source: trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/GaussianProcess/CovarianceFunctions/CovarianceNoise.cs @ 13784

Last change on this file since 13784 was 13784, checked in by pfleck, 5 years ago

#2591 Made the creation of a GaussianProcessModel faster by avoiding additional iterators during calculation of the hyperparameter gradients.
The gradients of the hyperparameters are now calculated in one sweep and returned as IList, instead of returning an iterator (with yield return).
This avoids a large amount of Move-calls of the iterator, especially for covariance functions with a lot of hyperparameters.
Besides, the signature of the CovarianceGradientFunctionDelegate is changed, to return an IList instead of an IEnumerable to avoid unnececary ToList or ToArray calls.

File size: 3.5 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2015 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 HeuristicLab.Common;
24using HeuristicLab.Core;
25using HeuristicLab.Data;
26using HeuristicLab.Parameters;
27using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
28
29namespace HeuristicLab.Algorithms.DataAnalysis {
30  [StorableClass]
31  [Item(Name = "CovarianceNoise",
32    Description = "Noise covariance function for Gaussian processes.")]
33  public sealed class CovarianceNoise : ParameterizedNamedItem, ICovarianceFunction {
34    public IValueParameter<DoubleValue> ScaleParameter {
35      get { return (IValueParameter<DoubleValue>)Parameters["Scale"]; }
36    }
37    private bool HasFixedScaleParameter {
38      get { return ScaleParameter.Value != null; }
39    }
40
41    [StorableConstructor]
42    private CovarianceNoise(bool deserializing)
43      : base(deserializing) {
44    }
45
46    private CovarianceNoise(CovarianceNoise original, Cloner cloner)
47      : base(original, cloner) {
48    }
49
50    public CovarianceNoise()
51      : base() {
52      Name = ItemName;
53      Description = ItemDescription;
54
55      Parameters.Add(new OptionalValueParameter<DoubleValue>("Scale", "The scale of noise."));
56    }
57
58    public override IDeepCloneable Clone(Cloner cloner) {
59      return new CovarianceNoise(this, cloner);
60    }
61
62    public int GetNumberOfParameters(int numberOfVariables) {
63      return HasFixedScaleParameter ? 0 : 1;
64    }
65
66    public void SetParameter(double[] p) {
67      double scale;
68      GetParameterValues(p, out scale);
69      ScaleParameter.Value = new DoubleValue(scale);
70    }
71
72    private void GetParameterValues(double[] p, out double scale) {
73      int c = 0;
74      // gather parameter values
75      if (HasFixedScaleParameter) {
76        scale = ScaleParameter.Value.Value;
77      } else {
78        scale = Math.Exp(2 * p[c]);
79        c++;
80      }
81      if (p.Length != c) throw new ArgumentException("The length of the parameter vector does not match the number of free parameters for CovarianceNoise", "p");
82    }
83
84    public ParameterizedCovarianceFunction GetParameterizedCovarianceFunction(double[] p, int[] columnIndices) {
85      double scale;
86      GetParameterValues(p, out scale);
87      var fixedScale = HasFixedScaleParameter;
88      // create functions
89      var cov = new ParameterizedCovarianceFunction();
90      cov.Covariance = (x, i, j) => i == j ? scale : 0.0;
91      cov.CrossCovariance = (x, xt, i, j) => Util.SqrDist(x, i, xt, j, columnIndices, 1.0) < 1e-9 ? scale : 0.0;
92      if (fixedScale)
93        cov.CovarianceGradient = (x, i, j) => new double[0];
94      else
95        cov.CovarianceGradient = (x, i, j) => new double[1] { i == j ? 2.0 * scale : 0.0 };
96      return cov;
97    }
98  }
99}
Note: See TracBrowser for help on using the repository browser.