Changeset 13941 for trunk/sources/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression

Timestamp:

06/28/16 13:33:17 (8 years ago)

Author:

mkommend

Message:

#2604:

• Base classes for data analysis, classification, and regression models
• Added target variable to classification and regression models
• Switched parameter order in data analysis solutions (model, problemdata)

Location:

trunk/sources/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression

Files:

• ConstantRegressionModel.cs (modified) (3 diffs)
• RegressionEnsembleModel.cs (modified) (8 diffs)
• RegressionModel.cs (added)

trunk/sources/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/ConstantRegressionModel.cs

@@ -32 +32 @@
   [Item("Constant Regression Model", "A model that always returns the same constant value regardless of the presented input data.")]
   [Obsolete]
-  public class ConstantRegressionModel : NamedItem, IRegressionModel, IStringConvertibleValue {
-    public IEnumerable<string> VariablesUsedForPrediction { get { return Enumerable.Empty<string>(); } }
-
-    [Storable]
-    private readonly string targetVariable;
-    public string TargetVariable {
-      get { return targetVariable; }
-    }
+  public class ConstantRegressionModel : RegressionModel, IStringConvertibleValue {
+    public override IEnumerable<string> VariablesUsedForPrediction { get { return Enumerable.Empty<string>(); } }
 
     [Storable]
@@ -53 +47 @@
       : base(original, cloner) {
       this.constant = original.constant;
-      this.targetVariable = original.targetVariable;
     }
 
@@ -59 +52 @@
 
     public ConstantRegressionModel(double constant, string targetVariable = "Target")
-      : base() {
+      : base(targetVariable) {
       this.name = ItemName;
       this.description = ItemDescription;
       this.constant = constant;
       this.ReadOnly = true; // changing a constant regression model is not supported
-      this.targetVariable = targetVariable;
     }
 
-    public IEnumerable<double> GetEstimatedValues(IDataset dataset, IEnumerable<int> rows) {
+    public override IEnumerable<double> GetEstimatedValues(IDataset dataset, IEnumerable<int> rows) {
       return rows.Select(row => Constant);
     }
 
-    public IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
+    public override IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
       return new ConstantRegressionSolution(new ConstantModel(constant), new RegressionProblemData(problemData));
     }

trunk/sources/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/RegressionEnsembleModel.cs

@@ -33 +33 @@
   [StorableClass]
   [Item("RegressionEnsembleModel", "A regression model that contains an ensemble of multiple regression models")]
-  public sealed class RegressionEnsembleModel : NamedItem, IRegressionEnsembleModel {
-    public IEnumerable<string> VariablesUsedForPrediction {
+  public sealed class RegressionEnsembleModel : RegressionModel, IRegressionEnsembleModel {
+    public override IEnumerable<string> VariablesUsedForPrediction {
       get { return models.SelectMany(x => x.VariablesUsedForPrediction).Distinct().OrderBy(x => x); }
     }
@@ -41 +41 @@
     public IEnumerable<IRegressionModel> Models {
       get { return new List<IRegressionModel>(models); }
-    }
-
-    [Storable]
-    private readonly string target;
-    public string TargetVariable {
-      get { return models.First().TargetVariable; }
     }
 
@@ -109 +103 @@
     public RegressionEnsembleModel(IEnumerable<IRegressionModel> models) : this(models, models.Select(m => 1.0)) { }
     public RegressionEnsembleModel(IEnumerable<IRegressionModel> models, IEnumerable<double> modelWeights)
-      : base() {
+      : base(string.Empty) {
       this.name = ItemName;
       this.description = ItemDescription;
 
-
       this.models = new List<IRegressionModel>(models);
       this.modelWeights = new List<double>(modelWeights);
+
+      if (this.models.Any()) this.TargetVariable = this.models.First().TargetVariable;
     }
 
     public void Add(IRegressionModel model) {
+      if (string.IsNullOrEmpty(TargetVariable)) TargetVariable = model.TargetVariable;
       Add(model, 1.0);
     }
     public void Add(IRegressionModel model, double weight) {
+      if (string.IsNullOrEmpty(TargetVariable)) TargetVariable = model.TargetVariable;
+
       models.Add(model);
       modelWeights.Add(weight);
@@ -131 +129 @@
     }
     public void AddRange(IEnumerable<IRegressionModel> models, IEnumerable<double> weights) {
+      if (string.IsNullOrEmpty(TargetVariable)) TargetVariable = models.First().TargetVariable;
+
       this.models.AddRange(models);
       modelWeights.AddRange(weights);
@@ -140 +140 @@
       models.RemoveAt(index);
       modelWeights.RemoveAt(index);
+
+      if (!models.Any()) TargetVariable = string.Empty;
       OnChanged();
     }
@@ -148 +150 @@
         modelWeights.RemoveAt(index);
       }
+
+      if (!models.Any()) TargetVariable = string.Empty;
       OnChanged();
     }
@@ -174 +178 @@
     }
 
-    public IEnumerable<double> GetEstimatedValues(IDataset dataset, IEnumerable<int> rows) {
+    public override IEnumerable<double> GetEstimatedValues(IDataset dataset, IEnumerable<int> rows) {
       double weightsSum = modelWeights.Sum();
       var summedEstimates = from estimatedValuesVector in GetEstimatedValueVectors(dataset, rows)
@@ -222 +226 @@
 
 
-    public RegressionEnsembleSolution CreateRegressionSolution(IRegressionProblemData problemData) {
+    public override IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
       return new RegressionEnsembleSolution(this, new RegressionEnsembleProblemData(problemData));
-    }
-    IRegressionSolution IRegressionModel.CreateRegressionSolution(IRegressionProblemData problemData) {
-      return CreateRegressionSolution(problemData);
     }
   }