Changeset 9208 for trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification

Timestamp:

02/06/13 12:30:13 (12 years ago)

Author:

sforsten

Message:

#1941:

• added wisconsin breast cancer problem instance
• corrected iris dataset
• changed classification data descriptors to be able to set training and test partition as well as input and target variables (in the same way as it is done in regression)

Location:

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

Files:

• ClassificationDataDescriptor.cs (added)
• Data/UCI.zip (modified) (previous)
• ResourceClassificationDataDescriptor.cs (modified) (1 diff)
• ResourceClassificationInstanceProvider.cs (modified) (3 diffs)
• UCI/IUCIDataDescriptor.cs (deleted)
• UCI/Iris.cs (modified) (2 diffs)
• UCI/Mammography.cs (modified) (2 diffs)
• UCI/Parkinson.cs (modified) (2 diffs)
• UCI/Thyroid.cs (modified) (2 diffs)
• UCI/UCIDataDescriptor.cs (added)
• UCI/UCIInstanceProvider.cs (modified) (3 diffs)
• UCI/Vertebral_3C.cs (modified) (2 diffs)
• UCI/Wine.cs (modified) (2 diffs)
• UCI/WisconsinDiagnosticBreastCancer.cs (added)

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification/ResourceClassificationDataDescriptor.cs

@@ -20 +20 @@
 #endregion
 
+using System.Collections.Generic;
+using System.Linq;
 
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  internal class ResourceClassificationDataDescriptor : IDataDescriptor {
-    public string Name { get; internal set; }
-    public string Description { get; internal set; }
+  public abstract class ResourceClassificationDataDescriptor : ClassificationDataDescriptor {
+    internal string ResourceName { get; set; }
 
-    internal string ResourceName { get; set; }
-    internal ResourceClassificationDataDescriptor(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/Classification/ResourceClassificationInstanceProvider.cs

@@ -21 +21 @@
 
 using System;
-using System.Collections.Generic;
 using System.Globalization;
 using System.IO;
@@ -34 +33 @@
 
     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 ResourceClassificationDataDescriptor(Path.GetFileNameWithoutExtension(entry), Description, entry);
-          }
-        }
-      }
-    }
 
     public override IClassificationProblemData LoadData(IDataDescriptor id) {
@@ -70 +53 @@
 
         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.");
+        }
 
-        ClassificationProblemData claData = new ClassificationProblemData(dataset, allowedInputVars, targetVar);
-
-        int trainingPartEnd = csvFileParser.Rows * 2 / 3;
-        claData.TrainingPartition.Start = 0;
-        claData.TrainingPartition.End = trainingPartEnd;
-        claData.TestPartition.Start = trainingPartEnd;
-        claData.TestPartition.End = csvFileParser.Rows;
-
-        claData.Name = descriptor.Name;
-        claData.Description = descriptor.Description;
-        return claData;
+        return descriptor.GenerateClassificationData(dataset);
       }
     }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification/UCI/Iris.cs

@@ -23 +23 @@
 using System;
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  public class Iris : IUCIDataDescriptor {
-    public string Name { get { return "Iris"; } }
-    public string Description {
+  public class Iris : UCIDataDescriptor {
+    public override string Filename { get { return "Iris"; } }
+    public override string Description {
       get {
         return "Data Set Information:" + Environment.NewLine
@@ -35 +35 @@
         + "other 2; the latter are NOT linearly separable from each other." + Environment.NewLine
         + "Predicted attribute: class of iris plant." + Environment.NewLine
-        + "This is an exceedingly simple domain.";
+        + "This is an exceedingly simple domain." + Environment.NewLine + Environment.NewLine
+        + "The classes have been converted in the following way" + Environment.NewLine
+        + "Iris-setosa     = 0" + Environment.NewLine
+        + "Iris-versicolor = 1" + Environment.NewLine
+        + "Iris-virginica  = 2";
       }
     }
-    public string Donor { get { return "M. Marshall"; } }
-    public int Year { get { return 1988; } }
+    public override string Donor { get { return "M. Marshall"; } }
+    public override int Year { get { return 1988; } }
+
+    protected override string TargetVariable { get { return "class"; } }
+    protected override string[] VariableNames {
+      get { return new string[] { "sepal_length", "sepal_width", "petal_length", "petal_width", "class" }; }
+    }
+    protected override string[] AllowedInputVariables {
+      get { return new string[] { "sepal_length", "sepal_width", "petal_length", "petal_width" }; }
+    }
+    protected override int TrainingPartitionStart { get { return 0; } }
+    protected override int TrainingPartitionEnd { get { return 100; } }
+    protected override int TestPartitionStart { get { return 100; } }
+    protected override int TestPartitionEnd { get { return 150; } }
   }
 }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification/UCI/Mammography.cs

