Free cookie consent management tool by TermsFeed Policy Generator

source: branches/2892_LR-prediction-intervals/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression.Views/3.4/SymbolicRegressionSolutionErrorCharacteristicsCurveView.cs @ 16388

Last change on this file since 16388 was 16388, checked in by gkronber, 5 years ago

#2892: Merging r15750:16382 (HEAD) from trunk to branch, resolving conflicts

File size: 4.8 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2018 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
4 *
5 * This file is part of HeuristicLab.
6 *
7 * HeuristicLab is free software: you can redistribute it and/or modify
8 * it under the terms of the GNU General Public License as published by
9 * the Free Software Foundation, either version 3 of the License, or
10 * (at your option) any later version.
11 *
12 * HeuristicLab is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
15 * GNU General Public License for more details.
16 *
17 * You should have received a copy of the GNU General Public License
18 * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
19 */
20#endregion
21
22using System;
23using System.Collections;
24using System.Collections.Generic;
25using System.Linq;
26using HeuristicLab.Algorithms.DataAnalysis;
27using HeuristicLab.MainForm;
28using HeuristicLab.Problems.DataAnalysis.Views;
29
30namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Regression.Views {
31  [View("Error Characteristics Curve")]
32  [Content(typeof(ISymbolicRegressionSolution))]
33  public partial class SymbolicRegressionSolutionErrorCharacteristicsCurveView : RegressionSolutionErrorCharacteristicsCurveView {
34    public SymbolicRegressionSolutionErrorCharacteristicsCurveView() {
35      InitializeComponent();
36    }
37
38    public new ISymbolicRegressionSolution Content {
39      get { return (ISymbolicRegressionSolution)base.Content; }
40      set { base.Content = value; }
41    }
42
43    private IRegressionSolution CreateLinearRegressionSolution() {
44      if (Content == null) throw new InvalidOperationException();
45      double rmse, cvRmsError;
46      var problemData = (IRegressionProblemData)ProblemData.Clone();
47      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)
48
49      var usedVariables = Content.Model.VariablesUsedForPrediction;
50
51      var usedDoubleVariables = usedVariables
52        .Where(name => problemData.Dataset.VariableHasType<double>(name))
53      .Distinct();
54
55      var usedFactorVariables = usedVariables
56        .Where(name => problemData.Dataset.VariableHasType<string>(name))
57        .Distinct();
58
59      // gkronber: for binary factors we actually produce a binary variable in the new dataset
60      // but only if the variable is not used as a full factor anyway (LR creates binary columns anyway)
61      var usedBinaryFactors =
62        Content.Model.SymbolicExpressionTree.IterateNodesPostfix().OfType<BinaryFactorVariableTreeNode>()
63        .Where(node => !usedFactorVariables.Contains(node.VariableName))
64        .Select(node => Tuple.Create(node.VariableValue, node.VariableValue));
65
66      // create a new problem and dataset
67      var variableNames =
68        usedDoubleVariables
69        .Concat(usedFactorVariables)
70        .Concat(usedBinaryFactors.Select(t => t.Item1 + "=" + t.Item2))
71        .Concat(new string[] { problemData.TargetVariable })
72        .ToArray();
73      var variableValues =
74        usedDoubleVariables.Select(name => (IList)problemData.Dataset.GetDoubleValues(name).ToList())
75        .Concat(usedFactorVariables.Select(name => problemData.Dataset.GetStringValues(name).ToList()))
76        .Concat(
77          // create binary variable
78          usedBinaryFactors.Select(t => problemData.Dataset.GetReadOnlyStringValues(t.Item1).Select(val => val == t.Item2 ? 1.0 : 0.0).ToList())
79        )
80        .Concat(new[] { problemData.Dataset.GetDoubleValues(problemData.TargetVariable).ToList() });
81
82      var newDs = new Dataset(variableNames, variableValues);
83      var newProblemData = new RegressionProblemData(newDs, variableNames.Take(variableNames.Length - 1), variableNames.Last());
84      newProblemData.TrainingPartition.Start = problemData.TrainingPartition.Start;
85      newProblemData.TrainingPartition.End = problemData.TrainingPartition.End;
86      newProblemData.TestPartition.Start = problemData.TestPartition.Start;
87      newProblemData.TestPartition.End = problemData.TestPartition.End;
88
89      var solution = LinearRegression.CreateLinearRegressionSolution(newProblemData, out rmse, out cvRmsError);
90      solution.Name = "Baseline (linear subset)";
91      return solution;
92    }
93
94
95    protected override IEnumerable<IRegressionSolution> CreateBaselineSolutions() {
96      foreach (var sol in base.CreateBaselineSolutions()) yield return sol;
97
98      // does not support lagged variables
99      if (Content.Model.SymbolicExpressionTree.IterateNodesPrefix().OfType<LaggedVariableTreeNode>().Any()) yield break;
100
101      yield return CreateLinearRegressionSolution();
102    }
103  }
104}
Note: See TracBrowser for help on using the repository browser.