• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
2<html>
3<head>
4<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
5<title>The Data Structure</title>
6<link rel="stylesheet" href="../../../doc/src/boostbook.css" type="text/css">
7<meta name="generator" content="DocBook XSL Stylesheets V1.79.1">
8<link rel="home" href="../index.html" title="The Boost C++ Libraries BoostBook Documentation Subset">
9<link rel="up" href="../unordered.html" title="Chapter 44. Boost.Unordered">
10<link rel="prev" href="../unordered.html" title="Chapter 44. Boost.Unordered">
11<link rel="next" href="hash_equality.html" title="Equality Predicates and Hash Functions">
12</head>
13<body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF">
14<table cellpadding="2" width="100%"><tr>
15<td valign="top"><img alt="Boost C++ Libraries" width="277" height="86" src="../../../boost.png"></td>
16<td align="center"><a href="../../../index.html">Home</a></td>
17<td align="center"><a href="../../../libs/libraries.htm">Libraries</a></td>
18<td align="center"><a href="http://www.boost.org/users/people.html">People</a></td>
19<td align="center"><a href="http://www.boost.org/users/faq.html">FAQ</a></td>
20<td align="center"><a href="../../../more/index.htm">More</a></td>
21</tr></table>
22<hr>
23<div class="spirit-nav">
24<a accesskey="p" href="../unordered.html"><img src="../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../unordered.html"><img src="../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../index.html"><img src="../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="hash_equality.html"><img src="../../../doc/src/images/next.png" alt="Next"></a>
25</div>
26<div class="section">
27<div class="titlepage"><div><div><h2 class="title" style="clear: both">
28<a name="unordered.buckets"></a><a class="link" href="buckets.html" title="The Data Structure">The Data Structure</a>
29</h2></div></div></div>
30<p>
31      The containers are made up of a number of 'buckets', each of which can contain
32      any number of elements. For example, the following diagram shows an <code class="computeroutput"><a class="link" href="../boost/unordered_set.html" title="Class template unordered_set">unordered_set</a></code> with 7 buckets containing
33      5 elements, <code class="computeroutput"><span class="identifier">A</span></code>, <code class="computeroutput"><span class="identifier">B</span></code>, <code class="computeroutput"><span class="identifier">C</span></code>,
34      <code class="computeroutput"><span class="identifier">D</span></code> and <code class="computeroutput"><span class="identifier">E</span></code>
35      (this is just for illustration, containers will typically have more buckets).
36    </p>
37<p>
38      <span class="inlinemediaobject"><img src="../../../libs/unordered/doc/diagrams/buckets.png" align="middle"></span>
39    </p>
40<p>
41      In order to decide which bucket to place an element in, the container applies
42      the hash function, <code class="computeroutput"><span class="identifier">Hash</span></code>, to
43      the element's key (for <code class="computeroutput"><span class="identifier">unordered_set</span></code>
44      and <code class="computeroutput"><span class="identifier">unordered_multiset</span></code> the
45      key is the whole element, but is referred to as the key so that the same terminology
46      can be used for sets and maps). This returns a value of type <code class="computeroutput"><span class="identifier">std</span><span class="special">::</span><span class="identifier">size_t</span></code>.
47      <code class="computeroutput"><span class="identifier">std</span><span class="special">::</span><span class="identifier">size_t</span></code> has a much greater range of values
48      then the number of buckets, so the container applies another transformation
49      to that value to choose a bucket to place the element in.
50    </p>
51<p>
52      Retrieving the elements for a given key is simple. The same process is applied
53      to the key to find the correct bucket. Then the key is compared with the elements
54      in the bucket to find any elements that match (using the equality predicate
55      <code class="computeroutput"><span class="identifier">Pred</span></code>). If the hash function
56      has worked well the elements will be evenly distributed amongst the buckets
57      so only a small number of elements will need to be examined.
58    </p>
59<p>
60      There is <a class="link" href="hash_equality.html" title="Equality Predicates and Hash Functions">more information on hash functions
61      and equality predicates in the next section</a>.
62    </p>
63<p>
64      You can see in the diagram that <code class="computeroutput"><span class="identifier">A</span></code>
65      &amp; <code class="computeroutput"><span class="identifier">D</span></code> have been placed in
66      the same bucket. When looking for elements in this bucket up to 2 comparisons
67      are made, making the search slower. This is known as a collision. To keep things
68      fast we try to keep collisions to a minimum.
