Re: Retrieving data with largest count

my %largest;

while (<>) {
   my ($item, $group_id, $count) = split;

   next if $largest{$group_id}
        && $largest{$group_id}[2] >= $count;

   $largest{$group_id} = [ $item, $group_id, $count, $_ ];
}

for (values(%largest)) {
   print($_->[-1]);
}
[download]

Keeps first on ties.

Update: Added note at bottom.

Comment on Re: Retrieving data with largest count Download Code

Replies are listed 'Best First'.
Re^2: Retrieving data with largest count by Limbic~Region (Chancellor) on Jun 02, 2009 at 15:25 UTC
ikegami, I know that it is unlikely in this situation that there would be negative or even nil counts, but this code does the wrong thing in the following situation: `ASDF 1 0 ASDF 1 -3` [download] Specifically, it first sets the high water mark to 0 and then to -3. (Originally, I didn't have the group the same - corrected) Cheers - L~R	[reply] [d/l]
Re^3: Retrieving data with largest count by ikegami (Patriarch) on Jun 02, 2009 at 15:43 UTC
No it doesn't??? In: `1ct9B 3 1 1f7uA 2 3 1gaxA 2 1 1gpmA 3 5 1ihoA 6 4 1mopA 6 3 1ileA 2 5 1iq0A 2 3 1vlhB 5 3 1jhdA 5 4 ASDF 1 0 ASDF 1 -3` [download] Out: `1ihoA 6 4 ASDF 1 0 1gpmA 3 5 1ileA 2 5 1jhdA 5 4` [download] `0 >= -3`, so `next` is executed, so the record for `-3` is ignored. The item with the highest value for group 1 (`ASDF 1 0`) is printed.	[reply] [d/l] [select]
Re^4: Retrieving data with largest count by Limbic~Region (Chancellor) on Jun 02, 2009 at 17:23 UTC
ikegami, ~~You replied to yourself and not me~~ I didn't see the response initially. I must be having a second Monday. I read the following code: `next if $largest{$group_id} && $largest{$group_id}[2] >= $count;` [download] as `next if $largest{$group_id}[2] && $largest{$group_id}[2] >= $count;` [download] Obviously not having a computer at home for a month and a trip to Vegas to boot has shrunk what little brain cells I have. Cheers - L~R	[reply] [d/l] [select]
Re^5: Retrieving data with largest count by ikegami (Patriarch) on Jun 02, 2009 at 17:29 UTC