- Timestamp:
- 07/06/17 09:22:48 (7 years ago)
- Location:
- stable
- Files:
-
- 17 edited
- 10 copied
Legend:
- Unmodified
- Added
- Removed
-
stable/HeuristicLab.Algorithms.DataAnalysis.Views/3.4/HeuristicLab.Algorithms.DataAnalysis.Views-3.4.csproj
r14166 r15127 125 125 </ItemGroup> 126 126 <ItemGroup> 127 <Compile Include="RandomForestClassificationSolutionView.cs"> 128 <SubType>UserControl</SubType> 129 </Compile> 130 <Compile Include="RandomForestClassificationSolutionView.Designer.cs"> 131 <DependentUpon>RandomForestClassificationSolutionView.cs</DependentUpon> 132 </Compile> 133 <Compile Include="RandomForestModelView.cs"> 134 <SubType>UserControl</SubType> 135 </Compile> 136 <Compile Include="RandomForestModelView.Designer.cs"> 137 <DependentUpon>RandomForestModelView.cs</DependentUpon> 138 </Compile> 139 <Compile Include="RandomForestRegressionSolutionView.cs"> 140 <SubType>UserControl</SubType> 141 </Compile> 142 <Compile Include="RandomForestRegressionSolutionView.Designer.cs"> 143 <DependentUpon>RandomForestRegressionSolutionView.cs</DependentUpon> 144 </Compile> 145 <Compile Include="GradientBoostedTreesModelView.cs"> 146 <SubType>UserControl</SubType> 147 </Compile> 148 <Compile Include="GradientBoostedTreesModelView.Designer.cs"> 149 <DependentUpon>GradientBoostedTreesModelView.cs</DependentUpon> 150 </Compile> 127 151 <Compile Include="MeanProdView.cs"> 128 152 <SubType>UserControl</SubType> … … 192 216 <Compile Include="SupportVectorMachineModelView.Designer.cs"> 193 217 <DependentUpon>SupportVectorMachineModelView.cs</DependentUpon> 218 </Compile> 219 <Compile Include="GradientBoostedTreesSolutionView.cs"> 220 <SubType>UserControl</SubType> 221 </Compile> 222 <Compile Include="GradientBoostedTreesSolutionView.Designer.cs"> 223 <DependentUpon>GradientBoostedTreesSolutionView.cs</DependentUpon> 194 224 </Compile> 195 225 </ItemGroup> … … 245 275 <Private>False</Private> 246 276 </ProjectReference> 277 <ProjectReference Include="..\..\HeuristicLab.Encodings.SymbolicExpressionTreeEncoding\3.4\HeuristicLab.Encodings.SymbolicExpressionTreeEncoding-3.4.csproj"> 278 <Project>{06D4A186-9319-48A0-BADE-A2058D462EEA}</Project> 279 <Name>HeuristicLab.Encodings.SymbolicExpressionTreeEncoding-3.4</Name> 280 </ProjectReference> 247 281 <ProjectReference Include="..\..\HeuristicLab.MainForm.WindowsForms\3.3\HeuristicLab.MainForm.WindowsForms-3.3.csproj"> 248 282 <Project>{AB687BBE-1BFE-476B-906D-44237135431D}</Project> … … 269 303 <Name>HeuristicLab.PluginInfrastructure-3.3</Name> 270 304 <Private>False</Private> 305 </ProjectReference> 306 <ProjectReference Include="..\..\HeuristicLab.Problems.DataAnalysis.Symbolic.Classification\3.4\HeuristicLab.Problems.DataAnalysis.Symbolic.Classification-3.4.csproj"> 307 <Project>{05BAE4E1-A9FA-4644-AA77-42558720159E}</Project> 308 <Name>HeuristicLab.Problems.DataAnalysis.Symbolic.Classification-3.4</Name> 309 </ProjectReference> 310 <ProjectReference Include="..\..\HeuristicLab.Problems.DataAnalysis.Symbolic.Regression\3.4\HeuristicLab.Problems.DataAnalysis.Symbolic.Regression-3.4.csproj"> 311 <Project>{5AC82412-911B-4FA2-A013-EDC5E3F3FCC2}</Project> 312 <Name>HeuristicLab.Problems.DataAnalysis.Symbolic.Regression-3.4</Name> 313 </ProjectReference> 314 <ProjectReference Include="..\..\HeuristicLab.Problems.DataAnalysis.Symbolic.Views\3.4\HeuristicLab.Problems.DataAnalysis.Symbolic.Views-3.4.csproj"> 315 <Project>{7a2531ce-3f7c-4f13-bcca-ed6dc27a7086}</Project> 316 <Name>HeuristicLab.Problems.DataAnalysis.Symbolic.Views-3.4</Name> 317 </ProjectReference> 318 <ProjectReference Include="..\..\HeuristicLab.Problems.DataAnalysis.Symbolic\3.4\HeuristicLab.Problems.DataAnalysis.Symbolic-3.4.csproj"> 319 <Project>{3d28463f-ec96-4d82-afee-38be91a0ca00}</Project> 320 <Name>HeuristicLab.Problems.DataAnalysis.Symbolic-3.4</Name> 271 321 </ProjectReference> 272 322 <ProjectReference Include="..\..\HeuristicLab.Problems.DataAnalysis.Views\3.4\HeuristicLab.Problems.DataAnalysis.Views-3.4.csproj"> -