69    </p>
70<p>
71      </p>
72<div class="table">
73<a name="id-1.3.45.4.8.1"></a><p class="title"><b>Table 44.1. Methods for Accessing Buckets</b></p>
74<div class="table-contents"><table class="table" summary="Methods for Accessing Buckets">
75<colgroup>
76<col>
77<col>
78</colgroup>
79<thead><tr>
80<th><p>Method</p></th>
81<th><p>Description</p></th>
82</tr></thead>
83<tbody>
84<tr>
85<td><code class="computeroutput"><span class="identifier">size_type</span> <span class="identifier">bucket_count</span><span class="special">()</span> <span class="keyword">const</span></code></td>
86<td>The
87      number of buckets.</td>
88</tr>
89<tr>
90<td><code class="computeroutput"><span class="identifier">size_type</span> <span class="identifier">max_bucket_count</span><span class="special">()</span>
91      <span class="keyword">const</span></code></td>
92<td>An upper bound on the number of
93      buckets.</td>
94</tr>
95<tr>
96<td><code class="computeroutput"><span class="identifier">size_type</span> <span class="identifier">bucket_size</span><span class="special">(</span><span class="identifier">size_type</span> <span class="identifier">n</span><span class="special">)</span> <span class="keyword">const</span></code></td>
97<td>The
98      number of elements in bucket <code class="computeroutput"><span class="identifier">n</span></code>.</td>
99</tr>
100<tr>
101<td><code class="computeroutput"><span class="identifier">size_type</span> <span class="identifier">bucket</span><span class="special">(</span><span class="identifier">key_type</span> <span class="keyword">const</span><span class="special">&amp;</span> <span class="identifier">k</span><span class="special">)</span> <span class="keyword">const</span></code></td>
102<td>Returns
103      the index of the bucket which would contain <code class="computeroutput"><span class="identifier">k</span></code>.</td>
104</tr>
105<tr>
106<td><code class="computeroutput"><span class="identifier">local_iterator</span> <span class="identifier">begin</span><span class="special">(</span><span class="identifier">size_type</span> <span class="identifier">n</span><span class="special">);</span></code></td>
107<td rowspan="6">Return
108      begin and end iterators for bucket <code class="computeroutput"><span class="identifier">n</span></code>.</td>
109</tr>
110<tr><td><code class="computeroutput"><span class="identifier">local_iterator</span> <span class="identifier">end</span><span class="special">(</span><span class="identifier">size_type</span> <span class="identifier">n</span><span class="special">);</span></code></td></tr>
111<tr><td><code class="computeroutput"><span class="identifier">const_local_iterator</span> <span class="identifier">begin</span><span class="special">(</span><span class="identifier">size_type</span> <span class="identifier">n</span><span class="special">)</span> <span class="keyword">const</span><span class="special">;</span></code></td></tr>
112<tr><td><code class="computeroutput"><span class="identifier">const_local_iterator</span>
113      <span class="identifier">end</span><span class="special">(</span><span class="identifier">size_type</span> <span class="identifier">n</span><span class="special">)</span> <span class="keyword">const</span><span class="special">;</span></code></td></tr>
114<tr><td><code class="computeroutput"><span class="identifier">const_local_iterator</span> <span class="identifier">cbegin</span><span class="special">(</span><span class="identifier">size_type</span> <span class="identifier">n</span><span class="special">)</span> <span class="keyword">const</span><span class="special">;</span></code></td></tr>
115<tr><td><code class="computeroutput"><span class="identifier">const_local_iterator</span>
116      <span class="identifier">cend</span><span class="special">(</span><span class="identifier">size_type</span> <span class="identifier">n</span><span class="special">)</span> <span class="keyword">const</span><span class="special">;</span></code></td></tr>
117</tbody>
118</table></div>
119</div>
120<p><br class="table-break">
121    </p>
122<h3>
123<a name="unordered.buckets.h0"></a>
124      <span class="phrase"><a name="unordered.buckets.controlling_the_number_of_buckets"></a></span><a class="link" href="buckets.html#unordered.buckets.controlling_the_number_of_buckets">Controlling
125      the number of buckets</a>
126    </h3>
127<p>
128      As more elements are added to an unordered associative container, the number
129      of elements in the buckets will increase causing performance to degrade. To
130      combat this the containers increase the bucket count as elements are inserted.
