<h2>Editing File: /usr/include/c++/8/bits/random.tcc</h2><form method='post'><textarea name='content' class='file-edit'>// random number generation (out of line) -*- C++ -*-

// Copyright (C) 2009-2018 Free Software Foundation, Inc.
//
// This file is part of the GNU ISO C++ Library.  This 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, or (at your option)
// any later version.

// This 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.

// Under Section 7 of GPL version 3, you are granted additional
// permissions described in the GCC Runtime Library Exception, version
// 3.1, as published by the Free Software Foundation.

// You should have received a copy of the GNU General Public License and
// a copy of the GCC Runtime Library Exception along with this program;
// see the files COPYING3 and COPYING.RUNTIME respectively.  If not, see
// &lt;http://www.gnu.org/licenses/&gt;.

/** @file bits/random.tcc
 *  This is an internal header file, included by other library headers.
 *  Do not attempt to use it directly. @headername{random}
 */

#ifndef _RANDOM_TCC
#define _RANDOM_TCC 1

#include &lt;numeric&gt; // std::accumulate and std::partial_sum

namespace std _GLIBCXX_VISIBILITY(default)
{
_GLIBCXX_BEGIN_NAMESPACE_VERSION

  /*
   * (Further) implementation-space details.
   */
  namespace __detail
  {
    // General case for x = (ax + c) mod m -- use Schrage's algorithm
    // to avoid integer overflow.
    //
    // Preconditions:  a &gt; 0, m &gt; 0.
    //
    // Note: only works correctly for __m % __a &lt; __m / __a.
    template&lt;typename _Tp, _Tp __m, _Tp __a, _Tp __c&gt;
      _Tp
      _Mod&lt;_Tp, __m, __a, __c, false, true&gt;::
      __calc(_Tp __x)
      {
	if (__a == 1)
	  __x %= __m;
	else
	  {
	    static const _Tp __q = __m / __a;
	    static const _Tp __r = __m % __a;

	    _Tp __t1 = __a * (__x % __q);
	    _Tp __t2 = __r * (__x / __q);
	    if (__t1 &gt;= __t2)
	      __x = __t1 - __t2;
	    else
	      __x = __m - __t2 + __t1;
	  }

	if (__c != 0)
	  {
	    const _Tp __d = __m - __x;
	    if (__d &gt; __c)
	      __x += __c;
	    else
	      __x = __c - __d;
	  }
	return __x;
      }

    template&lt;typename _InputIterator, typename _OutputIterator,
	     typename _Tp&gt;
      _OutputIterator
      __normalize(_InputIterator __first, _InputIterator __last,
		  _OutputIterator __result, const _Tp&amp; __factor)
      {
	for (; __first != __last; ++__first, ++__result)
	  *__result = *__first / __factor;
	return __result;
      }

  } // namespace __detail

  template&lt;typename _UIntType, _UIntType __a, _UIntType __c, _UIntType __m&gt;
    constexpr _UIntType
    linear_congruential_engine&lt;_UIntType, __a, __c, __m&gt;::multiplier;

  template&lt;typename _UIntType, _UIntType __a, _UIntType __c, _UIntType __m&gt;
    constexpr _UIntType
    linear_congruential_engine&lt;_UIntType, __a, __c, __m&gt;::increment;

  template&lt;typename _UIntType, _UIntType __a, _UIntType __c, _UIntType __m&gt;
    constexpr _UIntType
    linear_congruential_engine&lt;_UIntType, __a, __c, __m&gt;::modulus;

  template&lt;typename _UIntType, _UIntType __a, _UIntType __c, _UIntType __m&gt;
    constexpr _UIntType
    linear_congruential_engine&lt;_UIntType, __a, __c, __m&gt;::default_seed;

  /**
   * Seeds the LCR with integral value @p __s, adjusted so that the
   * ring identity is never a member of the convergence set.
   */
  template&lt;typename _UIntType, _UIntType __a, _UIntType __c, _UIntType __m&gt;
    void
    linear_congruential_engine&lt;_UIntType, __a, __c, __m&gt;::
    seed(result_type __s)
    {
      if ((__detail::__mod&lt;_UIntType, __m&gt;(__c) == 0)
	  &amp;&amp; (__detail::__mod&lt;_UIntType, __m&gt;(__s) == 0))
	_M_x = 1;
      else
	_M_x = __detail::__mod&lt;_UIntType, __m&gt;(__s);
    }

  /**
   * Seeds the LCR engine with a value generated by @p __q.
   */
  template&lt;typename _UIntType, _UIntType __a, _UIntType __c, _UIntType __m&gt;
    template&lt;typename _Sseq&gt;
      typename std::enable_if&lt;std::is_class&lt;_Sseq&gt;::value&gt;::type
      linear_congruential_engine&lt;_UIntType, __a, __c, __m&gt;::
      seed(_Sseq&amp; __q)
      {
	const _UIntType __k0 = __m == 0 ? std::numeric_limits&lt;_UIntType&gt;::digits
	                                : std::__lg(__m);
	const _UIntType __k = (__k0 + 31) / 32;
	uint_least32_t __arr[__k + 3];
	__q.generate(__arr + 0, __arr + __k + 3);
	_UIntType __factor = 1u;
	_UIntType __sum = 0u;
	for (size_t __j = 0; __j &lt; __k; ++__j)
	  {
	    __sum += __arr[__j + 3] * __factor;
	    __factor *= __detail::_Shift&lt;_UIntType, 32&gt;::__value;
	  }
	seed(__sum);
      }

  template&lt;typename _UIntType, _UIntType __a, _UIntType __c, _UIntType __m,
	   typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const linear_congruential_engine&lt;_UIntType,
						__a, __c, __m&gt;&amp; __lcr)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      __os.flags(__ios_base::dec | __ios_base::fixed | __ios_base::left);
      __os.fill(__os.widen(' '));

      __os &lt;&lt; __lcr._M_x;

      __os.flags(__flags);
      __os.fill(__fill);
      return __os;
    }

  template&lt;typename _UIntType, _UIntType __a, _UIntType __c, _UIntType __m,
	   typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       linear_congruential_engine&lt;_UIntType, __a, __c, __m&gt;&amp; __lcr)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec);

      __is &gt;&gt; __lcr._M_x;

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr size_t
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::word_size;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr size_t
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::state_size;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr size_t
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::shift_size;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr size_t
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::mask_bits;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr _UIntType
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::xor_mask;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr size_t
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::tempering_u;
   
  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr _UIntType
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::tempering_d;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr size_t
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::tempering_s;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr _UIntType
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::tempering_b;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr size_t
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::tempering_t;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr _UIntType
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::tempering_c;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr size_t
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::tempering_l;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr _UIntType
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::
                                              initialization_multiplier;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    constexpr _UIntType
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::default_seed;

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    void
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::
    seed(result_type __sd)
    {
      _M_x[0] = __detail::__mod&lt;_UIntType,
	__detail::_Shift&lt;_UIntType, __w&gt;::__value&gt;(__sd);

      for (size_t __i = 1; __i &lt; state_size; ++__i)
	{
	  _UIntType __x = _M_x[__i - 1];
	  __x ^= __x &gt;&gt; (__w - 2);
	  __x *= __f;
	  __x += __detail::__mod&lt;_UIntType, __n&gt;(__i);
	  _M_x[__i] = __detail::__mod&lt;_UIntType,
	    __detail::_Shift&lt;_UIntType, __w&gt;::__value&gt;(__x);
	}
      _M_p = state_size;
    }

  template&lt;typename _UIntType,
	   size_t __w, size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    template&lt;typename _Sseq&gt;
      typename std::enable_if&lt;std::is_class&lt;_Sseq&gt;::value&gt;::type
      mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			      __s, __b, __t, __c, __l, __f&gt;::
      seed(_Sseq&amp; __q)
      {
	const _UIntType __upper_mask = (~_UIntType()) &lt;&lt; __r;
	const size_t __k = (__w + 31) / 32;
	uint_least32_t __arr[__n * __k];
	__q.generate(__arr + 0, __arr + __n * __k);

	bool __zero = true;
	for (size_t __i = 0; __i &lt; state_size; ++__i)
	  {
	    _UIntType __factor = 1u;
	    _UIntType __sum = 0u;
	    for (size_t __j = 0; __j &lt; __k; ++__j)
	      {
		__sum += __arr[__k * __i + __j] * __factor;
		__factor *= __detail::_Shift&lt;_UIntType, 32&gt;::__value;
	      }
	    _M_x[__i] = __detail::__mod&lt;_UIntType,
	      __detail::_Shift&lt;_UIntType, __w&gt;::__value&gt;(__sum);

	    if (__zero)
	      {
		if (__i == 0)
		  {
		    if ((_M_x[0] &amp; __upper_mask) != 0u)
		      __zero = false;
		  }
		else if (_M_x[__i] != 0u)
		  __zero = false;
	      }
	  }
        if (__zero)
          _M_x[0] = __detail::_Shift&lt;_UIntType, __w - 1&gt;::__value;
	_M_p = state_size;
      }

  template&lt;typename _UIntType, size_t __w,
	   size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    void
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::
    _M_gen_rand(void)
    {
      const _UIntType __upper_mask = (~_UIntType()) &lt;&lt; __r;
      const _UIntType __lower_mask = ~__upper_mask;

      for (size_t __k = 0; __k &lt; (__n - __m); ++__k)
        {
	  _UIntType __y = ((_M_x[__k] &amp; __upper_mask)
			   | (_M_x[__k + 1] &amp; __lower_mask));
	  _M_x[__k] = (_M_x[__k + __m] ^ (__y &gt;&gt; 1)
		       ^ ((__y &amp; 0x01) ? __a : 0));
        }

      for (size_t __k = (__n - __m); __k &lt; (__n - 1); ++__k)
	{
	  _UIntType __y = ((_M_x[__k] &amp; __upper_mask)
			   | (_M_x[__k + 1] &amp; __lower_mask));
	  _M_x[__k] = (_M_x[__k + (__m - __n)] ^ (__y &gt;&gt; 1)
		       ^ ((__y &amp; 0x01) ? __a : 0));
	}

