1  #region License Information


2  /* HeuristicLab


3  * Copyright (C) 20022012 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.Data;


28  using HeuristicLab.Parameters;


29  using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;


30 


31  namespace 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  }

