Changeset 15281 for branches/Async/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression.Views/3.4/SymbolicRegressionSolutionErrorCharacteristicsCurveView.cs
- Timestamp:
- 07/23/17 11:17:18 (7 years ago)
- Location:
- branches/Async
- Files:
-
- 3 edited
Legend:
- Unmodified
- Added
- Removed
-
branches/Async
- Property svn:ignore
-
old new 24 24 protoc.exe 25 25 obj 26 .vs
-
- Property svn:mergeinfo changed
- Property svn:ignore
-
branches/Async/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression.Views
- Property svn:mergeinfo changed
-
branches/Async/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression.Views/3.4/SymbolicRegressionSolutionErrorCharacteristicsCurveView.cs
r13003 r15281 1 1 #region License Information 2 2 /* HeuristicLab 3 * Copyright (C) 2002-201 5Heuristic and Evolutionary Algorithms Laboratory (HEAL)3 * Copyright (C) 2002-2016 Heuristic and Evolutionary Algorithms Laboratory (HEAL) 4 4 * 5 5 * This file is part of HeuristicLab. … … 21 21 22 22 using System; 23 using System.Collections; 23 24 using System.Collections.Generic; 25 using System.Diagnostics.Contracts; 24 26 using System.Linq; 25 27 using HeuristicLab.Algorithms.DataAnalysis; … … 46 48 if (!problemData.TrainingIndices.Any()) return null; // don't create an LR model if the problem does not have a training set (e.g. loaded into an existing model) 47 49 48 //clear checked inputVariables 49 foreach (var inputVariable in problemData.InputVariables.CheckedItems) { 50 problemData.InputVariables.SetItemCheckedState(inputVariable.Value, false); 51 } 50 var usedVariables = Content.Model.VariablesUsedForPrediction; 52 51 53 //check inputVariables used in the symbolic regression model 54 var usedVariables = 55 Content.Model.SymbolicExpressionTree.IterateNodesPostfix().OfType<VariableTreeNode>().Select( 56 node => node.VariableName).Distinct(); 57 foreach (var variable in usedVariables) { 58 problemData.InputVariables.SetItemCheckedState( 59 problemData.InputVariables.First(x => x.Value == variable), true); 60 } 52 var usedDoubleVariables = usedVariables 53 .Where(name => problemData.Dataset.VariableHasType<double>(name)) 54 .Distinct(); 61 55 62 var solution = LinearRegression.CreateLinearRegressionSolution(problemData, out rmse, out cvRmsError); 56 var usedFactorVariables = usedVariables 57 .Where(name => problemData.Dataset.VariableHasType<string>(name)) 58 .Distinct(); 59 60 // gkronber: for binary factors we actually produce a binary variable in the new dataset 61 // but only if the variable is not used as a full factor anyway (LR creates binary columns anyway) 62 var usedBinaryFactors = 63 Content.Model.SymbolicExpressionTree.IterateNodesPostfix().OfType<BinaryFactorVariableTreeNode>() 64 .Where(node => !usedFactorVariables.Contains(node.VariableName)) 65 .Select(node => Tuple.Create(node.VariableValue, node.VariableValue)); 66 67 // create a new problem and dataset 68 var variableNames = 69 usedDoubleVariables 70 .Concat(usedFactorVariables) 71 .Concat(usedBinaryFactors.Select(t => t.Item1 + "=" + t.Item2)) 72 .Concat(new string[] { problemData.TargetVariable }) 73 .ToArray(); 74 var variableValues = 75 usedDoubleVariables.Select(name => (IList)problemData.Dataset.GetDoubleValues(name).ToList()) 76 .Concat(usedFactorVariables.Select(name => problemData.Dataset.GetStringValues(name).ToList())) 77 .Concat( 78 // create binary variable 79 usedBinaryFactors.Select(t => problemData.Dataset.GetReadOnlyStringValues(t.Item1).Select(val => val == t.Item2 ? 1.0 : 0.0).ToList()) 80 ) 81 .Concat(new[] { problemData.Dataset.GetDoubleValues(problemData.TargetVariable).ToList() }); 82 83 var newDs = new Dataset(variableNames, variableValues); 84 var newProblemData = new RegressionProblemData(newDs, variableNames.Take(variableNames.Length - 1), variableNames.Last()); 85 newProblemData.TrainingPartition.Start = problemData.TrainingPartition.Start; 86 newProblemData.TrainingPartition.End = problemData.TrainingPartition.End; 87 newProblemData.TestPartition.Start = problemData.TestPartition.Start; 88 newProblemData.TestPartition.End = problemData.TestPartition.End; 89 90 var solution = LinearRegression.CreateLinearRegressionSolution(newProblemData, out rmse, out cvRmsError); 63 91 solution.Name = "Baseline (linear subset)"; 64 92 return solution; … … 68 96 protected override IEnumerable<IRegressionSolution> CreateBaselineSolutions() { 69 97 foreach (var sol in base.CreateBaselineSolutions()) yield return sol; 98 99 // does not support lagged variables 100 if (Content.Model.SymbolicExpressionTree.IterateNodesPrefix().OfType<LaggedVariableTreeNode>().Any()) yield break; 101 70 102 yield return CreateLinearRegressionSolution(); 71 103 }
Note: See TracChangeset
for help on using the changeset viewer.