Changeset 13994 for branches/HeuristicLab.RegressionSolutionGradientView

Timestamp:

07/04/16 23:40:44 (8 years ago)

Author:

bburlacu

Message:

#2597: Merged trunk changes.

Location:

branches/HeuristicLab.RegressionSolutionGradientView

Files:

• HeuristicLab.Algorithms.DataAnalysis (modified) (1 prop)
• HeuristicLab.Algorithms.DataAnalysis.Views (modified) (1 prop)
• HeuristicLab.Algorithms.DataAnalysis/3.4/BaselineClassifiers/ZeroR.cs (modified) (1 diff)
• HeuristicLab.Algorithms.DataAnalysis/3.4/GBM/GradientBoostingRegressionAlgorithm.cs (modified) (2 diffs)
• HeuristicLab.Algorithms.DataAnalysis/3.4/GaussianProcess/StudentTProcessRegressionModelCreator.cs (deleted)
• HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/RegressionTreeBuilder.cs (modified) (4 diffs)
• HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/RegressionTreeModel.cs (modified) (1 diff)
• HeuristicLab.Algorithms.DataAnalysis/3.4/HeuristicLab.Algorithms.DataAnalysis-3.4.csproj (modified) (1 diff)
• HeuristicLab.Problems.DataAnalysis (modified) (1 prop)
• HeuristicLab.Problems.DataAnalysis.Views (modified) (1 prop)
• HeuristicLab.Problems.DataAnalysis.Views/3.4/Regression/RegressionSolutionErrorCharacteristicsCurveView.cs (modified) (1 diff)
• HeuristicLab.Problems.DataAnalysis.Views/3.4/Regression/RegressionSolutionScatterPlotView.cs (modified) (2 diffs)
• HeuristicLab.Problems.DataAnalysis/3.4/HeuristicLab.Problems.DataAnalysis-3.4.csproj (modified) (2 diffs)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Classification/ClassificationModel.cs (modified) (1 diff)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/DataAnalysisSolution.cs (modified) (1 diff)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/ConstantRegressionModel.cs (modified) (1 diff)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/RegressionModel.cs (modified) (1 diff)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/RegressionSolutionVariableImpactsCalculator.cs (modified) (3 diffs)
• HeuristicLab.Problems.DataAnalysis/3.4/Plugin.cs.frame (modified) (1 diff)

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Algorithms.DataAnalysis/3.4/BaselineClassifiers/ZeroR.cs

@@ -64 +64 @@
         .MaxItems(kvp => kvp.Value).Select(x => x.Key).First();
 
-      var model = new ConstantModel(dominantClass);
+      var model = new ConstantModel(dominantClass, target);
       var solution = model.CreateClassificationSolution(problemData);
       return solution;

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Algorithms.DataAnalysis/3.4/GBM/GradientBoostingRegressionAlgorithm.cs

@@ -181 +181 @@
     public GradientBoostingRegressionAlgorithm() {
       Problem = new RegressionProblem(); // default problem
-      var mctsSymbReg = new MctsSymbolicRegressionAlgorithm();
-      mctsSymbReg.Iterations = 10000;
-      mctsSymbReg.StoreAlgorithmInEachRun = false;
-      var sgp = CreateOSGP();
+      var osgp = CreateOSGP();
       var regressionAlgs = new ItemSet<IAlgorithm>(new IAlgorithm[] {
         new RandomForestRegression(),
-        sgp,
-        mctsSymbReg
+        osgp,
       });
       foreach (var alg in regressionAlgs) alg.Prepare();
@@ -208 +204 @@
         new DoubleValue(0.5)));
       Parameters.Add(new ConstrainedValueParameter<IAlgorithm>(RegressionAlgorithmParameterName,
-        "The regression algorithm to use as a base learner", regressionAlgs, mctsSymbReg));
+        "The regression algorithm to use as a base learner", regressionAlgs, osgp));
       Parameters.Add(new FixedValueParameter<StringValue>(RegressionAlgorithmSolutionResultParameterName,
         "The name of the solution produced by the regression algorithm", new StringValue("Solution")));

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/RegressionTreeBuilder.cs

@@ -22 +22 @@
 
 using System;
-using System.Collections;
 using System.Collections.Generic;
 using System.Diagnostics;
@@ -129 +128 @@
 
       // y and curPred are changed in gradient boosting
-      this.y = y; 
-      this.curPred = curPred; 
+      this.y = y;
+      this.curPred = curPred;
 