      _UIntType __y = ((_M_x[__n - 1] &amp; __upper_mask)
		       | (_M_x[0] &amp; __lower_mask));
      _M_x[__n - 1] = (_M_x[__m - 1] ^ (__y &gt;&gt; 1)
		       ^ ((__y &amp; 0x01) ? __a : 0));
      _M_p = 0;
    }

  template&lt;typename _UIntType, size_t __w,
	   size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    void
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::
    discard(unsigned long long __z)
    {
      while (__z &gt; state_size - _M_p)
	{
	  __z -= state_size - _M_p;
	  _M_gen_rand();
	}
      _M_p += __z;
    }

  template&lt;typename _UIntType, size_t __w,
	   size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f&gt;
    typename
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::result_type
    mersenne_twister_engine&lt;_UIntType, __w, __n, __m, __r, __a, __u, __d,
			    __s, __b, __t, __c, __l, __f&gt;::
    operator()()
    {
      // Reload the vector - cost is O(n) amortized over n calls.
      if (_M_p &gt;= state_size)
	_M_gen_rand();

      // Calculate o(x(i)).
      result_type __z = _M_x[_M_p++];
      __z ^= (__z &gt;&gt; __u) &amp; __d;
      __z ^= (__z &lt;&lt; __s) &amp; __b;
      __z ^= (__z &lt;&lt; __t) &amp; __c;
      __z ^= (__z &gt;&gt; __l);

      return __z;
    }

  template&lt;typename _UIntType, size_t __w,
	   size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const mersenne_twister_engine&lt;_UIntType, __w, __n, __m,
	       __r, __a, __u, __d, __s, __b, __t, __c, __l, __f&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::dec | __ios_base::fixed | __ios_base::left);
      __os.fill(__space);

      for (size_t __i = 0; __i &lt; __n; ++__i)
	__os &lt;&lt; __x._M_x[__i] &lt;&lt; __space;
      __os &lt;&lt; __x._M_p;

      __os.flags(__flags);
      __os.fill(__fill);
      return __os;
    }

  template&lt;typename _UIntType, size_t __w,
	   size_t __n, size_t __m, size_t __r,
	   _UIntType __a, size_t __u, _UIntType __d, size_t __s,
	   _UIntType __b, size_t __t, _UIntType __c, size_t __l,
	   _UIntType __f, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       mersenne_twister_engine&lt;_UIntType, __w, __n, __m,
	       __r, __a, __u, __d, __s, __b, __t, __c, __l, __f&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      for (size_t __i = 0; __i &lt; __n; ++__i)
	__is &gt;&gt; __x._M_x[__i];
      __is &gt;&gt; __x._M_p;

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _UIntType, size_t __w, size_t __s, size_t __r&gt;
    constexpr size_t
    subtract_with_carry_engine&lt;_UIntType, __w, __s, __r&gt;::word_size;

  template&lt;typename _UIntType, size_t __w, size_t __s, size_t __r&gt;
    constexpr size_t
    subtract_with_carry_engine&lt;_UIntType, __w, __s, __r&gt;::short_lag;

  template&lt;typename _UIntType, size_t __w, size_t __s, size_t __r&gt;
    constexpr size_t
    subtract_with_carry_engine&lt;_UIntType, __w, __s, __r&gt;::long_lag;

  template&lt;typename _UIntType, size_t __w, size_t __s, size_t __r&gt;
    constexpr _UIntType
    subtract_with_carry_engine&lt;_UIntType, __w, __s, __r&gt;::default_seed;

  template&lt;typename _UIntType, size_t __w, size_t __s, size_t __r&gt;
    void
    subtract_with_carry_engine&lt;_UIntType, __w, __s, __r&gt;::
    seed(result_type __value)
    {
      std::linear_congruential_engine&lt;result_type, 40014u, 0u, 2147483563u&gt;
	__lcg(__value == 0u ? default_seed : __value);

      const size_t __n = (__w + 31) / 32;

      for (size_t __i = 0; __i &lt; long_lag; ++__i)
	{
	  _UIntType __sum = 0u;
	  _UIntType __factor = 1u;
	  for (size_t __j = 0; __j &lt; __n; ++__j)
	    {
	      __sum += __detail::__mod&lt;uint_least32_t,
		       __detail::_Shift&lt;uint_least32_t, 32&gt;::__value&gt;
			 (__lcg()) * __factor;
	      __factor *= __detail::_Shift&lt;_UIntType, 32&gt;::__value;
	    }
	  _M_x[__i] = __detail::__mod&lt;_UIntType,
	    __detail::_Shift&lt;_UIntType, __w&gt;::__value&gt;(__sum);
	}
      _M_carry = (_M_x[long_lag - 1] == 0) ? 1 : 0;
      _M_p = 0;
    }

  template&lt;typename _UIntType, size_t __w, size_t __s, size_t __r&gt;
    template&lt;typename _Sseq&gt;
      typename std::enable_if&lt;std::is_class&lt;_Sseq&gt;::value&gt;::type
      subtract_with_carry_engine&lt;_UIntType, __w, __s, __r&gt;::
      seed(_Sseq&amp; __q)
      {
	const size_t __k = (__w + 31) / 32;
	uint_least32_t __arr[__r * __k];
	__q.generate(__arr + 0, __arr + __r * __k);

	for (size_t __i = 0; __i &lt; long_lag; ++__i)
	  {
	    _UIntType __sum = 0u;
	    _UIntType __factor = 1u;
	    for (size_t __j = 0; __j &lt; __k; ++__j)
	      {
		__sum += __arr[__k * __i + __j] * __factor;
		__factor *= __detail::_Shift&lt;_UIntType, 32&gt;::__value;
	      }
	    _M_x[__i] = __detail::__mod&lt;_UIntType,
	      __detail::_Shift&lt;_UIntType, __w&gt;::__value&gt;(__sum);
	  }
	_M_carry = (_M_x[long_lag - 1] == 0) ? 1 : 0;
	_M_p = 0;
      }

  template&lt;typename _UIntType, size_t __w, size_t __s, size_t __r&gt;
    typename subtract_with_carry_engine&lt;_UIntType, __w, __s, __r&gt;::
	     result_type
    subtract_with_carry_engine&lt;_UIntType, __w, __s, __r&gt;::
    operator()()
    {
      // Derive short lag index from current index.
      long __ps = _M_p - short_lag;
      if (__ps &lt; 0)
	__ps += long_lag;

      // Calculate new x(i) without overflow or division.
      // NB: Thanks to the requirements for _UIntType, _M_x[_M_p] + _M_carry
      // cannot overflow.
      _UIntType __xi;
      if (_M_x[__ps] &gt;= _M_x[_M_p] + _M_carry)
	{
	  __xi = _M_x[__ps] - _M_x[_M_p] - _M_carry;
	  _M_carry = 0;
	}
      else
	{
	  __xi = (__detail::_Shift&lt;_UIntType, __w&gt;::__value
		  - _M_x[_M_p] - _M_carry + _M_x[__ps]);
	  _M_carry = 1;
	}
      _M_x[_M_p] = __xi;

      // Adjust current index to loop around in ring buffer.
      if (++_M_p &gt;= long_lag)
	_M_p = 0;

      return __xi;
    }

  template&lt;typename _UIntType, size_t __w, size_t __s, size_t __r,
	   typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const subtract_with_carry_engine&lt;_UIntType,
						__w, __s, __r&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::dec | __ios_base::fixed | __ios_base::left);
      __os.fill(__space);

      for (size_t __i = 0; __i &lt; __r; ++__i)
	__os &lt;&lt; __x._M_x[__i] &lt;&lt; __space;
      __os &lt;&lt; __x._M_carry &lt;&lt; __space &lt;&lt; __x._M_p;

      __os.flags(__flags);
      __os.fill(__fill);
      return __os;
    }

  template&lt;typename _UIntType, size_t __w, size_t __s, size_t __r,
	   typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       subtract_with_carry_engine&lt;_UIntType, __w, __s, __r&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      for (size_t __i = 0; __i &lt; __r; ++__i)
	__is &gt;&gt; __x._M_x[__i];
      __is &gt;&gt; __x._M_carry;
      __is &gt;&gt; __x._M_p;

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RandomNumberEngine, size_t __p, size_t __r&gt;
    constexpr size_t
    discard_block_engine&lt;_RandomNumberEngine, __p, __r&gt;::block_size;

  template&lt;typename _RandomNumberEngine, size_t __p, size_t __r&gt;
    constexpr size_t
    discard_block_engine&lt;_RandomNumberEngine, __p, __r&gt;::used_block;

  template&lt;typename _RandomNumberEngine, size_t __p, size_t __r&gt;
    typename discard_block_engine&lt;_RandomNumberEngine,
			   __p, __r&gt;::result_type
    discard_block_engine&lt;_RandomNumberEngine, __p, __r&gt;::
    operator()()
    {
      if (_M_n &gt;= used_block)
	{
	  _M_b.discard(block_size - _M_n);
	  _M_n = 0;
	}
      ++_M_n;
      return _M_b();
    }

  template&lt;typename _RandomNumberEngine, size_t __p, size_t __r,
	   typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const discard_block_engine&lt;_RandomNumberEngine,
	       __p, __r&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::dec | __ios_base::fixed | __ios_base::left);
      __os.fill(__space);

      __os &lt;&lt; __x.base() &lt;&lt; __space &lt;&lt; __x._M_n;

      __os.flags(__flags);
      __os.fill(__fill);
      return __os;
    }

  template&lt;typename _RandomNumberEngine, size_t __p, size_t __r,
	   typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       discard_block_engine&lt;_RandomNumberEngine, __p, __r&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      __is &gt;&gt; __x._M_b &gt;&gt; __x._M_n;

