benchmark.ipynb

{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import random\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "N_POINTS = 1000"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def generate_points_as_dicts(count):\n",
      "    random.seed(42)\n",
      "    points = []\n",
      "    for __ in range(count):\n",
      "        x = random.random()\n",
      "        y = random.random()\n",
      "        points.append({'x' : x, 'y': y})\n",
      "    return points\n",
      "    "
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "class Point(object):\n",
      "    x = None\n",
      "    y = None\n",
      "    def __init__(self, x, y):\n",
      "        self.x, self.y = x, y\n",
      "        \n",
      "def generate_points_as_classes(count):\n",
      "    random.seed(42)\n",
      "    points = []\n",
      "    for __ in range(count):\n",
      "        x = random.random()\n",
      "        y = random.random()\n",
      "        points.append(Point(x,y))\n",
      "    return points"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "class PointWithSlots(object):\n",
      "    __slots__ = ['x', 'y']\n",
      "    def __init__(self, x, y):\n",
      "        self.x, self.y = x, y\n",
      "\n",
      "def generate_points_as_classes_with_slots(count):\n",
      "    random.seed(42)\n",
      "    points = []\n",
      "    for __ in range(count):\n",
      "        x = random.random()\n",
      "        y = random.random()\n",
      "        points.append(PointWithSlots(x,y))\n",
      "    return points"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from collections import namedtuple\n",
      "PointTuple = namedtuple('PointTuple', ['x', 'y'])\n",
      "\n",
      "def generate_points_as_namedtuples(count):\n",
      "    random.seed(42)\n",
      "    points = []\n",
      "    for __ in range(count):\n",
      "        x = random.random()\n",
      "        y = random.random()\n",
      "        points.append(PointTuple(x,y))\n",
      "    return points"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit generate_points_as_dicts(N_POINTS)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit generate_points_as_classes(N_POINTS)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit generate_points_as_classes_with_slots(N_POINTS)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit generate_points_as_namedtuples(N_POINTS)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def sum_all_points_as_dict(points):\n",
      "    sum_x, sum_y = 0, 0\n",
      "    for point in points:\n",
      "        sum_x += point['x']\n",
      "        sum_y += point['y']\n",
      "    return sum_x, sum_y"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def sum_all_points_as_class(points):\n",
      "    sum_x, sum_y = 0, 0\n",
      "    for point in points:\n",
      "        sum_x += point.x\n",
      "        sum_y += point.y\n",
      "    return sum_x, sum_y"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def sum_all_points_as_namedtuple(points):\n",
      "    sum_x, sum_y = 0, 0\n",
      "    for point in points:\n",
      "        sum_x += point.x\n",
      "        sum_y += point.y\n",
      "    return sum_x, sum_y"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print \"Dicts:\"\n",
      "points_dict = generate_points_as_dicts(N_POINTS)\n",
      "%timeit sum_all_points_as_dict(points_dict)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print \"Classes: \"\n",
      "points_class = generate_points_as_classes(N_POINTS)\n",
      "%timeit sum_all_points_as_class(points_class)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print \"Classes with __slots__: \"\n",
      "points_class_with_slots = generate_points_as_classes_with_slots(N_POINTS)\n",
      "%timeit sum_all_points_as_class(points_class_with_slots)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print \"Namedtuples: \"\n",
      "points_namedtuple = generate_points_as_namedtuples(N_POINTS)\n",
      "%timeit sum_all_points_as_namedtuple(points_namedtuple)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}