| 1 |
424 |
jeremybenn |
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
|
| 2 |
|
|
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
|
| 3 |
|
|
|
| 4 |
|
|
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
|
| 5 |
|
|
<head>
|
| 6 |
|
|
<meta name="generator" content="HTML Tidy for Linux/x86 (vers 12 April 2005), see www.w3.org" />
|
| 7 |
|
|
<title>Tree Split Join Timing Test</title>
|
| 8 |
|
|
<meta http-equiv="Content-Type" content="text/html; charset=us-ascii" />
|
| 9 |
|
|
</head>
|
| 10 |
|
|
<body>
|
| 11 |
|
|
<div id="page">
|
| 12 |
|
|
<h1>Tree Split-Join Timing Test</h1>
|
| 13 |
|
|
<h2><a name="description" id="description">Description</a></h2>
|
| 14 |
|
|
<p>This test a container, inserts into a number of values,
|
| 15 |
|
|
splits the container at the median, and joins the two
|
| 16 |
|
|
containers. (If the containers are one of <tt>pb_ds</tt> 's
|
| 17 |
|
|
trees, it splits and joins with the <tt>split</tt> and
|
| 18 |
|
|
<tt>join</tt> method; otherwise, it uses the <tt>erase</tt> and
|
| 19 |
|
|
<tt>insert</tt> methods.) It measures the time for splitting
|
| 20 |
|
|
and joining the containers as a function of the number of
|
| 21 |
|
|
values inserted.</p>
|
| 22 |
|
|
<p>(The test was executed with <a href="http://gcc.gnu.org/viewcvs/*checkout*/trunk/libstdc%2B%2B-v3/testsuite/performance/ext/pb_ds/tree_split_join_timing.cc"><tt>tree_split_join_timing_test</tt></a>
|
| 23 |
|
|
200 200 2100)</p>
|
| 24 |
|
|
<h2><a name="purpose" id="purpose">Purpose</a></h2>
|
| 25 |
|
|
<p>The test checks the performance difference of <tt>join</tt>
|
| 26 |
|
|
as opposed to a sequence of <tt>insert</tt> operations; by
|
| 27 |
|
|
implication, this test checks the most efficient way to erase a
|
| 28 |
|
|
sub-sequence from a tree-like-based container, since this can
|
| 29 |
|
|
always be performed by a small sequence of splits and joins
|
| 30 |
|
|
(see <a href="motivation.html#assoc_split_join_methods">Motivation::Associative
|
| 31 |
|
|
Containers::Slightly Different Methods::Methods Related to
|
| 32 |
|
|
Split and Join</a> and <a href="tree_based_containers.html#add_methods">Design::Associative
|
| 33 |
|
|
Containers::Tree-Based Containers::Additional Methods</a>
|
| 34 |
|
|
.)</p>
|
| 35 |
|
|
<h2><a name="results" id="results">Results</a></h2>
|
| 36 |
|
|
<p>Figures <a href="#NTG">NTG</a>, <a href="#NTM">NTM</a>, and
|
| 37 |
|
|
<a href="#NTL">NTL</a> show the results for the native and
|
| 38 |
|
|
tree-based containers in <a href="assoc_performance_tests.html#gcc"><u>g++</u></a>, <a href="assoc_performance_tests.html#msvc"><u>msvc++</u></a>, and
|
| 39 |
|
|
<a href="assoc_performance_tests.html#local"><u>local</u></a>,
|
| 40 |
|
|
respectively.</p>
|
| 41 |
|
|
<div id="NTG_res_div">
|
| 42 |
|
|
<div id="NTG_gcc">
|
| 43 |
|
|
<div id="NTG_tree_split_join_timing_test">
|
| 44 |
|
|
<div id="NTG_assoc">
|
| 45 |
|
|
