Changeset 7759 for branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression

Timestamp:

04/24/12 16:37:40 (13 years ago)

Author:

sforsten

Message:

#1784:

• deleted ClassificationData and RegressionData. RegressionProblemData and ClassificationProblemData are used instead
• deleted not needed Transformer
• ValueGenerator is now a static class and yield return is used return IEnumerable

Location:

branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression/3.4

Files:

• ArtificialRegressionDataDescriptor.cs (modified) (2 diffs)
• ArtificialRegressionInstanceProvider.cs (modified) (1 diff)
• HeuristicLab.Problems.Instances.Regression-3.4.csproj (modified) (2 diffs)
• RealWorld/RealWorldInstanceProvider.cs (modified) (1 diff)
• RegressionInstanceProvider.cs (modified) (5 diffs)
• ResourceRegressionInstanceProvider.cs (modified) (4 diffs)
• TrentMcConaghy/TrentMcConaghyInstanceProvider.cs (modified) (2 diffs)
• ValueGenerator.cs (modified) (1 diff)

branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression/3.4/ArtificialRegressionDataDescriptor.cs

@@ -21 +21 @@
 
 using System.Collections.Generic;
+using HeuristicLab.Problems.DataAnalysis;
+
 namespace HeuristicLab.Problems.Instances.Regression {
   public abstract class ArtificialRegressionDataDescriptor : IDataDescriptor {
@@ -34 +36 @@
     protected abstract int TestPartitionEnd { get; }
 
-    public RegressionData GenerateRegressionData() {
-      RegressionData regData = new RegressionData();
+    public RegressionProblemData GenerateRegressionData() {
+      Dataset dataset = new Dataset(InputVariables, this.GenerateValues());
+
+      RegressionProblemData regData = new RegressionProblemData(dataset, AllowedInputVariables, TargetVariable);
       regData.Name = this.Name;
       regData.Description = this.Description;
-      regData.InputVariables = this.InputVariables;
-      regData.AllowedInputVariables = this.AllowedInputVariables;
-      regData.TargetVariable = this.TargetVariable;
-      regData.Values = Transformer.Transformation(this.GenerateValues());
-      regData.TrainingPartitionStart = this.TrainingPartitionStart;
-      regData.TrainingPartitionEnd = this.TrainingPartitionEnd;
-      regData.TestPartitionStart = this.TestPartitionStart;
-      regData.TestPartitionEnd = this.TestPartitionEnd;
+      regData.TrainingPartition.Start = this.TrainingPartitionStart;
+      regData.TrainingPartition.End = this.TrainingPartitionEnd;
+      regData.TestPartition.Start = this.TestPartitionStart;
+      regData.TestPartition.End = this.TestPartitionEnd;
       return regData;
     }

branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression/3.4/ArtificialRegressionInstanceProvider.cs

@@ -20 +20 @@
 #endregion
 
+using HeuristicLab.Problems.DataAnalysis;
 
 namespace HeuristicLab.Problems.Instances.Regression {
   public abstract class ArtificialRegressionInstanceProvider : RegressionInstanceProvider {
-    public override RegressionData LoadData(IDataDescriptor descriptor) {
+    public override RegressionProblemData LoadData(IDataDescriptor descriptor) {
       return ((ArtificialRegressionDataDescriptor)descriptor).GenerateRegressionData();
     }

branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression/3.4/HeuristicLab.Problems.Instances.Regression-3.4.csproj

@@ -38 +38 @@
   </PropertyGroup>
   <ItemGroup>
+    <Reference Include="HeuristicLab.Collections-3.3, Version=3.3.0.0, Culture=neutral, PublicKeyToken=ba48961d6f65dcec, processorArchitecture=MSIL">
+      <Private>False</Private>
+    </Reference>
     <Reference Include="HeuristicLab.Common-3.3">
       <HintPath>..\..\..\..\trunk\sources\bin\HeuristicLab.Common-3.3.dll</HintPath>
@@ -44 +47 @@
     <Reference Include="HeuristicLab.Core-3.3">
       <HintPath>..\..\..\..\trunk\sources\bin\HeuristicLab.Core-3.3.dll</HintPath>
+      <Private>False</Private>
+    </Reference>
+    <Reference Include="HeuristicLab.Data-3.3, Version=3.3.0.0, Culture=neutral, PublicKeyToken=ba48961d6f65dcec, processorArchitecture=MSIL">
       <Private>False</Private>
     </Reference>

branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression/3.4/RealWorld/RealWorldInstanceProvider.cs