stable/HeuristicLab.Algorithms.DataAnalysis.Views/3.4/KMeansClusteringModelView.cs
r14186 r15127 19 19 */ 20 20 #endregion 21 using System;22 21 using System.Linq; 23 using System.IO;24 using System.Windows.Forms;25 22 using HeuristicLab.MainForm; 26 23 using HeuristicLab.MainForm.WindowsForms; -
stable/HeuristicLab.Algorithms.DataAnalysis.Views/3.4/Plugin.cs.frame
r14190 r15127 37 37 [PluginDependency("HeuristicLab.Data", "3.3")] 38 38 [PluginDependency("HeuristicLab.Data.Views", "3.3")] 39 [PluginDependency("HeuristicLab.Encodings.SymbolicExpressionTreeEncoding", "3.4")] 39 40 [PluginDependency("HeuristicLab.LibSVM", "3.12")] 40 41 [PluginDependency("HeuristicLab.MainForm", "3.3")] … … 44 45 [PluginDependency("HeuristicLab.Problems.DataAnalysis", "3.4")] 45 46 [PluginDependency("HeuristicLab.Problems.DataAnalysis.Views", "3.4")] 47 [PluginDependency("HeuristicLab.Problems.DataAnalysis.Symbolic", "3.4")] 48 [PluginDependency("HeuristicLab.Problems.DataAnalysis.Symbolic.Regression", "3.4")] 49 [PluginDependency("HeuristicLab.Problems.DataAnalysis.Symbolic.Classification", "3.4")] 46 50 public class HeuristicLabAlgorithmsDataAnalysisViewsPlugin : PluginBase { 47 51 } -
stable/HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/GradientBoostedTreesAlgorithm.cs
r15061 r15127 270 270 } else { 271 271 // otherwise we produce a regression solution 272 Results.Add(new Result("Solution", new RegressionSolution(model, problemData)));272 Results.Add(new Result("Solution", new GradientBoostedTreesSolution(model, problemData))); 273 273 } 274 274 } -
stable/HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/GradientBoostedTreesSolution.cs
r14186 r15127 20 20 #endregion 21 21 22 using System.Collections.Generic;23 using System.Linq;24 22 using HeuristicLab.Common; 25 23 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable; -
stable/HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/RegressionTreeModel.cs
r14186 r15127 28 28 using HeuristicLab.Common; 29 29 using HeuristicLab.Core; 30 using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding; 30 31 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable; 31 32 using HeuristicLab.Problems.DataAnalysis; 33 using HeuristicLab.Problems.DataAnalysis.Symbolic; 34 using HeuristicLab.Problems.DataAnalysis.Symbolic.Regression; 32 35 33 36 namespace HeuristicLab.Algorithms.DataAnalysis { … … 210 213 } 211 214 215 /// <summary> 216 /// Transforms the tree model to a symbolic regression solution 217 /// </summary> 218 /// <param name="problemData"></param> 219 /// <returns>A new symbolic regression solution which matches the tree model</returns> 220 public ISymbolicRegressionSolution CreateSymbolicRegressionSolution(IRegressionProblemData problemData) { 221 var rootSy = new ProgramRootSymbol(); 222 var startSy = new StartSymbol(); 223 var varCondSy = new VariableCondition() { IgnoreSlope = true }; 224 var constSy = new Constant(); 225 226 var startNode = startSy.CreateTreeNode(); 227 startNode.AddSubtree(CreateSymbolicRegressionTreeRecursive(tree, 0, varCondSy, constSy)); 228 var rootNode = rootSy.CreateTreeNode(); 229 rootNode.AddSubtree(startNode); 230 var model = new SymbolicRegressionModel(TargetVariable, new SymbolicExpressionTree(rootNode), new SymbolicDataAnalysisExpressionTreeLinearInterpreter()); 231 return model.CreateRegressionSolution(problemData); 232 } 233 234 private ISymbolicExpressionTreeNode CreateSymbolicRegressionTreeRecursive(TreeNode[] treeNodes, int nodeIdx, VariableCondition varCondSy, Constant constSy) { 235 var curNode = treeNodes[nodeIdx]; 236 if (curNode.VarName == TreeNode.NO_VARIABLE) { 237 var node = (ConstantTreeNode)constSy.CreateTreeNode(); 238 node.Value = curNode.Val; 239 return node; 240 } else { 241 var node = (VariableConditionTreeNode)varCondSy.CreateTreeNode(); 242 node.VariableName = curNode.VarName; 243 node.Threshold = curNode.Val; 244 245 var left = CreateSymbolicRegressionTreeRecursive(treeNodes, curNode.LeftIdx, varCondSy, constSy); 246 var right = CreateSymbolicRegressionTreeRecursive(treeNodes, curNode.RightIdx, varCondSy, constSy); 247 node.AddSubtree(left); 248 node.AddSubtree(right); 249 return node; 250 } 251 } 252 253 212 254 private string TreeToString(int idx, string part) { 213 255 var n = tree[idx]; -
