source: branches/2971_named_intervals/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/RegressionProblemData.cs @ 16774

Last change on this file since 16774 was 16774, checked in by chaider, 11 months ago

#2971 Changed Parser and IntevalConstraint View

File size: 10.7 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2019 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.Generic;
24using System.Linq;
25using HeuristicLab.Common;
26using HeuristicLab.Core;
27using HeuristicLab.Data;
28using HeuristicLab.Parameters;
29using HEAL.Attic;
30using HeuristicLab.Problems.DataAnalysis.Implementation;
31
32namespace HeuristicLab.Problems.DataAnalysis {
33  [StorableType("EE612297-B1AF-42D2-BF21-AF9A2D42791C")]
34  [Item("RegressionProblemData", "Represents an item containing all data defining a regression problem.")]
35  public class RegressionProblemData : DataAnalysisProblemData, IRegressionProblemData, IStorableContent {
36    protected const string TargetVariableParameterName = "TargetVariable";
37    protected const string VariableRangesParameterName = "VariableRanges";
38    protected const string IntervalConstraintsParameterName = "IntervalConstraints";
39    public string Filename { get; set; }
40
41    #region default data
42    private static double[,] kozaF1 = new double[,] {
43          {2.017885919, -1.449165046},
44          {1.30060506,  -1.344523885},
45          {1.147134798, -1.317989331},
46          {0.877182504, -1.266142284},
47          {0.852562452, -1.261020794},
48          {0.431095788, -1.158793317},
49          {0.112586002, -1.050908405},
50          {0.04594507,  -1.021989402},
51          {0.042572879, -1.020438113},
52          {-0.074027291,  -0.959859562},
53          {-0.109178553,  -0.938094706},
54          {-0.259721109,  -0.803635355},
55          {-0.272991057,  -0.387519561},
56          {-0.161978191,  -0.193611001},
57          {-0.102489983,  -0.114215349},
58          {-0.01469968, -0.014918985},
59          {-0.008863365,  -0.008942626},
60          {0.026751057, 0.026054094},
61          {0.166922436, 0.14309643},
62          {0.176953808, 0.1504144},
63          {0.190233418, 0.159916534},
64          {0.199800708, 0.166635331},
65          {0.261502822, 0.207600348},
66          {0.30182879,  0.232370249},
67          {0.83763905,  0.468046718}
68    };
69    private static readonly Dataset defaultDataset;
70    private static readonly IEnumerable<string> defaultAllowedInputVariables;
71    private static readonly string defaultTargetVariable;
72
73    private static readonly RegressionProblemData emptyProblemData;
74    public static RegressionProblemData EmptyProblemData {
75      get { return emptyProblemData; }
76    }
77
78    static RegressionProblemData() {
79      defaultDataset = new Dataset(new string[] { "y", "x" }, kozaF1);
80      defaultDataset.Name = "Fourth-order Polynomial Function Benchmark Dataset";
81      defaultDataset.Description = "f(x) = x^4 + x^3 + x^2 + x^1";
82      defaultAllowedInputVariables = new List<string>() { "x" };
83      defaultTargetVariable = "y";
84
85      var problemData = new RegressionProblemData();
86      problemData.Parameters.Clear();
87      problemData.Name = "Empty Regression ProblemData";
88      problemData.Description = "This ProblemData acts as place holder before the correct problem data is loaded.";
89      problemData.isEmpty = true;
90
91      problemData.Parameters.Add(new FixedValueParameter<Dataset>(DatasetParameterName, "", new Dataset()));
92      problemData.Parameters.Add(new FixedValueParameter<ReadOnlyCheckedItemList<StringValue>>(InputVariablesParameterName, ""));
93      problemData.Parameters.Add(new FixedValueParameter<IntRange>(TrainingPartitionParameterName, "", (IntRange)new IntRange(0, 0).AsReadOnly()));
94      problemData.Parameters.Add(new FixedValueParameter<IntRange>(TestPartitionParameterName, "", (IntRange)new IntRange(0, 0).AsReadOnly()));
95      problemData.Parameters.Add(new ConstrainedValueParameter<StringValue>(TargetVariableParameterName, new ItemSet<StringValue>()));
96      problemData.Parameters.Add(new FixedValueParameter<NamedIntervals>(VariableRangesParameterName, "", new NamedIntervals()));
97      problemData.Parameters.Add(new FixedValueParameter<ParsedConstraint>(IntervalConstraintsParameterName, "", new ParsedConstraint()));
98      emptyProblemData = problemData;
99    }
100    #endregion
101
102    public IConstrainedValueParameter<StringValue> TargetVariableParameter {
103      get { return (IConstrainedValueParameter<StringValue>)Parameters[TargetVariableParameterName]; }
104    }
105
106    public IFixedValueParameter<NamedIntervals> VariableRangesParameter => (IFixedValueParameter<NamedIntervals>)Parameters[VariableRangesParameterName];
107
108    public NamedIntervals VariableRanges => VariableRangesParameter.Value;
109
110    public IFixedValueParameter<ParsedConstraint> IntervalConstraintsParameter => (IFixedValueParameter<ParsedConstraint>)Parameters[IntervalConstraintsParameterName];
111    public ParsedConstraint IntervalConstraints => IntervalConstraintsParameter.Value;
112
113    public string TargetVariable {
114      get { return TargetVariableParameter.Value.Value; }
115      set {
116        if (value == null) throw new ArgumentNullException("targetVariable", "The provided value for the targetVariable is null.");