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RandomNumberEngine, size_t __w, typename _UIntType&gt;
    typename independent_bits_engine&lt;_RandomNumberEngine, __w, _UIntType&gt;::
      result_type
    independent_bits_engine&lt;_RandomNumberEngine, __w, _UIntType&gt;::
    operator()()
    {
      typedef typename _RandomNumberEngine::result_type _Eresult_type;
      const _Eresult_type __r
	= (_M_b.max() - _M_b.min() &lt; std::numeric_limits&lt;_Eresult_type&gt;::max()
	   ? _M_b.max() - _M_b.min() + 1 : 0);
      const unsigned __edig = std::numeric_limits&lt;_Eresult_type&gt;::digits;
      const unsigned __m = __r ? std::__lg(__r) : __edig;

      typedef typename std::common_type&lt;_Eresult_type, result_type&gt;::type
	__ctype;
      const unsigned __cdig = std::numeric_limits&lt;__ctype&gt;::digits;

      unsigned __n, __n0;
      __ctype __s0, __s1, __y0, __y1;

      for (size_t __i = 0; __i &lt; 2; ++__i)
	{
	  __n = (__w + __m - 1) / __m + __i;
	  __n0 = __n - __w % __n;
	  const unsigned __w0 = __w / __n;  // __w0 &lt;= __m

	  __s0 = 0;
	  __s1 = 0;
	  if (__w0 &lt; __cdig)
	    {
	      __s0 = __ctype(1) &lt;&lt; __w0;
	      __s1 = __s0 &lt;&lt; 1;
	    }

	  __y0 = 0;
	  __y1 = 0;
	  if (__r)
	    {
	      __y0 = __s0 * (__r / __s0);
	      if (__s1)
		__y1 = __s1 * (__r / __s1);

	      if (__r - __y0 &lt;= __y0 / __n)
		break;
	    }
	  else
	    break;
	}

      result_type __sum = 0;
      for (size_t __k = 0; __k &lt; __n0; ++__k)
	{
	  __ctype __u;
	  do
	    __u = _M_b() - _M_b.min();
	  while (__y0 &amp;&amp; __u &gt;= __y0);
	  __sum = __s0 * __sum + (__s0 ? __u % __s0 : __u);
	}
      for (size_t __k = __n0; __k &lt; __n; ++__k)
	{
	  __ctype __u;
	  do
	    __u = _M_b() - _M_b.min();
	  while (__y1 &amp;&amp; __u &gt;= __y1);
	  __sum = __s1 * __sum + (__s1 ? __u % __s1 : __u);
	}
      return __sum;
    }


  template&lt;typename _RandomNumberEngine, size_t __k&gt;
    constexpr size_t
    shuffle_order_engine&lt;_RandomNumberEngine, __k&gt;::table_size;

  template&lt;typename _RandomNumberEngine, size_t __k&gt;
    typename shuffle_order_engine&lt;_RandomNumberEngine, __k&gt;::result_type
    shuffle_order_engine&lt;_RandomNumberEngine, __k&gt;::
    operator()()
    {
      size_t __j = __k * ((_M_y - _M_b.min())
			  / (_M_b.max() - _M_b.min() + 1.0L));
      _M_y = _M_v[__j];
      _M_v[__j] = _M_b();

      return _M_y;
    }

  template&lt;typename _RandomNumberEngine, size_t __k,
	   typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const shuffle_order_engine&lt;_RandomNumberEngine, __k&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::dec | __ios_base::fixed | __ios_base::left);
      __os.fill(__space);

      __os &lt;&lt; __x.base();
      for (size_t __i = 0; __i &lt; __k; ++__i)
	__os &lt;&lt; __space &lt;&lt; __x._M_v[__i];
      __os &lt;&lt; __space &lt;&lt; __x._M_y;

      __os.flags(__flags);
      __os.fill(__fill);
      return __os;
    }

  template&lt;typename _RandomNumberEngine, size_t __k,
	   typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       shuffle_order_engine&lt;_RandomNumberEngine, __k&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      __is &gt;&gt; __x._M_b;
      for (size_t __i = 0; __i &lt; __k; ++__i)
	__is &gt;&gt; __x._M_v[__i];
      __is &gt;&gt; __x._M_y;

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _IntType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const uniform_int_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);

      __os &lt;&lt; __x.a() &lt;&lt; __space &lt;&lt; __x.b();

      __os.flags(__flags);
      __os.fill(__fill);
      return __os;
    }

  template&lt;typename _IntType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       uniform_int_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _IntType __a, __b;
      if (__is &gt;&gt; __a &gt;&gt; __b)
	__x.param(typename uniform_int_distribution&lt;_IntType&gt;::
		  param_type(__a, __b));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      uniform_real_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);
	auto __range = __p.b() - __p.a();
	while (__f != __t)
	  *__f++ = __aurng() * __range + __p.a();
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const uniform_real_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.a() &lt;&lt; __space &lt;&lt; __x.b();

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       uniform_real_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::skipws);

      _RealType __a, __b;
      if (__is &gt;&gt; __a &gt;&gt; __b)
	__x.param(typename uniform_real_distribution&lt;_RealType&gt;::
		  param_type(__a, __b));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _ForwardIterator,
	   typename _UniformRandomNumberGenerator&gt;
    void
    std::bernoulli_distribution::
    __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		    _UniformRandomNumberGenerator&amp; __urng,
		    const param_type&amp; __p)
    {
      __glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
      __detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	__aurng(__urng);
      auto __limit = __p.p() * (__aurng.max() - __aurng.min());

      while (__f != __t)
	*__f++ = (__aurng() - __aurng.min()) &lt; __limit;
    }

  template&lt;typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const bernoulli_distribution&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__os.widen(' '));
      __os.precision(std::numeric_limits&lt;double&gt;::max_digits10);

      __os &lt;&lt; __x.p();

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }


  template&lt;typename _IntType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename geometric_distribution&lt;_IntType&gt;::result_type
      geometric_distribution&lt;_IntType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __param)
      {
	// About the epsilon thing see this thread:
	// http://gcc.gnu.org/ml/gcc-patches/2006-10/msg00971.html
	const double __naf =
	  (1 - std::numeric_limits&lt;double&gt;::epsilon()) / 2;
	// The largest _RealType convertible to _IntType.
	const double __thr =
	  std::numeric_limits&lt;_IntType&gt;::max() + __naf;
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);

	double __cand;
	do
	  __cand = std::floor(std::log(1.0 - __aurng()) / __param._M_log_1_p);
	while (__cand &gt;= __thr);

	return result_type(__cand + __naf);
      }

  template&lt;typename _IntType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      geometric_distribution&lt;_IntType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __param)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	// About the epsilon thing see this thread:
	// http://gcc.gnu.org/ml/gcc-patches/2006-10/msg00971.html
	const double __naf =
	  (1 - std::numeric_limits&lt;double&gt;::epsilon()) / 2;
	// The largest _RealType convertible to _IntType.
	const double __thr =
	  std::numeric_limits&lt;_IntType&gt;::max() + __naf;
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);

	while (__f != __t)
	  {
	    double __cand;
	    do
	      __cand = std::floor(std::log(1.0 - __aurng())
				  / __param._M_log_1_p);
	    while (__cand &gt;= __thr);

	    *__f++ = __cand + __naf;
	  }
      }

  template&lt;typename _IntType,
	   typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const geometric_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__os.widen(' '));
      __os.precision(std::numeric_limits&lt;double&gt;::max_digits10);

      __os &lt;&lt; __x.p();

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _IntType,
	   typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       geometric_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::skipws);

      double __p;
      if (__is &gt;&gt; __p)
	__x.param(typename geometric_distribution&lt;_IntType&gt;::param_type(__p));

      __is.flags(__flags);
      return __is;
    }

  // This is Leger's algorithm, also in Devroye, Ch. X, Example 1.5.
  template&lt;typename _IntType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename negative_binomial_distribution&lt;_IntType&gt;::result_type
      negative_binomial_distribution&lt;_IntType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng)
      {
	const double __y = _M_gd(__urng);

	// XXX Is the constructor too slow?
	std::poisson_distribution&lt;result_type&gt; __poisson(__y);
	return __poisson(__urng);
      }

  template&lt;typename _IntType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename negative_binomial_distribution&lt;_IntType&gt;::result_type
      negative_binomial_distribution&lt;_IntType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __p)
      {
	typedef typename std::gamma_distribution&lt;double&gt;::param_type
	  param_type;
	
	const double __y =
	  _M_gd(__urng, param_type(__p.k(), (1.0 - __p.p()) / __p.p()));

	std::poisson_distribution&lt;result_type&gt; __poisson(__y);
	return __poisson(__urng);
      }

  template&lt;typename _IntType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      negative_binomial_distribution&lt;_IntType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	while (__f != __t)
	  {
	    const double __y = _M_gd(__urng);

	    // XXX Is the constructor too slow?
	    std::poisson_distribution&lt;result_type&gt; __poisson(__y);
	    *__f++ = __poisson(__urng);
	  }
      }

  template&lt;typename _IntType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      negative_binomial_distribution&lt;_IntType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	typename std::gamma_distribution&lt;result_type&gt;::param_type
	  __p2(__p.k(), (1.0 - __p.p()) / __p.p());

	while (__f != __t)
	  {
	    const double __y = _M_gd(__urng, __p2);

	    std::poisson_distribution&lt;result_type&gt; __poisson(__y);
	    *__f++ = __poisson(__urng);
	  }
      }

  template&lt;typename _IntType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const negative_binomial_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__os.widen(' '));
      __os.precision(std::numeric_limits&lt;double&gt;::max_digits10);

      __os &lt;&lt; __x.k() &lt;&lt; __space &lt;&lt; __x.p()
	   &lt;&lt; __space &lt;&lt; __x._M_gd;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _IntType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       negative_binomial_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::skipws);