@@ -39 +39 @@
     }
 
-    protected override string FileName { get { return "RealWorld"; } }
+    protected override string FileExtension { get { return "RealWorld"; } }
   }
 }

branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression/3.4/RegressionInstanceProvider.cs

@@ -26 +26 @@
 using System.Linq;
 using System.Text;
+using HeuristicLab.Problems.DataAnalysis;
 
 namespace HeuristicLab.Problems.Instances.Regression {
-  public abstract class RegressionInstanceProvider : IProblemInstanceProvider<RegressionData> {
-    public RegressionData LoadData(string path) {
+  public abstract class RegressionInstanceProvider : IProblemInstanceProvider<RegressionProblemData> {
+    public RegressionProblemData LoadData(string path) {
       TableFileParser csvFileParser = new TableFileParser();
       csvFileParser.Parse(path);
 
-      RegressionData regData = new RegressionData();
+      Dataset dataset = new Dataset(csvFileParser.VariableNames, csvFileParser.Values);
+      string targetVar = csvFileParser.VariableNames.Last();
+      IEnumerable<string> allowedInputVars = csvFileParser.VariableNames.Where(x => !x.Equals(targetVar));
+
+      RegressionProblemData regData = new RegressionProblemData(dataset, allowedInputVars, targetVar);
       int pos = path.LastIndexOf('\\');
       if (pos < 0)
@@ -41 +46 @@
         regData.Name = path.Substring(pos, path.Length - pos);
       }
-      regData.InputVariables = new List<string>(csvFileParser.VariableNames).ToArray();
-      regData.TargetVariable = csvFileParser.VariableNames.Last();
-      regData.AllowedInputVariables = regData.InputVariables.Where(x => !x.Equals(regData.TargetVariable)).ToArray();
-      //convert to multidimensional array
-      List<IList> values = csvFileParser.Values;
-      regData.Values = new double[values.First().Count, values.Count];
-      for (int i = 0; i < values.Count; i++) {
-        for (int j = 0; j < values.First().Count; j++) {
-          regData.Values[j, i] = (double)values[i][j];
-        }
-      }
 
       int trainingPartEnd = csvFileParser.Rows * 2 / 3;
-      regData.TrainingPartitionStart = 0;
-      regData.TrainingPartitionEnd = trainingPartEnd;
-      regData.TestPartitionStart = trainingPartEnd;
-      regData.TestPartitionEnd = csvFileParser.Rows;
+      regData.TrainingPartition.Start = 0;
+      regData.TrainingPartition.End = trainingPartEnd;
+      regData.TestPartition.Start = trainingPartEnd;
+      regData.TestPartition.End = csvFileParser.Rows;
       return regData;
     }
 
-    public void SaveData(RegressionData instance, string path) {
+    public void SaveData(RegressionProblemData instance, string path) {
       StringBuilder strBuilder = new StringBuilder();
 
@@ -70 +64 @@
       strBuilder.AppendLine();
 
-      double[,] values = instance.Values;
+      Dataset dataset = instance.Dataset;
 
-      for (int i = 0; i < values.GetLength(0); i++) {
-        for (int j = 0; j < values.GetLength(1); j++) {
-          strBuilder.Append(values[i, j] + ";");
+      for (int i = 0; i < dataset.Rows; i++) {
+        for (int j = 0; j < dataset.Columns; j++) {
+          strBuilder.Append(dataset.GetValue(i, j) + ";");
         }
         strBuilder.Remove(strBuilder.Length - 1, 1);
@@ -86 +80 @@
 
     public abstract IEnumerable<IDataDescriptor> GetDataDescriptors();
-    public abstract RegressionData LoadData(IDataDescriptor descriptor);
+    public abstract RegressionProblemData LoadData(IDataDescriptor descriptor);
 
     public abstract string Name { get; }
@@ -92 +86 @@
     public abstract Uri WebLink { get; }
     public abstract string ReferencePublication { get; }
-
-    public IProblemInstanceConsumer<RegressionData> Consumer { get; set; }
   }
 }

branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression/3.4/ResourceRegressionInstanceProvider.cs

@@ -26 +26 @@
 using System.Reflection;
 using System.Text.RegularExpressions;
+using HeuristicLab.Problems.DataAnalysis;
 using ICSharpCode.SharpZipLib.Zip;
 
@@ -31 +32 @@
   public abstract class ResourceRegressionInstanceProvider : RegressionInstanceProvider {
 
