GumbelMaxLikelihoodFitter.h

Go to the documentation of this file.

// --------------------------------------------------------------------------
//                   OpenMS -- Open-Source Mass Spectrometry
// --------------------------------------------------------------------------
// Copyright The OpenMS Team -- Eberhard Karls University Tuebingen,
// ETH Zurich, and Freie Universitaet Berlin 2002-2020.
//
// This software is released under a three-clause BSD license:
//  * Redistributions of source code must retain the above copyright
//    notice, this list of conditions and the following disclaimer.
//  * Redistributions in binary form must reproduce the above copyright
//    notice, this list of conditions and the following disclaimer in the
//    documentation and/or other materials provided with the distribution.
//  * Neither the name of any author or any participating institution
//    may be used to endorse or promote products derived from this software
//    without specific prior written permission.
// For a full list of authors, refer to the file AUTHORS.
// --------------------------------------------------------------------------
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
// AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
// ARE DISCLAIMED. IN NO EVENT SHALL ANY OF THE AUTHORS OR THE CONTRIBUTING
// INSTITUTIONS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
// OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
// WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
// OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
// ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
//
// --------------------------------------------------------------------------
// $Maintainer: Julianus Pfeuffer $
// $Authors: Julianus Pfeuffer $
// --------------------------------------------------------------------------
//
#pragma once
 
#include <OpenMS/DATASTRUCTURES/String.h>
#include <OpenMS/DATASTRUCTURES/DPosition.h>
#include <unsupported/Eigen/NonLinearOptimization>
#include <vector>
 
 
namespace OpenMS
{
  namespace Math
  {
    class OPENMS_DLLAPI GumbelMaxLikelihoodFitter
    {
 
public:
 
      struct GumbelDistributionFitResult
      {
        GumbelDistributionFitResult(double local_a, double local_b) :
          a(local_a),
          b(local_b)
        {
        }
 
        double a;
        double b;
 
        double log_eval_no_normalize(double x) const;
      };
 
      GumbelMaxLikelihoodFitter();
      GumbelMaxLikelihoodFitter(GumbelDistributionFitResult init);
      virtual ~GumbelMaxLikelihoodFitter();
 
      void setInitialParameters(const GumbelDistributionFitResult & result);
 
      // Generic functor
      template<typename _Scalar, int NX = Eigen::Dynamic, int NY = Eigen::Dynamic>
      struct Functor
      {
        typedef _Scalar Scalar;
        enum {
          InputsAtCompileTime = NX,
          ValuesAtCompileTime = NY
        };
        typedef Eigen::Matrix<Scalar,InputsAtCompileTime,1> InputType;
        typedef Eigen::Matrix<Scalar,ValuesAtCompileTime,1> ValueType;
        typedef Eigen::Matrix<Scalar,ValuesAtCompileTime,InputsAtCompileTime> JacobianType;
 
        int m_inputs, m_values;
 
        Functor() : m_inputs(InputsAtCompileTime), m_values(ValuesAtCompileTime) {}
        Functor(int inputs, int values) : m_inputs(inputs), m_values(values) {}
 
        int inputs() const { return m_inputs; }
        int values() const { return m_values; }
 
      };
 
      struct GumbelDistributionFunctor : Functor<double>
      {
 
        GumbelDistributionFunctor(const std::vector<double>& data, const std::vector<double>& weights):
            Functor<double>(2,2),
            m_data(data), m_weights(weights)
        {
        }
 
        int operator()(const Eigen::VectorXd &x, Eigen::VectorXd &fvec) const
        {
          fvec(0) = 0.0;
          double sigma = fabs(x(1));
          double logsigma = log(sigma);
          auto wit = m_weights.cbegin();
          for (auto it = m_data.cbegin(); it != m_data.cend(); ++it, ++wit)
          {
            double diff = (*it - x(0)) / sigma;
            fvec(0) += *wit * (-logsigma - diff - exp(-diff));
          }
          double foo = -fvec(0);
          fvec(0) = foo;
          fvec(1) = 0.0;
          return 0;
        }
        const std::vector<double>& m_data;
        const std::vector<double>& m_weights;
      };
 
      GumbelDistributionFitResult fitWeighted(const std::vector<double> & x, const std::vector<double> & w)
      {
 
        Eigen::VectorXd x_init (2);
        x_init(0) = init_param_.a;
        x_init(1) = init_param_.b;
        GumbelDistributionFunctor functor (x, w);
        Eigen::NumericalDiff<GumbelDistributionFunctor> numDiff(functor);
        Eigen::LevenbergMarquardt<Eigen::NumericalDiff<GumbelDistributionFunctor>,double> lm(numDiff);
        Eigen::LevenbergMarquardtSpace::Status status = lm.minimize(x_init);
 
        //the states are poorly documented. after checking the source, we believe that
        //all states except NotStarted, Running and ImproperInputParameters are good
        //termination states.
        if (status <= Eigen::LevenbergMarquardtSpace::Status::ImproperInputParameters)
        {
          throw Exception::UnableToFit(__FILE__, __LINE__, OPENMS_PRETTY_FUNCTION, "UnableToFit-GumbelMaxLikelihoodFitter", "Could not fit the gumbel distribution to the data");
        }
 
        #ifdef GUMBEL_DISTRIBUTION_FITTER_VERBOSE
        // build a formula with the fitted parameters for gnuplot
      stringstream formula;
      formula << "f(x)=" << "(1/" << x_init(1) << ") * " << "exp(( " << x_init(0) << "- x)/" << x_init(1) << ") * exp(-exp((" << x_init(0) << " - x)/" << x_init(1) << "))";
      cout << formula.str() << endl;
        #endif
        init_param_.a = x_init(0);
        init_param_.b = fabs(x_init(1));
 
        return {x_init(0), fabs(x_init(1))};
      }
 
protected:
 
      GumbelDistributionFitResult init_param_;
 
private:
      GumbelMaxLikelihoodFitter(const GumbelMaxLikelihoodFitter & rhs);
      GumbelMaxLikelihoodFitter & operator=(const GumbelMaxLikelihoodFitter & rhs);
    };
  }
}