/usr/share/doc/mcl/html/clmmate.html

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<html>
<!-- Copyright (c) 2014 Stijn van Dongen -->
<head>
<meta name="keywords" content="manual">
<style type="text/css">
/* START aephea.base.css */
body
{ text-align: justify;
margin-left: 0%;
margin-right: 0%;
}
a:link { text-decoration: none; }
a:active { text-decoration: none; }
a:visited { text-decoration: none; }
a:link { color: #1111aa; }
a:active { color: #1111aa; }
a:visited { color: #111166; }
a.local:link { color: #11aa11; }
a.local:active { color: #11aa11; }
a.local:visited { color: #116611; }
a.intern:link { color: #1111aa; }
a.intern:active { color: #1111aa; }
a.intern:visited { color: #111166; }
a.extern:link { color: #aa1111; }
a.extern:active { color: #aa1111; }
a.extern:visited { color: #661111; }
a.quiet:link { color: black; }
a.quiet:active { color: black; }
a.quiet:visited { color: black; }
div.verbatim
{ font-family: monospace;
margin-top: 1em;
margin-bottom: 1em;
font-size: 10pt;
margin-left: 2em;
white-space: pre;
}
div.indent
{ margin-left: 8%;
margin-right: 0%;
}
.right { text-align: right; }
.left { text-align: left; }
.nowrap { white-space: nowrap; }
.item_leader
{ position: relative;
margin-left: 8%;
}
.item_compact { position: absolute; vertical-align: baseline; }
.item_cascade { position: relative; }
.item_leftalign { text-align: left; }
.item_rightalign
{ width: 2em;
text-align: right;
}
.item_compact .item_rightalign
{ position: absolute;
width: 52em;
right: -2em;
text-align: right;
}
.item_text
{ position: relative;
margin-left: 3em;
}
.smallcaps { font-size: smaller; text-transform: uppercase }
/* END aephea.base.css */
body { font-family: "Garamond", "Gill Sans", "Verdana", sans-serif; }
body
{ text-align: justify;
margin-left: 8%;
margin-right: 8%;
}
</style>
<title>The clm mate manual</title>
</head>
<body>
<p style="text-align:right">
16 May 2014&nbsp;&nbsp;&nbsp;
<a class="local" href="clmmate.ps"><b>clm mate</b></a>
14-137
</p>
<div class=" itemize " style="margin-top:1em; font-size:100%">
<div class=" item_compact"><div class=" item_rightalign nowrap " style="right:-3em">1.</div></div>
<div class=" item_text " style="margin-left:4em">
<a class="intern" href="#name">NAME</a>
</div>
<div class=" item_compact"><div class=" item_rightalign nowrap " style="right:-3em">2.</div></div>
<div class=" item_text " style="margin-left:4em">
<a class="intern" href="#synopsis">SYNOPSIS</a>
</div>
<div class=" item_compact"><div class=" item_rightalign nowrap " style="right:-3em">3.</div></div>
<div class=" item_text " style="margin-left:4em">
<a class="intern" href="#description">DESCRIPTION</a>
</div>
<div class=" item_compact"><div class=" item_rightalign nowrap " style="right:-3em">4.</div></div>
<div class=" item_text " style="margin-left:4em">
<a class="intern" href="#options">OPTIONS</a>
</div>
<div class=" item_compact"><div class=" item_rightalign nowrap " style="right:-3em">5.</div></div>
<div class=" item_text " style="margin-left:4em">
<a class="intern" href="#author">AUTHOR</a>
</div>
<div class=" item_compact"><div class=" item_rightalign nowrap " style="right:-3em">6.</div></div>
<div class=" item_text " style="margin-left:4em">
<a class="intern" href="#seealso">SEE ALSO</a>
</div>
</div>

<a name="name"></a>
<h2>NAME</h2>
<p style="margin-bottom:0" class="asd_par">
clm_mate &mdash; compute best matches between two clusterings</p>
<p style="margin-bottom:0" class="asd_par">clmmate is not in actual fact a program. This manual
page documents the behaviour and options of the clm program when
invoked in mode <i>mate</i>. The options <b>-h</b>, <b>--apropos</b>,
<b>--version</b>, <b>-set</b>, <b>--nop</b> are accessible
in all <b>clm</b> modes. They are described
in the <a class="local sibling" href="clm.html">clm</a> manual page.</p>

<a name="synopsis"></a>
<h2>SYNOPSIS</h2>
<p style="margin-bottom:0" class="asd_par">
<b>clm mate</b>
<a class="intern" href="#opt-o"><b>[-o</b> fname (<i>output file name</i>)<b>]</b></a>
<a class="intern" href="#opt-b"><b>[-b</b> (<i>omit headers</i>)<b>]</b></a>
<a class="intern" href="#opt--one-to-many"><b>[--one-to-many</b> (<i>require multiple hits in &lt;clfile1&gt;</i>)<b>]</b></a>
<a class="intern" href="#opt-h"><b>[-h</b> (<i>print synopsis, exit</i>)<b>]</b></a>
<a class="intern" href="#opt--apropos"><b>[--apropos</b> (<i>print synopsis, exit</i>)<b>]</b></a>
<a class="intern" href="#opt--version"><b>[--version</b> (<i>print version, exit</i>)<b>]</b></a>
&lt;clfile1&gt; &lt;clfile2&gt;
</p>