<div id="NTG_Native_and_tree-based_container_splits_and_joins"><div style="border-style: dotted; border-width: 1px; border-color: lightgray"><h6 class="c1"><a name="NTG" id="NTG"><img src="tree_split_join_timing_test_gcc.png" alt="no image" /></a></h6>NTG: Native and tree-based container splits and joins - <a href="assoc_performance_tests.html#gcc">g++</a><p>In the above figure, the names in the legends have the following meaning:</p>
|
| 46 |
|
|
<ol>
|
| 47 |
|
|
<li>
|
| 48 |
|
|
n_set-
|
| 49 |
|
|
<tt>std::set</tt></li>
|
| 50 |
|
|
<li>
|
| 51 |
|
|
splay_tree_set-
|
| 52 |
|
|
<a href="tree.html"><tt>tree</tt></a>
|
| 53 |
|
|
with <tt>Tag</tt> = <a href="splay_tree_tag.html"><tt>splay_tree_tag</tt></a>
|
| 54 |
|
|
, and <tt>Node_Update</tt> = <a href="null_tree_node_update.html"><tt>null_tree_node_update</tt></a>
|
| 55 |
|
|
</li>
|
| 56 |
|
|
<li>
|
| 57 |
|
|
rb_tree_set-
|
| 58 |
|
|
<a href="tree.html"><tt>tree</tt></a>
|
| 59 |
|
|
with <tt>Tag</tt> = <a href="rb_tree_tag.html"><tt>rb_tree_tag</tt></a>
|
| 60 |
|
|
, and <tt>Node_Update</tt> = <a href="null_tree_node_update.html"><tt>null_tree_node_update</tt></a>
|
| 61 |
|
|
</li>
|
| 62 |
|
|
<li>
|
| 63 |
|
|
ov_tree_set-
|
| 64 |
|
|
<a href="tree.html"><tt>tree</tt></a>
|
| 65 |
|
|
with <tt>Tag</tt> = <a href="ov_tree_tag.html"><tt>ov_tree_tag</tt></a>
|
| 66 |
|
|
, and <tt>Node_Update</tt> = <a href="null_tree_node_update.html"><tt>null_tree_node_update</tt></a>
|
| 67 |
|
|
</li>
|
| 68 |
|
|
</ol>
|
| 69 |
|
|
</div><div style="width: 100%; height: 20px"></div></div>
|
| 70 |
|
|
</div>
|
| 71 |
|
|
</div>
|
| 72 |
|
|
</div>
|
| 73 |
|
|
</div>
|
| 74 |
|
|
<div id="NTM_res_div">
|
| 75 |
|
|
<div id="NTM_msvc">
|
| 76 |
|
|
<div id="NTM_tree_split_join_timing_test">
|
| 77 |
|
|
<div id="NTM_assoc">
|
| 78 |
|
|
<div id="NTM_Native_and_tree-based_container_splits_and_joins"><div style="border-style: dotted; border-width: 1px; border-color: lightgray"><h6 class="c1"><a name="NTM" id="NTM"><img src="tree_split_join_timing_test_msvc.png" alt="no image" /></a></h6>NTM: Native and tree-based container splits and joins - <a href="assoc_performance_tests.html#msvc">msvc++</a><p>In the above figure, the names in the legends have the following meaning:</p>
|
| 79 |
|
|
<ol>
|
| 80 |
|
|
<li>
|
| 81 |
|
|
n_set-
|
| 82 |
|
|
<tt>std::set</tt></li>
|
| 83 |
|
|
<li>
|
| 84 |
|
|
splay_tree_set-
|
| 85 |
|
|
<a href="tree.html"><tt>tree</tt></a>
|
| 86 |
|
|
with <tt>Tag</tt> = <a href="splay_tree_tag.html"><tt>splay_tree_tag</tt></a>
|
| 87 |
|
|
, and <tt>Node_Update</tt> = <a href="null_tree_node_update.html"><tt>null_tree_node_update</tt></a>
|
| 88 |
|
|
</li>
|
| 89 |
|
|
<li>
|
| 90 |
|
|
rb_tree_set-
|
| 91 |
|
|
<a href="tree.html"><tt>tree</tt></a>
|
| 92 |
|
|
with <tt>Tag</tt> = <a href="rb_tree_tag.html"><tt>rb_tree_tag</tt></a>
|
| 93 |
|
|
, and <tt>Node_Update</tt> = <a href="null_tree_node_update.html"><tt>null_tree_node_update</tt></a>
|
| 94 |
|
|
</li>
|
| 95 |
|
|
<li>
|
| 96 |
|
|
ov_tree_set-
|
| 97 |
|
|
<a href="tree.html"><tt>tree</tt></a>
|
| 98 |
|
|
with <tt>Tag</tt> = <a href="ov_tree_tag.html"><tt>ov_tree_tag</tt></a>