       // shuffle row idx
@@ -137 +136 @@
       int nRows = idx.Count();
 
-      // shuffle variable idx
+      // shuffle variable names
       HeuristicLab.Random.ListExtensions.ShuffleInPlace(allowedVariables, random);
 
@@ -176 +175 @@
       CreateRegressionTreeFromQueue(maxSize, lossFunction);
 
-      return new RegressionTreeModel(tree.ToArray());
+      return new RegressionTreeModel(tree.ToArray(), problemData.TargetVariable);
     }
 

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/RegressionTreeModel.cs

@@ -149 +149 @@
     #endregion
 
-
-
-
     [StorableConstructor]
     private RegressionTreeModel(bool serializing) : base(serializing) { }

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Algorithms.DataAnalysis/3.4/HeuristicLab.Algorithms.DataAnalysis-3.4.csproj

@@ -240 +240 @@
     </Compile>
     <Compile Include="FixedDataAnalysisAlgorithm.cs" />
-    <Compile Include="GaussianProcess\StudentTProcessRegressionModelCreator.cs" />
-    <Compile Include="GaussianProcess\StudentTProcessModel.cs" />
     <Compile Include="GaussianProcess\CovarianceFunctions\CovarianceSpectralMixture.cs" />
     <Compile Include="GaussianProcess\CovarianceFunctions\CovariancePiecewisePolynomial.cs" />

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Problems.DataAnalysis.Views/3.4/Regression/RegressionSolutionErrorCharacteristicsCurveView.cs

@@ -292 +292 @@
     private IRegressionSolution CreateConstantSolution() {
       double averageTrainingTarget = ProblemData.Dataset.GetDoubleValues(ProblemData.TargetVariable, ProblemData.TrainingIndices).Average();
-      var model = new ConstantModel(averageTrainingTarget);
+      var model = new ConstantModel(averageTrainingTarget, ProblemData.TargetVariable);
       var solution = model.CreateRegressionSolution(ProblemData);
       solution.Name = "Baseline (constant)";

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Problems.DataAnalysis.Views/3.4/Regression/RegressionSolutionScatterPlotView.cs

@@ -36 +36 @@
     private const string TEST_SERIES = "Test samples";
 
+    private const int OPACITY_LEVEL = 150;
+
     public new IRegressionSolution Content {
       get { return (IRegressionSolution)base.Content; }
@@ -60 +62 @@
       this.chart.TextAntiAliasingQuality = TextAntiAliasingQuality.High;
       this.chart.AxisViewChanged += new EventHandler<System.Windows.Forms.DataVisualization.Charting.ViewEventArgs>(chart_AxisViewChanged);
+
+      //make series colors semi transparent
+      this.chart.ApplyPaletteColors();
+      this.chart.Series[ALL_SERIES].Color = Color.FromArgb(OPACITY_LEVEL, this.chart.Series[ALL_SERIES].Color);
+      this.chart.Series[TRAINING_SERIES].Color = Color.FromArgb(OPACITY_LEVEL, this.chart.Series[TRAINING_SERIES].Color);
+      this.chart.Series[TEST_SERIES].Color = Color.FromArgb(OPACITY_LEVEL, this.chart.Series[TEST_SERIES].Color);
+
+      //change all markers to circles
+      this.chart.Series[ALL_SERIES].MarkerStyle = MarkerStyle.Circle;
+      this.chart.Series[TRAINING_SERIES].MarkerStyle = MarkerStyle.Circle;
+      this.chart.Series[TEST_SERIES].MarkerStyle = MarkerStyle.Circle;
 
       //configure axis 

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Problems.DataAnalysis/3.4/HeuristicLab.Problems.DataAnalysis-3.4.csproj

