Yes I caught that. My data does have whitespace in the organism column. Thanks. I have been trying to work through the code and have another question. How can I add column 2(text) of the original data to the output and keep it referenced to organism?

Yes you are correct Chris. I was having difficulty keeping column 2(accession) referenced to column 3(organism). I understand the code as the values are pushed into the array but I did not understand how to keep the above referenced to each other. I assume a hash but I was not sure how to implement it along with the other hashes. I was also having a problem with the organism names being too long. I came up with a solution for both. I would rather learn than be given a solution so let me know if this is acceptable practice. Again, thanks for your help.

use strict;
use warnings;
use Acme::Tools;
use Text::Table;

my %data;
my %freq;
my $ref_filelist = $ARGV[0];
open(BLASTFILE, $ref_filelist )
   or die "Could not open Reference filelist...($!)";
   
while (<BLASTFILE>) {
    chomp;
    
    my ($accession,$organism, @vals) = (split /\t/)[1..5];
    my $organismcut = substr( $organism, 0,75 );
    my $tot =$accession . $organismcut;
    #print "$tot\n";
    
    $freq{ $tot }++;
    
    my $col = 4;
    for my $val (@vals) {
        push @{ $data{$tot} {$col} }, $val;
        $col++;
    }
}

my @headers =  qw/ Organism Freq Median_Eval Med_Contig_Length Med_Map
+ped_Length /;

my $tb = Text::Table->new( map {title => $_}, @headers);

for my $test (sort {$freq{ $b } <=> $freq{ $a }} keys %freq) {
    my @row = ($test, $freq{ $test });
    for my $col (sort keys %{ $data{$test} }) {
        push @row, median(@{ $data{$test}{$col} });
        
    } 
    $tb->load( [@row] );
    #print "@row\n"; 
    
}
print $tb;
[download]

I don't think Text::Table is the right tool for this application. It is only really useful for about a page worth of reading. Your first column will be 75 plus ? spaces alone. How do you plan to view the data? A couple of ideas that occured to me would be to create a comma separated values file or use Perl6::Form (like I did in Re: Formatting text, eg long lines). (You could wrap that long first column so it wouldn't run across the page).

With a comma separated values file format, you could open the files in Excel, (if you are on a Windows machine), and, if there is a large dataset, (more than a couple of pages), you can freeze the headers in Excel while still scrolling your data.

If your results will have many lines of results, with Perl6::Form, you could arrange headers to print after so many lines.

I say that about Text::Table because I'm guessing your results will be more than a couple of pages. If not, Text::Table would be OK.

If there are alot of rows to print, you might want to print your header for every (50?) lines or so to aid your readers. A way to do it below:

Instead of print $tb at the end of your script, the following would repeat the header every 50 lines.

my $rows = $tb->body_height(); 
$pagelines = 50;

for my $i (0 .. $rows-1) {
    print $tb->title() if $i % $pagelines == 0;
    print $tb->body($i);
    print "\n" if $i % $pagelines == $pagelines-1;
}
[download]

Chris

Update: set rows to the number of rows in table.
I was getting the number of items from keys %freq.