<a name="description"></a>
<h2>DESCRIPTION</h2>
<p style="margin-bottom:0" class="asd_par">
<b>clm mate</b> computes for each cluster <tt>X</tt> in <tt>clfile1</tt> all clusters
<tt>Y</tt> in <tt>clfile2</tt> that have non-empty intersection and outputs
a line with the data points listed below.</p>
<div class="verbatim">   overlap(X,Y)               # 2 * size(meet(X,Y)) / (size(X)+size(Y))
   index(X)                   # name of cluster
   index(Y)                   # name of cluster
   size(meet(X,Y))
   size(X-Y)                  # size of left difference
   size(Y-X)                  # size of right difference
   size(X)
   size(Y)
   projection(X, clfile2)     # see below
   projection(Y, clfile1)     # see below
</div>
<p style="margin-bottom:0" class="asd_par">
The projected size of a cluster <tt>X</tt> relative to a clustering <tt>K</tt> is
simply the sum of all the nodes shared between any cluster <tt>Y</tt> in <tt>K</tt>
and <tt>X</tt>, duplications allowed. For example, the projected size of
<tt>(0,1)</tt> relative to <tt>{(0,2,4), (1,4,9), (1,3,5)}</tt> equals <tt>3</tt>.</p>
<p style="margin-bottom:0" class="asd_par">
The overlap between <tt>X</tt> and <tt>Y</tt> is exactly
1.0 if the two clusters are identical, and for nearly identical
clusterings the score will be close to 1.0.</p>
<p style="margin-bottom:0" class="asd_par">
All of this information can also be obtained from the
contingency matrix defined for two clusterings.
The <tt>[i,j]</tt> row-column entry in a contigency matrix between
to clusterings gives the number of entries in the intersection
between cluster&nbsp;<tt>i</tt> and cluster&nbsp;<tt>j</tt> from the respective
clusterings. The other information is implicitly present;
the total number of nodes in clusters&nbsp;<tt>i</tt> and&nbsp;<tt>j</tt>
for example can be obtained as the sum of entries in row&nbsp;<tt>i</tt>
and column&nbsp;<tt>j</tt> respectively, and the difference counts
can then be obtained by substracting the intersection count.
The contingency matrix can easily be computed using <b>mcx</b>;
e.g.</p>
<div class="verbatim">
mcx /clfile2 lm /clfile1 lm tp mul /ting wm</div>
<p style="margin-top:0em; margin-bottom:0em">
will create the contingency matrix in mcl matrix format
in the file <tt>ting</tt>, where columns range over the clusters
in <tt>clfile1</tt>.</p>
<p style="margin-bottom:0" class="asd_par">
The output can be put to good use by sorting it numerically on
that first score field. It is advisable to use a stable sort routine
(use the <b>-s</b> option for UNIX sort)
From this information one can quickly extract the closest
clusters between two clusterings.</p>

<a name="options"></a>
<h2>OPTIONS</h2>
<div class=" itemize " style="margin-top:1em; font-size:100%">
<div class=" item_cascade"><div class=" item_leftalign nowrap " ><a name="opt-o"></a><b>-o</b> fname (<i>output file name</i>)</div></div>
<div class=" item_text " style="margin-left:2em">
<p style="margin-top:0em; margin-bottom:0em">Specify the name of the output file.</p>
</div>
<div style="margin-top:0em">&nbsp;</div><div class=" item_cascade"><div class=" item_leftalign nowrap " ><a name="opt-b"></a><b>-b</b> (<i>omit headers</i>)</div></div>
<div class=" item_text " style="margin-left:2em">
<p style="margin-top:0em; margin-bottom:0em">Batch mode, omit column names.</p>
</div>
<div style="margin-top:0em">&nbsp;</div><div class=" item_cascade"><div class=" item_leftalign nowrap " ><a name="opt--one-to-many"></a><b>--one-to-many</b> (<i>require multiple hits in &lt;clfile1&gt;</i>)</div></div>
<div class=" item_text " style="margin-left:2em">
<p style="margin-top:0em; margin-bottom:0em">Do not output information for clusters in the first file
that are subset of a cluster in the second file.</p>
</div>
</div>

<a name="author"></a>
<h2>AUTHOR</h2>
<p style="margin-bottom:0" class="asd_par">
Stijn van Dongen.
</p>

<a name="seealso"></a>
<h2>SEE ALSO</h2>
<p style="margin-bottom:0" class="asd_par">
<a class="local sibling" href="mclfamily.html">mclfamily</a> for an overview of all the documentation
and the utilities in the mcl family.
</p>
</body>
</html>
mcl-doc 1:14-137-1 / usr / share / doc / mcl / html / clmmate.html
