summaryrefslogtreecommitdiff
path: root/boost/sort/spreadsort/integer_sort.hpp
blob: 0727ccd4a021e4b80d49b490f4be44747c62445e (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
//Templated Spreadsort-based implementation of integer_sort

//          Copyright Steven J. Ross 2001 - 2014.
// Distributed under the Boost Software License, Version 1.0.
//    (See accompanying file LICENSE_1_0.txt or copy at
//          http://www.boost.org/LICENSE_1_0.txt)

// See http://www.boost.org/libs/sort/ for library home page.

/*
Some improvements suggested by:
Phil Endecott and Frank Gennari

Doxygen comments by Paul A. Bristow Jan 2015

*/

#ifndef BOOST_INTEGER_SORT_HPP
#define BOOST_INTEGER_SORT_HPP
#include <algorithm>
#include <vector>
#include <cstring>
#include <limits>
#include <boost/static_assert.hpp>
#include <boost/sort/spreadsort/detail/constants.hpp>
#include <boost/sort/spreadsort/detail/integer_sort.hpp>

namespace boost {
namespace sort {
namespace spreadsort {
  //Top-level sorting call for integers.


/*! \brief Integer sort algorithm using random access iterators.
  (All variants fall back to @c std::sort if the data size is too small, < @c detail::min_sort_size).

  \details @c integer_sort is a fast templated in-place hybrid radix/comparison algorithm,
which in testing tends to be roughly 50% to 2X faster than @c std::sort for large tests (>=100kB).\n
Worst-case performance is <em>  O(N * (lg(range)/s + s)) </em>,
so @c integer_sort is asymptotically faster
than pure comparison-based algorithms. @c s is @c max_splits, which defaults to 11,
so its worst-case with default settings for 32-bit integers is
<em> O(N * ((32/11) </em> slow radix-based iterations fast comparison-based iterations).\n\n
Some performance plots of runtime vs. n and log(range) are provided:\n
   <a href="../../doc/graph/windows_integer_sort.htm"> windows_integer_sort</a>
   \n
   <a href="../../doc/graph/osx_integer_sort.htm"> osx_integer_sort</a>

   \param[in] first Iterator pointer to first element.
   \param[in] last Iterator pointing to one beyond the end of data.

   \pre [@c first, @c last) is a valid range.
   \pre @c RandomAccessIter @c value_type is mutable.
   \pre @c RandomAccessIter @c value_type is <a href="http://en.cppreference.com/w/cpp/concept/LessThanComparable">LessThanComparable</a>
   \pre @c RandomAccessIter @c value_type supports the @c operator>>,
   which returns an integer-type right-shifted a specified number of bits.
   \post The elements in the range [@c first, @c last) are sorted in ascending order.

   \throws std::exception Propagates exceptions if any of the element comparisons, the element swaps (or moves),
   the right shift, subtraction of right-shifted elements, functors, or any operations on iterators throw.

   \warning Throwing an exception may cause data loss. This will also throw if a small vector resize throws, in which case there will be no data loss.
   \warning Invalid arguments cause undefined behaviour.
   \note @c spreadsort function provides a wrapper that calls the fastest sorting algorithm available for a data type,
   enabling faster generic-programming.

   \remark The lesser of <em> O(N*log(N)) </em> comparisons and <em> O(N*log(K/S + S)) </em>operations worst-case, where:
   \remark  *  N is @c last - @c first,
   \remark  *  K is the log of the range in bits (32 for 32-bit integers using their full range),
   \remark  *  S is a constant called max_splits, defaulting to 11 (except for strings where it is the log of the character size).

*/
  template <class RandomAccessIter>
  inline void integer_sort(RandomAccessIter first, RandomAccessIter last)
  {
    // Don't sort if it's too small to optimize.
    if (last - first < detail::min_sort_size)
      std::sort(first, last);
    else
      detail::integer_sort(first, last, *first >> 0);
  }

/*! \brief Integer sort algorithm using random access iterators with both right-shift and user-defined comparison operator.
  (All variants fall back to @c std::sort if the data size is too small, < @c detail::min_sort_size).

  \details @c integer_sort is a fast templated in-place hybrid radix/comparison algorithm,
which in testing tends to be roughly 50% to 2X faster than @c std::sort for large tests (>=100kB).\n
Worst-case performance is <em>  O(N * (lg(range)/s + s)) </em>,
so @c integer_sort is asymptotically faster
than pure comparison-based algorithms. @c s is @c max_splits, which defaults to 11,
so its worst-case with default settings for 32-bit integers is
<em> O(N * ((32/11) </em> slow radix-based iterations fast comparison-based iterations).\n\n
Some performance plots of runtime vs. n and log(range) are provided:\n
   <a href="../../doc/graph/windows_integer_sort.htm"> windows_integer_sort</a>
   \n
   <a href="../../doc/graph/osx_integer_sort.htm"> osx_integer_sort</a>

   \param[in] first Iterator pointer to first element.
   \param[in] last Iterator pointing to one beyond the end of data.
   \param[in] shift Functor that returns the result of shifting the value_type right a specified number of bits.
   \param[in] comp A binary functor that returns whether the first element passed to it should go before the second in order.

