エントロピーを計算するツール

entropy

#!/usr/bin/env ruby

# Usage
# ./entropy file
# echo "aaaaaaaabbbbbbbbbcc" | ./entropy

require 'optparse'

options = {}

OptionParser.new do |opts|
  opts.banner = "Usage: entropy [options] file"

  opts.on("-v", "--[no-]verbose", "Run verbosely") do |v|
    options[:verbose] = v
  end
end.parse!

def entropy(str, options)
  count = str.size
  list={}
  list.default=0
  str.split("").each {|c| list[c]+=1}
  sum = 0

  if options[:verbose]
    puts "記号: 確率 : 値"
  end

  list.each do |key, val|
    i = (val.to_f / count.to_f)
    tmp_e = -(i) * Math.log(i, 2)
    if options[:verbose]
      puts "#{key}: #{val} / #{count}: #{tmp_e}"
    end
    sum += tmp_e
  end
  sum
end

if $stdin.tty?
  file = ARGV[0]
  str  = File.read(file)
else
  str = ARGF.read.chomp
end

e = entropy(str, options)

if options[:verbose]
  puts ""
  puts "サイズ エントロピー 下限値"
  puts "#{str.bytesize} #{e} #{((str.bytesize * e) / 8).to_i}"
else
  puts e
end