<?xml version="1.0" encoding="windows-1252"?>
<node id="545901" title="How do I find peaks in noisy data?" created="2006-04-26 19:09:58" updated="2006-04-26 15:09:58">
<type id="115">
perlquestion</type>
<author id="520832">
tamaguchi</author>
<data>
<field name="doctext">
I have to make a primitive pattern identifier.. &lt;BR&gt;
There is a list of values the values are  but most of them are insignificant random of noice of various magnitude.
&lt;BR&gt;
Like this for example 101, 203, 321, 45, 67, 156 etc.&lt;BR&gt;
or for example 521, 313, 31, 145, 167, 56 etc.&lt;BR&gt;
&lt;BR&gt;
Among this values are "islands" of values where the values are much higher then the surrounding noice. In this fashion:&lt;BR&gt;
&lt;BR&gt;
101, 203, 321, 45, 67, 156, 203, &lt;B&gt;502, 899, 2003, 5007, 8020, 7301, 5030, 3045, 1243, 567, &lt;/B&gt; 321, 234, 45, 123 453 etc..&lt;BR&gt;
&lt;BR&gt;
I have to make some algorithm to identify this "islands" which stick up from the surrounding noice. The noice could be at different levels so I can not just sum up values over a general noice level.v
Do you have any ideas how to do this in the best way?
I would be very happy if you knew about a book about how to do similar things?&lt;BR&gt;
&lt;BR&gt;
Than you very much for your help.&lt;BR&gt;
&lt;BR&gt;



&lt;p&gt;&lt;small&gt;2006-04-27 Retitled by [GrandFather], as per Monastery [id://341118|guidelines] &lt;br /&gt;Original title: 'primitive pattern identifier'&lt;/small&gt;&lt;/p&gt;</field>
</data>
</node>
