duplicate_columns.rb

require 'benchmark'
require 'matrix'

v1 = [1, 1, 1, 1]
v2 = [2, 3, 2, 3]
v3 = [1, 1, 1, 1 ]

def sol_1(a,b,c)
  sum = 0
  zip = a.zip(b, c)
  zip.group_by { |e| e}
     .select { |_, value| value.size > 1 }
     .each_value { |value| sum += (value.size - 1) }
  return sum
end


def sol_2(a,b,c)
  zip = a.zip(b, c)
  hash = Hash.new(0)
  zip.each { |e| hash.store(e, hash[e]+1) }
  hash.each{|e, _| hash[e] -= 1}
  return hash.sum {|e, _| hash[e] }
end

def sol_3(matrix)
  Matrix.
    columns(matrix).
    to_a.
    each_with_object({}) { |e, a|
      digest = e.hash
      a[digest] = a[digest].nil? ? 1 : a[digest] + 1
    }.sum { |_, v| v > 1 ? 1 : 0 }
end

n=1_000
Benchmark.bmbm do |x|
  x.report("sol_1"){n.times{sol_1(v1, v2, v3)} }
  x.report("sol_2"){n.times{sol_2(v1, v2, v3)} }
  x.report("sol_3"){n.times{sol_3([v1, v2, v3])} }
end