http://www.perlmonks.org?node_id=11135135


in reply to efficient perl code to count, rank

Since you didn't provide a small test data set I had to guess about it and then fake it (I may have guessed wrong).

You were doing the ranking sort for each column for each line when it only had to be done once per column.

I am reading through the file twice so that the whole file does not have to be stored in memory. Because of that I had to change your file handle names.

I had to change your I/O because a) I'm not on windows, and b) debug. I think I labeled all those changes with FIXME.

Of course, comment out the Data::Dump before running on a large file. You'll find Data::Dump is extremely slow on huge data structures.

As I've said to others here, at least my code does not fail any of your provided test cases :)

#!/usr/bin/perl use strict; # https://perlmonks.org/?node_id=11135101 use warnings; use List::Util qw( uniq ); my $x="Room_reserve.csv"; my $in = "D:\\package properties\\${x}.csv"; $in = "D:\\package properties\\${x}.csv"; # ????????? duplicate line?? +? my $out = "D:\\package properties\\output\\${x}_output.csv"; print my $testdata = <<END; # FIXME x,x,x,x,x,x,x,one,two,three x,x,x,x,x,x,x,foo,bar,baz x,x,x,x,x,x,x,foo2,bar7,baz3 x,x,x,x,x,x,x,foo2,bar7,baz3 x,x,x,x,x,x,x,foo,bax,baz x,x,x,x,x,x,x,foo,bar,baz x,x,x,x,x,x,x,foo,bar,baz END $in = \$testdata; #FIXME open(my $infh, '<', $in) or die "Could not open file '$in' $!"; my @columns; my $lines = 0; <$infh>; # skip first line while( <$infh> ) { my @rows = split /[,\n]/; $columns[$_]{$rows[$_]}++ for 0..$#rows; $lines++; } my @ranksbycolumn; for ( @columns ) { my @dat_val = sort { $b <=> $a } uniq values %$_; my %ranks; @ranks{ @dat_val } = 1 .. @dat_val; push @ranksbycolumn, \%ranks; } use Data::Dump 'dd'; dd 'columns', \@columns, 'ranksbycolumn', \@ranks +bycolumn; # FIXME my $first = 1; #output #open(my $outfh, '>', $out) or die "Could not open file '$out' $!"; #F +IXME open(my $outfh, ">&STDOUT") or die "Could not open file STDOUT$!"; #FI +XME seek $infh, 0, 0 or die "Could not seek file '$in' $!"; while( <$infh> ) { my @rows = split /[,\n]/; foreach my $i (0..$#rows) { if ($i > 6) { if ( $first == 1 ) # for modifying name { my $line = join( ",", "Rank_$rows[$i]", "Percent_$rows[$i]", "Count_$rows[$i]", $rows[$i]); print $outfh "$line,"; if ( $i == $#rows ) { $first = 0; } } else { my $cnt = $columns[$i]{$rows[$i]}; my $rank = $ranksbycolumn[$i]{$cnt}; # my $ave = ($cnt / 14000000) * 100; # FIXME my $ave = int +($cnt / $lines) * 100; # FIXME my $line = join( ",", $rank, $ave, $cnt, $rows[$i]); print $outfh "$line,"; } } else { print $outfh "$rows[$i],"; } } print $outfh "\n"; } close $outfh;

Outputs:

x,x,x,x,x,x,x,one,two,three x,x,x,x,x,x,x,foo,bar,baz x,x,x,x,x,x,x,foo2,bar7,baz3 x,x,x,x,x,x,x,foo2,bar7,baz3 x,x,x,x,x,x,x,foo,bax,baz x,x,x,x,x,x,x,foo,bar,baz x,x,x,x,x,x,x,foo,bar,baz ( "columns", [ { x => 6 }, { x => 6 }, { x => 6 }, { x => 6 }, { x => 6 }, { x => 6 }, { x => 6 }, { foo => 4, foo2 => 2 }, { bar => 3, bar7 => 2, bax => 1 }, { baz => 4, baz3 => 2 }, ], "ranksbycolumn", [ { 6 => 1 }, { 6 => 1 }, { 6 => 1 }, { 6 => 1 }, { 6 => 1 }, { 6 => 1 }, { 6 => 1 }, { 2 => 2, 4 => 1 }, { 1 => 3, 2 => 2, 3 => 1 }, { 2 => 2, 4 => 1 }, ], ) x,x,x,x,x,x,x,Rank_one,Percent_one,Count_one,one,Rank_two,Percent_two, +Count_two,two,Rank_three,Percent_three,Count_three,three, x,x,x,x,x,x,x,1,66,4,foo,1,50,3,bar,1,66,4,baz, x,x,x,x,x,x,x,2,33,2,foo2,2,33,2,bar7,2,33,2,baz3, x,x,x,x,x,x,x,2,33,2,foo2,2,33,2,bar7,2,33,2,baz3, x,x,x,x,x,x,x,1,66,4,foo,3,16,1,bax,1,66,4,baz, x,x,x,x,x,x,x,1,66,4,foo,1,50,3,bar,1,66,4,baz, x,x,x,x,x,x,x,1,66,4,foo,1,50,3,bar,1,66,4,baz,

Replies are listed 'Best First'.
Re^2: efficient perl code to count, rank
by LanX (Sage) on Jul 18, 2021 at 19:28 UTC
    Thanks for showing example in and output!

    So the OP wants to add a rank and percentage to every field of his 1100+ columns in 14m rows and blow up his 63GB file by factor 3?

    Interesting ... :)

    Cheers Rolf
    (addicted to the Perl Programming Language :)
    Wikisyntax for the Monastery

      You're Welcome!

      Also count, for a factor of four. Imagine the fun of looking through a 14000000 row by 4400+ column spreadsheet :)

        > Imagine the fun of looking through a 14000000 row by 4400+ column spreadsheet :)

        • "Our language barely has words for the fun that has been wrought here.

          -Angela Merkel (almost)

        ;)

        Cheers Rolf
        (addicted to the Perl Programming Language :)
        Wikisyntax for the Monastery