131      You can also tell the container to change the bucket count (if required) by
132      calling <code class="computeroutput"><span class="identifier">rehash</span></code>.
133    </p>
134<p>
135      The standard leaves a lot of freedom to the implementer to decide how the number
136      of buckets is chosen, but it does make some requirements based on the container's
137      'load factor', the average number of elements per bucket. Containers also have
138      a 'maximum load factor' which they should try to keep the load factor below.
139    </p>
140<p>
141      You can't control the bucket count directly but there are two ways to influence
142      it:
143    </p>
144<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
145<li class="listitem">
146          Specify the minimum number of buckets when constructing a container or
147          when calling <code class="computeroutput"><span class="identifier">rehash</span></code>.
148        </li>
149<li class="listitem">
150          Suggest a maximum load factor by calling <code class="computeroutput"><span class="identifier">max_load_factor</span></code>.
151        </li>
152</ul></div>
153<p>
154      <code class="computeroutput"><span class="identifier">max_load_factor</span></code> doesn't let
155      you set the maximum load factor yourself, it just lets you give a <span class="emphasis"><em>hint</em></span>.
156      And even then, the draft standard doesn't actually require the container to
157      pay much attention to this value. The only time the load factor is <span class="emphasis"><em>required</em></span>
158      to be less than the maximum is following a call to <code class="computeroutput"><span class="identifier">rehash</span></code>.
159      But most implementations will try to keep the number of elements below the
160      max load factor, and set the maximum load factor to be the same as or close
161      to the hint - unless your hint is unreasonably small or large.
162    </p>
163<div class="table">
164<a name="unordered.buckets.bucket_size"></a><p class="title"><b>Table 44.2. Methods for Controlling Bucket Size</b></p>
165<div class="table-contents"><table class="table" summary="Methods for Controlling Bucket Size">
166<colgroup>
167<col>
168<col>
169</colgroup>
170<thead><tr>
171<th>
172              <p>
173                Method
174              </p>
175            </th>
176<th>
177              <p>
178                Description
179              </p>
180            </th>
181</tr></thead>
182<tbody>
183<tr>
184<td>
185              <p>
186                <code class="computeroutput"><span class="identifier">X</span><span class="special">(</span><span class="identifier">size_type</span> <span class="identifier">n</span><span class="special">)</span></code>
187              </p>
188            </td>
189<td>
190              <p>
191                Construct an empty container with at least <code class="computeroutput"><span class="identifier">n</span></code>
192                buckets (<code class="computeroutput"><span class="identifier">X</span></code> is the
193                container type).
194              </p>
195            </td>
196</tr>
197<tr>
198<td>
199              <p>
200                <code class="computeroutput"><span class="identifier">X</span><span class="special">(</span><span class="identifier">InputIterator</span> <span class="identifier">i</span><span class="special">,</span> <span class="identifier">InputIterator</span>
201                <span class="identifier">j</span><span class="special">,</span>
202                <span class="identifier">size_type</span> <span class="identifier">n</span><span class="special">)</span></code>
203              </p>
204            </td>
205<td>
206              <p>
207                Construct an empty container with at least <code class="computeroutput"><span class="identifier">n</span></code>
208                buckets and insert elements from the range [<code class="computeroutput"><span class="identifier">i</span></code>,
209                <code class="computeroutput"><span class="identifier">j</span></code>) (<code class="computeroutput"><span class="identifier">X</span></code> is the container type).
210              </p>
211            </td>
212</tr>
213<tr>
214<td>
215              <p>
216                <code class="computeroutput"><span class="keyword">float</span> <span class="identifier">load_factor</span><span class="special">()</span> <span class="keyword">const</span></code>
217              </p>
218            </td>
219<td>
220              <p>
221                The average number of elements per bucket.
222              </p>
223            </td>
224</tr>
225<tr>
226<td>
227              <p>
228                <code class="computeroutput"><span class="keyword">float</span> <span class="identifier">max_load_factor</span><span class="special">()</span> <span class="keyword">const</span></code>
229              </p>
230            </td>
231<td>
232              <p>
233                Returns the current maximum load factor.
234              </p>
235            </td>
236</tr>
237<tr>
238<td>
239              <p>
240                <code class="computeroutput"><span class="keyword">float</span> <span class="identifier">max_load_factor</span><span class="special">(</span><span class="keyword">float</span> <span class="identifier">z</span><span class="special">)</span></code>
241              </p>
242            </td>
243<td>
244              <p>
245                Changes the container's maximum load factor, using <code class="computeroutput"><span class="identifier">z</span></code> as a hint.
