Changeset 13941 for trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest

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.Algorithms.DataAnalysis/3.4/RandomForest

Files:

• RandomForestClassification.cs (modified) (1 diff)
• RandomForestClassificationSolution.cs (modified) (1 diff)
• RandomForestModel.cs (modified) (7 diffs)
• RandomForestRegression.cs (modified) (2 diffs)
• RandomForestRegressionSolution.cs (modified) (1 diff)

trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestClassification.cs

@@ -143 +143 @@
 
       if (CreateSolution) {
-        var solution = new RandomForestClassificationSolution((IClassificationProblemData)Problem.ProblemData.Clone(), model);
+        var solution = new RandomForestClassificationSolution(model, (IClassificationProblemData)Problem.ProblemData.Clone());
         Results.Add(new Result(RandomForestClassificationModelResultName, "The random forest classification solution.", solution));
       }
     }
-    
+
     // keep for compatibility with old API
     public static RandomForestClassificationSolution CreateRandomForestClassificationSolution(IClassificationProblemData problemData, int nTrees, double r, double m, int seed,
       out double rmsError, out double relClassificationError, out double outOfBagRmsError, out double outOfBagRelClassificationError) {
       var model = CreateRandomForestClassificationModel(problemData, nTrees, r, m, seed, out rmsError, out relClassificationError, out outOfBagRmsError, out outOfBagRelClassificationError);
-      return new RandomForestClassificationSolution((IClassificationProblemData)problemData.Clone(), model);
+      return new RandomForestClassificationSolution(model, (IClassificationProblemData)problemData.Clone());
     }
 

trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestClassificationSolution.cs

@@ -43 +43 @@
       : base(original, cloner) {
     }
-    public RandomForestClassificationSolution(IClassificationProblemData problemData, IRandomForestModel randomForestModel)
+    public RandomForestClassificationSolution(IRandomForestModel randomForestModel, IClassificationProblemData problemData)
       : base(randomForestModel, problemData) {
     }

trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestModel.cs

@@ -34 +34 @@
   [StorableClass]
   [Item("RandomForestModel", "Represents a random forest for regression and classification.")]
-  public sealed class RandomForestModel : NamedItem, IRandomForestModel {
+  public sealed class RandomForestModel : ClassificationModel, IRandomForestModel {
     // not persisted
     private alglib.decisionforest randomForest;
@@ -45 +45 @@
     }
 
-    public IEnumerable<string> VariablesUsedForPrediction {
+    public override IEnumerable<string> VariablesUsedForPrediction {
       get { return originalTrainingData.AllowedInputVariables; }
     }
 
-    public string TargetVariable {
-      get {
-        var regressionProblemData = originalTrainingData as IRegressionProblemData;
-        var classificationProblemData = originalTrainingData as IClassificationProblemData;
-        if (classificationProblemData != null)
-          return classificationProblemData.TargetVariable;
-        if (regressionProblemData != null)
-          return regressionProblemData.TargetVariable;
-        throw new InvalidOperationException("Getting the target variable requires either a regression or a classification problem data.");
-      }
-    }
 
     // instead of storing the data of the model itself
@@ -107 +96 @@
 
     // random forest models can only be created through the static factory methods CreateRegressionModel and CreateClassificationModel
-    private RandomForestModel(alglib.decisionforest randomForest,
+    private RandomForestModel(string targetVariable, alglib.decisionforest randomForest,
       int seed, IDataAnalysisProblemData originalTrainingData,
       int nTrees, double r, double m, double[] classValues = null)
-      : base() {
+      : base(targetVariable) {
       this.name = ItemName;
       this.description = ItemDescription;
@@ -163 +152 @@
     }
 
-    public IEnumerable<double> GetEstimatedClassValues(IDataset dataset, IEnumerable<int> rows) {
+    public override IEnumerable<double> GetEstimatedClassValues(IDataset dataset, IEnumerable<int> rows) {
       double[,] inputData = AlglibUtil.PrepareInputMatrix(dataset, AllowedInputVariables, rows);
       AssertInputMatrix(inputData);
@@ -190 +179 @@
     }
 
-    public IRandomForestRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
-      return new RandomForestRegressionSolution(new RegressionProblemData(problemData), this);
-    }
-    IRegressionSolution IRegressionModel.CreateRegressionSolution(IRegressionProblemData problemData) {
-      return CreateRegressionSolution(problemData);
-    }
-    public IRandomForestClassificationSolution CreateClassificationSolution(IClassificationProblemData problemData) {
-      return new RandomForestClassificationSolution(new ClassificationProblemData(problemData), this);
-    }
-    IClassificationSolution IClassificationModel.CreateClassificationSolution(IClassificationProblemData problemData) {
-      return CreateClassificationSolution(problemData);
+
+    public IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
+      return new RandomForestRegressionSolution(this, new RegressionProblemData(problemData));
+    }
+    public override IClassificationSolution CreateClassificationSolution(IClassificationProblemData problemData) {
+      return new RandomForestClassificationSolution(this, new ClassificationProblemData(problemData));
     }
 
@@ -221 +205 @@
       outOfBagRmsError = rep.oobrmserror;
 
-      return new RandomForestModel(dForest, seed, problemData, nTrees, r, m);
+      return new RandomForestModel(problemData.TargetVariable, dForest, seed, problemData, nTrees, r, m);
     }
 
@@ -258 +242 @@
       outOfBagRelClassificationError = rep.oobrelclserror;
 
-      return new RandomForestModel(dForest, seed, problemData, nTrees, r, m, classValues);
+      return new RandomForestModel(problemData.TargetVariable, dForest, seed, problemData, nTrees, r, m, classValues);
     }
 

trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestRegression.cs

@@ -143 +143 @@
 
       if (CreateSolution) {
-        var solution = new RandomForestRegressionSolution((IRegressionProblemData)Problem.ProblemData.Clone(), model);
+        var solution = new RandomForestRegressionSolution(model, (IRegressionProblemData)Problem.ProblemData.Clone());
         Results.Add(new Result(RandomForestRegressionModelResultName, "The random forest regression solution.", solution));
       }
@@ -153 +153 @@
       var model = CreateRandomForestRegressionModel(problemData, nTrees, r, m, seed,
         out rmsError, out avgRelError, out outOfBagRmsError, out outOfBagAvgRelError);
-      return new RandomForestRegressionSolution((IRegressionProblemData)problemData.Clone(), model);
+      return new RandomForestRegressionSolution(model, (IRegressionProblemData)problemData.Clone());
     }
 

trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestRegressionSolution.cs

@@ -43 +43 @@
       : base(original, cloner) {
     }
-    public RandomForestRegressionSolution(IRegressionProblemData problemData, IRandomForestModel randomForestModel)
+    public RandomForestRegressionSolution(IRandomForestModel randomForestModel, IRegressionProblemData problemData)
       : base(randomForestModel, problemData) {
     }