Initial pass at calculating the median for CV measures

gistfile1.rb

require 'moped'

db = Moped::Session.new(["localhost:27017"])
db.use('foo')

wrapped_result = db.command(:aggregate => 'patient_cache', :pipeline => [
  {'$unwind' => "$values"},
  {'$group' => {'_id' => '$values', 'count' => {"$sum" => 1}}}
  ])

puts wrapped_result

result = {}

wrapped_result['result'].each do |freq_count_pair|
  result[freq_count_pair['_id']] = freq_count_pair['count']
end

set_size = result.values.reduce(0) {|memo, freq| memo + freq}
sorted_values = result.keys.sort
median_position = (set_size / 2) + 1

current_position = 0
median = 0
sorted_values.each do |value|
  current_position += result[value]
  if current_position >= median_position
    median = value
    break
  end
end

puts "Median: #{median}"
puts "Set size: #{set_size}"