Changeset 9133 for trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis

Timestamp:

01/09/13 14:56:57 (12 years ago)

Author:

sforsten

Message:

#2001:

• removed IRealWorldRegressionDataDescriptor and RealWorldResourceRegressionDataDescriptor
• simplified regression data descriptors as suggested by mkommend

Location:

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3

Files:

• HeuristicLab.Problems.Instances.DataAnalysis-3.3.csproj (modified) (1 diff)
• Regression/ArtificialRegressionDataDescriptor.cs (modified) (1 diff)
• Regression/RealWorld/ChemicalOne.cs (added)
• Regression/RealWorld/Housing.cs (added)
• Regression/RealWorld/IRealWorldRegressionDataDescriptor.cs (deleted)
• Regression/RealWorld/RealWorldRegressionInstanceProvider.cs (modified) (3 diffs)
• Regression/RealWorld/RealWorldResourceRegressionDataDescriptor.cs (deleted)
• Regression/RealWorld/Tower.cs (modified) (2 diffs)
• Regression/RegressionDataDescriptor.cs (added)
• Regression/ResourceRegressionDataDescriptor.cs (modified) (1 diff)
• Regression/ResourceRegressionInstanceProvider.cs (modified) (3 diffs)

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/HeuristicLab.Problems.Instances.DataAnalysis-3.3.csproj

@@ -185 +185 @@
     <Compile Include="Regression\Nguyen\NguyenFunctionTwo.cs" />
     <Compile Include="Regression\Nguyen\NguyenInstanceProvider.cs" />
-    <Compile Include="Regression\RealWorld\IRealWorldRegressionDataDescriptor.cs" />
+    <Compile Include="Regression\RealWorld\ChemicalOne.cs" />
+    <Compile Include="Regression\RealWorld\Housing.cs" />
     <Compile Include="Regression\RealWorld\RealWorldRegressionInstanceProvider.cs" />
-    <Compile Include="Regression\RealWorld\RealWorldResourceRegressionDataDescriptor.cs" />
     <Compile Include="Regression\RealWorld\Tower.cs" />
+    <Compile Include="Regression\RegressionDataDescriptor.cs" />
     <Compile Include="Regression\RegressionImportType.cs" />
     <Compile Include="Regression\RegressionInstanceProvider.cs" />

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Regression/ArtificialRegressionDataDescriptor.cs

@@ -24 +24 @@
 
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  public abstract class ArtificialRegressionDataDescriptor : IDataDescriptor {
-    public abstract string Name { get; }
-    public abstract string Description { get; }
-
-    protected abstract string TargetVariable { get; }
-    protected abstract string[] VariableNames { get; }
-    protected abstract string[] AllowedInputVariables { get; }
-    protected abstract int TrainingPartitionStart { get; }
-    protected abstract int TrainingPartitionEnd { get; }
-    protected abstract int TestPartitionStart { get; }
-    protected abstract int TestPartitionEnd { get; }
-
+  public abstract class ArtificialRegressionDataDescriptor : RegressionDataDescriptor {
     public IRegressionProblemData GenerateRegressionData() {
       Dataset dataset = new Dataset(VariableNames, this.GenerateValues());
-
-      RegressionProblemData regData = new RegressionProblemData(dataset, AllowedInputVariables, TargetVariable);
-      regData.Name = this.Name;
-      regData.Description = this.Description;
-      regData.TrainingPartition.Start = this.TrainingPartitionStart;
-      regData.TrainingPartition.End = this.TrainingPartitionEnd;
-      regData.TestPartition.Start = this.TestPartitionStart;
-      regData.TestPartition.End = this.TestPartitionEnd;
-      return regData;
+      return GenerateRegressionData(dataset);
     }
 

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Regression/RealWorld/RealWorldRegressionInstanceProvider.cs

@@ -24 +24 @@
 using System.IO;
 using System.Linq;
-using HeuristicLab.Problems.DataAnalysis;
 using ICSharpCode.SharpZipLib.Zip;
 