@@ -151 +151 @@
       <SpecificVersion>False</SpecificVersion>
       <HintPath>..\..\..\..\trunk\sources\bin\HeuristicLab.Problems.Instances-3.3.dll</HintPath>
+      <Private>False</Private>
+    </Reference>
+    <Reference Include="HeuristicLab.Random-3.3, Version=3.3.0.0, Culture=neutral, PublicKeyToken=ba48961d6f65dcec, processorArchitecture=MSIL">
+      <SpecificVersion>False</SpecificVersion>
+      <HintPath>..\..\..\..\trunk\sources\bin\HeuristicLab.Random-3.3.dll</HintPath>
       <Private>False</Private>
     </Reference>
@@ -317 +322 @@
     </BootstrapperPackage>
   </ItemGroup>
+  <ItemGroup>
+    <ProjectReference Include="..\..\HeuristicLab.Collections\3.3\HeuristicLab.Collections-3.3.csproj">
+      <Project>{958B43BC-CC5C-4FA2-8628-2B3B01D890B6}</Project>
+      <Name>HeuristicLab.Collections-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+    <ProjectReference Include="..\..\HeuristicLab.Common.Resources\3.3\HeuristicLab.Common.Resources-3.3.csproj">
+      <Project>{0E27A536-1C4A-4624-A65E-DC4F4F23E3E1}</Project>
+      <Name>HeuristicLab.Common.Resources-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+    <ProjectReference Include="..\..\HeuristicLab.Common\3.3\HeuristicLab.Common-3.3.csproj">
+      <Project>{A9AD58B9-3EF9-4CC1-97E5-8D909039FF5C}</Project>
+      <Name>HeuristicLab.Common-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+    <ProjectReference Include="..\..\HeuristicLab.Core\3.3\HeuristicLab.Core-3.3.csproj">
+      <Project>{C36BD924-A541-4A00-AFA8-41701378DDC5}</Project>
+      <Name>HeuristicLab.Core-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+    <ProjectReference Include="..\..\HeuristicLab.Data\3.3\HeuristicLab.Data-3.3.csproj">
+      <Project>{BBAB9DF5-5EF3-4BA8-ADE9-B36E82114937}</Project>
+      <Name>HeuristicLab.Data-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+    <ProjectReference Include="..\..\HeuristicLab.Optimization\3.3\HeuristicLab.Optimization-3.3.csproj">
+      <Project>{14AB8D24-25BC-400C-A846-4627AA945192}</Project>
+      <Name>HeuristicLab.Optimization-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+    <ProjectReference Include="..\..\HeuristicLab.Parameters\3.3\HeuristicLab.Parameters-3.3.csproj">
+      <Project>{56F9106A-079F-4C61-92F6-86A84C2D84B7}</Project>
+      <Name>HeuristicLab.Parameters-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+    <ProjectReference Include="..\..\HeuristicLab.Persistence\3.3\HeuristicLab.Persistence-3.3.csproj">
+      <Project>{102BC7D3-0EF9-439C-8F6D-96FF0FDB8E1B}</Project>
+      <Name>HeuristicLab.Persistence-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+    <ProjectReference Include="..\..\HeuristicLab.PluginInfrastructure\3.3\HeuristicLab.PluginInfrastructure-3.3.csproj">
+      <Project>{94186A6A-5176-4402-AE83-886557B53CCA}</Project>
+      <Name>HeuristicLab.PluginInfrastructure-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+    <ProjectReference Include="..\..\HeuristicLab.Problems.Instances\3.3\HeuristicLab.Problems.Instances-3.3.csproj">
+      <Project>{3540E29E-4793-49E7-8EE2-FEA7F61C3994}</Project>
+      <Name>HeuristicLab.Problems.Instances-3.3</Name>
+      <Private>False</Private>
+    </ProjectReference>
+  </ItemGroup>
   <ItemGroup />
   <Import Project="$(MSBuildToolsPath)\Microsoft.CSharp.targets" />

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Classification/ClassificationModel.cs

@@ -55 +55 @@
     }
 
+    [StorableHook(HookType.AfterDeserialization)]
+    private void AfterDeserialization() {
+      // BackwardsCompatibility3.3
+      #region Backwards compatible code, remove with 3.4
+      targetVariable = string.Empty;
+      #endregion
+    }
+
     public abstract IEnumerable<double> GetEstimatedClassValues(IDataset dataset, IEnumerable<int> rows);
     public abstract IClassificationSolution CreateClassificationSolution(IClassificationProblemData problemData);

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/DataAnalysisSolution.cs

@@ -89 +89 @@
     protected abstract void RecalculateResults();
 
+    public virtual IEnumerable<string> GetUsedVariablesForPrediction() {
+      return this.ProblemData.AllowedInputVariables;
+    }
+
     private void ProblemData_Changed(object sender, EventArgs e) {
       OnProblemDataChanged();

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/ConstantRegressionModel.cs

@@ -64 +64 @@
 
     public override IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
-      return new ConstantRegressionSolution(new ConstantModel(constant), new RegressionProblemData(problemData));
+      return new ConstantRegressionSolution(new ConstantModel(constant, TargetVariable), new RegressionProblemData(problemData));
     }
 

