Self-Looping over hash - how to remove duplicate

monkfan has asked for the wisdom of the Perl Monks concerning the following question:

Hi,
Suppose I have this array:


 my %hash2 =(
           '1' => 2,
           '3' => 5,
           '4' => 2,);
[download]

Where the key is index, and the values is the count of that index.
How can I create the new set hash, so that it gives:

$VAR1  = 
{
          '3-3' => 5,
          '3-4' => 10,
          '1-1' => 2,
          '3-1' => 10,
          '1-4' => 4,
          '4-4' => 2,#ends here
};
[download]

My current code:

use strict;
use warnings;
use Data::Dumper;

#print Dumper \%hash2;
my %spair = ();
my $prod;
foreach my $si (keys %hash2)
  {
    foreach my $sn (keys %hash2)
      {    
           if ($si == $sn)
         {
               $prod = $hash2{$si};
         }
           else
         { $prod = $hash2{$si}*$hash2{$sn};
         }
           $spair{$si."-".$sn} = $prod;  
      }
  }
print Dumper \%spair;
[download]

give this which is not as desired:

$VAR1 = {
          '3-3' => 5,
          '3-4' => 10,
          '1-1' => 2,
          '3-1' => 10,
          '1-4' => 4,
          '4-4' => 2,
          '4-1' => 4, #repeat
          '1-3' => 10,#repeat
          '4-3' => 10 #repeat
        };
[download]

As you can see it repeats 1-4 and 4-1, etc. which actually should only occur once. How can I modify my code so that we can get the expected result?

Regards,
Edward WIJAYA

Comment on Self-Looping over hash - how to remove duplicate Select or Download Code

Replies are listed 'Best First'.
Re: Self-Looping over hash - how to remove duplicate by tachyon (Chancellor) on Oct 05, 2004 at 08:07 UTC
Just check if the A-B exists before adding B-A. `my %hash2 =( '1' => 2, '3' => 5, '4' => 2, ); my %spair = (); for my $si (keys %hash2) { for my $sn (keys %hash2) { next if exists $spair{$sn."-".$si}; $spair{$si."-".$sn} = ($si == $sn)? $hash2{$si} : $hash2{$si}* +$hash2{$sn}; } } use Data::Dumper; print Dumper \%spair; __DATA__ $VAR1 = { '3-3' => 5, '1-1' => 2, '3-4' => '10', '1-3' => '10', '1-4' => '4', '4-4' => 2 };` [download] cheers tachyon	[reply] [d/l]
Re: Self-Looping over hash - how to remove duplicate by gothic_mallard (Pilgrim) on Oct 05, 2004 at 08:16 UTC
The problem is that, although keys like `1-4` and `4-1` technically mean the same thing in context, as far as Perl is concerned they're different strings, so they're different keys. What you need to do is ensure that your keys are generated consistently. Your current code doesn't actually work as written as you've cut a few bits out (like the declaration of `%hash2`), but try this: Replace the line: `$spair{$si."-".$sn} = $prod;` With: `if ( $si >= $sn ) { $spair{$si."-".$sn} = $prod; } else { $spair{$sn."-".$si} = $prod; }` [download] That ensures that the keys are always created in the same order (i.e. highest value first in this case) - which means that your keys should then map as you're expecting. Running the following complete code should give you want you want: `use strict; use warnings; use Data::Dumper; my %hash2 =( '1' => 2, '3' => 5, '4' => 2,); my %spair = (); my $prod; foreach my $si (keys %hash2) { foreach my $sn (keys %hash2) { if ($si == $sn) { $prod = $hash2{$si}; } else { $prod = $hash2{$si}$hash2{$sn}; } if ( $si >= $sn ) { $spair{$si."-".$sn} = $prod; } else { $spair{$sn."-".$si} = $prod; } } } print Dumper \%spair;` [download] Produces: `$VAR1 = { '3-1' => '10', '3-3' => 5, '1-1' => 2, '4-1' => '4', '4-3' => '10', '4-4' => 2 };` [download] As always TMTOWTDI and there's almost certainly a more elegant solution but this albeit off-the-cuff should do the job. Hope this helps. --- Jay* All code is untested unless otherwise stated.	[reply] [d/l] [select]
Re: Self-Looping over hash - how to remove duplicate by gaal (Parson) on Oct 05, 2004 at 08:07 UTC
You could have both iterations go over the hash in an ordered manner, and then quit the inner loop after processing the key against itself. This is not tested: `use strict; use warnings; #print Dumper \%hash2; my %spair = (); my $prod; foreach my $si (sort { $a <=> $b } keys %hash2) { INNER: foreach my $sn (sort { $a <=> $b } keys %hash2) { if ($si == $sn) { $prod = $hash2{$si}; $spair{$si."-".$sn} = $prod; #: $hash2{$si}\n"; } else { $prod = $hash2{$si}*$hash2{$sn}; $spair{$si."-".$sn} = $prod; #: $hash2{$si}\n"; last INNER; } } } print Dumper \%spair;` [download]	[reply] [d/l]
Re^2: Self-Looping over hash - how to remove duplicate by monkfan (Curate) on Oct 05, 2004 at 08:20 UTC
Hi, thanks for the reply, but your code gives: `$VAR1 = { '1-1' => 2, '3-1' => 10, '4-1' => 4, '1-3' => 10 };` [download] Notice that 2-2, 3-3, and 4-4 is not included	[reply] [d/l]
Re: Self-Looping over hash - how to remove duplicate by atcroft (Abbot) on Oct 05, 2004 at 08:08 UTC
Probably the easiest way to do this would be to change the following line: `$spair{$si."-".$sn} = $prod;` [download] to something along the lines of: `if ((! defined($spair{$si."-".$sn})) and (! defined($spair{$sn."-".$si}))) { $spair{$si."-".$sn} = $prod; }` [download] This will only add the entry for $si."-".$sn if there is no entry currently in %spair for either $si."-".$sn nor $sn."-".$si. Hope that helps.	[reply] [d/l] [select]
Re: Self-Looping over hash - how to remove duplicate by Hena (Friar) on Oct 05, 2004 at 08:14 UTC
If you accept that you get always same order (eg, not 3-1, instead 1-3) then you can print smaller as first index and larger as second. `if ($si == $sn) { $prod = $hash2{$si}; } else { $prod = $hash2{$si}*$hash2{$sn}; } my $tmp = $si; $si = $si <= $sn ? $si : $sn; $sn = $si == $sn ? $tmp : $sn; $spair{"$si-$sn"} = $prod;` [download]	[reply] [d/l]
Re^2: Self-Looping over hash - how to remove duplicate by Jasper (Chaplain) on Oct 05, 2004 at 09:20 UTC
Easier way to swap two values: `($a,$b)=($b,$a);` [download] So we get: `($si,$sn) = sort {$a<=>$b} $si, $sn` [download] I believe this is straight out of 'Learning Perl'	[reply] [d/l] [select]
Re: Self-Looping over hash - how to remove duplicate by Anonymous Monk on Oct 05, 2004 at 12:25 UTC
`my %hash1; my %hash2 = (1 => 2, 3 => 5, 4 => 2,); my @keys = keys %hash2; for (my $i=0; $i<@keys; $i++) { $hash1{"$keys[$i]-$keys[$i]"}=$hash2{$keys[$i]}; for (my $j=$i+1; $j<@keys; $j++) { $hash1{"$keys[$i]-$keys[$j]"} = $hash2{$keys[$i]}*$hash2{$keys +[$j]}; } }` [download]	[reply] [d/l]