azymnis

import com.twitter.algebird.{Aggregator, Semigroup}
import com.twitter.scalding._

import scala.util.Random

/**
 * This job is a tutorial of sorts for scalding's Execution[T] abstraction.
 * It is a simple implementation of Lloyd's algorithm for k-means on 2D data.
 *
 * http://en.wikipedia.org/wiki/K-means_clustering

azymnis

import com.twitter.scalding._
import com.twitter.algebird.{ MinHasher, MinHasher32, MinHashSignature }

/**
 * Computes similar items (with a string itemId), based on approximate
 * Jaccard similarity, using LSH.
 *
 * Assumes an input data TSV file of the following format:
 *
 *    itemId   userId

azymnis

#!/usr/bin/env ruby
$LOAD_PATH << File.join(File.expand_path(File.dirname(File.symlink?(__FILE__) ? File.readlink(__FILE__) : __FILE__)), 'lib')

require 'fileutils'
require 'open-uri'
require 'thread'
require 'trollop'
require 'yaml'

USAGE = <<END