Re^4: clustering pairs

Thanks a lot, that was of a great help with the comments explaining in detail...Just thought, y not possible with arrays. Check out the program below :)

#!/usr/bin/perl
open FILE,"sampledata";
@arr = <FILE>;
chomp @arr;
close(FILE);
local $,="\n";
while(@arr) {
my @reslt;
@str = shift @arr;
push(@reslt,$str[0]);
while (@str) {
  $flag = 0;
  $str = shift @str;
  $s1,$s2,$flag) = split(/ /,$str);
  my $count = -1;
  my $acount = 0; #to arrange o/p
  foreach(@arr) {
        $count++;
        if($_ =~ /$s1|$s2/) {
              $acount++;
              if($acount == 2 || $flag == 1) {
                  unshift(@reslt,$_);
                  unshift(@str,$_." 1");
              }
              else {
                   push(@reslt,$_);
                   push(@str,$_);
              }
              splice(@arr,$count,1);
         }
    }
 }
 print @reslt,"\n";
}
[download]

Comment on Re^4: clustering pairs Download Code

Replies are listed 'Best First'.
Re^5: clustering pairs by GrandFather (Saint) on Dec 03, 2008 at 02:27 UTC
Sadly there are a few issues with that code. Some are transcription errors, some are coding convention related, some are bugs and one is at odds with a PerlMonks' convention. First the PerlMonks' convention: runnable stand alone code. By removing the file handling and using DATA instead it's easy to make your code stand alone. Coding conventions: always use strictures (use strict; use warnings;). Use a consistent indentation style (Perl tends toward K&R with 4 character indents). Use the three parameter version of open and test for errors (`open ... or die "... $!\n"` by convention). Don't slurp (`my @arr = <DATA>;`). Use blank lines to break your code up into "paragraphs". Comment tricky stuff (your use of $flag and the appended 1 for example). Bugs: it doesn't work! I get two one row clusters then everything else in one cluster. With strictures on there are "Use of uninitialized value" warnings. Perl's payment curve coincides with its learning curve.	[reply] [d/l] [select]
Re^6: clustering pairs by sugar (Beadle) on Dec 03, 2008 at 02:56 UTC
well, i will tell you the reason. There is a small change in my input data (sorry for that, earlier),so now the ID is completely similar. It used to be similar only after the '.' before. So, i am splitting it with space now. you can try running the program with the input data given below: `SID5141.C1665 SID5141.C2448 SID5141.C1253 SID5141.C2039 SID5141.C1596 SID5144.C1956 SID5141.C1906 SID5144.C2149 SID5142.C1221 SID5144.C1956 SID5144.C2149 SID5141.C2386 SID5141.C2039 SID5142.C1221 SID5141.C5887 SID5141.C7685 SID5141.C1005 SID5142.C2808 SID5141.C1046 SID5141.C1596 SID5141.C2386 SID5141.C4990 SID5141.C7685 SID5141.C4888` [download] apart from this, i have noted down your other suggestions. i am sure i will improve and take care of all what you have mentioned :)	[reply] [d/l]