|
| 99 |
|
|
, and <tt>Node_Update</tt> = <a href="null_tree_node_update.html"><tt>null_tree_node_update</tt></a>
|
| 100 |
|
|
</li>
|
| 101 |
|
|
</ol>
|
| 102 |
|
|
</div><div style="width: 100%; height: 20px"></div></div>
|
| 103 |
|
|
</div>
|
| 104 |
|
|
</div>
|
| 105 |
|
|
</div>
|
| 106 |
|
|
</div>
|
| 107 |
|
|
<div id="NTL_res_div">
|
| 108 |
|
|
<div id="NTL_local">
|
| 109 |
|
|
<div id="NTL_tree_split_join_timing_test">
|
| 110 |
|
|
<div id="NTL_assoc">
|
| 111 |
|
|
<div id="NTL_Native_and_tree-based_container_splits_and_joins"><div style = "border-style: dotted; border-width: 1px; border-color: lightgray"><h6 class="c1"><a name="NTL" id= "NTL"><img src="tree_split_join_timing_test_local.png" alt="no image" /></a></h6>NTL: Native and tree-based container splits and joins - <a href = "assoc_performance_tests.html#local">local</a></div><div style = "width: 100%; height: 20px"></div></div>
|
| 112 |
|
|
</div>
|
| 113 |
|
|
</div>
|
| 114 |
|
|
</div>
|
| 115 |
|
|
</div>
|
| 116 |
|
|
<h2><a name="observations" id="observations">Observations</a></h2>
|
| 117 |
|
|
<p>In this test, the native red-black trees must be split and
|
| 118 |
|
|
joined externally, through a sequence of <tt>erase</tt> and
|
| 119 |
|
|
<tt>insert</tt> operations. This is clearly super-linear, and
|
| 120 |
|
|
it is not that surprising that the cost is high.</p>
|
| 121 |
|
|
<p><tt>pb_ds</tt> 's tree-based containers use in this test the
|
| 122 |
|
|
<tt>split</tt> and <tt>join</tt> methods, which have lower
|
| 123 |
|
|
complexity: the <tt>join</tt> method of a splay tree ( <a href="tree.html"><tt>tree</tt></a>
|
| 124 |
|
|
with <tt>Tag =</tt> <a href="splay_tree_tag.html"><tt>splay_tree_tag</tt></a> ) is
|
| 125 |
|
|
quadratic in the length of the longest root-leaf path, and
|
| 126 |
|
|
linear in the total number of elements; the <tt>join</tt>
|
| 127 |
|
|
method of a red-black tree ( <a href="tree.html"><tt>tree</tt></a>
|
| 128 |
|
|
with <tt>Tag =</tt> <a href="rb_tree_tag.html"><tt>rb_tree_tag</tt></a> ) or an
|
| 129 |
|
|
ordered-vector tree ( <a href="tree.html"><tt>tree</tt></a>
|
| 130 |
|
|
with <tt>Tag =</tt> <a href="ov_tree_tag.html"><tt>ov_tree_tag</tt></a> ) is linear
|
| 131 |
|
|
in the number of elements.</p>
|
| 132 |
|
|
<p>Asides from orders of growth, <tt>pb_ds</tt> 's trees access
|
| 133 |
|
|
their allocator very little in these operations, and some of
|
| 134 |
|
|
them do not access it at all. This leads to lower constants in
|
| 135 |
|
|
their complexity, and, for some containers, to exception-free
|
| 136 |
|
|
splits and joins (which can be determined via <a href="assoc_container_traits.html"><tt>container_traits</tt></a>).</p>
|
| 137 |
|
|
<p>It is important to note that <tt>split</tt> and
|
| 138 |
|
|
<tt>join</tt> are not esoteric methods - they are the most
|
| 139 |
|
|
efficient means of erasing a contiguous range of values from a
|
| 140 |
|
|
tree based container.</p>
|
| 141 |
|
|
</div>
|
| 142 |
|
|
</body>
|
| 143 |
|
|
</html>
|
