id,summary,reporter,owner,description,type,status,milestone,component,version,severity,resolution,keywords,cc
9393,quantile for hypergeometric distribution - precision/rounding issue,anonymous,John Maddock,"Hi (John),

This is the same as Ticket #9183 (https://svn.boost.org/trac/boost/ticket/9183) but applies to the hypergeometric distribution.  For example,

{{{
#define BOOST_MATH_DISCRETE_QUANTILE_POLICY integer_round_up

#include <cstdio>
#include <boost/math/distributions/hypergeometric.hpp>

int main()
{
   unsigned int xmin = 71;
   unsigned int length = 9;

   boost::math::hypergeometric_distribution<> dist( 79u, 101u, 109u );

   for ( unsigned int i = 0; i < length; ++i )
   {
      double x = xmin + i;
      double y =
         boost::math::cdf( dist, x );
         // boost::math::cdf( boost::math::complement( dist, x ) );

      std::printf(
            "" x = %g\tCDF(x) = %f\tinverse-CDF(CDF(x)) = %g\n"",
            x, y, boost::math::quantile( dist, y ) );
   }

   return 0;
}
}}}

returns

{{{
 x = 71	CDF(x) = 0.068671	inverse-CDF(CDF(x)) = 72
 x = 72	CDF(x) = 0.297575	inverse-CDF(CDF(x)) = 72
 x = 73	CDF(x) = 0.615846	inverse-CDF(CDF(x)) = 73
 x = 74	CDF(x) = 0.856699	inverse-CDF(CDF(x)) = 74
 x = 75	CDF(x) = 0.965083	inverse-CDF(CDF(x)) = 75
 x = 76	CDF(x) = 0.994746	inverse-CDF(CDF(x)) = 77
 x = 77	CDF(x) = 0.999561	inverse-CDF(CDF(x)) = 77
 x = 78	CDF(x) = 0.999985	inverse-CDF(CDF(x)) = 78
 x = 79	CDF(x) = 1.000000	inverse-CDF(CDF(x)) = 79
}}}

I was hoping/expecting to obtain:

{{{
       x          y  z
 [1,] 71 0.06867117 71
 [2,] 72 0.29757505 72
 [3,] 73 0.61584552 73
 [4,] 74 0.85669885 74
 [5,] 75 0.96508285 75
 [6,] 76 0.99474584 76
 [7,] 77 0.99956126 77
 [8,] 78 0.99998459 78
 [9,] 79 1.00000000 79
}}}
which I was used to on R.

Thank you!",Feature Requests,new,To Be Determined,math,Boost Development Trunk,Showstopper,,,