117        if (value == TargetVariable) return;
118
119        var matchingParameterValue = TargetVariableParameter.ValidValues.FirstOrDefault(v => v.Value == value);
120        if (matchingParameterValue == null) throw new ArgumentException("The provided value is not valid as the targetVariable.", "targetVariable");
121        TargetVariableParameter.Value = matchingParameterValue;
122      }
123    }
124
125    public IEnumerable<double> TargetVariableValues {
126      get { return Dataset.GetDoubleValues(TargetVariable); }
127    }
128    public IEnumerable<double> TargetVariableTrainingValues {
129      get { return Dataset.GetDoubleValues(TargetVariable, TrainingIndices); }
130    }
131    public IEnumerable<double> TargetVariableTestValues {
132      get { return Dataset.GetDoubleValues(TargetVariable, TestIndices); }
133    }
134
135
136    [StorableConstructor]
137    protected RegressionProblemData(StorableConstructorFlag _) : base(_) { }
138    [StorableHook(HookType.AfterDeserialization)]
139    private void AfterDeserialization() {
140      RegisterParameterEvents();
141    }
142
143    protected RegressionProblemData(RegressionProblemData original, Cloner cloner)
144      : base(original, cloner) {
145      RegisterParameterEvents();
146    }
147    public override IDeepCloneable Clone(Cloner cloner) {
148      if (this == emptyProblemData) return emptyProblemData;
149      return new RegressionProblemData(this, cloner);
150    }
151
152    public RegressionProblemData()
153      : this(defaultDataset, defaultAllowedInputVariables, defaultTargetVariable) {
154    }
155    public RegressionProblemData(IRegressionProblemData regressionProblemData)
156      : this(regressionProblemData.Dataset, regressionProblemData.AllowedInputVariables, regressionProblemData.TargetVariable) {
157      TrainingPartition.Start = regressionProblemData.TrainingPartition.Start;
158      TrainingPartition.End = regressionProblemData.TrainingPartition.End;
159      TestPartition.Start = regressionProblemData.TestPartition.Start;
160      TestPartition.End = regressionProblemData.TestPartition.End;
161    }
162
163    public RegressionProblemData(IDataset dataset, IEnumerable<string> allowedInputVariables, string targetVariable, IEnumerable<ITransformation> transformations = null)
164      : base(dataset, allowedInputVariables, transformations ?? Enumerable.Empty<ITransformation>()) {
165      var variables = InputVariables.Select(x => x.AsReadOnly()).ToList();
166      Parameters.Add(new ConstrainedValueParameter<StringValue>(TargetVariableParameterName, new ItemSet<StringValue>(variables), variables.Where(x => x.Value == targetVariable).First()));
167     
168      NamedIntervals namedIntervals = new NamedIntervals();
169      foreach (var variable in variables) {
170        if (!dataset.DoubleVariables.Contains(variable.Value)) continue; // intervals are only possible for double variables
171        var variableInterval = Interval.GetInterval(dataset.GetDoubleValues(variable.Value));
172        namedIntervals.Add(variable.Value, variableInterval);
173      }
174     
175      var commentMessage = "#Example of a target variable constraint" + Environment.NewLine +
176                           "#F1: TargetVariable" + Environment.NewLine +
177                           "#in: keyword has to follow" + Environment.NewLine +
178                           "#bracket: open or closed regarding to open or closed interval definition" + Environment.NewLine +
179                           "#double: lowerbound of the interval" + Environment.NewLine +
180                           "#..: keyword has to follow to seperate lowerbound from upperbound" + Environment.NewLine +
181                           "#double: upperbound of the interval" + Environment.NewLine +
182                           "#bracket: open or closed regarding to open or closed interval definition" + Environment.NewLine +
183                           "#Y2 in [0 .. 1.2]" + Environment.NewLine + Environment.NewLine +
184                           "#Constraints on model parameters:" + Environment.NewLine +
185                           "#derivation symbol: d or ∂" + Environment.NewLine +
186                           "#derivation in leibniz notation" + Environment.NewLine +
187                           "#in and interval syntax is the same like above" + Environment.NewLine +
188                           "#∂Y2/∂X1 in [0 .. inf.[" + Environment.NewLine +
189                           "#∂²Y2/∂X1² in [0 .. inf.[" + Environment.NewLine +
190                           "#∂Y2/∂X9 in [-100 .. 0]" + Environment.NewLine +
191                           "#∂²Y2/∂X9² in [0 .. inf.[";
192
193      Parameters.Add(new FixedValueParameter<NamedIntervals>(VariableRangesParameterName, namedIntervals));
194      Parameters.Add(new FixedValueParameter<ParsedConstraint>(IntervalConstraintsParameterName, new ParsedConstraint(commentMessage, this)));
195      RegisterParameterEvents();
196    }
197
198    private void RegisterParameterEvents() {
199      TargetVariableParameter.ValueChanged += new EventHandler(TargetVariableParameter_ValueChanged);
200    }
201    private void TargetVariableParameter_ValueChanged(object sender, EventArgs e) {
202      OnChanged();
203    }
204  }
205}
Note: See TracBrowser for help on using the repository browser.