It's a lot more timely to worry about efficiency after you have some working code.

I've seen a couple of these in the past week or so, where the poster has no code to start off with, as if no attempt has been made to answer the question for themselves yet. It sort of feels like we're being asked to complete homework assignments!

Sorry guys... I am new to the forum, will post my code alongwith any questions i have in the future...To clarify, I just wanted an idea to start off my code with, not any ready-made code to copy-paste into any hypothetical homework..

#!/usr/bin/perl
use strict;
use warnings;

# ----------------------------------------------
# make dummy data
# ----------------------------------------------
my @data;
foreach (1..500) {
    push @data,int(rand(50)+1);
}

# ----------------------------------------------
# analyze data
# ----------------------------------------------

# what are the limits of my data, and define hash 
# (all in one!)
my $min;
my $max;
my %hist;
my $most=0;
foreach my $datum (@data) {
    $min = $datum if ! defined $min || $min > $datum;    
    $max = $datum if ! defined $max || $max < $datum;    
    $hist{$datum}++;
    $most = $hist{$datum} if $most < $hist{$datum};
}

# ----------------------------------------------
# print the histogram
# ----------------------------------------------
print "\n\n";
no warnings 'uninitialized';            # UPDATED
foreach my $m (0 .. $most-1) {
    printf "%05d   ",$most-$m;
    foreach my $i ($min .. $max) {      # UPDATED
        if ($hist{$i} >= ($most-$m)) {
            print "x";
        }
        else {
            print " ";
        }
    }
    print "\n";
}
print "     $min"," "x($max-$min),"$max\n";
[download]

UPDATE

Note that there was no attempt to scale the output. I leave that up to you.

#!/usr/bin/env clj
(ns histosample
    (:use [incanter core charts]
          [clojure.contrib.duck-streams :only (read-lines)]))
(def data 
    (if *command-line-args*
        (map #(Integer/parseInt %) (read-lines (first *command-line-ar
+gs*)))
        (take 500
            (iterate (fn [_] (int (rand 100))) 0))))
(let
    [max (reduce max data) min (reduce min data) nbins (Math/abs (- ma
+x min))]
    (view (histogram data :nbins nbins :x-label "Value")))
[download]