@@ -47 +46 @@
 
     public override IEnumerable<IDataDescriptor> GetDataDescriptors() {
-      List<IRealWorldRegressionDataDescriptor> descriptorList = new List<IRealWorldRegressionDataDescriptor>();
+      List<ResourceRegressionDataDescriptor> descriptorList = new List<ResourceRegressionDataDescriptor>();
+      descriptorList.Add(new ChemicalOne());
+      descriptorList.Add(new Housing());
       descriptorList.Add(new Tower());
       var solutionsArchiveName = GetResourceName(FileName + @"\.zip");
@@ -59 +60 @@
           foreach (var entry in entries.OrderBy(x => x)) {
             string prettyName = Path.GetFileNameWithoutExtension(entry);
-            IRealWorldRegressionDataDescriptor desc = descriptorList.Where(x => x.Name.Equals(prettyName)).FirstOrDefault();
+            ResourceRegressionDataDescriptor desc = descriptorList.Where(x => x.Name.Equals(prettyName)).FirstOrDefault();
             if (desc != null) {
-              yield return new RealWorldResourceRegressionDataDescriptor(prettyName, desc.Description, entry, desc.Training, desc.Test);
+              desc.ResourceName = entry;
+              yield return desc;
             } else
-              yield return new ResourceRegressionDataDescriptor(prettyName, Description, entry);
+              throw new ArgumentNullException("No Descriptor could be found for this entry.");
           }
         }
       }
     }
-
-    public override IRegressionProblemData LoadData(IDataDescriptor id) {
-      var problem = base.LoadData(id);
-
-      var descriptor = id as RealWorldResourceRegressionDataDescriptor;
-
-      if (descriptor == null) {
-        return problem;
-      }
-
-      problem.TrainingPartition.Start = descriptor.Training.Start;
-      problem.TrainingPartition.End = descriptor.Training.End;
-      problem.TestPartition.Start = descriptor.Test.Start;
-      problem.TestPartition.End = descriptor.Test.End;
-
-      return problem;
-    }
   }
 }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Regression/RealWorld/Tower.cs

@@ -20 +20 @@
 #endregion
 
+using System;
 
-using System;
-using HeuristicLab.Data;
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  public class Tower : IRealWorldRegressionDataDescriptor {
-    public string Name { get { return "Tower"; } }
-    public string Description {
+  public class Tower : ResourceRegressionDataDescriptor {
+    public override string Name { get { return "Tower"; } }
+    public override string Description {
       get {
         return "Publication: Order of Nonlinearity as a Complexity Measure for Models " +
@@ -33 +32 @@
       }
     }
-    public IntRange Training { get { return new IntRange(0, 3135); } }
-    public IntRange Test { get { return new IntRange(3135, 4999); } }
+    protected override string TargetVariable { get { return "towerResponse"; } }
+    protected override string[] VariableNames {
+      get { return new string[] { "x1", "x2", "x3", "x4", "x5", "x6", "x7", "x8", "x9", "x10", "x11", "x12", "x13", "x14", "x15", "x16", "x17", "x18", "x19", "x20", "x21", "x22", "x23", "x24", "x25", "towerResponse" }; }
+    }
+    protected override string[] AllowedInputVariables {
+      get { return new string[] { "x1", "x2", "x3", "x4", "x5", "x6", "x7", "x8", "x9", "x10", "x11", "x12", "x13", "x14", "x15", "x16", "x17", "x18", "x19", "x20", "x21", "x22", "x23", "x24", "x25" }; }
+    }
+    protected override int TrainingPartitionStart { get { return 0; } }
+    protected override int TrainingPartitionEnd { get { return 3135; } }
+    protected override int TestPartitionStart { get { return 3135; } }
+    protected override int TestPartitionEnd { get { return 4999; } }
   }
 }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Regression/ResourceRegressionDataDescriptor.cs

