Changeset 15865 for branches/2906_Transformations/HeuristicLab.DataPreprocessing/3.4

Timestamp:

03/27/18 15:46:12 (7 years ago)

Author:

pfleck

Message:

#2906 Added PreprocessingTransformation as a custom view-model for transformations in preprocessing.

Location:

branches/2906_Transformations/HeuristicLab.DataPreprocessing/3.4

Files:

• Content/TransformationContent.cs (modified) (4 diffs)
• Data/FilteredPreprocessingData.cs (modified) (1 diff)
• Data/IPreprocessingData.cs (modified) (1 diff)
• Data/PreprocessingData.cs (modified) (8 diffs)
• Data/PreprocessingTransformation.cs (added)
• HeuristicLab.DataPreprocessing-3.4.csproj (modified) (2 diffs)
• ProblemDataCreator.cs (modified) (7 diffs)

branches/2906_Transformations/HeuristicLab.DataPreprocessing/3.4/Content/TransformationContent.cs

@@ -29 +29 @@
 
 namespace HeuristicLab.DataPreprocessing {
+
   [Item("Transformation", "Represents the transformation grid.")]
   [StorableClass]
@@ -37 +38 @@
 
     [Storable]
-    public ICheckedItemList<IDataAnalysisTransformation> CheckedTransformationList { get; private set; }
+    public IItemList<PreprocessingTransformation> TransformationList { get; private set; }
 
     #region Constructor, Cloning & Persistence
     public TransformationContent(IFilteredPreprocessingData preprocessingData)
       : base(preprocessingData) {
-      CheckedTransformationList = new CheckedItemList<IDataAnalysisTransformation>();
+      TransformationList = new ItemList<PreprocessingTransformation>();
     }
 
     public TransformationContent(TransformationContent original, Cloner cloner)
       : base(original, cloner) {
-      CheckedTransformationList = cloner.Clone(original.CheckedTransformationList);
+      TransformationList = cloner.Clone(original.TransformationList);
     }
     public override IDeepCloneable Clone(Cloner cloner) {
@@ -59 +60 @@
 
     public bool ApplyTransformations(out IEnumerable<string> errorMessages) {
-      var transformations = CheckedTransformationList.CheckedItems.Select(x => x.Value);
-
       bool success = true;
       var errors = new List<string>();
       errorMessages = errors;
 
-      foreach (var transformation in transformations) {
+      foreach (var transformation in TransformationList.Where(x => !x.IsApplied)) {
         var sourceVariable = transformation.OriginalVariable;
         var targetVariable = transformation.TransformedVariable ?? sourceVariable + " Transformed";
@@ -86 +85 @@
 
           PreprocessingData.Transformations.Add(transformation);
-          CheckedTransformationList.SetItemCheckedState(transformation, false);
+          transformation.IsApplied = true;
+          // TODO: remove unused valid values in constrainedParameters
         } else {
           success = false;

branches/2906_Transformations/HeuristicLab.DataPreprocessing/3.4/Data/FilteredPreprocessingData.cs

@@ -204 +204 @@
 
     #region Transformations
-    public IList<IDataAnalysisTransformation> Transformations {
+    public IList<PreprocessingTransformation> Transformations {
       get { return originalData.Transformations; }
     }

branches/2906_Transformations/HeuristicLab.DataPreprocessing/3.4/Data/IPreprocessingData.cs

@@ -78 +78 @@
 
     #region Transformations
-    IList<IDataAnalysisTransformation> Transformations { get; }
+    IList<PreprocessingTransformation> Transformations { get; }
     #endregion
 

branches/2906_Transformations/HeuristicLab.DataPreprocessing/3.4/Data/PreprocessingData.cs

@@ -47 +47 @@
       Name = "Preprocessing Data";
 
-      Transformations = new List<IDataAnalysisTransformation>();
+      Transformations = new List<PreprocessingTransformation>();
       selection = new Dictionary<int, IList<int>>();
 
