source: branches/HeuristicLab.Problems.GaussianProcessTuning/HeuristicLab.Problems.Instances.DataAnalysis.GaussianProcessRegression/GaussianProcessPolyTen.cs @ 11010

#region License Information
/* HeuristicLab
 * Copyright (C) 2002-2012 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
 *
 * This file is part of HeuristicLab.
 *
 * HeuristicLab is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * HeuristicLab is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
 */
#endregion

using System.Collections.Generic;
using System.Linq;
using HeuristicLab.Algorithms.DataAnalysis;
using HeuristicLab.Data;
using HeuristicLab.Random;

namespace HeuristicLab.Problems.Instances.DataAnalysis {
  public class GaussianProcessPolyTen : ArtificialRegressionDataDescriptor {

    public override string Name {
      get {
        return "Gaussian Process Poly-10 y = GP(0, CovSEIso(X1)*CovSEIso(X2) + " +
               "CovSEIso(X3)*CovSEIso(X4) + CovSEIso(X5)*CovSEIso(X6) + CovSEIso(X1)*CovSEIso(X7)*CovSEIso(X9) + CovSEIso(X3)*CovSEIso(X6)*CovSEIso(X10)";
      }
    }
    public override string Description {
      get { return ""; }
    }
    protected override string TargetVariable { get { return "Y"; } }
    protected override string[] VariableNames { get { return new string[] { "X1", "X2", "X3", "X4", "X5", "X6", "X7", "X8", "X9", "X10", "Y" }; } }
    protected override string[] AllowedInputVariables { get { return new string[] { "X1", "X2", "X3", "X4", "X5", "X6", "X7", "X8", "X9", "X10" }; } }
    protected override int TrainingPartitionStart { get { return 0; } }
    protected override int TrainingPartitionEnd { get { return 500; } }
    protected override int TestPartitionStart { get { return 500; } }
    protected override int TestPartitionEnd { get { return 1000; } }

    protected override List<List<double>> GenerateValues() {
      var mt = new MersenneTwister(31415);

      List<List<double>> data = new List<List<double>>();
      for (int i = 0; i < AllowedInputVariables.Count(); i++) {
        data.Add(ValueGenerator.GenerateUniformDistributedValues(TestPartitionEnd, -1, 1).ToList());
      }


      var hyp = new double[]
        {
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          0.0, 0.0,
          -5.0 // noise
        };


      var covarianceFunction = new CovarianceSum();
      var t1 = new CovarianceProduct();
      var m1 = new CovarianceMask();
      m1.SelectedDimensionsParameter.Value = new IntArray(new int[] { 0 });
      var m2 = new CovarianceMask();
      m2.SelectedDimensionsParameter.Value = new IntArray(new int[] { 1 });
      t1.Factors.Add(m1);
      t1.Factors.Add(m2);

      var t2 = new CovarianceProduct();
      var m3 = new CovarianceMask();
      m3.SelectedDimensionsParameter.Value = new IntArray(new int[] { 2 });
      var m4 = new CovarianceMask();
      m4.SelectedDimensionsParameter.Value = new IntArray(new int[] { 3 });
      t2.Factors.Add(m3);
      t2.Factors.Add(m4);

      var t3 = new CovarianceProduct();
      var m5 = new CovarianceMask();
      m5.SelectedDimensionsParameter.Value = new IntArray(new int[] { 4 });
      var m6 = new CovarianceMask();
      m6.SelectedDimensionsParameter.Value = new IntArray(new int[] { 5 });
      t3.Factors.Add(m5);
      t3.Factors.Add(m6);

      var t4 = new CovarianceProduct();
      var m1_ = new CovarianceMask();
      m1_.SelectedDimensionsParameter.Value = new IntArray(new int[] { 0 });
      var m7 = new CovarianceMask();
      m7.SelectedDimensionsParameter.Value = new IntArray(new int[] { 6 });
      var m9 = new CovarianceMask();
      m9.SelectedDimensionsParameter.Value = new IntArray(new int[] { 8 });
      t4.Factors.Add(m1_);
      t4.Factors.Add(m7);
      t4.Factors.Add(m9);

      var t5 = new CovarianceProduct();
      var m3_ = new CovarianceMask();
      m3_.SelectedDimensionsParameter.Value = new IntArray(new int[] { 2 });
      var m6_ = new CovarianceMask();
      m6_.SelectedDimensionsParameter.Value = new IntArray(new int[] { 5 });
      var m10 = new CovarianceMask();
      m10.SelectedDimensionsParameter.Value = new IntArray(new int[] { 9 });
      t5.Factors.Add(m3);
      t5.Factors.Add(m6_);
      t5.Factors.Add(m10);

      covarianceFunction.Terms.Add(t1);
      covarianceFunction.Terms.Add(t2);
      covarianceFunction.Terms.Add(t3);
      covarianceFunction.Terms.Add(t4);
      covarianceFunction.Terms.Add(t5);

      var cov = covarianceFunction.GetParameterizedCovarianceFunction(hyp, Enumerable.Range(0, 10));


      var target = Util.SampleGaussianProcess(mt, cov, data);
      data.Add(target);

      return data;
    }
  }
}