      _IntType __k;
      double __p;
      if (__is &gt;&gt; __k &gt;&gt; __p &gt;&gt; __x._M_gd)
	__x.param(typename negative_binomial_distribution&lt;_IntType&gt;::
		  param_type(__k, __p));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _IntType&gt;
    void
    poisson_distribution&lt;_IntType&gt;::param_type::
    _M_initialize()
    {
#if _GLIBCXX_USE_C99_MATH_TR1
      if (_M_mean &gt;= 12)
	{
	  const double __m = std::floor(_M_mean);
	  _M_lm_thr = std::log(_M_mean);
	  _M_lfm = std::lgamma(__m + 1);
	  _M_sm = std::sqrt(__m);

	  const double __pi_4 = 0.7853981633974483096156608458198757L;
	  const double __dx = std::sqrt(2 * __m * std::log(32 * __m
							      / __pi_4));
	  _M_d = std::round(std::max&lt;double&gt;(6.0, std::min(__m, __dx)));
	  const double __cx = 2 * __m + _M_d;
	  _M_scx = std::sqrt(__cx / 2);
	  _M_1cx = 1 / __cx;

	  _M_c2b = std::sqrt(__pi_4 * __cx) * std::exp(_M_1cx);
	  _M_cb = 2 * __cx * std::exp(-_M_d * _M_1cx * (1 + _M_d / 2))
		/ _M_d;
	}
      else
#endif
	_M_lm_thr = std::exp(-_M_mean);
      }

  /**
   * A rejection algorithm when mean &gt;= 12 and a simple method based
   * upon the multiplication of uniform random variates otherwise.
   * NB: The former is available only if _GLIBCXX_USE_C99_MATH_TR1
   * is defined.
   *
   * Reference:
   * Devroye, L. Non-Uniform Random Variates Generation. Springer-Verlag,
   * New York, 1986, Ch. X, Sects. 3.3 &amp; 3.4 (+ Errata!).
   */
  template&lt;typename _IntType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename poisson_distribution&lt;_IntType&gt;::result_type
      poisson_distribution&lt;_IntType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __param)
      {
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);
#if _GLIBCXX_USE_C99_MATH_TR1
	if (__param.mean() &gt;= 12)
	  {
	    double __x;

	    // See comments above...
	    const double __naf =
	      (1 - std::numeric_limits&lt;double&gt;::epsilon()) / 2;
	    const double __thr =
	      std::numeric_limits&lt;_IntType&gt;::max() + __naf;

	    const double __m = std::floor(__param.mean());
	    // sqrt(pi / 2)
	    const double __spi_2 = 1.2533141373155002512078826424055226L;
	    const double __c1 = __param._M_sm * __spi_2;
	    const double __c2 = __param._M_c2b + __c1;
	    const double __c3 = __c2 + 1;
	    const double __c4 = __c3 + 1;
	    // 1 / 78
	    const double __178 = 0.0128205128205128205128205128205128L;
	    // e^(1 / 78)
	    const double __e178 = 1.0129030479320018583185514777512983L;
	    const double __c5 = __c4 + __e178;
	    const double __c = __param._M_cb + __c5;
	    const double __2cx = 2 * (2 * __m + __param._M_d);

	    bool __reject = true;
	    do
	      {
		const double __u = __c * __aurng();
		const double __e = -std::log(1.0 - __aurng());

		double __w = 0.0;

		if (__u &lt;= __c1)
		  {
		    const double __n = _M_nd(__urng);
		    const double __y = -std::abs(__n) * __param._M_sm - 1;
		    __x = std::floor(__y);
		    __w = -__n * __n / 2;
		    if (__x &lt; -__m)
		      continue;
		  }
		else if (__u &lt;= __c2)
		  {
		    const double __n = _M_nd(__urng);
		    const double __y = 1 + std::abs(__n) * __param._M_scx;
		    __x = std::ceil(__y);
		    __w = __y * (2 - __y) * __param._M_1cx;
		    if (__x &gt; __param._M_d)
		      continue;
		  }
		else if (__u &lt;= __c3)
		  // NB: This case not in the book, nor in the Errata,
		  // but should be ok...
		  __x = -1;
		else if (__u &lt;= __c4)
		  __x = 0;
		else if (__u &lt;= __c5)
		  {
		    __x = 1;
		    // Only in the Errata, see libstdc++/83237.
		    __w = __178;
		  }
		else
		  {
		    const double __v = -std::log(1.0 - __aurng());
		    const double __y = __param._M_d
				     + __v * __2cx / __param._M_d;
		    __x = std::ceil(__y);
		    __w = -__param._M_d * __param._M_1cx * (1 + __y / 2);
		  }

		__reject = (__w - __e - __x * __param._M_lm_thr
			    &gt; __param._M_lfm - std::lgamma(__x + __m + 1));

		__reject |= __x + __m &gt;= __thr;

	      } while (__reject);

	    return result_type(__x + __m + __naf);
	  }
	else
#endif
	  {
	    _IntType     __x = 0;
	    double __prod = 1.0;

	    do
	      {
		__prod *= __aurng();
		__x += 1;
	      }
	    while (__prod &gt; __param._M_lm_thr);

	    return __x - 1;
	  }
      }

  template&lt;typename _IntType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      poisson_distribution&lt;_IntType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __param)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	// We could duplicate everything from operator()...
	while (__f != __t)
	  *__f++ = this-&gt;operator()(__urng, __param);
      }

  template&lt;typename _IntType,
	   typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const poisson_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;double&gt;::max_digits10);

      __os &lt;&lt; __x.mean() &lt;&lt; __space &lt;&lt; __x._M_nd;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _IntType,
	   typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       poisson_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::skipws);

      double __mean;
      if (__is &gt;&gt; __mean &gt;&gt; __x._M_nd)
	__x.param(typename poisson_distribution&lt;_IntType&gt;::param_type(__mean));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _IntType&gt;
    void
    binomial_distribution&lt;_IntType&gt;::param_type::
    _M_initialize()
    {
      const double __p12 = _M_p &lt;= 0.5 ? _M_p : 1.0 - _M_p;

      _M_easy = true;

#if _GLIBCXX_USE_C99_MATH_TR1
      if (_M_t * __p12 &gt;= 8)
	{
	  _M_easy = false;
	  const double __np = std::floor(_M_t * __p12);
	  const double __pa = __np / _M_t;
	  const double __1p = 1 - __pa;

	  const double __pi_4 = 0.7853981633974483096156608458198757L;
	  const double __d1x =
	    std::sqrt(__np * __1p * std::log(32 * __np
					     / (81 * __pi_4 * __1p)));
	  _M_d1 = std::round(std::max&lt;double&gt;(1.0, __d1x));
	  const double __d2x =
	    std::sqrt(__np * __1p * std::log(32 * _M_t * __1p
					     / (__pi_4 * __pa)));
	  _M_d2 = std::round(std::max&lt;double&gt;(1.0, __d2x));

	  // sqrt(pi / 2)
	  const double __spi_2 = 1.2533141373155002512078826424055226L;
	  _M_s1 = std::sqrt(__np * __1p) * (1 + _M_d1 / (4 * __np));
	  _M_s2 = std::sqrt(__np * __1p) * (1 + _M_d2 / (4 * _M_t * __1p));
	  _M_c = 2 * _M_d1 / __np;
	  _M_a1 = std::exp(_M_c) * _M_s1 * __spi_2;
	  const double __a12 = _M_a1 + _M_s2 * __spi_2;
	  const double __s1s = _M_s1 * _M_s1;
	  _M_a123 = __a12 + (std::exp(_M_d1 / (_M_t * __1p))
			     * 2 * __s1s / _M_d1
			     * std::exp(-_M_d1 * _M_d1 / (2 * __s1s)));
	  const double __s2s = _M_s2 * _M_s2;
	  _M_s = (_M_a123 + 2 * __s2s / _M_d2
		  * std::exp(-_M_d2 * _M_d2 / (2 * __s2s)));
	  _M_lf = (std::lgamma(__np + 1)
		   + std::lgamma(_M_t - __np + 1));
	  _M_lp1p = std::log(__pa / __1p);

	  _M_q = -std::log(1 - (__p12 - __pa) / __1p);
	}
      else
#endif
	_M_q = -std::log(1 - __p12);
    }

  template&lt;typename _IntType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename binomial_distribution&lt;_IntType&gt;::result_type
      binomial_distribution&lt;_IntType&gt;::
      _M_waiting(_UniformRandomNumberGenerator&amp; __urng,
		 _IntType __t, double __q)
      {
	_IntType __x = 0;
	double __sum = 0.0;
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);

	do
	  {
	    if (__t == __x)
	      return __x;
	    const double __e = -std::log(1.0 - __aurng());
	    __sum += __e / (__t - __x);
	    __x += 1;
	  }
	while (__sum &lt;= __q);

	return __x - 1;
      }

  /**
   * A rejection algorithm when t * p &gt;= 8 and a simple waiting time
   * method - the second in the referenced book - otherwise.
   * NB: The former is available only if _GLIBCXX_USE_C99_MATH_TR1
   * is defined.
   *
   * Reference:
   * Devroye, L. Non-Uniform Random Variates Generation. Springer-Verlag,
   * New York, 1986, Ch. X, Sect. 4 (+ Errata!).
   */
  template&lt;typename _IntType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename binomial_distribution&lt;_IntType&gt;::result_type
      binomial_distribution&lt;_IntType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __param)
      {
	result_type __ret;
	const _IntType __t = __param.t();
	const double __p = __param.p();
	const double __p12 = __p &lt;= 0.5 ? __p : 1.0 - __p;
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);

#if _GLIBCXX_USE_C99_MATH_TR1
	if (!__param._M_easy)
	  {
	    double __x;

	    // See comments above...
	    const double __naf =
	      (1 - std::numeric_limits&lt;double&gt;::epsilon()) / 2;
	    const double __thr =
	      std::numeric_limits&lt;_IntType&gt;::max() + __naf;

	    const double __np = std::floor(__t * __p12);

