It's the full (Google's) pagerank algorithm as a procedure ... want to make it into a general class

pagerank_jruby.rb

# script doesn't include classes, yet

require 'MDArray'





size = 4
matrix_as_array = [ 0 , 2.0/5 , 0 , 0 , 4.0/15 , 0, 0, 2.0/5, 4.0/15, 0. 4/5.0 , 2/5.0, 4.0/15, 2/5.0, 0 , 0] ]
beta = 0.8
epsilon = 0.000000001
delta = 1.0
it = 0

    # The total number — N — of webpages and nodes
    n = size.to_f

    # The initial rank score vector.
    # Each cell is set to (1/N).
    pagerank = MDMatrix.double([n,1]).fill(1/n)

    # The input graph, M.
    # M must have the cells filled
    # with (1/d_i), representing the link weights.
    # The "matrix_as_array" is a flattened Array,
    # like [1,2,3  ,  4,5,6   ,7,8,9] for a 3x3 matrix.
    matrix = MDMatrix.double([n,n],matrix_as_array)



    # until delta <= epsilon do

    6.times do

      pagerank_old = pagerank


      # (1-beta/N) vector.
      teleport = (1.0 - beta)

      constant = MDMatrix.double([n,1]).fill(teleport/n)

      # Actual page rank calculation.
      x = matrix * pagerank * beta + constant

      # # Re-insert leaked PageRanks.

      # Absolute sum of the pagerank scores.
      s = x.norm1

      s1 = 1.0-s

      # Vector correcting dead ends and spider traps.
      anti_leak = MDMatrix.double([n,1]).fill(s1/n)

      zz = x + anti_leak

      pagerank = zz
      pagerank.print
      # # MDMatrix does not allow subtraction by minus sign.
      # # I use a plus b times negative one.
      # # #norm1 makes the sum of the column vector's absolute values.
      neg_pager = pagerank * -1.0

      delta = pagerank_old + neg_pager

      # Absolute sum of the scores elements in the delta vector.
      delta = delta.norm1

      it = it + 1
      puts "Power Iteration\t #{it}:\t #{delta.round(5)}"

    end

pagerank.print