What have you tried thus far? This feels like homework.

Nevertheless, here's a recent thread where a monk was trying to accomplish a similar thing: Hash of Arrays

[tauser02@gfmwsp02lds bin]$ cat x
use strict;
use warnings;

my %data;
while (<DATA>) {
    my ($cluster, $unit) = split /\s+/;
    push(@{$data{$cluster}}, $unit);
}
for my $cluster (sort keys %data) {
    print $cluster, ' ', join(' ', @{$data{$cluster}}), "\n";
}

__DATA__
Cluster1 unit1
Cluster1 unit2
Cluster1 unit3
Cluster1 unit4
Cluster1 unit5
Cluster2 unit6
Cluster2 unit7
Cluster2 unit8
Cluster2 unit9
Cluster2 unit10 
[tauser02@gfmwsp02lds bin]$ perl x
Cluster1 unit1 unit2 unit3 unit4 unit5
Cluster2 unit6 unit7 unit8 unit9 unit10
[download]

As you can see from mifflin’s post, Perl actually makes this sort of thing quite easy.

You clearly want the computer to build a “list of” the items (second column of input data) that appear under “keys” (first column).   Perl provides array (and list) data structures that allow you to easily manage “a list of zero-or-more things” and a hash data structure that lets you file things under arbitrary “keys.”

The magic that really brings all of this together is the idea of references.   So, a hash can contain (for each key that it stores...) “a reference to” an array which contains all of the items belonging to that key.

One of the most important features of a language like Perl is that it gives you a well-chosen set of memory data structures, plus a “garbage-collector based” memory management system that lets you string together these complicated data structures (almost...) without worrying about it.

perl -ane "push @{$x{shift @F}},@F}{print qq|$_: @{$x{$_}}\n| for sort
+ keys %x" INPUTFILE.txt
[download]