@@ -20 +20 @@
 #endregion
 
+using System.Collections.Generic;
+using System.Linq;
 
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  internal class ResourceRegressionDataDescriptor : IDataDescriptor {
-    public string Name { get; internal set; }
-    public string Description { get; internal set; }
+  public abstract class ResourceRegressionDataDescriptor : RegressionDataDescriptor {
+    internal string ResourceName { get; set; }
 
-    internal string ResourceName { get; set; }
-    internal ResourceRegressionDataDescriptor(string name, string description, string resourceName) {
-      Name = name;
-      Description = description;
-      ResourceName = resourceName;
+    public bool CheckVariableNames(IEnumerable<string> VariableNames) {
+      return this.VariableNames.All(x => VariableNames.Contains(x));
     }
   }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Regression/ResourceRegressionInstanceProvider.cs

@@ -21 +21 @@
 
 using System;
-using System.Collections.Generic;
 using System.Globalization;
 using System.IO;
@@ -35 +34 @@
     protected abstract string FileName { get; }
 
-    public override IEnumerable<IDataDescriptor> GetDataDescriptors() {
-      var solutionsArchiveName = GetResourceName(FileName + @"\.zip");
-      if (!String.IsNullOrEmpty(solutionsArchiveName)) {
-        using (var solutionsZipFile = new ZipInputStream(GetType().Assembly.GetManifestResourceStream(solutionsArchiveName))) {
-          IList<string> entries = new List<string>();
-          ZipEntry curEntry;
-          while ((curEntry = solutionsZipFile.GetNextEntry()) != null) {
-            entries.Add(curEntry.Name);
-          }
-          foreach (var entry in entries.OrderBy(x => x)) {
-            yield return new ResourceRegressionDataDescriptor(Path.GetFileNameWithoutExtension(entry), Description, entry);
-          }
-        }
-      }
-    }
+    //public override IEnumerable<IDataDescriptor> GetDataDescriptors() {
+    //  var solutionsArchiveName = GetResourceName(FileName + @"\.zip");
+    //  if (!String.IsNullOrEmpty(solutionsArchiveName)) {
+    //    using (var solutionsZipFile = new ZipInputStream(GetType().Assembly.GetManifestResourceStream(solutionsArchiveName))) {
+    //      IList<string> entries = new List<string>();
+    //      ZipEntry curEntry;
+    //      while ((curEntry = solutionsZipFile.GetNextEntry()) != null) {
+    //        entries.Add(curEntry.Name);
+    //      }
+    //      foreach (var entry in entries.OrderBy(x => x)) {
+    //        yield return new ResourceRegressionDataDescriptor(Path.GetFileNameWithoutExtension(entry), Description, entry);
+    //      }
+    //    }
+    //  }
+    //}
 
     public override IRegressionProblemData LoadData(IDataDescriptor id) {
@@ -70 +69 @@
 
         Dataset dataset = new Dataset(csvFileParser.VariableNames, csvFileParser.Values);
-        string targetVar = csvFileParser.VariableNames.Where(x => dataset.DoubleVariables.Contains(x)).Last();
-        IEnumerable<string> allowedInputVars = dataset.DoubleVariables.Where(x => !x.Equals(targetVar));
+        if (!descriptor.CheckVariableNames(csvFileParser.VariableNames)) {
+          throw new ArgumentException("Parsed file contains variables which are not in the descriptor.");
+        }
 
-        IRegressionProblemData regData = new RegressionProblemData(dataset, allowedInputVars, targetVar);
-
-        int trainingPartEnd = csvFileParser.Rows * 2 / 3;
-        regData.TrainingPartition.Start = 0;
-        regData.TrainingPartition.End = trainingPartEnd;
-        regData.TestPartition.Start = trainingPartEnd;
-        regData.TestPartition.End = csvFileParser.Rows;
-
-        regData.Name = descriptor.Name;
-        regData.Description = descriptor.Description;
-        return regData;
+        return descriptor.GenerateRegressionData(dataset);
       }
     }