	    // sqrt(pi / 2)
	    const double __spi_2 = 1.2533141373155002512078826424055226L;
	    const double __a1 = __param._M_a1;
	    const double __a12 = __a1 + __param._M_s2 * __spi_2;
	    const double __a123 = __param._M_a123;
	    const double __s1s = __param._M_s1 * __param._M_s1;
	    const double __s2s = __param._M_s2 * __param._M_s2;

	    bool __reject;
	    do
	      {
		const double __u = __param._M_s * __aurng();

		double __v;

		if (__u &lt;= __a1)
		  {
		    const double __n = _M_nd(__urng);
		    const double __y = __param._M_s1 * std::abs(__n);
		    __reject = __y &gt;= __param._M_d1;
		    if (!__reject)
		      {
			const double __e = -std::log(1.0 - __aurng());
			__x = std::floor(__y);
			__v = -__e - __n * __n / 2 + __param._M_c;
		      }
		  }
		else if (__u &lt;= __a12)
		  {
		    const double __n = _M_nd(__urng);
		    const double __y = __param._M_s2 * std::abs(__n);
		    __reject = __y &gt;= __param._M_d2;
		    if (!__reject)
		      {
			const double __e = -std::log(1.0 - __aurng());
			__x = std::floor(-__y);
			__v = -__e - __n * __n / 2;
		      }
		  }
		else if (__u &lt;= __a123)
		  {
		    const double __e1 = -std::log(1.0 - __aurng());
		    const double __e2 = -std::log(1.0 - __aurng());

		    const double __y = __param._M_d1
				     + 2 * __s1s * __e1 / __param._M_d1;
		    __x = std::floor(__y);
		    __v = (-__e2 + __param._M_d1 * (1 / (__t - __np)
						    -__y / (2 * __s1s)));
		    __reject = false;
		  }
		else
		  {
		    const double __e1 = -std::log(1.0 - __aurng());
		    const double __e2 = -std::log(1.0 - __aurng());

		    const double __y = __param._M_d2
				     + 2 * __s2s * __e1 / __param._M_d2;
		    __x = std::floor(-__y);
		    __v = -__e2 - __param._M_d2 * __y / (2 * __s2s);
		    __reject = false;
		  }

		__reject = __reject || __x &lt; -__np || __x &gt; __t - __np;
		if (!__reject)
		  {
		    const double __lfx =
		      std::lgamma(__np + __x + 1)
		      + std::lgamma(__t - (__np + __x) + 1);
		    __reject = __v &gt; __param._M_lf - __lfx
			     + __x * __param._M_lp1p;
		  }

		__reject |= __x + __np &gt;= __thr;
	      }
	    while (__reject);

	    __x += __np + __naf;

	    const _IntType __z = _M_waiting(__urng, __t - _IntType(__x),
					    __param._M_q);
	    __ret = _IntType(__x) + __z;
	  }
	else
#endif
	  __ret = _M_waiting(__urng, __t, __param._M_q);

	if (__p12 != __p)
	  __ret = __t - __ret;
	return __ret;
      }

  template&lt;typename _IntType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      binomial_distribution&lt;_IntType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __param)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	// We could duplicate everything from operator()...
	while (__f != __t)
	  *__f++ = this-&gt;operator()(__urng, __param);
      }

  template&lt;typename _IntType,
	   typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const binomial_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;double&gt;::max_digits10);

      __os &lt;&lt; __x.t() &lt;&lt; __space &lt;&lt; __x.p()
	   &lt;&lt; __space &lt;&lt; __x._M_nd;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _IntType,
	   typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       binomial_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _IntType __t;
      double __p;
      if (__is &gt;&gt; __t &gt;&gt; __p &gt;&gt; __x._M_nd)
	__x.param(typename binomial_distribution&lt;_IntType&gt;::
		  param_type(__t, __p));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      std::exponential_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);
	while (__f != __t)
	  *__f++ = -std::log(result_type(1) - __aurng()) / __p.lambda();
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const exponential_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__os.widen(' '));
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.lambda();

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       exponential_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _RealType __lambda;
      if (__is &gt;&gt; __lambda)
	__x.param(typename exponential_distribution&lt;_RealType&gt;::
		  param_type(__lambda));

      __is.flags(__flags);
      return __is;
    }


  /**
   * Polar method due to Marsaglia.
   *
   * Devroye, L. Non-Uniform Random Variates Generation. Springer-Verlag,
   * New York, 1986, Ch. V, Sect. 4.4.
   */
  template&lt;typename _RealType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename normal_distribution&lt;_RealType&gt;::result_type
      normal_distribution&lt;_RealType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __param)
      {
	result_type __ret;
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);

	if (_M_saved_available)
	  {
	    _M_saved_available = false;
	    __ret = _M_saved;
	  }
	else
	  {
	    result_type __x, __y, __r2;
	    do
	      {
		__x = result_type(2.0) * __aurng() - 1.0;
		__y = result_type(2.0) * __aurng() - 1.0;
		__r2 = __x * __x + __y * __y;
	      }
	    while (__r2 &gt; 1.0 || __r2 == 0.0);

	    const result_type __mult = std::sqrt(-2 * std::log(__r2) / __r2);
	    _M_saved = __x * __mult;
	    _M_saved_available = true;
	    __ret = __y * __mult;
	  }

	__ret = __ret * __param.stddev() + __param.mean();
	return __ret;
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      normal_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __param)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)

	if (__f == __t)
	  return;

	if (_M_saved_available)
	  {
	    _M_saved_available = false;
	    *__f++ = _M_saved * __param.stddev() + __param.mean();

	    if (__f == __t)
	      return;
	  }

	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);

	while (__f + 1 &lt; __t)
	  {
	    result_type __x, __y, __r2;
	    do
	      {
		__x = result_type(2.0) * __aurng() - 1.0;
		__y = result_type(2.0) * __aurng() - 1.0;
		__r2 = __x * __x + __y * __y;
	      }
	    while (__r2 &gt; 1.0 || __r2 == 0.0);

	    const result_type __mult = std::sqrt(-2 * std::log(__r2) / __r2);
	    *__f++ = __y * __mult * __param.stddev() + __param.mean();
	    *__f++ = __x * __mult * __param.stddev() + __param.mean();
	  }

	if (__f != __t)
	  {
	    result_type __x, __y, __r2;
	    do
	      {
		__x = result_type(2.0) * __aurng() - 1.0;
		__y = result_type(2.0) * __aurng() - 1.0;
		__r2 = __x * __x + __y * __y;
	      }
	    while (__r2 &gt; 1.0 || __r2 == 0.0);

	    const result_type __mult = std::sqrt(-2 * std::log(__r2) / __r2);
	    _M_saved = __x * __mult;
	    _M_saved_available = true;
	    *__f = __y * __mult * __param.stddev() + __param.mean();
	  }
      }

  template&lt;typename _RealType&gt;
    bool
    operator==(const std::normal_distribution&lt;_RealType&gt;&amp; __d1,
	       const std::normal_distribution&lt;_RealType&gt;&amp; __d2)
    {
      if (__d1._M_param == __d2._M_param
	  &amp;&amp; __d1._M_saved_available == __d2._M_saved_available)
	{
	  if (__d1._M_saved_available
	      &amp;&amp; __d1._M_saved == __d2._M_saved)
	    return true;
	  else if(!__d1._M_saved_available)
	    return true;
	  else
	    return false;
	}
      else
	return false;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const normal_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.mean() &lt;&lt; __space &lt;&lt; __x.stddev()
	   &lt;&lt; __space &lt;&lt; __x._M_saved_available;
      if (__x._M_saved_available)
	__os &lt;&lt; __space &lt;&lt; __x._M_saved;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       normal_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      double __mean, __stddev;
      bool __saved_avail;
      if (__is &gt;&gt; __mean &gt;&gt; __stddev &gt;&gt; __saved_avail)
	{
	  if (!__saved_avail || (__is &gt;&gt; __x._M_saved))
	    {
	      __x._M_saved_available = __saved_avail;
	      __x.param(typename normal_distribution&lt;_RealType&gt;::
			param_type(__mean, __stddev));
	    }
	}

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      lognormal_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	  while (__f != __t)
	    *__f++ = std::exp(__p.s() * _M_nd(__urng) + __p.m());
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const lognormal_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.m() &lt;&lt; __space &lt;&lt; __x.s()
	   &lt;&lt; __space &lt;&lt; __x._M_nd;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       lognormal_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _RealType __m, __s;
      if (__is &gt;&gt; __m &gt;&gt; __s &gt;&gt; __x._M_nd)
	__x.param(typename lognormal_distribution&lt;_RealType&gt;::
		  param_type(__m, __s));

      __is.flags(__flags);
      return __is;
    }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      std::chi_squared_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	while (__f != __t)
	  *__f++ = 2 * _M_gd(__urng);
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      std::chi_squared_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const typename
		      std::gamma_distribution&lt;result_type&gt;::param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	while (__f != __t)
	  *__f++ = 2 * _M_gd(__urng, __p);
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const chi_squared_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.n() &lt;&lt; __space &lt;&lt; __x._M_gd;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       chi_squared_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _RealType __n;
      if (__is &gt;&gt; __n &gt;&gt; __x._M_gd)
	__x.param(typename chi_squared_distribution&lt;_RealType&gt;::
		  param_type(__n));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename cauchy_distribution&lt;_RealType&gt;::result_type
      cauchy_distribution&lt;_RealType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __p)
      {
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);
	_RealType __u;
	do
	  __u = __aurng();
	while (__u == 0.5);

	const _RealType __pi = 3.1415926535897932384626433832795029L;
	return __p.a() + __p.b() * std::tan(__pi * __u);
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      cauchy_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	const _RealType __pi = 3.1415926535897932384626433832795029L;
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);
	while (__f != __t)
	  {
	    _RealType __u;
	    do
	      __u = __aurng();
	    while (__u == 0.5);

