Toree, Jupyter and Spark

toree-example

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Toree on Mac\n",
    "\n",
    "## Installing\n",
    "\n",
    "This seems to be easy. First note that my `Jupyter` is installed using `Anaconda` on my user's spaces. As indicated on the project's [GitHub](https://github.com/apache/incubator-toree/blob/master/README.md) the installation is simple:\n",
    "\n",
    "1. `pip install toree`\n",
    "2. `jupyter toree install`\n",
    "\n",
    "My problem was, that since my Jupyter is local, the second step yielded a permission error. The way around it is:\n",
    "\n",
    "   jupyter toree install --spark_home=/Users/username/Applications/spark --user\n",
    "   \n",
    "Next, I could simply run `jupyter notebook` and start use `Scala/Spark` within `Jupyter`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Example --- Compute Pi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "val NUM_SAMPLES = 10000000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pi is roughly 3.1416408\n"
     ]
    }
   ],
   "source": [
    "val count = sc.parallelize(1 to NUM_SAMPLES).map{i =>\n",
    "  val x = Math.random()\n",
    "  val y = Math.random()\n",
    "  if (x*x + y*y < 1) 1 else 0\n",
    "}.reduce(_ + _)\n",
    "println(\"Pi is roughly \" + 4.0 * count / NUM_SAMPLES)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Toree",
   "language": "",
   "name": "toree"
  },
  "language_info": {
   "name": "scala"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}