@@ -23 +23 @@
 using System;
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  public class Mammography : IUCIDataDescriptor {
-    public string Name { get { return "Mammography"; } }
-    public string Description {
+  public class Mammography : UCIDataDescriptor {
+    public override string Filename { get { return "Mammography"; } }
+    public override string Description {
       get {
         return "Data Set Information: Features are computed from a digitized image of a fine needle aspirate (FNA) of a breast mass."
@@ -46 +46 @@
       }
     }
-    public string Donor { get { return "M. Elter"; } }
-    public int Year { get { return 2007; } }
+    public override string Donor { get { return "M. Elter"; } }
+    public override int Year { get { return 2007; } }
+
+    protected override string TargetVariable { get { return "Severity"; } }
+    protected override string[] VariableNames {
+      get { return new string[] { "BI-RADS", "Age", "Shape", "Margin", "Density", "Severity" }; }
+    }
+    protected override string[] AllowedInputVariables {
+      get { return new string[] { "BI-RADS", "Age", "Shape", "Margin", "Density" }; }
+    }
+    protected override int TrainingPartitionStart { get { return 0; } }
+    protected override int TrainingPartitionEnd { get { return 641; } }
+    protected override int TestPartitionStart { get { return 641; } }
+    protected override int TestPartitionEnd { get { return 961; } }
   }
 }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification/UCI/Parkinson.cs

@@ -23 +23 @@
 
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  public class Parkinson : IUCIDataDescriptor {
-    public string Name { get { return "Parkinson"; } }
-    public string Description {
+  public class Parkinson : UCIDataDescriptor {
+    public override string Filename { get { return "Parkinson"; } }
+    public override string Description {
       get {
         return "Data Set Information:" + Environment.NewLine
@@ -40 +40 @@
       }
     }
-    public string Donor { get { return "M. Little"; } }
-    public int Year { get { return 2008; } }
+    public override string Donor { get { return "M. Little"; } }
+    public override int Year { get { return 2008; } }
+
+    protected override string TargetVariable { get { return "status"; } }
+    protected override string[] VariableNames {
+      get { return new string[] { "MDVP:Fo(Hz)", "MDVP:Fhi(Hz)", "MDVP:Flo(Hz)", "MDVP:Jitter(%)", "MDVP:Jitter(Abs)", "MDVP:RAP", "MDVP:PPQ", "Jitter:DDP", "MDVP:Shimmer", "MDVP:Shimmer(dB)", "Shimmer:APQ3", "Shimmer:APQ5", "MDVP:APQ", "Shimmer:DDA", "NHR", "HNR", "RPDE", "DFA", "spread1", "spread2", "D2", "PPE", "status" }; }
+    }
+    protected override string[] AllowedInputVariables {
+      get { return new string[] { "MDVP:Fo(Hz)", "MDVP:Fhi(Hz)", "MDVP:Flo(Hz)", "MDVP:Jitter(%)", "MDVP:Jitter(Abs)", "MDVP:RAP", "MDVP:PPQ", "Jitter:DDP", "MDVP:Shimmer", "MDVP:Shimmer(dB)", "Shimmer:APQ3", "Shimmer:APQ5", "MDVP:APQ", "Shimmer:DDA", "NHR", "HNR", "RPDE", "DFA", "spread1", "spread2", "D2", "PPE" }; }
+    }
+    protected override int TrainingPartitionStart { get { return 0; } }
+    protected override int TrainingPartitionEnd { get { return 130; } }
+    protected override int TestPartitionStart { get { return 130; } }
+    protected override int TestPartitionEnd { get { return 195; } }
   }
 }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification/UCI/Thyroid.cs

@@ -23 +23 @@
 using System;
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  public class Thyroid : IUCIDataDescriptor {
-    public string Name { get { return "Thyroid"; } }
-    public string Description {
+  public class Thyroid : UCIDataDescriptor {
+    public override string Filename { get { return "Thyroid"; } }
+    public override string Description {
       get {
         return "Thyroid gland data. ('normal', hypo and hyper functioning)" + Environment.NewLine + Environment.NewLine +
@@ -39 +39 @@
       }
     }
-    public string Donor { get { return "S. Aeberhard"; } }
-    public int Year { get { return 1992; } }
+    public override string Donor { get { return "S. Aeberhard"; } }
+    public override int Year { get { return 1992; } }
+
+    protected override string TargetVariable { get { return "X000"; } }
+    protected override string[] VariableNames {
+      get { return new string[] { "X001", "X002", "X003", "X004", "X005", "X000" }; }
+    }
+    protected override string[] AllowedInputVariables {
+      get { return new string[] { "X001", "X002", "X003", "X004", "X005" }; }
+    }
+    protected override int TrainingPartitionStart { get { return 0; } }
+    protected override int TrainingPartitionEnd { get { return 143; } }
+    protected override int TestPartitionStart { get { return 143; } }
+    protected override int TestPartitionEnd { get { return 215; } }
   }
 }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification/UCI/UCIInstanceProvider.cs