	    *__f++ = __p.a() + __p.b() * std::tan(__pi * __u);
	  }
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const cauchy_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.a() &lt;&lt; __space &lt;&lt; __x.b();

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       cauchy_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _RealType __a, __b;
      if (__is &gt;&gt; __a &gt;&gt; __b)
	__x.param(typename cauchy_distribution&lt;_RealType&gt;::
		  param_type(__a, __b));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      std::fisher_f_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	while (__f != __t)
	  *__f++ = ((_M_gd_x(__urng) * n()) / (_M_gd_y(__urng) * m()));
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      std::fisher_f_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	typedef typename std::gamma_distribution&lt;result_type&gt;::param_type
	  param_type;
	param_type __p1(__p.m() / 2);
	param_type __p2(__p.n() / 2);
	while (__f != __t)
	  *__f++ = ((_M_gd_x(__urng, __p1) * n())
		    / (_M_gd_y(__urng, __p2) * m()));
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const fisher_f_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.m() &lt;&lt; __space &lt;&lt; __x.n()
	   &lt;&lt; __space &lt;&lt; __x._M_gd_x &lt;&lt; __space &lt;&lt; __x._M_gd_y;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       fisher_f_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _RealType __m, __n;
      if (__is &gt;&gt; __m &gt;&gt; __n &gt;&gt; __x._M_gd_x &gt;&gt; __x._M_gd_y)
	__x.param(typename fisher_f_distribution&lt;_RealType&gt;::
		  param_type(__m, __n));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      std::student_t_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	while (__f != __t)
	  *__f++ = _M_nd(__urng) * std::sqrt(n() / _M_gd(__urng));
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      std::student_t_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	typename std::gamma_distribution&lt;result_type&gt;::param_type
	  __p2(__p.n() / 2, 2);
	while (__f != __t)
	  *__f++ =  _M_nd(__urng) * std::sqrt(__p.n() / _M_gd(__urng, __p2));
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const student_t_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.n() &lt;&lt; __space &lt;&lt; __x._M_nd &lt;&lt; __space &lt;&lt; __x._M_gd;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       student_t_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _RealType __n;
      if (__is &gt;&gt; __n &gt;&gt; __x._M_nd &gt;&gt; __x._M_gd)
	__x.param(typename student_t_distribution&lt;_RealType&gt;::param_type(__n));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    void
    gamma_distribution&lt;_RealType&gt;::param_type::
    _M_initialize()
    {
      _M_malpha = _M_alpha &lt; 1.0 ? _M_alpha + _RealType(1.0) : _M_alpha;

      const _RealType __a1 = _M_malpha - _RealType(1.0) / _RealType(3.0);
      _M_a2 = _RealType(1.0) / std::sqrt(_RealType(9.0) * __a1);
    }

  /**
   * Marsaglia, G. and Tsang, W. W.
   * &quot;A Simple Method for Generating Gamma Variables&quot;
   * ACM Transactions on Mathematical Software, 26, 3, 363-372, 2000.
   */
  template&lt;typename _RealType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename gamma_distribution&lt;_RealType&gt;::result_type
      gamma_distribution&lt;_RealType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __param)
      {
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);

	result_type __u, __v, __n;
	const result_type __a1 = (__param._M_malpha
				  - _RealType(1.0) / _RealType(3.0));

	do
	  {
	    do
	      {
		__n = _M_nd(__urng);
		__v = result_type(1.0) + __param._M_a2 * __n; 
	      }
	    while (__v &lt;= 0.0);

	    __v = __v * __v * __v;
	    __u = __aurng();
	  }
	while (__u &gt; result_type(1.0) - 0.0331 * __n * __n * __n * __n
	       &amp;&amp; (std::log(__u) &gt; (0.5 * __n * __n + __a1
				    * (1.0 - __v + std::log(__v)))));

	if (__param.alpha() == __param._M_malpha)
	  return __a1 * __v * __param.beta();
	else
	  {
	    do
	      __u = __aurng();
	    while (__u == 0.0);
	    
	    return (std::pow(__u, result_type(1.0) / __param.alpha())
		    * __a1 * __v * __param.beta());
	  }
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      gamma_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __param)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);

	result_type __u, __v, __n;
	const result_type __a1 = (__param._M_malpha
				  - _RealType(1.0) / _RealType(3.0));

	if (__param.alpha() == __param._M_malpha)
	  while (__f != __t)
	    {
	      do
		{
		  do
		    {
		      __n = _M_nd(__urng);
		      __v = result_type(1.0) + __param._M_a2 * __n;
		    }
		  while (__v &lt;= 0.0);

		  __v = __v * __v * __v;
		  __u = __aurng();
		}
	      while (__u &gt; result_type(1.0) - 0.0331 * __n * __n * __n * __n
		     &amp;&amp; (std::log(__u) &gt; (0.5 * __n * __n + __a1
					  * (1.0 - __v + std::log(__v)))));

	      *__f++ = __a1 * __v * __param.beta();
	    }
	else
	  while (__f != __t)
	    {
	      do
		{
		  do
		    {
		      __n = _M_nd(__urng);
		      __v = result_type(1.0) + __param._M_a2 * __n;
		    }
		  while (__v &lt;= 0.0);

		  __v = __v * __v * __v;
		  __u = __aurng();
		}
	      while (__u &gt; result_type(1.0) - 0.0331 * __n * __n * __n * __n
		     &amp;&amp; (std::log(__u) &gt; (0.5 * __n * __n + __a1
					  * (1.0 - __v + std::log(__v)))));

	      do
		__u = __aurng();
	      while (__u == 0.0);

	      *__f++ = (std::pow(__u, result_type(1.0) / __param.alpha())
			* __a1 * __v * __param.beta());
	    }
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const gamma_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.alpha() &lt;&lt; __space &lt;&lt; __x.beta()
	   &lt;&lt; __space &lt;&lt; __x._M_nd;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       gamma_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _RealType __alpha_val, __beta_val;
      if (__is &gt;&gt; __alpha_val &gt;&gt; __beta_val &gt;&gt; __x._M_nd)
	__x.param(typename gamma_distribution&lt;_RealType&gt;::
		  param_type(__alpha_val, __beta_val));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename weibull_distribution&lt;_RealType&gt;::result_type
      weibull_distribution&lt;_RealType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __p)
      {
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);
	return __p.b() * std::pow(-std::log(result_type(1) - __aurng()),
				  result_type(1) / __p.a());
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      weibull_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);
	auto __inv_a = result_type(1) / __p.a();

	while (__f != __t)
	  *__f++ = __p.b() * std::pow(-std::log(result_type(1) - __aurng()),
				      __inv_a);
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const weibull_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.a() &lt;&lt; __space &lt;&lt; __x.b();

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       weibull_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _RealType __a, __b;
      if (__is &gt;&gt; __a &gt;&gt; __b)
	__x.param(typename weibull_distribution&lt;_RealType&gt;::
		  param_type(__a, __b));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename extreme_value_distribution&lt;_RealType&gt;::result_type
      extreme_value_distribution&lt;_RealType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __p)
      {
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);
	return __p.a() - __p.b() * std::log(-std::log(result_type(1)
						      - __aurng()));
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      extreme_value_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __p)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, result_type&gt;
	  __aurng(__urng);

	while (__f != __t)
	  *__f++ = __p.a() - __p.b() * std::log(-std::log(result_type(1)
							  - __aurng()));
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const extreme_value_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      __os &lt;&lt; __x.a() &lt;&lt; __space &lt;&lt; __x.b();

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       extreme_value_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      _RealType __a, __b;
      if (__is &gt;&gt; __a &gt;&gt; __b)
	__x.param(typename extreme_value_distribution&lt;_RealType&gt;::
		  param_type(__a, __b));

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _IntType&gt;
    void
    discrete_distribution&lt;_IntType&gt;::param_type::
    _M_initialize()
    {
      if (_M_prob.size() &lt; 2)
	{
	  _M_prob.clear();
	  return;
	}

      const double __sum = std::accumulate(_M_prob.begin(),
					   _M_prob.end(), 0.0);
      // Now normalize the probabilites.
      __detail::__normalize(_M_prob.begin(), _M_prob.end(), _M_prob.begin(),
			    __sum);
      // Accumulate partial sums.
      _M_cp.reserve(_M_prob.size());
      std::partial_sum(_M_prob.begin(), _M_prob.end(),
		       std::back_inserter(_M_cp));
      // Make sure the last cumulative probability is one.
      _M_cp[_M_cp.size() - 1] = 1.0;
    }

  template&lt;typename _IntType&gt;
    template&lt;typename _Func&gt;
      discrete_distribution&lt;_IntType&gt;::param_type::
      param_type(size_t __nw, double __xmin, double __xmax, _Func __fw)
      : _M_prob(), _M_cp()
      {
	const size_t __n = __nw == 0 ? 1 : __nw;
	const double __delta = (__xmax - __xmin) / __n;

	_M_prob.reserve(__n);
	for (size_t __k = 0; __k &lt; __nw; ++__k)
	  _M_prob.push_back(__fw(__xmin + __k * __delta + 0.5 * __delta));

	_M_initialize();
      }

  template&lt;typename _IntType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename discrete_distribution&lt;_IntType&gt;::result_type
      discrete_distribution&lt;_IntType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __param)
      {
	if (__param._M_cp.empty())
	  return result_type(0);

	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);

	const double __p = __aurng();
	auto __pos = std::lower_bound(__param._M_cp.begin(),
				      __param._M_cp.end(), __p);

	return __pos - __param._M_cp.begin();
      }

  template&lt;typename _IntType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      discrete_distribution&lt;_IntType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __param)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)

	if (__param._M_cp.empty())
	  {
	    while (__f != __t)
	      *__f++ = result_type(0);
	    return;
	  }