stable/HeuristicLab.Algorithms.DataAnalysis/3.4/Interfaces/IRandomForestClassificationSolution.cs
r14186 r15127 30 30 public interface IRandomForestClassificationSolution : IClassificationSolution { 31 31 new IRandomForestModel Model { get; } 32 int NumberOfTrees { get; } 32 33 } 33 34 } -
stable/HeuristicLab.Algorithms.DataAnalysis/3.4/Interfaces/IRandomForestModel.cs
r14822 r15127 20 20 #endregion 21 21 22 using HeuristicLab. Optimization;22 using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding; 23 23 using HeuristicLab.Problems.DataAnalysis; 24 using HeuristicLab.Core; 25 using System.Collections.Generic; 24 26 25 27 26 namespace HeuristicLab.Algorithms.DataAnalysis { … … 30 29 /// </summary> 31 30 public interface IRandomForestModel : IConfidenceRegressionModel, IClassificationModel { 31 int NumberOfTrees { get; } 32 ISymbolicExpressionTree ExtractTree(int treeIdx); // returns a specific tree from the random forest as a ISymbolicRegressionModel 32 33 } 33 34 } -
stable/HeuristicLab.Algorithms.DataAnalysis/3.4/Interfaces/IRandomForestRegressionSolution.cs
r14822 r15127 20 20 #endregion 21 21 22 using HeuristicLab.Optimization;23 22 using HeuristicLab.Problems.DataAnalysis; 24 using HeuristicLab. Core;23 using HeuristicLab.Problems.DataAnalysis.Symbolic.Regression; 25 24 26 25 namespace HeuristicLab.Algorithms.DataAnalysis { … … 30 29 public interface IRandomForestRegressionSolution : IConfidenceRegressionSolution { 31 30 new IRandomForestModel Model { get; } 31 int NumberOfTrees { get; } 32 32 } 33 33 } -
stable/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestClassificationSolution.cs
r14186 r15127 38 38 } 39 39 40 public int NumberOfTrees { 41 get { return Model.NumberOfTrees; } 42 } 43 40 44 [StorableConstructor] 41 45 private RandomForestClassificationSolution(bool deserializing) : base(deserializing) { } -
stable/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestModel.cs
r14822 r15127 25 25 using HeuristicLab.Common; 26 26 using HeuristicLab.Core; 27 using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding; 27 28 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable; 28 29 using HeuristicLab.Problems.DataAnalysis; 30 using HeuristicLab.Problems.DataAnalysis.Symbolic; 29 31 30 32 namespace HeuristicLab.Algorithms.DataAnalysis { … … 49 51 } 50 52 53 public int NumberOfTrees { 54 get { return nTrees; } 55 } 51 56 52 57 // instead of storing the data of the model itself … … 64 69 [Storable] 65 70 private double m; 66 67 71 68 72 [StorableConstructor] … … 197 201 } 198 202 203 public ISymbolicExpressionTree ExtractTree(int treeIdx) { 204 var rf = RandomForest; 205 // hoping that the internal representation of alglib is stable 206 207 // TREE FORMAT 208 // W[Offs] - size of sub-array (for the tree) 209 // node info: 210 // W[K+0] - variable number (-1 for leaf mode) 211 // W[K+1] - threshold (class/value for leaf node) 212 // W[K+2] - ">=" branch index (absent for leaf node) 213 214 // skip irrelevant trees 215 int offset = 0; 216 for (int i = 0; i < treeIdx - 1; i++) { 217 offset = offset + (int)Math.Round(rf.innerobj.trees[offset]); 218 } 219 220 var constSy = new Constant(); 221 var varCondSy = new VariableCondition() { IgnoreSlope = true }; 222 223 var node = CreateRegressionTreeRec(rf.innerobj.trees, offset, offset + 1, constSy, varCondSy); 224 225 var startNode = new StartSymbol().CreateTreeNode(); 226 