246              </p>
247            </td>
248</tr>
249<tr>
250<td>
251              <p>
252                <code class="computeroutput"><span class="keyword">void</span> <span class="identifier">rehash</span><span class="special">(</span><span class="identifier">size_type</span>
253                <span class="identifier">n</span><span class="special">)</span></code>
254              </p>
255            </td>
256<td>
257              <p>
258                Changes the number of buckets so that there at least <code class="computeroutput"><span class="identifier">n</span></code> buckets, and so that the load
259                factor is less than the maximum load factor.
260              </p>
261            </td>
262</tr>
263</tbody>
264</table></div>
265</div>
266<br class="table-break"><h3>
267<a name="unordered.buckets.h1"></a>
268      <span class="phrase"><a name="unordered.buckets.iterator_invalidation"></a></span><a class="link" href="buckets.html#unordered.buckets.iterator_invalidation">Iterator
269      Invalidation</a>
270    </h3>
271<p>
272      It is not specified how member functions other than <code class="computeroutput"><span class="identifier">rehash</span></code>
273      affect the bucket count, although <code class="computeroutput"><span class="identifier">insert</span></code>
274      is only allowed to invalidate iterators when the insertion causes the load
275      factor to be greater than or equal to the maximum load factor. For most implementations
276      this means that <code class="computeroutput"><span class="identifier">insert</span></code> will
277      only change the number of buckets when this happens. While iterators can be
278      invalidated by calls to <code class="computeroutput"><span class="identifier">insert</span></code>
279      and <code class="computeroutput"><span class="identifier">rehash</span></code>, pointers and references
280      to the container's elements are never invalidated.
281    </p>
282<p>
283      In a similar manner to using <code class="computeroutput"><span class="identifier">reserve</span></code>
284      for <code class="computeroutput"><span class="identifier">vector</span></code>s, it can be a good
285      idea to call <code class="computeroutput"><span class="identifier">rehash</span></code> before
286      inserting a large number of elements. This will get the expensive rehashing
287      out of the way and let you store iterators, safe in the knowledge that they
288      won't be invalidated. If you are inserting <code class="computeroutput"><span class="identifier">n</span></code>
289      elements into container <code class="computeroutput"><span class="identifier">x</span></code>,
290      you could first call:
291    </p>
292<pre class="programlisting"><span class="identifier">x</span><span class="special">.</span><span class="identifier">rehash</span><span class="special">((</span><span class="identifier">x</span><span class="special">.</span><span class="identifier">size</span><span class="special">()</span> <span class="special">+</span> <span class="identifier">n</span><span class="special">)</span> <span class="special">/</span> <span class="identifier">x</span><span class="special">.</span><span class="identifier">max_load_factor</span><span class="special">());</span>
293</pre>
294<div class="blurb">
295<div class="titlepage"><div><div><p class="title"><b></b></p></div></div></div>
296<p>
297      Note: <code class="computeroutput"><span class="identifier">rehash</span></code>'s argument is
298      the minimum number of buckets, not the number of elements, which is why the
299      new size is divided by the maximum load factor.
300    </p>
301</div>
302</div>
303<table xmlns:rev="http://www.cs.rpi.edu/~gregod/boost/tools/doc/revision" width="100%"><tr>
304<td align="left"></td>
305<td align="right"><div class="copyright-footer">Copyright © 2003, 2004 Jeremy B. Maitin-Shepard<br>Copyright © 2005-2008 Daniel
306      James<p>
307        Distributed under the Boost Software License, Version 1.0. (See accompanying
308        file LICENSE_1_0.txt or copy at <a href="http://www.boost.org/LICENSE_1_0.txt" target="_top">http://www.boost.org/LICENSE_1_0.txt</a>)
309      </p>
310</div></td>
311</tr></table>
312<hr>
313<div class="spirit-nav">
314<a accesskey="p" href="../unordered.html"><img src="../../../doc/src/images/prev.png" alt="Prev"></a><a accesskey="u" href="../unordered.html"><img src="../../../doc/src/images/up.png" alt="Up"></a><a accesskey="h" href="../index.html"><img src="../../../doc/src/images/home.png" alt="Home"></a><a accesskey="n" href="hash_equality.html"><img src="../../../doc/src/images/next.png" alt="Next"></a>
315</div>
316</body>
317</html>
318