(testRMSERow);
         Results.Add(new Result(qualityTable.Name, qualityTable.Name + " for all restarts", qualityTable));
-        if (SetSeedRandomly) Seed = (new System.Random()).Next();
+        if (SetSeedRandomly) Seed = RandomSeedGenerator.GetSeed();
         var rand = new MersenneTwister((uint)Seed);
         bestSolution = CreateRegressionSolution(Problem.ProblemData, ModelStructure, Iterations, ApplyLinearScaling, rand);

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestClassification.cs

@@ -135 +135 @@
     protected override void Run(CancellationToken cancellationToken) {
       double rmsError, relClassificationError, outOfBagRmsError, outOfBagRelClassificationError;
-      if (SetSeedRandomly) Seed = new System.Random().Next();
+      if (SetSeedRandomly) Seed = Random.RandomSeedGenerator.GetSeed();
 
       var model = CreateRandomForestClassificationModel(Problem.ProblemData, NumberOfTrees, R, M, Seed, out rmsError, out relClassificationError, out outOfBagRmsError, out outOfBagRelClassificationError);

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestModel.cs

@@ -286 +286 @@
     }
 
+    public bool IsProblemDataCompatible(IRegressionProblemData problemData, out string errorMessage) {
+      return RegressionModel.IsProblemDataCompatible(this, problemData, out errorMessage);
+    }
+
+    public override bool IsProblemDataCompatible(IDataAnalysisProblemData problemData, out string errorMessage) {
+      if (problemData == null) throw new ArgumentNullException("problemData", "The provided problemData is null.");
+
+      var regressionProblemData = problemData as IRegressionProblemData;
+      if (regressionProblemData != null)
+        return IsProblemDataCompatible(regressionProblemData, out errorMessage);
+
+      var classificationProblemData = problemData as IClassificationProblemData;
+      if (classificationProblemData != null)
+        return IsProblemDataCompatible(classificationProblemData, out errorMessage);
+
+      throw new ArgumentException("The problem data is not a regression nor a classification problem data. Instead a " + problemData.GetType().GetPrettyName() + " was provided.", "problemData");
+    }
+
     public static RandomForestModel CreateRegressionModel(IRegressionProblemData problemData, int nTrees, double r, double m, int seed,
       out double rmsError, out double outOfBagRmsError, out double avgRelError, out double outOfBagAvgRelError) {

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestRegression.cs

@@ -134 +134 @@
     protected override void Run(CancellationToken cancellationToken) {
       double rmsError, avgRelError, outOfBagRmsError, outOfBagAvgRelError;
-      if (SetSeedRandomly) Seed = new System.Random().Next();
+      if (SetSeedRandomly) Seed = Random.RandomSeedGenerator.GetSeed();
       var model = CreateRandomForestRegressionModel(Problem.ProblemData, NumberOfTrees, R, M, Seed,
         out rmsError, out avgRelError, out outOfBagRmsError, out outOfBagAvgRelError);

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/SupportVectorMachine/SupportVectorMachineModel.cs

@@ -126 +126 @@
       return new SupportVectorRegressionSolution(this, new RegressionProblemData(problemData));
     }
-    #endregion
+
+    public bool IsProblemDataCompatible(IRegressionProblemData problemData, out string errorMessage) {
+      return RegressionModel.IsProblemDataCompatible(this, problemData, out errorMessage);
+    }
+    #endregion
+
+    public override bool IsProblemDataCompatible(IDataAnalysisProblemData problemData, out string errorMessage) {
+      if (problemData == null) throw new ArgumentNullException("problemData", "The provided problemData is null.");
+
+      var regressionProblemData = problemData as IRegressionProblemData;
+      if (regressionProblemData != null)
+        return IsProblemDataCompatible(regressionProblemData, out errorMessage);
+
+      var classificationProblemData = problemData as IClassificationProblemData;
+      if (classificationProblemData != null)
+        return IsProblemDataCompatible(classificationProblemData, out errorMessage);
+
+      throw new ArgumentException("The problem data is not a regression nor a classification problem data. Instead a " + problemData.GetType().GetPrettyName() + " was provided.", "problemData");
+    }
 
     #region IClassificationModel Members
@@ -153 +171 @@
     }
     #endregion
+
     private IEnumerable<double> GetEstimatedValuesHelper(IDataset dataset, IEnumerable<int> rows) {
       // calculate predictions for the currently requested rows

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/TSNE/TSNEAlgorithm.cs

@@ -275 +275 @@
       if (wdist != null) wdist.Initialize(problemData);
       if (state == null) {
-        if (SetSeedRandomly) Seed = new System.Random().Next();
+        if (SetSeedRandomly) Seed = RandomSeedGenerator.GetSeed();
         var random = new MersenneTwister((uint)Seed);
         var dataset = problemData.Dataset;

branches/2925_AutoDiffForDynamicalModels/HeuristicLab.Algorithms.DataAnalysis/3.4/kMeans/KMeansClusteringModel.cs

@@ -20 +20 @@
 #endregion
 
+using System;
 using System.Collections.Generic;
 using System.Drawing;
@@ -34 +35 @@
   [StorableClass]
   [Item("KMeansClusteringModel", "Represents a k-Means clustering model.")]
-  public sealed class KMeansClusteringModel : NamedItem, IClusteringModel {
+  public sealed class KMeansClusteringModel : DataAnalysisModel, IClusteringModel {
     public static new Image StaticItemImage {
       get { return HeuristicLab.Common.Resources.VSImageLibrary.Function; }
     }
 
-    public IEnumerable<string> VariablesUsedForPrediction {
+    public override IEnumerable<string> VariablesUsedForPrediction {
       get { return allowedInputVariables; }
     }
@@ -84 +85 @@
     }
 
+    public override bool IsProblemDataCompatible(IDataAnalysisProblemData problemData, out string errorMessage) {
+      if (problemData == null) throw new ArgumentNullException("problemData", "The provided problemData is null.");
+      return IsDatasetCompatible(problemData.Dataset, out errorMessage);
+    }
+
 
     public IEnumerable<int> GetClusterValues(IDataset dataset, IEnumerable<int> rows) {