in reply to Re^2: Combining 3 files
in thread Combining 3 files

That is easy, the line is confusing :)

It is also confusing $dataset for a hashref

perltidy

my @results = ( $col[0], $col[3] ); for my $dataset (@data) { push @results, ( sort { my $diff_a = $col[2] - $a->[1]; $diff_a *= -1 if $diff_a < 0; my $diff_b = $col[2] - $b->[1]; $diff_b *= -1 if $diff_b < 0; $diff_a <=> $diff_b; } @{ $dataset->{ $col[0] } } )[0]->[2]; }
is like
my @results = ( $col[0], $col[3] ); for my $dataset (@data) { my ( $first ) = sort { my $diff_a = $col[2] - $a->[1]; $diff_a *= -1 if $diff_a < 0; my $diff_b = $col[2] - $b->[1]; $diff_b *= -1 if $diff_b < 0; $diff_a <=> $diff_b; } @{ $dataset->{ $col[0] } } push @results, $first->[2]; }
is like
my @results = ( $col[0], $col[3] ); for my $dataset (@data) { my @beef = @{ $dataset->{ $col[0] } }; @beef = sort { my $diff_a = $col[2] - $a->[1]; $diff_a *= -1 if $diff_a < 0; my $diff_b = $col[2] - $b->[1]; $diff_b *= -1 if $diff_b < 0; $diff_a <=> $diff_b; } @beef; push @results, $beef[0]->[2]; }

$a and $b are globals, they're how sort works

Is it still confusing?

Replies are listed 'Best First'.
Re^4: Combining 3 files
by garyboyd (Acolyte) on Jun 27, 2011 at 09:11 UTC

    Hi anonymous monk, yes I'm still finding this confusing. You mentioned it is confusing $dataset for a hashref. I've been playing around with this for a while and just keep getting syntax errors.

      Hi anonymous monk, yes I'm still finding this confusing.

      Well hold onto your pants :)

      You mentioned it is confusing $dataset for a hashref.

      Look at the @beef equivalent example, it is this line that dies with Can't use an undefined value as an ARRAY reference

      my @beef = @{ $dataset->{ $col[0] } };
      This line is really
      my $key = $col[0] ; my $meat = $dataset->{ $key }; my @beef = @{ $meat };
      I misspoke when I said it is confusing $dataset for a hashref , in actuality, $dataset is a hashref.

      The problem is, $key is not in $dataset, so $meat is undef

      Since meat is undef, trying to treat $meat as an array by de-referencing it, triggers a warning if you have warnings on, and triggers an error, if you have strict on.

      ## no warnings or errors $ perl -e "print @{ undef() }; print 6" 6 ## a warning is issued, but 6 still gets printed $ perl -we "print @{ undef() } Use of uninitialized value in array dereference at -e line 1. 6 ## with strict the warning is fatal, 6 not printed cause program died $ perl -Mstrict -we "print @{ undef() } Can't use an undefined value as an ARRAY reference at -e line 1.
      See References quick reference

      To avoid this error, you might add

      next unless $meat;
      You don't want to disable strict :)

      Lets take apart a simpler example, one that builds @data

        Thanks for the detailed post. I'm still having problems with the "Use of uninitialized value in array element" error. I adjusted the code, however I'm not sure where I need to insert the equivalent of the

        next unless $meat;

        I can get it to work by commenting out use strict;

        ##!/usr/bin/perl #22/06/2011 use strict; use warnings; use File::Slurp; use Data::Dumper; my @data; my @col; my $dataset; my @fields; my %out; my $Hashref; my $fileCount; my @out; open INFILE, "<Primer-Rev1" or die $!; open my $outfh, '>', "outputfile.txt" or die $!; for my $nr (1..2) { for my $line (read_file('Primer-For'.$nr)) { my @col = split(/\t/,$line); push @{$data[$nr - 1]->{shift(@col)}},\@col; } } while (<INFILE>){ @col = split(/\t+/, $_); chomp (@col); my ($header, $length, $tm, $sequence) = @col[0..3]; # expecting file3 line in @col my @results = ( $col[1], $col[2] ); #print Dumper (\@results) for my $dataset (@data) { my @out = push @{ $data[ $fileCount ]->{ shift @col } }, \@col; #print Dumper (\@data); #} @out = sort { my $diff_a = $col[2] - $a->[1]; $diff_a *= -1 if $diff_a < 0; my $diff_b = $col[2] - $b->[1]; $diff_b *= -1 if $diff_b < 0; $diff_a <=> $diff_b; } @out; print Dumper (\@out); push @results, $out[0]->[2]; } #foreach (@results){ # print $_."\n"} }