-    protected abstract string FileName { get; }
+    protected abstract string FileExtension { get; }
 
     public override IEnumerable<IDataDescriptor> GetDataDescriptors() {
-      var solutionsArchiveName = GetResourceName(FileName + @"\.zip");
+      var solutionsArchiveName = GetResourceName(FileExtension + @"\.zip");
       if (!String.IsNullOrEmpty(solutionsArchiveName)) {
         using (var solutionsZipFile = new ZipInputStream(GetType().Assembly.GetManifestResourceStream(solutionsArchiveName))) {
@@ -43 +44 @@
     }
 
-    public override RegressionData LoadData(IDataDescriptor id) {
+    public override RegressionProblemData LoadData(IDataDescriptor id) {
       var descriptor = (ResourceRegressionDataDescriptor)id;
 
-      RegressionData regData = LoadData(GetTempFileForResource(descriptor.ResourceName));
+      RegressionProblemData regData = LoadData(GetTempFileForResource(descriptor.ResourceName));
       regData.Name = descriptor.Name;
       regData.Description = descriptor.Description;
@@ -66 +67 @@
 
     private string GetTempFileForResource(string resourceName) {
-      var instanceArchiveName = GetResourceName(FileName + @"\.zip");
+      var instanceArchiveName = GetResourceName(FileExtension + @"\.zip");
       using (var instancesZipFile = new ZipFile(GetType().Assembly.GetManifestResourceStream(instanceArchiveName))) {
         var entry = instancesZipFile.GetEntry(resourceName);

branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression/3.4/TrentMcConaghy/TrentMcConaghyInstanceProvider.cs

@@ -21 +21 @@
 
 using System;
+using System.Collections.Generic;
 using System.Linq;
+using HeuristicLab.Problems.DataAnalysis;
 
 namespace HeuristicLab.Problems.Instances.Regression {
@@ -43 +45 @@
     }
 
-    protected override string FileName { get { return "TrentMcConaghy"; } }
+    protected override string FileExtension { get { return "TrentMcConaghy"; } }
 
-    public override RegressionData LoadData(IDataDescriptor id) {
-      RegressionData regData = base.LoadData(id);
-
-      regData.TargetVariable = regData.InputVariables.First();
-      regData.AllowedInputVariables = regData.InputVariables.Where(x => !x.Equals(regData.TargetVariable)).ToArray();
+    public override RegressionProblemData LoadData(IDataDescriptor id) {
+      RegressionProblemData regData = base.LoadData(id);
+      string targetVar = regData.InputVariables.First().Value;
+      IEnumerable<string> allowedInputVars = regData.InputVariables.Where(x => !x.Equals(regData.TargetVariable)).Select(x => x.Value);
+      RegressionProblemData newRegData = new RegressionProblemData(regData.Dataset, allowedInputVars, targetVar);
 
       return regData;

branches/ProblemInstancesRegressionAndClassification/HeuristicLab.Problems.Instances.Regression/3.4/ValueGenerator.cs

@@ -26 +26 @@
 using HeuristicLab.Random;
 namespace HeuristicLab.Problems.Instances.Regression {
-  public class ValueGenerator {
-    protected static FastRandom rand = new FastRandom();
+  public static class ValueGenerator {
+    private static FastRandom rand = new FastRandom();
 
     public static IEnumerable<double> GenerateSteps(double start, double end, double stepWidth) {
-      return Enumerable.Range(0, (int)Math.Round(((end - start) / stepWidth) + 1))
-                                      .Select(i => (start + i * stepWidth));
+      int steps = (int)Math.Round(((end - start) / stepWidth) + 1);
+      for (int i = 0; i < steps; i++)
+        yield return start + i * stepWidth;
     }
 
     public static IEnumerable<double> GenerateUniformDistributedValues(int amount, double start, double end) {
-      List<double> values = new List<double>();
-      for (int i = 0; i < amount; i++) {
-        values.Add(rand.NextDouble() * (end - start) + start);
-      }
-      return values;
+      for (int i = 0; i < amount; i++)
+        yield return rand.NextDouble() * (end - start) + start;
     }
 
     public static IEnumerable<double> GenerateNormalDistributedValues(int amount, double mu, double sigma) {
-      List<double> values = new List<double>();
-      for (int i = 0; i < amount; i++) {
-        values.Add(NormalDistributedRandom.NextDouble(rand, mu, sigma));
-      }
-      return values;
+      for (int i = 0; i < amount; i++)
+        yield return NormalDistributedRandom.NextDouble(rand, mu, sigma);
     }