N-grams counter

ngram.rb

require 'pp'

WORDS = %w(correct horse battery staple)
LENGTH = 100
SOURCE = (0..LENGTH).map { WORDS.sample }

puts "Source: #{SOURCE.join(', ')}"

###

NGRAM_LENGTH = 2
TOP_LENGTH = 10

###

ngram = -> (i) { SOURCE[i, NGRAM_LENGTH].join(' ') }
ngrams = (0..(LENGTH - NGRAM_LENGTH + 1)).map(&ngram)

counters = ngrams.each_with_object(Hash.new(0)) { |item, hash| hash[item] += 1 }
pp Hash[counters.sort_by { |_key, value| value }.last(TOP_LENGTH).reverse]

result.txt

Source: battery, correct, correct, correct, staple, battery, correct, staple, staple, correct, correct, correct, correct, horse, staple, staple, correct, correct, staple, horse, staple, correct, horse, staple, staple, staple, battery, horse, correct, battery, horse, battery, correct, horse, correct, correct, horse, horse, correct, battery, battery, correct, correct, battery, horse, battery, correct, staple, battery, battery, battery, correct, battery, correct, battery, staple, horse, staple, correct, correct, correct, horse, battery, correct, horse, correct, correct, battery, battery, battery, correct, staple, correct, staple, battery, staple, staple, battery, battery, battery, staple, correct, horse, battery, horse, correct, battery, horse, battery, battery, horse, correct, horse, battery, horse, horse, horse, staple, correct, correct, staple

{"correct correct"=>12,
 "battery correct"=>9,
 "battery battery"=>8,
 "correct horse"=>8,
 "battery horse"=>7,
 "correct staple"=>7,
 "correct battery"=>7,
 "staple correct"=>7,
 "horse correct"=>6,
 "horse battery"=>6}