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 |
|
---|
22 | using System;
|
---|
23 | using System.Collections.Generic;
|
---|
24 | using System.Linq;
|
---|
25 | using HeuristicLab.Common;
|
---|
26 | using HeuristicLab.Core;
|
---|
27 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
28 |
|
---|
29 | namespace HeuristicLab.Algorithms.DataAnalysis {
|
---|
30 | [StorableClass]
|
---|
31 | [Item(Name = "CovarianceProd",
|
---|
32 | Description = "Product covariance function for Gaussian processes.")]
|
---|
33 | public class CovarianceProd : Item, ICovarianceFunction {
|
---|
34 | [Storable]
|
---|
35 | private ItemList<ICovarianceFunction> factors;
|
---|
36 |
|
---|
37 | [Storable]
|
---|
38 | private int numberOfVariables;
|
---|
39 | public ItemList<ICovarianceFunction> Factors {
|
---|
40 | get { return factors; }
|
---|
41 | }
|
---|
42 |
|
---|
43 | [StorableConstructor]
|
---|
44 | protected CovarianceProd(bool deserializing)
|
---|
45 | : base(deserializing) {
|
---|
46 | }
|
---|
47 |
|
---|
48 | protected CovarianceProd(CovarianceProd original, Cloner cloner)
|
---|
49 | : base(original, cloner) {
|
---|
50 | this.factors = cloner.Clone(original.factors);
|
---|
51 | this.numberOfVariables = original.numberOfVariables;
|
---|
52 | AttachEventHandlers();
|
---|
53 | }
|
---|
54 |
|
---|
55 | public CovarianceProd()
|
---|
56 | : base() {
|
---|
57 | this.factors = new ItemList<ICovarianceFunction>();
|
---|
58 | AttachEventHandlers();
|
---|
59 | }
|
---|
60 |
|
---|
61 | private void AttachEventHandlers() {
|
---|
62 | this.factors.CollectionReset += (sender, args) => ClearCache();
|
---|
63 | this.factors.ItemsAdded += (sender, args) => ClearCache();
|
---|
64 | this.factors.ItemsRemoved += (sender, args) => ClearCache();
|
---|
65 | this.factors.ItemsReplaced += (sender, args) => ClearCache();
|
---|
66 | this.factors.ItemsMoved += (sender, args) => ClearCache();
|
---|
67 | }
|
---|
68 |
|
---|
69 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
70 | return new CovarianceProd(this, cloner);
|
---|
71 | }
|
---|
72 |
|
---|
73 | public int GetNumberOfParameters(int numberOfVariables) {
|
---|
74 | this.numberOfVariables = numberOfVariables;
|
---|
75 | return factors.Select(f => f.GetNumberOfParameters(numberOfVariables)).Sum();
|
---|
76 | }
|
---|
77 |
|
---|
78 | public void SetParameter(double[] hyp) {
|
---|
79 | if (factors.Count == 0) throw new ArgumentException("at least one factor is necessary for the product covariance function.");
|
---|
80 | int offset = 0;
|
---|
81 | foreach (var t in factors) {
|
---|
82 | var numberOfParameters = t.GetNumberOfParameters(numberOfVariables);
|
---|
83 | t.SetParameter(hyp.Skip(offset).Take(numberOfParameters).ToArray());
|
---|
84 | offset += numberOfParameters;
|
---|
85 | }
|
---|
86 | }
|
---|
87 |
|
---|
88 | public double GetCovariance(double[,] x, int i, int j) {
|
---|
89 | return factors.Select(f => f.GetCovariance(x, i, j)).Aggregate((a, b) => a * b);
|
---|
90 | }
|
---|
91 |
|
---|
92 | public IEnumerable<double> GetGradient(double[,] x, int i, int j) {
|
---|
93 | //if (cachedParameterMap == null) {
|
---|
94 | // CalculateParameterMap();
|
---|
95 | //}
|
---|
96 | //int ti = cachedParameterMap[k].Item1;
|
---|
97 | //k = cachedParameterMap[k].Item2;
|
---|
98 | //double gradient = 1.0;
|
---|
99 | //for (int ii = 0; ii < factors.Count; ii++) {
|
---|
100 | // var f = factors[ii];
|
---|
101 | // if (ii == ti) {
|
---|
102 | // gradient *= f.GetGradient(x, i, j, k);
|
---|
103 | // } else {
|
---|
104 | // gradient *= f.GetCovariance(x, i, j);
|
---|
105 | // }
|
---|
106 | //}
|
---|
107 | //return gradient;
|
---|
108 | var covariances = factors.Select(f => f.GetCovariance(x, i, j)).ToArray();
|
---|
109 | for (int ii = 0; ii < factors.Count; ii++) {
|
---|
110 | foreach (var g in factors[ii].GetGradient(x, i, j)) {
|
---|
111 | double res = g;
|
---|
112 | for (int jj = 0; jj < covariances.Length; jj++)
|
---|
113 | if (ii != jj) res *= covariances[jj];
|
---|
114 | yield return res;
|
---|
115 | }
|
---|
116 | }
|
---|
117 | }
|
---|
118 |
|
---|
119 | public double GetCrossCovariance(double[,] x, double[,] xt, int i, int j) {
|
---|
120 | return factors.Select(f => f.GetCrossCovariance(x, xt, i, j)).Aggregate((a, b) => a * b);
|
---|
121 | }
|
---|
122 |
|
---|
123 | private Dictionary<int, Tuple<int, int>> cachedParameterMap;
|
---|
124 | private void ClearCache() {
|
---|
125 | cachedParameterMap = null;
|
---|
126 | }
|
---|
127 |
|
---|
128 | private void CalculateParameterMap() {
|
---|
129 | cachedParameterMap = new Dictionary<int, Tuple<int, int>>();
|
---|
130 | int k = 0;
|
---|
131 | for (int ti = 0; ti < factors.Count; ti++) {
|
---|
132 | for (int ti_k = 0; ti_k < factors[ti].GetNumberOfParameters(numberOfVariables); ti_k++) {
|
---|
133 | cachedParameterMap[k++] = Tuple.Create(ti, ti_k);
|
---|
134 | }
|
---|
135 | }
|
---|
136 | }
|
---|
137 | }
|
---|
138 | }
|
---|