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/RegressionModel.cs

@@ -56 +56 @@
     }
 
+    [StorableHook(HookType.AfterDeserialization)]
+    private void AfterDeserialization() {
+      // BackwardsCompatibility3.3
+      #region Backwards compatible code, remove with 3.4
+      targetVariable = string.Empty;
+      #endregion
+    }
     public abstract IEnumerable<double> GetEstimatedValues(IDataset dataset, IEnumerable<int> rows);
     public abstract IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData);

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/RegressionSolutionVariableImpactsCalculator.cs

@@ -30 +30 @@
 using HeuristicLab.Parameters;
 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
+using HeuristicLab.Random;
 
 namespace HeuristicLab.Problems.DataAnalysis {
   [StorableClass]
-  [Item("RegressionSolution Impacts Calculator", "Calculation of the impacts of input variables for a concrete ")]
+  [Item("RegressionSolution Impacts Calculator", "Calculation of the impacts of input variables for any regression solution")]
   public sealed class RegressionSolutionVariableImpactsCalculator : ParameterizedNamedItem {
     public enum ReplacementMethodEnum {
       Median,
-      Average
+      Average,
+      Shuffle,
+      Noise
     }
 
@@ -77 +80 @@
       : base() {
       Parameters.Add(new FixedValueParameter<EnumValue<ReplacementMethodEnum>>(ReplacementParameterName, "The replacement method for variables during impact calculation.", new EnumValue<ReplacementMethodEnum>(ReplacementMethodEnum.Median)));
-      Parameters.Add(new FixedValueParameter<EnumValue<DataPartitionEnum>>(ReplacementParameterName, "The data partition on which the impacts are calculated.", new EnumValue<DataPartitionEnum>(DataPartitionEnum.Training)));
+      Parameters.Add(new FixedValueParameter<EnumValue<DataPartitionEnum>>(DataPartitionParameterName, "The data partition on which the impacts are calculated.", new EnumValue<DataPartitionEnum>(DataPartitionEnum.Training)));
     }
 
@@ -138 +141 @@
       var originalValues = dataset.GetReadOnlyDoubleValues(variable).ToList();
       double replacementValue;
+      List<double> replacementValues;
+      IRandom rand;
 
       switch (replacement) {
         case ReplacementMethodEnum.Median:
           replacementValue = rows.Select(r => originalValues[r]).Median();
+          replacementValues = Enumerable.Repeat(replacementValue, dataset.Rows).ToList();
           break;
         case ReplacementMethodEnum.Average:
           replacementValue = rows.Select(r => originalValues[r]).Average();
+          replacementValues = Enumerable.Repeat(replacementValue, dataset.Rows).ToList();
           break;
+        case ReplacementMethodEnum.Shuffle:
+          // new var has same empirical distribution but the relation to y is broken
+          rand = new FastRandom(31415);
+          replacementValues = rows.Select(r => originalValues[r]).Shuffle(rand).ToList();
+          break;
+        case ReplacementMethodEnum.Noise:
+          var avg = rows.Select(r => originalValues[r]).Average();
+          var stdDev = rows.Select(r => originalValues[r]).StandardDeviation();
+          rand = new FastRandom(31415);
+          replacementValues = rows.Select(_ => NormalDistributedRandom.NextDouble(rand, avg, stdDev)).ToList();
+          break;
+
         default:
           throw new ArgumentException(string.Format("ReplacementMethod {0} cannot be handled.", replacement));
       }
 
-      dataset.ReplaceVariable(variable, Enumerable.Repeat(replacementValue, dataset.Rows).ToList());
+      dataset.ReplaceVariable(variable, replacementValues);
       //mkommend: ToList is used on purpose to avoid lazy evaluation that could result in wrong estimates due to variable replacements
       var estimates = model.GetEstimatedValues(dataset, rows).ToList();

branches/HeuristicLab.RegressionSolutionGradientView/HeuristicLab.Problems.DataAnalysis/3.4/Plugin.cs.frame

@@ -38 +38 @@
   [PluginDependency("HeuristicLab.Persistence", "3.3")]
   [PluginDependency("HeuristicLab.Problems.Instances", "3.3")]  
+  [PluginDependency("HeuristicLab.Random", "3.3")]
   public class HeuristicLabProblemsDataAnalysisPlugin : PluginBase {
   }