<?xml version="1.0" encoding="windows-1252"?>
<node id="69570" title="(Golf) Nearest Neighbors" created="2001-04-04 07:10:51" updated="2005-08-13 03:30:02">
<type id="1588">
obfuscated</type>
<author id="17251">
MeowChow</author>
<data>
<field name="doctext">
The specification is simple: given a list of numbers, return the two which are nearest to one-another. Assume that the list contains at least two numbers, and the return order is irrelevent. Some foolhardy attempts of my own follow:
&lt;code&gt;
# 114 chars
sub nn1 {
  my@p;my$d;for(0..$#_-1){for my$n ($_+1..$#_){my$e=abs$_[$_]-$_[$n];if(!defined$d or$e&lt;$d){$d=$e;@p=@_[$_,$n];}}}@p
}

# 100 chars
sub nn2 {
  my@l=sort@_;$l[$_]=[@l[$_,$_+1]]for 0..$#l;pop@l;@{(sort{abs$$a[0]-$$a[1]&lt;=&gt;abs$$b[0]-$$b[1]}@l)[0]}
}
  
# 87 chars
sub nn3 {
  @_=sort@_;my%d;$d{abs$_[$_]-$_[$_+1]}=[@_[$_,$_+1]]for 0..$#_-1;@{$d{(sort keys%d)[0]}}
}
&lt;/code&gt;
&lt;b&gt;Update&lt;/b&gt;: regarding the issue of duplicates, assume they are pairs like any other, so you would return duplicate numbers if they were present in the list.
&lt;pre&gt;
  &lt;a href="/index.pl?node=MeowChow"&gt; MeowChow                                   &lt;/a&gt;
               s aamecha.s a..a\u$&amp;owag.print&lt;/pre&gt;</field>
</data>
</node>