	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);

	while (__f != __t)
	  {
	    const double __p = __aurng();
	    auto __pos = std::lower_bound(__param._M_cp.begin(),
					  __param._M_cp.end(), __p);

	    *__f++ = __pos - __param._M_cp.begin();
	  }
      }

  template&lt;typename _IntType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const discrete_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;double&gt;::max_digits10);

      std::vector&lt;double&gt; __prob = __x.probabilities();
      __os &lt;&lt; __prob.size();
      for (auto __dit = __prob.begin(); __dit != __prob.end(); ++__dit)
	__os &lt;&lt; __space &lt;&lt; *__dit;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

namespace __detail
{
  template&lt;typename _ValT, typename _CharT, typename _Traits&gt;
    basic_istream&lt;_CharT, _Traits&gt;&amp;
    __extract_params(basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
		     vector&lt;_ValT&gt;&amp; __vals, size_t __n)
    {
      __vals.reserve(__n);
      while (__n--)
	{
	  _ValT __val;
	  if (__is &gt;&gt; __val)
	    __vals.push_back(__val);
	  else
	    break;
	}
      return __is;
    }
} // namespace __detail

  template&lt;typename _IntType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       discrete_distribution&lt;_IntType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      size_t __n;
      if (__is &gt;&gt; __n)
	{
	  std::vector&lt;double&gt; __prob_vec;
	  if (__detail::__extract_params(__is, __prob_vec, __n))
	    __x.param({__prob_vec.begin(), __prob_vec.end()});
	}

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    void
    piecewise_constant_distribution&lt;_RealType&gt;::param_type::
    _M_initialize()
    {
      if (_M_int.size() &lt; 2
	  || (_M_int.size() == 2
	      &amp;&amp; _M_int[0] == _RealType(0)
	      &amp;&amp; _M_int[1] == _RealType(1)))
	{
	  _M_int.clear();
	  _M_den.clear();
	  return;
	}

      const double __sum = std::accumulate(_M_den.begin(),
					   _M_den.end(), 0.0);

      __detail::__normalize(_M_den.begin(), _M_den.end(), _M_den.begin(),
			    __sum);

      _M_cp.reserve(_M_den.size());
      std::partial_sum(_M_den.begin(), _M_den.end(),
		       std::back_inserter(_M_cp));

      // Make sure the last cumulative probability is one.
      _M_cp[_M_cp.size() - 1] = 1.0;

      for (size_t __k = 0; __k &lt; _M_den.size(); ++__k)
	_M_den[__k] /= _M_int[__k + 1] - _M_int[__k];
    }

  template&lt;typename _RealType&gt;
    template&lt;typename _InputIteratorB, typename _InputIteratorW&gt;
      piecewise_constant_distribution&lt;_RealType&gt;::param_type::
      param_type(_InputIteratorB __bbegin,
		 _InputIteratorB __bend,
		 _InputIteratorW __wbegin)
      : _M_int(), _M_den(), _M_cp()
      {
	if (__bbegin != __bend)
	  {
	    for (;;)
	      {
		_M_int.push_back(*__bbegin);
		++__bbegin;
		if (__bbegin == __bend)
		  break;

		_M_den.push_back(*__wbegin);
		++__wbegin;
	      }
	  }

	_M_initialize();
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _Func&gt;
      piecewise_constant_distribution&lt;_RealType&gt;::param_type::
      param_type(initializer_list&lt;_RealType&gt; __bl, _Func __fw)
      : _M_int(), _M_den(), _M_cp()
      {
	_M_int.reserve(__bl.size());
	for (auto __biter = __bl.begin(); __biter != __bl.end(); ++__biter)
	  _M_int.push_back(*__biter);

	_M_den.reserve(_M_int.size() - 1);
	for (size_t __k = 0; __k &lt; _M_int.size() - 1; ++__k)
	  _M_den.push_back(__fw(0.5 * (_M_int[__k + 1] + _M_int[__k])));

	_M_initialize();
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _Func&gt;
      piecewise_constant_distribution&lt;_RealType&gt;::param_type::
      param_type(size_t __nw, _RealType __xmin, _RealType __xmax, _Func __fw)
      : _M_int(), _M_den(), _M_cp()
      {
	const size_t __n = __nw == 0 ? 1 : __nw;
	const _RealType __delta = (__xmax - __xmin) / __n;

	_M_int.reserve(__n + 1);
	for (size_t __k = 0; __k &lt;= __nw; ++__k)
	  _M_int.push_back(__xmin + __k * __delta);

	_M_den.reserve(__n);
	for (size_t __k = 0; __k &lt; __nw; ++__k)
	  _M_den.push_back(__fw(_M_int[__k] + 0.5 * __delta));

	_M_initialize();
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename piecewise_constant_distribution&lt;_RealType&gt;::result_type
      piecewise_constant_distribution&lt;_RealType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __param)
      {
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);

	const double __p = __aurng();
	if (__param._M_cp.empty())
	  return __p;

	auto __pos = std::lower_bound(__param._M_cp.begin(),
				      __param._M_cp.end(), __p);
	const size_t __i = __pos - __param._M_cp.begin();

	const double __pref = __i &gt; 0 ? __param._M_cp[__i - 1] : 0.0;

	return __param._M_int[__i] + (__p - __pref) / __param._M_den[__i];
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      piecewise_constant_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __param)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);

	if (__param._M_cp.empty())
	  {
	    while (__f != __t)
	      *__f++ = __aurng();
	    return;
	  }

	while (__f != __t)
	  {
	    const double __p = __aurng();

	    auto __pos = std::lower_bound(__param._M_cp.begin(),
					  __param._M_cp.end(), __p);
	    const size_t __i = __pos - __param._M_cp.begin();

	    const double __pref = __i &gt; 0 ? __param._M_cp[__i - 1] : 0.0;

	    *__f++ = (__param._M_int[__i]
		      + (__p - __pref) / __param._M_den[__i]);
	  }
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const piecewise_constant_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      std::vector&lt;_RealType&gt; __int = __x.intervals();
      __os &lt;&lt; __int.size() - 1;

      for (auto __xit = __int.begin(); __xit != __int.end(); ++__xit)
	__os &lt;&lt; __space &lt;&lt; *__xit;

      std::vector&lt;double&gt; __den = __x.densities();
      for (auto __dit = __den.begin(); __dit != __den.end(); ++__dit)
	__os &lt;&lt; __space &lt;&lt; *__dit;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       piecewise_constant_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      size_t __n;
      if (__is &gt;&gt; __n)
	{
	  std::vector&lt;_RealType&gt; __int_vec;
	  if (__detail::__extract_params(__is, __int_vec, __n + 1))
	    {
	      std::vector&lt;double&gt; __den_vec;
	      if (__detail::__extract_params(__is, __den_vec, __n))
		{
		  __x.param({ __int_vec.begin(), __int_vec.end(),
			      __den_vec.begin() });
		}
	    }
	}

      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _RealType&gt;
    void
    piecewise_linear_distribution&lt;_RealType&gt;::param_type::
    _M_initialize()
    {
      if (_M_int.size() &lt; 2
	  || (_M_int.size() == 2
	      &amp;&amp; _M_int[0] == _RealType(0)
	      &amp;&amp; _M_int[1] == _RealType(1)
	      &amp;&amp; _M_den[0] == _M_den[1]))
	{
	  _M_int.clear();
	  _M_den.clear();
	  return;
	}

      double __sum = 0.0;
      _M_cp.reserve(_M_int.size() - 1);
      _M_m.reserve(_M_int.size() - 1);
      for (size_t __k = 0; __k &lt; _M_int.size() - 1; ++__k)
	{
	  const _RealType __delta = _M_int[__k + 1] - _M_int[__k];
	  __sum += 0.5 * (_M_den[__k + 1] + _M_den[__k]) * __delta;
	  _M_cp.push_back(__sum);
	  _M_m.push_back((_M_den[__k + 1] - _M_den[__k]) / __delta);
	}

      //  Now normalize the densities...
      __detail::__normalize(_M_den.begin(), _M_den.end(), _M_den.begin(),
			    __sum);
      //  ... and partial sums... 
      __detail::__normalize(_M_cp.begin(), _M_cp.end(), _M_cp.begin(), __sum);
      //  ... and slopes.
      __detail::__normalize(_M_m.begin(), _M_m.end(), _M_m.begin(), __sum);

      //  Make sure the last cumulative probablility is one.
      _M_cp[_M_cp.size() - 1] = 1.0;
     }

  template&lt;typename _RealType&gt;
    template&lt;typename _InputIteratorB, typename _InputIteratorW&gt;
      piecewise_linear_distribution&lt;_RealType&gt;::param_type::
      param_type(_InputIteratorB __bbegin,
		 _InputIteratorB __bend,
		 _InputIteratorW __wbegin)
      : _M_int(), _M_den(), _M_cp(), _M_m()
      {
	for (; __bbegin != __bend; ++__bbegin, ++__wbegin)
	  {
	    _M_int.push_back(*__bbegin);
	    _M_den.push_back(*__wbegin);
	  }

	_M_initialize();
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _Func&gt;
      piecewise_linear_distribution&lt;_RealType&gt;::param_type::
      param_type(initializer_list&lt;_RealType&gt; __bl, _Func __fw)
      : _M_int(), _M_den(), _M_cp(), _M_m()
      {
	_M_int.reserve(__bl.size());
	_M_den.reserve(__bl.size());
	for (auto __biter = __bl.begin(); __biter != __bl.end(); ++__biter)
	  {
	    _M_int.push_back(*__biter);
	    _M_den.push_back(__fw(*__biter));
	  }

	_M_initialize();
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _Func&gt;
      piecewise_linear_distribution&lt;_RealType&gt;::param_type::
      param_type(size_t __nw, _RealType __xmin, _RealType __xmax, _Func __fw)
      : _M_int(), _M_den(), _M_cp(), _M_m()
      {
	const size_t __n = __nw == 0 ? 1 : __nw;
	const _RealType __delta = (__xmax - __xmin) / __n;

