Changeset 11589 for branches/HeuristicLab.DatasetRefactor/sources

Timestamp:

11/26/14 16:13:11 (10 years ago)

Author:

bburlacu

Message:

#2276: Unsealed Dataset class, refactored GetDoubleValues, GetStringValues, GetDateTimeValues methods to internally use the same generic method, added ModifiableDataset class which adds the functionality of replacing, adding or removing rows in the dataset.

Location:

branches/HeuristicLab.DatasetRefactor/sources/HeuristicLab.Problems.DataAnalysis/3.4

Files:

• Dataset.cs (modified) (6 diffs)
• HeuristicLab.Problems.DataAnalysis-3.4.csproj (modified) (1 diff)
• Interfaces/IDataset.cs (modified) (2 diffs)
• ModifiableDataset.cs (added)
• SimpleDataset.cs (modified) (2 diffs)

branches/HeuristicLab.DatasetRefactor/sources/HeuristicLab.Problems.DataAnalysis/3.4/Dataset.cs

@@ -33 +33 @@
   [Item("Dataset", "Represents a dataset containing data that should be analyzed.")]
   [StorableClass]
-  public sealed class Dataset : NamedItem, IDataset {
+  public class Dataset : NamedItem, IDataset {
     [StorableConstructor]
-    private Dataset(bool deserializing) : base(deserializing) { }
-    private Dataset(Dataset original, Cloner cloner)
+    protected Dataset(bool deserializing) : base(deserializing) { }
+    protected Dataset(Dataset original, Cloner cloner)
       : base(original, cloner) {
       variableValues = new Dictionary<string, IList>(original.variableValues);
@@ -143 +143 @@
 
     [Storable(Name = "VariableValues")]
-    private Dictionary<string, IList> variableValues;
-
-    private List<string> variableNames;
+    protected Dictionary<string, IList> variableValues;
+
+    protected List<string> variableNames;
     [Storable]
     public IEnumerable<string> VariableNames {
       get { return variableNames; }
-      private set {
+      protected set {
         if (variableNames != null) throw new InvalidOperationException();
         variableNames = new List<string>(value);
       }
     }
-
     public IEnumerable<string> DoubleVariables {
       get { return variableValues.Where(p => p.Value is List<double>).Select(p => p.Key); }
     }
-
     public IEnumerable<double> GetDoubleValues(string variableName) {
+      return GetValues<double>(variableName);
+    }
+    public IEnumerable<string> GetStringValues(string variableName) {
+      return GetValues<string>(variableName);
+    }
+    public IEnumerable<DateTime> GetDateTimeValues(string variableName) {
+      return GetValues<DateTime>(variableName);
+    }
+    public ReadOnlyCollection<double> GetReadOnlyDoubleValues(string variableName) {
+      var values = GetValues<double>(variableName).ToList();
+      return values.AsReadOnly();
+    }
+    public double GetDoubleValue(string variableName, int row) {
+      var values = GetValues<double>(variableName) as List<double>;
+      return values[row];
+    }
+    public IEnumerable<double> GetDoubleValues(string variableName, IEnumerable<int> rows) {
+      return GetValues<double>(variableName, rows);
+    }
+    private IEnumerable<T> GetValues<T>(string variableName, IEnumerable<int> rows) {
+      var values = GetValues<T>(variableName) as List<T>;
+      return rows.Select(x => values[x]);
+    }
+    private IEnumerable<T> GetValues<T>(string variableName) {
       IList list;
       if (!variableValues.TryGetValue(variableName, out list))
         throw new ArgumentException("The variable " + variableName + " does not exist in the dataset.");
-      List<double> values = list as List<double>;
-      if (values == null) throw new ArgumentException("The variable " + variableName + " is not a double variable.");
-
-      //mkommend yield return used to enable lazy evaluation
-      foreach (double value in values)
-        yield return value;
-    }
-
-    public IEnumerable<string> GetStringValues(string variableName) {
-      IList list;
-      if (!variableValues.TryGetValue(variableName, out list))
-        throw new ArgumentException("The variable " + variableName + " does not exist in the dataset.");
-      List<string> values = list as List<string>;
-      if (values == null) throw new ArgumentException("The variable " + variableName + " is not a string variable.");
-
-      //mkommend yield return used to enable lazy evaluation
-      foreach (string value in values)
-        yield return value;
-    }
-
-    public IEnumerable<DateTime> GetDateTimeValues(string variableName) {
-      IList list;
-      if (!variableValues.TryGetValue(variableName, out list))
-        throw new ArgumentException("The variable " + variableName + " does not exist in the dataset.");
-      List<DateTime> values = list as List<DateTime>;
-      if (values == null) throw new ArgumentException("The variable " + variableName + " is not a datetime variable.");
-
-      //mkommend yield return used to enable lazy evaluation
-      foreach (DateTime value in values)
-        yield return value;
-    }
-
-    public ReadOnlyCollection<double> GetReadOnlyDoubleValues(string variableName) {
-      IList list;
-      if (!variableValues.TryGetValue(variableName, out list))
-        throw new ArgumentException("The variable " + variableName + " does not exist in the dataset.");
-      List<double> values = list as List<double>;
-      if (values == null) throw new ArgumentException("The variable " + variableName + " is not a double variable.");
-      return values.AsReadOnly();
-    }
-    public double GetDoubleValue(string variableName, int row) {
-      IList list;
-      if (!variableValues.TryGetValue(variableName, out list))
-        throw new ArgumentException("The variable " + variableName + " does not exist in the dataset.");
-      List<double> values = list as List<double>;
-      if (values == null) throw new ArgumentException("The variable " + variableName + " is not a double variable.");
-      return values[row];
-    }
-    public IEnumerable<double> GetDoubleValues(string variableName, IEnumerable<int> rows) {
-      IList list;
-      if (!variableValues.TryGetValue(variableName, out list))
-        throw new ArgumentException("The variable " + variableName + " does not exist in the dataset.");
-      List<double> values = list as List<double>;
-      if (values == null) throw new ArgumentException("The variable " + variableName + " is not a double variable.");
-
-      return rows.Select(index => values[index]);
-    }
-
+      List<T> values = list as List<T>;
+      if (values == null) throw new ArgumentException("The variable " + variableName + " is not a " + typeof(T) + " variable.");
+      return values;
+    }
     public bool VariableHasType<T>(string variableName) {
       return variableValues[variableName] is IList<T>;
@@ -227 +195 @@
     #region IStringConvertibleMatrix Members
     [Storable]
-    private int rows;
+    protected int rows;
     public int Rows {
       get { return rows; }
@@ -236 +204 @@
       set { throw new NotSupportedException(); }
     }
-
     public bool SortableView {
       get { return false; }
@@ -244 +211 @@
       get { return true; }
     }
-
     IEnumerable<string> IStringConvertibleMatrix.ColumnNames {
       get { return this.VariableNames; }
@@ -253 +219 @@
       set { throw new NotSupportedException(); }
     }
-
     public string GetValue(int rowIndex, int columnIndex) {
       return variableValues[variableNames[columnIndex]][rowIndex].ToString();