@@ -61 +61 @@
       TrainingPartition = (IntRange)original.TrainingPartition.Clone(cloner);
       TestPartition = (IntRange)original.TestPartition.Clone(cloner);
-      Transformations = new List<IDataAnalysisTransformation>(original.Transformations.Select(cloner.Clone));
+      Transformations = new List<PreprocessingTransformation>(original.Transformations.Select(cloner.Clone));
 
       InputVariables = new List<string>(original.InputVariables);
@@ -350 +350 @@
     #region Transformations
     [Storable]
-    public IList<IDataAnalysisTransformation> Transformations { get; protected set; }
+    public IList<PreprocessingTransformation> Transformations { get; protected set; }
     #endregion
 
@@ -414 +414 @@
       }
 
+      // TODO: set fixed constrained (allowed) values
+      foreach (var trans in problemData.Transformations) {
+        var newTrans = new PreprocessingTransformation(variableNames.Select(x => new StringValue(x))) {
+          OriginalVariable = trans.OriginalVariable,
+          TransformedVariable = trans.TransformedVariable,
+          IsApplied = true
+        };
+        var cloned = (ITransformation)trans.Transformation.Clone();
+        newTrans.TransformationParameter.ValidValues.Add(cloned);
+        newTrans.Transformation = cloned;
+        Transformations.Add(newTrans);
+      }
+
       TrainingPartition = new IntRange(problemData.TrainingPartition.Start, problemData.TrainingPartition.End);
       TestPartition = new IntRange(problemData.TestPartition.Start, problemData.TestPartition.End);
@@ -459 +472 @@
       public IntRange TrainingPartition { get; set; }
       public IntRange TestPartition { get; set; }
-      public IList<IDataAnalysisTransformation> Transformations { get; set; }
+      public IList<PreprocessingTransformation> Transformations { get; set; }
       public DataPreprocessingChangedEventType ChangedType { get; set; }
 
@@ -487 +500 @@
         TrainingPartition = new IntRange(TrainingPartition.Start, TrainingPartition.End),
         TestPartition = new IntRange(TestPartition.Start, TestPartition.End),
-        Transformations = new List<IDataAnalysisTransformation>(Transformations),
+        Transformations = new List<PreprocessingTransformation>(Transformations),
         ChangedType = changedType,
         ChangedColumn = column,
@@ -682 +695 @@
       l = 0;
       ir = n - 1;
-      for (;;) {
+      for (; ; ) {
         if (ir <= l + 1) {
           // Active partition contains 1 or 2 elements.
@@ -707 +720 @@
           j = ir;
           a = arr[l + 1]; // Partitioning element.
-          for (;;) { // Beginning of innermost loop.
+          for (; ; ) { // Beginning of innermost loop.
             do i++; while (arr[i].CompareTo(a) < 0); // Scan up to find element > a.
             do j--; while (arr[j].CompareTo(a) > 0); // Scan down to find element < a.

branches/2906_Transformations/HeuristicLab.DataPreprocessing/3.4/HeuristicLab.DataPreprocessing-3.4.csproj

@@ -107 +107 @@
       <Private>False</Private>
     </Reference>
+    <Reference Include="HeuristicLab.Parameters-3.3, Version=3.3.0.0, Culture=neutral, PublicKeyToken=ba48961d6f65dcec, processorArchitecture=MSIL">
+      <SpecificVersion>False</SpecificVersion>
+      <HintPath>..\..\..\..\trunk\bin\HeuristicLab.Parameters-3.3.dll</HintPath>
+    </Reference>
     <Reference Include="HeuristicLab.Persistence-3.3, Version=3.3.0.0, Culture=neutral, PublicKeyToken=ba48961d6f65dcec, processorArchitecture=MSIL">
       <SpecificVersion>False</SpecificVersion>
@@ -134 +138 @@
     <Compile Include="Content\MultiScatterPlotContent.cs" />
     <Compile Include="Content\PreprocessingContent.cs" />
+    <Compile Include="Data\PreprocessingTransformation.cs" />
     <Compile Include="Content\SingleScatterPlotContent.cs" />
     <Compile Include="Content\ScatterPlotContent.cs" />

branches/2906_Transformations/HeuristicLab.DataPreprocessing/3.4/ProblemDataCreator.cs

