source: trunk/yat/statistics/PearsonDistance.h @ 1703

#ifndef theplu_yat_statistics_pearson_distance_h
#define theplu_yat_statistics_pearson_distance_h

// $Id: PearsonDistance.h 1487 2008-09-10 08:41:36Z jari $

/*
  Copyright (C) 2007 Jari Häkkinen, Peter Johansson, Markus Ringnér
  Copyright (C) 2008 Peter Johansson, Markus Ringnér

  This file is part of the yat library, http://dev.thep.lu.se/yat

  The yat library 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.

  The yat library 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 yat. If not, see <http://www.gnu.org/licenses/>.
*/

#include "averager_traits.h"
#include "yat/utility/iterator_traits.h"

namespace theplu {
namespace yat {
namespace statistics {

  ///
  /// @brief Calculates the %Pearson correlation distance between two points given by elements of ranges. 
  ///
  /// This class is modelling the concept \ref concept_distance.
  ///
  struct PearsonDistance
  {
    /**
       \brief Calculates the %Pearson correlation distance between
       elements of two ranges.
    
       If elements of both ranges are unweighted the distance is
       calculated as \f$ 1-\mbox{C}(x,y) \f$, where \f$ x \f$ and \f$
       y \f$ are the two points and C is the %Pearson correlation.

       If elements of one or both of ranges have weights the distance
       is calculated as \f$ 1-[\sum w_{x,i}w_{y,i}(x_i-y_i)^2/(\sum
       w_{x,i}w_{y,i}(x_i-m_x)^2\sum w_{x,i}w_{y,i}(y_i-m_y)^2)] \f$,
       where and \f$ w_x \f$ and \f$ w_y \f$ are weights for the
       elements of the first and the second range, respectively, and
       \f$ m_x=\sum w_{x,i}w_{y,i}x_i/\sum w_{x,i}w_{y,i} \f$ and
       correspondingly for \f$ m_y \f$.  If the elements of one of the
       two ranges are unweighted, the weights for these elements are
       set to unity.
    */   
    template <typename Iter1, typename Iter2>
    double operator()
    (Iter1 beg1,Iter1 end1, Iter2 beg2) const
    {
      typename averager_pair<Iter1, Iter2>::type ap;
      add(ap,beg1,end1,beg2);
      return 1-ap.correlation();
    }
  };

}}} // of namespace statistics, yat, and theplu
    
#endif