   \pre [@c first, @c last) is a valid range.
   \pre @c RandomAccessIter @c value_type is mutable.
   \post The elements in the range [@c first, @c last) are sorted in ascending order.

   \return @c void.

   \throws std::exception Propagates exceptions if any of the element comparisons, the element swaps (or moves),
   the right shift, subtraction of right-shifted elements, functors,
   or any operations on iterators throw.

   \warning Throwing an exception may cause data loss. This will also throw if a small vector resize throws, in which case there will be no data loss.
   \warning Invalid arguments cause undefined behaviour.
   \note @c spreadsort function provides a wrapper that calls the fastest sorting algorithm available for a data type,
   enabling faster generic-programming.

   \remark The lesser of <em> O(N*log(N)) </em> comparisons and <em> O(N*log(K/S + S)) </em>operations worst-case, where:
   \remark  *  N is @c last - @c first,
   \remark  *  K is the log of the range in bits (32 for 32-bit integers using their full range),
   \remark  *  S is a constant called max_splits, defaulting to 11 (except for strings where it is the log of the character size).
*/
  template <class RandomAccessIter, class Right_shift, class Compare>
  inline void integer_sort(RandomAccessIter first, RandomAccessIter last,
                           Right_shift shift, Compare comp) {
    if (last - first < detail::min_sort_size)
      std::sort(first, last, comp);
    else
      detail::integer_sort(first, last, shift(*first, 0), shift, comp);
  }

/*! \brief Integer sort algorithm using random access iterators with just right-shift functor.
  (All variants fall back to @c std::sort if the data size is too small, < @c detail::min_sort_size).

  \details @c integer_sort is a fast templated in-place hybrid radix/comparison algorithm,
which in testing tends to be roughly 50% to 2X faster than @c std::sort for large tests (>=100kB).\n

\par Performance:
Worst-case performance is <em>  O(N * (lg(range)/s + s)) </em>,
so @c integer_sort is asymptotically faster
than pure comparison-based algorithms. @c s is @c max_splits, which defaults to 11,
so its worst-case with default settings for 32-bit integers is
<em> O(N * ((32/11) </em> slow radix-based iterations fast comparison-based iterations).\n\n
Some performance plots of runtime vs. n and log(range) are provided:\n
  * <a href="../../doc/graph/windows_integer_sort.htm"> windows_integer_sort</a>\n
  * <a href="../../doc/graph/osx_integer_sort.htm"> osx_integer_sort</a>

   \param[in] first Iterator pointer to first element.
   \param[in] last Iterator pointing to one beyond the end of data.
   \param[in] shift A functor that returns the result of shifting the value_type right a specified number of bits.

   \pre [@c first, @c last) is a valid range.
   \pre @c RandomAccessIter @c value_type is mutable.
   \pre @c RandomAccessIter @c value_type is <a href="http://en.cppreference.com/w/cpp/concept/LessThanComparable">LessThanComparable</a>
   \post The elements in the range [@c first, @c last) are sorted in ascending order.

   \throws std::exception Propagates exceptions if any of the element comparisons, the element swaps (or moves),
   the right shift, subtraction of right-shifted elements, functors,
   or any operations on iterators throw.

   \warning Throwing an exception may cause data loss. This will also throw if a small vector resize throws, in which case there will be no data loss.
   \warning Invalid arguments cause undefined behaviour.
   \note @c spreadsort function provides a wrapper that calls the fastest sorting algorithm available for a data type,
   enabling faster generic-programming.

   \remark The lesser of <em> O(N*log(N)) </em> comparisons and <em> O(N*log(K/S + S)) </em>operations worst-case, where:
   \remark  *  N is @c last - @c first,
   \remark  *  K is the log of the range in bits (32 for 32-bit integers using their full range),
   \remark  *  S is a constant called max_splits, defaulting to 11 (except for strings where it is the log of the character size).

*/
  template <class RandomAccessIter, class Right_shift>
  inline void integer_sort(RandomAccessIter first, RandomAccessIter last,
                           Right_shift shift) {
    if (last - first < detail::min_sort_size)
      std::sort(first, last);
    else
      detail::integer_sort(first, last, shift(*first, 0), shift);
  }
}
}
}

#endif