ContingencyTableGenerator.cpp

Go to the documentation of this file.

//
// File ContingencyTableGenerator.cpp
// Author: Julien Dutheil
// Created on: Fri Dec 10 2010 16:19
//


/*
Copyright or © or Copr. CNRS, (November 17, 2004)

This software is a computer program whose purpose is to provide classes
for numerical calculus.

This software is governed by the CeCILL  license under French law and
abiding by the rules of distribution of free software.  You can  use, 
modify and/ or redistribute the software under the terms of the CeCILL
license as circulated by CEA, CNRS and INRIA at the following URL
"http://www.cecill.info". 

As a counterpart to the access to the source code and  rights to copy,
modify and redistribute granted by the license, users are provided only
with a limited warranty  and the software's author,  the holder of the
economic rights,  and the successive licensors  have only  limited
liability. 

In this respect, the user's attention is drawn to the risks associated
with loading,  using,  modifying and/or developing or reproducing the
software by the user in light of its specific status of free software,
that may mean  that it is complicated to manipulate,  and  that  also
therefore means  that it is reserved for developers  and  experienced
professionals having in-depth computer knowledge. Users are therefore
encouraged to load and test the software's suitability as regards their
requirements in conditions enabling the security of their systems and/or 
data to be ensured and,  more generally, to use and operate it in the 
same conditions as regards security. 

The fact that you are presently reading this means that you have had
knowledge of the CeCILL license and that you accept its terms.
*/

#include "ContingencyTableGenerator.h"
#include "../VectorTools.h"

#include <iostream>

using namespace bpp;
using namespace std;

/**************************************************************************/

ContingencyTableGenerator::ContingencyTableGenerator(
    const std::vector<size_t>& nrowt,
    const std::vector<size_t>& ncolt):
  nrowt_(nrowt),
  ncolt_(ncolt),
  nrow_(nrowt.size()),
  ncol_(ncolt.size()),
  nrowm_(0),
  ncolm_(0),
  jwork_(ncolt.size()),
  ntot_(0),
  fact_(0)
{
  if (nrow_ < 2 || ncol_ < 2)
    throw Exception("ContingencyTableGenerator. Input marginals must have size greater than 1.");
  ntot_ = VectorTools::sum(nrowt_);
  if (ntot_ != VectorTools::sum(ncolt_))
    throw Exception("ContingencyTableGenerator. Marginal do not sum to the same value.");
  nrowm_ = nrow_ - 1;
  ncolm_ = ncol_ - 1;
  fact_.resize(ntot_ + 1);
  double x = 0.;
  fact_[0] = 0.;
  for (unsigned int i = 1; i <= ntot_; i++) {
    x = x + log(static_cast<double>(i));
    fact_[i] = x;
  } 
}

/* Algorithm AS 159 Applied Statistics (1981), vol. 30, no. 1
   original (C) Royal Statistical Society 1981

   Generate random two-way table with given marginal totals.

   Heavily pretty edited by Martin Maechler, Dec 2003
   use double precision for integer multiplication (against overflow);

   Taken from R source file rcont.c and adapted by Julien Dutheil, Dec 2010
*/

RowMatrix<size_t> ContingencyTableGenerator::rcont2(const RandomFactory& generator)
{
  RowMatrix<size_t> table(nrow_, ncol_); //Result
  size_t j, l, m, ia, ib, ic, jc, id, ie, ii, nll, nlm, nr_1, nc_1;
  long double x, y, dummy, sumprb;
  bool lsm, lsp;

  nr_1 = nrow_ - 1;
  nc_1 = ncol_ - 1;

  ib = 0; /* -Wall */

  /* Construct random matrix */
  for (j = 0; j < nc_1; ++j)
    jwork_[j] = ncolt_[j];

  jc = ntot_;

  for (l = 0; l < nr_1; ++l) { /* -----  matrix[ l, * ] ----- */
    ia = nrowt_[l];
    ic = jc;
    jc -= ia;/* = n_tot - sum(nr[0:l]) */

    for (m = 0; m < nc_1; ++m) {
      id = jwork_[m];
      ie = ic;
      ic -= id;
      ib = ie - ia;
      ii = ib - id;

      if (ie == 0) { /* Row [l,] is full, fill rest with zero entries */
        for (j = m; j < nc_1; ++j)
          table(l, j) = 0;
        ia = 0;
        break;
      }

      /* Generate pseudo-random number */
      dummy = generator.drawNumber();

      do {/* Outer Loop */

        /* Compute conditional expected value of MATRIX(L, M) */

        nlm = static_cast<size_t>(ia * (static_cast<long double>(id) / static_cast<long double>(ie)) + 0.5);
        x = exp(fact_[ia] + fact_[ib] + fact_[ic] + fact_[id]
          - fact_[ie] - fact_[nlm]
          - fact_[id - nlm] - fact_[ia - nlm] - fact_[ii + nlm]);
        if (x >= dummy)
          break;

        sumprb = x;
        y = x;
        nll = nlm;

        do {
          /* Increment entry in row L, column M */
          j = static_cast<size_t>((id - nlm) * static_cast<long double>(ia - nlm));
          lsp = (j == 0);
          if (!lsp) {
            ++nlm;
            x = x * j / (static_cast<long double>(nlm) * (ii + nlm));
            sumprb += x;
            if (sumprb >= dummy)
              goto L160;
          }

          do {
            /* Decrement entry in row L, column M */
            j = nll * (ii + nll);
            lsm = (j == 0);
            if (!lsm) {
              --nll;
              y = y * j / (static_cast<long double>(id - nll) * (ia - nll));
              sumprb += y;
              if (sumprb >= dummy) {
                nlm = nll;
                goto L160;
              }
              /* else */
              if (!lsp)
                break;/* to while (!lsp) */
            }
          } while (!lsm);
        } while (!lsp);

        dummy = sumprb * generator.drawNumber();

      } while (true);

L160:
      table(l, m) = nlm;
      ia -= nlm;
      jwork_[m] -= nlm;
    }
    table(l, nc_1) = ia;/* last column in row l */
  }

  /* Compute entries in last row of MATRIX */
  for (m = 0; m < nc_1; ++m)
    table(nr_1, m) = jwork_[m];

  table(nr_1, nc_1) = ib - table(nr_1, nc_1 - 1);

  return table;
}
  
/**************************************************************************/