@@ -34 +34 @@
     }
 
-    private IList<IDataAnalysisTransformation> Transformations {
+    private IList<PreprocessingTransformation> Transformations {
       get { return context.Data.Transformations; }
     }
@@ -76 +76 @@
         targetVariable = context.Data.VariableNames.First();
       var inputVariables = GetDoubleInputVariables(targetVariable);
-      var newProblemData = new TimeSeriesPrognosisProblemData(ExportedDataset, inputVariables, targetVariable, Transformations) {
+      var newProblemData = new TimeSeriesPrognosisProblemData(ExportedDataset, inputVariables, targetVariable, CreateDataAnalysisTransformation()) {
         TrainingHorizon = oldProblemData.TrainingHorizon,
         TestHorizon = oldProblemData.TestHorizon
@@ -85 +85 @@
     private IDataAnalysisProblemData CreateRegressionData(RegressionProblemData oldProblemData) {
       // TODO: transformations (additional inputs, target changed)
-      var targetVariable = RegressionProblemData.GetTransformedTragetVariable(oldProblemData.TargetVariable, Transformations);
+      var targetVariable = RegressionProblemData.GetTransformedTragetVariable(oldProblemData.TargetVariable, CreateDataAnalysisTransformation());
       if (!context.Data.VariableNames.Contains(targetVariable))
         targetVariable = context.Data.VariableNames.First();
       var inputVariables = GetDoubleInputVariables(targetVariable);
-      var newProblemData = new RegressionProblemData(ExportedDataset, inputVariables, targetVariable, Transformations);
+      var newProblemData = new RegressionProblemData(ExportedDataset, inputVariables, targetVariable, CreateDataAnalysisTransformation());
       return newProblemData;
     }
@@ -98 +98 @@
         targetVariable = context.Data.VariableNames.First();
       var inputVariables = GetDoubleInputVariables(targetVariable);
-      var newProblemData = new ClassificationProblemData(ExportedDataset, inputVariables, targetVariable, Transformations) {
+      var newProblemData = new ClassificationProblemData(ExportedDataset, inputVariables, targetVariable, CreateDataAnalysisTransformation()) {
         PositiveClass = oldProblemData.PositiveClass
       };
@@ -105 +105 @@
 
     private IDataAnalysisProblemData CreateClusteringData(ClusteringProblemData oldProblemData) {
-      return new ClusteringProblemData(ExportedDataset, GetDoubleInputVariables(String.Empty), Transformations);
+      return new ClusteringProblemData(ExportedDataset, GetDoubleInputVariables(String.Empty), CreateDataAnalysisTransformation());
     }
 
@@ -117 +117 @@
     }
 
-     void SetAllowedInputVariables(IDataAnalysisProblemData problemData, IEnumerable<string> oldInputVariables) {
-       var inputs = DataAnalysisProblemData.ExtendInputVariables(oldInputVariables, Transformations);
+    void SetAllowedInputVariables(IDataAnalysisProblemData problemData, IEnumerable<string> oldInputVariables) {
+      var inputs = DataAnalysisProblemData.ExtendInputVariables(oldInputVariables, problemData.Transformations);
 
       foreach (var input in problemData.InputVariables) {
@@ -142 +142 @@
       return context.Data.TrainingPartition.End - context.Data.TrainingPartition.Start > 1 || list.Range() > 0;
     }
+
+    private IEnumerable<IDataAnalysisTransformation> CreateDataAnalysisTransformation() {
+      return Transformations.Select(x => new DataAnalysisTransformation(x.OriginalVariable, x.TransformedVariable, (ITransformation)x.Transformation.Clone()));
+    }
   }
 }