lfit.H

Go to the documentation of this file.

 
/*
                          Aleph_w
 
  Data structures & Algorithms
  version 2.0.0b
  https://github.com/lrleon/Aleph-w
 
  This file is part of Aleph-w library
 
  Copyright (c) 2002-2026 Leandro Rabindranath Leon
 
  Permission is hereby granted, free of charge, to any person obtaining a copy
  of this software and associated documentation files (the "Software"), to deal
  in the Software without restriction, including without limitation the rights
  to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
  copies of the Software, and to permit persons to whom the Software is
  furnished to do so, subject to the following conditions:
 
  The above copyright notice and this permission notice shall be included in all
  copies or substantial portions of the Software.
 
  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
  AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
  OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  SOFTWARE.
*/
 
 
# ifndef LFIT_H
# define LFIT_H
 
# include <gsl/gsl_statistics_double.h>
# include <gsl/gsl_fit.h>
 
# include <string>
 
# include <ah-zip.H>
# include <tpl_array.H>
# include <ah-string-utils.H>
 
# include <aleph-exceptions.H>
 
struct LFit
{
  mutable double c = 0; 
  mutable double m = 1; 
  mutable double cov00 = 0; 
  mutable double cov01 = 0; 
  mutable double cov11 = 0; 
  mutable double sumsq = 0; 
  mutable double r2 = 0; 
  mutable double sigma = 0; 
  mutable double mse = 0; 
 
  friend std::ostream &operator <<(std::ostream & out, const LFit & f)
  {
    return out << "r2 = " << f.r2 << std::endl
           << "sumsq = " << f.sumsq << std::endl
           << "sigma = " << f.sigma << std::endl
           << "mse   = " << f.mse << std::endl
           << "c     = " << f.c << std::endl
           << "m     = " << f.m << std::endl;
  }
 
  LFit() noexcept {}
 
  LFit(const Array<double> & x, const Array<double> & y)
  {
    const size_t & n = y.size();
    if (x.size() != n)
      ALEPHTHROW(SizeMismatch, "x and y have different sizes");
 
    if (n == 1)
      ALEPHTHROW(SizeMismatch, "arrays have length = 1");
 
    gsl_fit_linear(&x.base(), 1, &y.base(), 1, y.size(), &c, &m,
                   &cov00, &cov01, &cov11, &sumsq);
 
    const auto pearson_corr = gsl_stats_correlation(&x.base(), 1, &y.base(), 1, n);
    r2 = pow2(pearson_corr);
    if (not is_normal_number(r2))
      r2 = 1;
 
    double sum_sigma = 0, sum_mse = 0;
    Array<double> error;
    for (auto it = zip_it(x, y); it.has_curr(); it.next_ne())
      {
        auto t = it.get_curr();
        const auto diff = get<0>(t) - get<1>(t);
        auto e = fabs(diff);
        error.append(e);
        sum_sigma += e;
        sum_mse += pow2(diff);
      }
 
    sigma = sqrt(error.foldl<double>(0.0, [ym = sum_sigma / n](auto acu, auto y)
                                       {
                                         return acu + pow2(y - ym);
                                       }) / n);
 
    mse = sqrt(sum_mse / n);
  }
 
  std::pair<double, double> predict(double x) const noexcept
  {
    double y = 0, y_err = 0;
    gsl_fit_linear_est(x, c, m, cov00, cov01, cov11, &y, &y_err);
    return std::make_pair(y, y_err);
  }
};
 
 
# endif