Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/GaussianProcess/CovarianceFunctions/CovarianceSquaredExponentialArd.cs @ 8982

Last change on this file since 8982 was 8982, checked in by gkronber, 12 years ago

#1902: removed class HyperParameter and changed implementations of covariance and mean functions to remove the parameter value caching and event handlers for parameter caching. Instead it is now possible to create the actual covariance and mean functions as Func from templates and specified parameter values. The instances of mean and covariance functions configured in the GUI are actually templates where the structure and fixed parameters can be specified.

File size: 5.3 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2012 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 HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
30
31namespace HeuristicLab.Algorithms.DataAnalysis {
32  [StorableClass]
33  [Item(Name = "CovarianceSquaredExponentialArd", Description = "Squared exponential covariance function with automatic relevance determination for Gaussian processes.")]
34  public sealed class CovarianceSquaredExponentialArd : ParameterizedNamedItem, ICovarianceFunction {
35    public IValueParameter<DoubleValue> ScaleParameter {
36      get { return (IValueParameter<DoubleValue>)Parameters["Scale"]; }
37    }
38
39    public IValueParameter<DoubleArray> InverseLengthParameter {
40      get { return (IValueParameter<DoubleArray>)Parameters["InverseLength"]; }
41    }
42
43    [StorableConstructor]
44    private CovarianceSquaredExponentialArd(bool deserializing) : base(deserializing) { }
45    private CovarianceSquaredExponentialArd(CovarianceSquaredExponentialArd original, Cloner cloner)
46      : base(original, cloner) {
47    }
48    public CovarianceSquaredExponentialArd()
49      : base() {
50      Name = ItemName;
51      Description = ItemDescription;
52
53      Parameters.Add(new OptionalValueParameter<DoubleValue>("Scale", "The scale parameter of the squared exponential covariance function with ARD."));
54      Parameters.Add(new OptionalValueParameter<DoubleArray>("InverseLength", "The inverse length parameter for automatic relevance determination."));
55    }
56
57    public override IDeepCloneable Clone(Cloner cloner) {
58      return new CovarianceSquaredExponentialArd(this, cloner);
59    }
60
61    public int GetNumberOfParameters(int numberOfVariables) {
62      return
63        (ScaleParameter.Value != null ? 0 : 1) +
64        (InverseLengthParameter.Value != null ? 0 : numberOfVariables);
65    }
66
67    public void SetParameter(double[] p) {
68      double scale;
69      double[] inverseLength;
70      GetParameterValues(p, out scale, out inverseLength);
71      ScaleParameter.Value = new DoubleValue(scale);
72      InverseLengthParameter.Value = new DoubleArray(inverseLength);
73    }
74
75    private void GetParameterValues(double[] p, out double scale, out double[] inverseLength) {
76      int c = 0;
77      // gather parameter values
78      if (ScaleParameter.Value != null) {
79        scale = ScaleParameter.Value.Value;
80      } else {
81        scale = Math.Exp(2 * p[c]);
82        c++;
83      }
84      if (InverseLengthParameter.Value != null) {
85        inverseLength = InverseLengthParameter.Value.ToArray();
86      } else {
87        inverseLength = p.Skip(1).Select(e => 1.0 / Math.Exp(e)).ToArray();
88        c += inverseLength.Length;
89      }
90      if (p.Length != c) throw new ArgumentException("The length of the parameter vector does not match the number of free parameters for CovarianceSquaredExponentialArd", "p");
91    }
92
93    public ParameterizedCovarianceFunction GetParameterizedCovarianceFunction(double[] p, IEnumerable<int> columnIndices) {
94      double scale;
95      double[] inverseLength;
96      GetParameterValues(p, out scale, out inverseLength);
97      // create functions
98      var cov = new ParameterizedCovarianceFunction();
99      cov.Covariance = (x, i, j) => {
100        double d = i == j
101                 ? 0.0
102                 : Util.SqrDist(x, i, j, inverseLength, columnIndices);
103        return scale * Math.Exp(-d / 2.0);
104      };
105      cov.CrossCovariance = (x, xt, i, j) => {
106        double d = Util.SqrDist(x, i, xt, j, inverseLength, columnIndices);
107        return scale * Math.Exp(-d / 2.0);
108      };
109      cov.CovarianceGradient = (x, i, j) => GetGradient(x, i, j, columnIndices, scale, inverseLength);
110      return cov;
111    }
112
113
114    private static IEnumerable<double> GetGradient(double[,] x, int i, int j, IEnumerable<int> columnIndices, double scale, double[] inverseLength) {
115      if (columnIndices == null) columnIndices = Enumerable.Range(0, x.GetLength(1));
116      double d = i == j
117                   ? 0.0
118                   : Util.SqrDist(x, i, j, inverseLength, columnIndices);
119      int k = 0;
120      foreach (var columnIndex in columnIndices) {
121        double sqrDist = Util.SqrDist(x[i, columnIndex] * inverseLength[k], x[j, columnIndex] * inverseLength[k]);
122        yield return scale * Math.Exp(-d / 2.0) * sqrDist;
123        k++;
124      }
125
126      yield return 2.0 * scale * Math.Exp(-d / 2.0);
127    }
128  }
129}
Note: See TracBrowser for help on using the repository browser.