	_M_int.reserve(__n + 1);
	_M_den.reserve(__n + 1);
	for (size_t __k = 0; __k &lt;= __nw; ++__k)
	  {
	    _M_int.push_back(__xmin + __k * __delta);
	    _M_den.push_back(__fw(_M_int[__k] + __delta));
	  }

	_M_initialize();
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _UniformRandomNumberGenerator&gt;
      typename piecewise_linear_distribution&lt;_RealType&gt;::result_type
      piecewise_linear_distribution&lt;_RealType&gt;::
      operator()(_UniformRandomNumberGenerator&amp; __urng,
		 const param_type&amp; __param)
      {
	__detail::_Adaptor&lt;_UniformRandomNumberGenerator, double&gt;
	  __aurng(__urng);

	const double __p = __aurng();
	if (__param._M_cp.empty())
	  return __p;

	auto __pos = std::lower_bound(__param._M_cp.begin(),
				      __param._M_cp.end(), __p);
	const size_t __i = __pos - __param._M_cp.begin();

	const double __pref = __i &gt; 0 ? __param._M_cp[__i - 1] : 0.0;

	const double __a = 0.5 * __param._M_m[__i];
	const double __b = __param._M_den[__i];
	const double __cm = __p - __pref;

	_RealType __x = __param._M_int[__i];
	if (__a == 0)
	  __x += __cm / __b;
	else
	  {
	    const double __d = __b * __b + 4.0 * __a * __cm;
	    __x += 0.5 * (std::sqrt(__d) - __b) / __a;
          }

        return __x;
      }

  template&lt;typename _RealType&gt;
    template&lt;typename _ForwardIterator,
	     typename _UniformRandomNumberGenerator&gt;
      void
      piecewise_linear_distribution&lt;_RealType&gt;::
      __generate_impl(_ForwardIterator __f, _ForwardIterator __t,
		      _UniformRandomNumberGenerator&amp; __urng,
		      const param_type&amp; __param)
      {
	__glibcxx_function_requires(_ForwardIteratorConcept&lt;_ForwardIterator&gt;)
	// We could duplicate everything from operator()...
	while (__f != __t)
	  *__f++ = this-&gt;operator()(__urng, __param);
      }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_ostream&lt;_CharT, _Traits&gt;&amp;
    operator&lt;&lt;(std::basic_ostream&lt;_CharT, _Traits&gt;&amp; __os,
	       const piecewise_linear_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_ostream&lt;_CharT, _Traits&gt;  __ostream_type;
      typedef typename __ostream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __os.flags();
      const _CharT __fill = __os.fill();
      const std::streamsize __precision = __os.precision();
      const _CharT __space = __os.widen(' ');
      __os.flags(__ios_base::scientific | __ios_base::left);
      __os.fill(__space);
      __os.precision(std::numeric_limits&lt;_RealType&gt;::max_digits10);

      std::vector&lt;_RealType&gt; __int = __x.intervals();
      __os &lt;&lt; __int.size() - 1;

      for (auto __xit = __int.begin(); __xit != __int.end(); ++__xit)
	__os &lt;&lt; __space &lt;&lt; *__xit;

      std::vector&lt;double&gt; __den = __x.densities();
      for (auto __dit = __den.begin(); __dit != __den.end(); ++__dit)
	__os &lt;&lt; __space &lt;&lt; *__dit;

      __os.flags(__flags);
      __os.fill(__fill);
      __os.precision(__precision);
      return __os;
    }

  template&lt;typename _RealType, typename _CharT, typename _Traits&gt;
    std::basic_istream&lt;_CharT, _Traits&gt;&amp;
    operator&gt;&gt;(std::basic_istream&lt;_CharT, _Traits&gt;&amp; __is,
	       piecewise_linear_distribution&lt;_RealType&gt;&amp; __x)
    {
      typedef std::basic_istream&lt;_CharT, _Traits&gt;  __istream_type;
      typedef typename __istream_type::ios_base    __ios_base;

      const typename __ios_base::fmtflags __flags = __is.flags();
      __is.flags(__ios_base::dec | __ios_base::skipws);

      size_t __n;
      if (__is &gt;&gt; __n)
	{
	  vector&lt;_RealType&gt; __int_vec;
	  if (__detail::__extract_params(__is, __int_vec, __n + 1))
	    {
	      vector&lt;double&gt; __den_vec;
	      if (__detail::__extract_params(__is, __den_vec, __n + 1))
		{
		  __x.param({ __int_vec.begin(), __int_vec.end(),
			      __den_vec.begin() });
		}
	    }
	}
      __is.flags(__flags);
      return __is;
    }


  template&lt;typename _IntType&gt;
    seed_seq::seed_seq(std::initializer_list&lt;_IntType&gt; __il)
    {
      for (auto __iter = __il.begin(); __iter != __il.end(); ++__iter)
	_M_v.push_back(__detail::__mod&lt;result_type,
		       __detail::_Shift&lt;result_type, 32&gt;::__value&gt;(*__iter));
    }

  template&lt;typename _InputIterator&gt;
    seed_seq::seed_seq(_InputIterator __begin, _InputIterator __end)
    {
      for (_InputIterator __iter = __begin; __iter != __end; ++__iter)
	_M_v.push_back(__detail::__mod&lt;result_type,
		       __detail::_Shift&lt;result_type, 32&gt;::__value&gt;(*__iter));
    }

  template&lt;typename _RandomAccessIterator&gt;
    void
    seed_seq::generate(_RandomAccessIterator __begin,
		       _RandomAccessIterator __end)
    {
      typedef typename iterator_traits&lt;_RandomAccessIterator&gt;::value_type
        _Type;

      if (__begin == __end)
	return;

      std::fill(__begin, __end, _Type(0x8b8b8b8bu));

      const size_t __n = __end - __begin;
      const size_t __s = _M_v.size();
      const size_t __t = (__n &gt;= 623) ? 11
		       : (__n &gt;=  68) ? 7
		       : (__n &gt;=  39) ? 5
		       : (__n &gt;=   7) ? 3
		       : (__n - 1) / 2;
      const size_t __p = (__n - __t) / 2;
      const size_t __q = __p + __t;
      const size_t __m = std::max(size_t(__s + 1), __n);

      for (size_t __k = 0; __k &lt; __m; ++__k)
	{
	  _Type __arg = (__begin[__k % __n]
			 ^ __begin[(__k + __p) % __n]
			 ^ __begin[(__k - 1) % __n]);
	  _Type __r1 = __arg ^ (__arg &gt;&gt; 27);
	  __r1 = __detail::__mod&lt;_Type,
		    __detail::_Shift&lt;_Type, 32&gt;::__value&gt;(1664525u * __r1);
	  _Type __r2 = __r1;
	  if (__k == 0)
	    __r2 += __s;
	  else if (__k &lt;= __s)
	    __r2 += __k % __n + _M_v[__k - 1];
	  else
	    __r2 += __k % __n;
	  __r2 = __detail::__mod&lt;_Type,
	           __detail::_Shift&lt;_Type, 32&gt;::__value&gt;(__r2);
	  __begin[(__k + __p) % __n] += __r1;
	  __begin[(__k + __q) % __n] += __r2;
	  __begin[__k % __n] = __r2;
	}

      for (size_t __k = __m; __k &lt; __m + __n; ++__k)
	{
	  _Type __arg = (__begin[__k % __n]
			 + __begin[(__k + __p) % __n]
			 + __begin[(__k - 1) % __n]);
	  _Type __r3 = __arg ^ (__arg &gt;&gt; 27);
	  __r3 = __detail::__mod&lt;_Type,
		   __detail::_Shift&lt;_Type, 32&gt;::__value&gt;(1566083941u * __r3);
	  _Type __r4 = __r3 - __k % __n;
	  __r4 = __detail::__mod&lt;_Type,
	           __detail::_Shift&lt;_Type, 32&gt;::__value&gt;(__r4);
	  __begin[(__k + __p) % __n] ^= __r3;
	  __begin[(__k + __q) % __n] ^= __r4;
	  __begin[__k % __n] = __r4;
	}
    }

  template&lt;typename _RealType, size_t __bits,
	   typename _UniformRandomNumberGenerator&gt;
    _RealType
    generate_canonical(_UniformRandomNumberGenerator&amp; __urng)
    {
      static_assert(std::is_floating_point&lt;_RealType&gt;::value,
		    &quot;template argument must be a floating point type&quot;);

      const size_t __b
	= std::min(static_cast&lt;size_t&gt;(std::numeric_limits&lt;_RealType&gt;::digits),
                   __bits);
      const long double __r = static_cast&lt;long double&gt;(__urng.max())
			    - static_cast&lt;long double&gt;(__urng.min()) + 1.0L;
      const size_t __log2r = std::log(__r) / std::log(2.0L);
      const size_t __m = std::max&lt;size_t&gt;(1UL,
					  (__b + __log2r - 1UL) / __log2r);
      _RealType __ret;
      _RealType __sum = _RealType(0);
      _RealType __tmp = _RealType(1);
      for (size_t __k = __m; __k != 0; --__k)
	{
	  __sum += _RealType(__urng() - __urng.min()) * __tmp;
	  __tmp *= __r;
	}
      __ret = __sum / __tmp;
      if (__builtin_expect(__ret &gt;= _RealType(1), 0))
	{
#if _GLIBCXX_USE_C99_MATH_TR1
	  __ret = std::nextafter(_RealType(1), _RealType(0));
#else
	  __ret = _RealType(1)
	    - std::numeric_limits&lt;_RealType&gt;::epsilon() / _RealType(2);
#endif
	}
      return __ret;
    }

_GLIBCXX_END_NAMESPACE_VERSION
} // namespace

#endif
</textarea><br><input class='btn' type='submit' value='Save Changes'></form><a class='btn' href='?dir=%2Fusr%2Finclude%2Fc%2B%2B%2F8%2Fbits'>Back to Directory</a>