Best 20 after sort

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

Replies are listed 'Best First'.
Re: Best 20 after sort by Taulmarill (Deacon) on Oct 20, 2008 at 14:24 UTC
Yes, like this: `foreach my $g( (sort {$Averages{$a} <=> $Averages{$b}} keys %Averages) +[0..19] ){ print $g,"\t",$Averages{$g},"\n"; }` [download] The parenthesis surrounding the sort turn the return value into a list which can be sliced like an array.	[reply] [d/l]
Re: Best 20 after sort by Limbic~Region (Chancellor) on Oct 20, 2008 at 14:23 UTC
seaver, You could easily use last to get out of your for loop after 20 items. You could also use Heap::Simple. I have been meaning to write a follow up to Better mousetrap (getting top N values from list X) for a long time explaining how mere mortals can do heap sorting - maybe someday. Cheers - L~R	[reply]
Re^2: Best 20 after sort by seaver (Pilgrim) on Oct 20, 2008 at 16:03 UTC
L~R I was using last, but I knew there'd be a more concise way of doing it. S	[reply]
Re: Best 20 after sort by salva (Canon) on Oct 21, 2008 at 07:51 UTC
You can use Sort::Key::Top, in particular if you want to extract the top 20 for a large array: `use Sort::Key::Top qw(nkeytopsort); my @top20 = nkeytopsort { $Averages{$_} } 20 => keys %Averages;` [download]	[reply] [d/l]
Re: Best 20 after sort by GrandFather (Saint) on Oct 20, 2008 at 20:08 UTC
`say join ("\n", map {"$_,\t$Averages{$_}"} (sort {$Averages{$a} <=> $A +verages{$b}} keys %Averages)[0..19]);` [download] Perl reduces RSI - it saves typing	[reply] [d/l]
Re: Best 20 after sort by johngg (Canon) on Oct 20, 2008 at 18:15 UTC
In the spirit of TIMTOWTDI, a different approach using sort, grep and map in a do block. `use strict; use warnings; my %Ave = map { $_ => rand } q{a} .. q{z}; print do { my $count = 0; map { qq{$_\t$Ave{ $_ }\n} } grep { $count ++ < 20 } sort { $Ave{ $a } <=> $Ave{ $b } } keys %Ave; };` [download] I hope this is of interest. Cheers, JohnGG	[reply] [d/l]