startNode.AddSubtree(node); 227 var root = new ProgramRootSymbol().CreateTreeNode(); 228 root.AddSubtree(startNode); 229 return new SymbolicExpressionTree(root); 230 } 231 232 private ISymbolicExpressionTreeNode CreateRegressionTreeRec(double[] trees, int offset, int k, Constant constSy, VariableCondition varCondSy) { 233 234 // alglib source for evaluation of one tree (dfprocessinternal) 235 // offs = 0 236 // 237 // Set pointer to the root 238 // 239 // k = offs + 1; 240 // 241 // // 242 // // Navigate through the tree 243 // // 244 // while (true) { 245 // if ((double)(df.trees[k]) == (double)(-1)) { 246 // if (df.nclasses == 1) { 247 // y[0] = y[0] + df.trees[k + 1]; 248 // } else { 249 // idx = (int)Math.Round(df.trees[k + 1]); 250 // y[idx] = y[idx] + 1; 251 // } 252 // break; 253 // } 254 // if ((double)(x[(int)Math.Round(df.trees[k])]) < (double)(df.trees[k + 1])) { 255 // k = k + innernodewidth; 256 // } else { 257 // k = offs + (int)Math.Round(df.trees[k + 2]); 258 // } 259 // } 260 261 if ((double)(trees[k]) == (double)(-1)) { 262 var constNode = (ConstantTreeNode)constSy.CreateTreeNode(); 263 constNode.Value = trees[k + 1]; 264 return constNode; 265 } else { 266 var condNode = (VariableConditionTreeNode)varCondSy.CreateTreeNode(); 267 condNode.VariableName = AllowedInputVariables[(int)Math.Round(trees[k])]; 268 condNode.Threshold = trees[k + 1]; 269 condNode.Slope = double.PositiveInfinity; 270 271 var left = CreateRegressionTreeRec(trees, offset, k + 3, constSy, varCondSy); 272 var right = CreateRegressionTreeRec(trees, offset, offset + (int)Math.Round(trees[k + 2]), constSy, varCondSy); 273 274 condNode.AddSubtree(left); // not 100% correct because interpreter uses: if(x <= thres) left() else right() and RF uses if(x < thres) left() else right() (see above) 275 condNode.AddSubtree(right); 276 return condNode; 277 } 278 } 279 199 280 200 281 public IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) { -
stable/HeuristicLab.Algorithms.DataAnalysis/3.4/RandomForest/RandomForestRegressionSolution.cs
r14822 r15127 20 20 #endregion 21 21 22 using System; 22 23 using HeuristicLab.Common; 23 24 using HeuristicLab.Core; 24 25 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable; 25 26 using HeuristicLab.Problems.DataAnalysis; 27 using HeuristicLab.Problems.DataAnalysis.Symbolic.Regression; 26 28 27 29 namespace HeuristicLab.Algorithms.DataAnalysis { … … 36 38 get { return (IRandomForestModel)base.Model; } 37 39 set { base.Model = value; } 40 } 41 42 public int NumberOfTrees { 43 get { return Model.NumberOfTrees; } 38 44 } 39 45 -
stable/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Interpreter/SymbolicDataAnalysisExpressionCompiledTreeInterpreter.cs
r14811 r15127 614 614 case OpCodes.VariableCondition: { 615 615 var variableConditionTreeNode = (VariableConditionTreeNode)node; 616 if (variableConditionTreeNode.Symbol.IgnoreSlope) throw new NotSupportedException("Strict variable conditionals are not supported"); 616 617 var variableName = variableConditionTreeNode.VariableName; 617 618 var indexExpr = Expression.Constant(variableIndices[variableName]); -
stable/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Interpreter/SymbolicDataAnalysisExpressionTreeInterpreter.cs
r14811 r15127 471 471 if (row < 0 || row >= dataset.Rows) return double.NaN; 472 472 var variableConditionTreeNode = (VariableConditionTreeNode)currentInstr.dynamicNode; 473 double variableValue = ((IList<double>)currentInstr.data)[row]; 474 double x = variableValue - variableConditionTreeNode.Threshold; 475 double p = 1 / (1 + Math.Exp(-variableConditionTreeNode.Slope * x)); 476 477 double trueBranch = Evaluate(dataset, ref row, state); 478 double falseBranch = Evaluate(dataset, ref row, state); 479 480 return trueBranch * p + falseBranch * (1 - p); 473 if (!variableConditionTreeNode.Symbol.IgnoreSlope) { 474 double variableValue = ((IList<double>)currentInstr.data)[row]; 475 double x = variableValue - variableConditionTreeNode.Threshold; 476 double p = 1 / (1 + Math.Exp(-variableConditionTreeNode.Slope * x)); 477 478 double trueBranch = Evaluate(dataset, ref row, state); 479 double falseBranch = Evaluate(dataset, ref row, state); 480 481 return trueBranch * p + falseBranch * (1 - p); 482 } else { 483 // strict threshold 484 double variableValue = ((IList<double>)currentInstr.data)[row]; 485 if (variableValue <= variableConditionTreeNode.Threshold) { 486 var left = Evaluate(dataset, ref row, state); 487 state.SkipInstructions(); 488 return left; 489 } else { 490 state.SkipInstructions(); 491 return Evaluate(dataset, ref row, state); 492 } 493 } 481 494 } 482 495 default: -
stable/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Interpreter/SymbolicDataAnalysisExpressionTreeLinearInterpreter.cs
r14974 r15127 126 126 private readonly object syncRoot = new object(); 127 127 public IEnumerable<double> GetSymbolicExpressionTreeValues(ISymbolicExpressionTree tree, IDataset dataset, IEnumerable<int> rows) { 128 if (!rows.Any()) return Enumerable.Empty<double>(); 128 129 if (CheckExpressionsWithIntervalArithmetic) 129 130 throw new NotSupportedException("Interval arithmetic is not yet supported in the symbolic data analysis interpreter."); … … 159 160 if (row < 0 || row >= dataset.Rows) instr.value = double.NaN; 160 161 var variableConditionTreeNode = (VariableConditionTreeNode)instr.dynamicNode; 161 double variableValue = ((IList<double>)instr.data)[row]; 162 double x = variableValue - variableConditionTreeNode.Threshold; 163 double p = 1 / (1 + Math.Exp(-variableConditionTreeNode.Slope * x)); 164 165 double trueBranch = code[instr.childIndex].value; 166 double falseBranch = code[instr.childIndex + 1].value; 167 168 instr.value = trueBranch * p + falseBranch * (1 - p); 162 if (!variableConditionTreeNode.Symbol.IgnoreSlope) { 163 double variableValue = ((IList<double>)instr.data)[row]; 164 double x = variableValue - variableConditionTreeNode.Threshold; 165 double p = 1 / (1 + Math.Exp(-variableConditionTreeNode.Slope * x)); 166 167 double trueBranch = code[instr.childIndex].value; 168 double falseBranch = code[instr.childIndex + 1].value; 169 170 instr.value = trueBranch * p + falseBranch * (1 - p); 171 } else { 172 double variableValue = ((IList<double>)instr.data)[row]; 173 if (variableValue <= variableConditionTreeNode.Threshold) { 174 instr.value = code[instr.childIndex].value; 175 } else { 176 instr.value = code[instr.childIndex + 1].value; 177 } 178 } 169 179 } else if (instr.opCode == OpCodes.Add) { 170 180 double s = code[instr.childIndex].value; … … 411 421 for (int j = 1; j != seq.Length; ++j) 412 422 seq[j].skip = true; 413 }414 break;423 break; 424 } 415 425 } 416 426 #endregion -
stable/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Symbols/VariableCondition.cs
r14811 r15127 149 149 } 150 150 } 151 } 152 153 /// <summary> 154 /// Flag to indicate if the interpreter should ignore the slope parameter (introduced for representation of expression trees) 155 /// </summary> 156 [Storable] 157 private bool ignoreSlope; 158 public bool IgnoreSlope { 159 get { return ignoreSlope; } 160 set { ignoreSlope = value; } 151 161 } 152 162 -
stable/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Symbols/VariableConditionTreeNode.cs
r14186 r15127 96 96 97 97 public override string ToString() { 98 if (slope.IsAlmost(0.0)) 98 if (slope.IsAlmost(0.0) || Symbol.IgnoreSlope) { 99 return variableName + " < " + threshold.ToString("E4"); 100 } else { 99 101 return variableName + " > " + threshold.ToString("E4") + Environment.NewLine + 100 "slope: " + slope.ToString("E4"); 101 else 102 return variableName + " > " + threshold.ToString("E4"); 102 "slope: " + slope.ToString("E4"); 103 } 103 104 } 104 105 }
Note: See TracChangeset
for help on using the changeset viewer.