@@ -46 +46 @@
 
     public override IEnumerable<IDataDescriptor> GetDataDescriptors() {
-      List<IUCIDataDescriptor> descriptorList = new List<IUCIDataDescriptor>();
+      List<UCIDataDescriptor> descriptorList = new List<UCIDataDescriptor>();
       descriptorList.Add(new Iris());
       descriptorList.Add(new Mammography());
@@ -53 +53 @@
       descriptorList.Add(new Vertebral_3C());
       descriptorList.Add(new Wine());
+      descriptorList.Add(new WisconsinDiagnosticBreastCancer());
       var solutionsArchiveName = GetResourceName(FileName + @"\.zip");
       if (!String.IsNullOrEmpty(solutionsArchiveName)) {
@@ -62 +63 @@
           }
           foreach (var entry in entries.OrderBy(x => x)) {
-            string prettyName = Path.GetFileNameWithoutExtension(entry);
-            IUCIDataDescriptor desc = descriptorList.Where(x => x.Name.Equals(prettyName)).FirstOrDefault();
+            string filename = Path.GetFileNameWithoutExtension(entry);
+            UCIDataDescriptor desc = descriptorList.Where(x => x.Filename.Equals(filename)).FirstOrDefault();
             if (desc != null) {
-              prettyName = String.Format("{0}, {1}, {2}", prettyName, desc.Donor, desc.Year);
-              yield return new ResourceClassificationDataDescriptor(prettyName, desc.Description, entry);
+              desc.ResourceName = entry;
+              yield return desc;
             } else
-              yield return new ResourceClassificationDataDescriptor(prettyName, Description, entry);
+              throw new ArgumentNullException("No Descriptor could be found for this entry.");
           }
         }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification/UCI/Vertebral_3C.cs

@@ -23 +23 @@
 
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  public class Vertebral_3C : IUCIDataDescriptor {
-    public string Name { get { return "Vertebral_3C"; } }
-    public string Description {
+  public class Vertebral_3C : UCIDataDescriptor {
+    public override string Filename { get { return "Vertebral_3C"; } }
+    public override string Description {
       get {
         return "Data Set Information:" + Environment.NewLine
@@ -40 +40 @@
       }
     }
-    public string Donor { get { return "H. da Mota"; } }
-    public int Year { get { return 2011; } }
+    public override string Donor { get { return "H. da Mota"; } }
+    public override int Year { get { return 2011; } }
+
+    protected override string TargetVariable { get { return "class"; } }
+    protected override string[] VariableNames {
+      get { return new string[] { "pelvic_incidence", "pelvic_tilt", "lumbar_lordosis_angle", "sacral_slope", "pelvic_radius", "degree_1", "class" }; }
+    }
+    protected override string[] AllowedInputVariables {
+      get { return new string[] { "pelvic_incidence", "pelvic_tilt", "lumbar_lordosis_angle", "sacral_slope", "pelvic_radius", "degree_1" }; }
+    }
+    protected override int TrainingPartitionStart { get { return 0; } }
+    protected override int TrainingPartitionEnd { get { return 207; } }
+    protected override int TestPartitionStart { get { return 207; } }
+    protected override int TestPartitionEnd { get { return 310; } }
   }
 }

trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification/UCI/Wine.cs

@@ -23 +23 @@
 using System;
 namespace HeuristicLab.Problems.Instances.DataAnalysis {
-  public class Wine : IUCIDataDescriptor {
-    public string Name { get { return "Wine"; } }
-    public string Description {
+  public class Wine : UCIDataDescriptor {
+    public override string Filename { get { return "Wine"; } }
+    public override string Description {
       get {
         return "These data are the results of a chemical analysis of wines grown in the same region " +
@@ -51 +51 @@
       }
     }
-    public string Donor { get { return "S. Aeberhard"; } }
-    public int Year { get { return 1991; } }
+    public override string Donor { get { return "S. Aeberhard"; } }
+    public override int Year { get { return 1991; } }
+
+    protected override string TargetVariable { get { return "Class"; } }
+    protected override string[] VariableNames {
+      get { return new string[] { "Alcohol", "Malic acid", "Ash", "Alcalinity of ash", "Magnesium", "Total phenols", "Flavanoids", "Nonflavanoid phenols", "Proanthocyanins", "Color intensity", "Hue", "OD280/OD315 of diluted wines", "Proline", "Class" }; }
+    }
+    protected override string[] AllowedInputVariables {
+      get { return new string[] { "Alcohol", "Malic acid", "Ash", "Alcalinity of ash", "Magnesium", "Total phenols", "Flavanoids", "Nonflavanoid phenols", "Proanthocyanins", "Color intensity", "Hue", "OD280/OD315 of diluted wines", "Proline" }; }
+    }
+    protected override int TrainingPartitionStart { get { return 0; } }
+    protected override int TrainingPartitionEnd { get { return 119; } }
+    protected override int TestPartitionStart { get { return 119; } }
+    protected override int TestPartitionEnd { get { return 178; } }
   }
 }