branches/HeuristicLab.DatasetRefactor/sources/HeuristicLab.Problems.DataAnalysis/3.4/HeuristicLab.Problems.DataAnalysis-3.4.csproj

@@ -172 +172 @@
     <Compile Include="Interfaces\TimeSeriesPrognosis\ITimeSeriesPrognosisProblemData.cs" />
     <Compile Include="Interfaces\TimeSeriesPrognosis\ITimeSeriesPrognosisSolution.cs" />
+    <Compile Include="ModifiableDataset.cs" />
     <Compile Include="OnlineCalculators\AutoCorrelationCalculator.cs" />
     <Compile Include="OnlineCalculators\DependencyCalculator\HoeffdingsDependenceCalculator.cs" />

branches/HeuristicLab.DatasetRefactor/sources/HeuristicLab.Problems.DataAnalysis/3.4/Interfaces/IDataset.cs

@@ -20 +20 @@
 #endregion
 
+using System;
 using System.Collections.Generic;
 using System.Collections.ObjectModel;
@@ -31 +32 @@
     ReadOnlyCollection<double> GetReadOnlyDoubleValues(string variableName);
     IEnumerable<string> GetStringValues(string variableName);
-
+    IEnumerable<DateTime> GetDateTimeValues(string variableName);
     IEnumerable<string> VariableNames { get; }
     IEnumerable<string> DoubleVariables { get; }

branches/HeuristicLab.DatasetRefactor/sources/HeuristicLab.Problems.DataAnalysis/3.4/SimpleDataset.cs

@@ -96 +96 @@
     }
 
+    public IEnumerable<T> GetValues<T>(string variableName, IEnumerable<int> rows) {
+      throw new NotImplementedException();
+    }
+
+    public IEnumerable<T> GetValues<T>(string variableName) {
+      throw new NotImplementedException();
+    }
+
     public IEnumerable<double> GetDoubleValues(string variableName, IEnumerable<int> rows) {
       yield return GetDoubleValue(variableName, 0);
@@ -110 +118 @@
     public IEnumerable<string> GetStringValues(string variableName) {
       yield return GetDoubleValue(variableName, 0).ToString(CultureInfo.InvariantCulture);
+    }
+
+    public IEnumerable<DateTime> GetDateTimeValues(string variableName) {
+      throw new NotImplementedException();
     }