Timing tests concerning IPython.parallel performance

pushing_vs_passing

{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Timing for Remote Parallel Randomisation"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import random\n",
      "\n",
      "import networkx as nx\n",
      "\n",
      "from __future__ import (division, print_function)\n",
      "\n",
      "from IPython.parallel import Client"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Our goal is to create random subgraphs from a (not too small) network."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "num_nodes = int(1E03)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "graph = nx.gnp_random_graph(num_nodes, 0.01)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "graph.number_of_edges()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 4,
       "text": [
        "5078"
       ]
      }
     ],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "nodes = sorted(graph.nodes_iter())"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "sample_size = int(num_nodes * 0.6)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def example_all_passed(net, nodes, size):\n",
      "    group = random.sample(nodes, size)\n",
      "    subnet = net.subgraph(group)\n",
      "    return subnet.size()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 7
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Time for doing this task in series."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "random_num = int(1E02)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 8
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit sum(example_all_passed(graph, nodes, size) for size in [sample_size] * random_num)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "1 loops, best of 3: 212 ms per loop\n"
       ]
      }
     ],
     "prompt_number": 9
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "random_num = int(1E04)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 29
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit sum(example_all_passed(graph, nodes, size) for size in [sample_size] * random_num)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "1 loops, best of 3: 21.3 s per loop\n"
       ]
      }
     ],
     "prompt_number": 30
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Can we do better in parallel?"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "rc = Client()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 10
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "len(rc)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 11,
       "text": [
        "8"
       ]
      }
     ],
     "prompt_number": 11
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "small_view = rc[:2]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 12
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "full_view = rc[:]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 13
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Times to import using `execute`."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit small_view.execute(\"import random\", block=True)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "100 loops, best of 3: 3.46 ms per loop\n"
       ]
      }
     ],
     "prompt_number": 14
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit full_view.execute(\"import random\", block=True)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "100 loops, best of 3: 6.84 ms per loop\n"
       ]
      }
     ],
     "prompt_number": 15
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Times for running function with all parameters passed."
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Smaller number of tasks."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "random_num = int(1E02)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 16
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%timeit\n",
      "results = small_view.map(example_all_passed, [graph] * random_num, [nodes] * random_num, [sample_size] * random_num, block=False)\n",
      "sum(results)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "10 loops, best of 3: 151 ms per loop\n"
       ]
      }
     ],
     "prompt_number": 17
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%timeit\n",
      "results = full_view.map(example_all_passed, [graph] * random_num, [nodes] * random_num, [sample_size] * random_num, block=False)\n",
      "sum(results)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "10 loops, best of 3: 107 ms per loop\n"
       ]
      }
     ],
     "prompt_number": 18
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Larger number of tasks."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "random_num = int(1E04)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 19
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%timeit\n",
      "results = small_view.map(example_all_passed, [graph] * random_num, [nodes] * random_num, [sample_size] * random_num, block=False)\n",
      "sum(results)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "1 loops, best of 3: 11.5 s per loop\n"
       ]
      }
     ],
     "prompt_number": 20
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%timeit\n",
      "results = full_view.map(example_all_passed, [graph] * random_num, [nodes] * random_num, [sample_size] * random_num, block=False)\n",
      "sum(results)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "1 loops, best of 3: 5.58 s per loop\n"
       ]
      }
     ],
     "prompt_number": 21
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "One would think that pushing the large objects to the remote kernels once and then operating on them would be faster."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def example_globals(size):\n",
      "    group = random.sample(nodes, size)\n",
      "    subnet = net.subgraph(group)\n",
      "    return subnet.size()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 22
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Times for pushing something to different number of remote kernels."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit small_view.push({\"nodes\": nodes, \"net\": graph}, block=True)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "100 loops, best of 3: 18 ms per loop\n"
       ]
      }
     ],
     "prompt_number": 31
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%timeit full_view.push({\"nodes\": nodes, \"net\": graph}, block=True)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "10 loops, best of 3: 49 ms per loop\n"
       ]
      }
     ],
     "prompt_number": 32
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Smaller number of tasks."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "random_num = int(1E02)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 23
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%timeit\n",
      "small_view.push({\"nodes\": nodes, \"net\": graph}, block=True)\n",
      "results = small_view.map(example_globals, [sample_size] * random_num, block=False)\n",
      "sum(results)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "10 loops, best of 3: 154 ms per loop\n"
       ]
      }
     ],
     "prompt_number": 24
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%timeit\n",
      "full_view.push({\"nodes\": nodes, \"net\": graph}, block=True)\n",
      "results = full_view.map(example_globals, [sample_size] * random_num, block=False)\n",
      "sum(results)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "10 loops, best of 3: 118 ms per loop\n"
       ]
      }
     ],
     "prompt_number": 25
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Larger number of tasks."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "random_num = int(1E04)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 26
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%timeit\n",
      "small_view.push({\"nodes\": nodes, \"net\": graph}, block=True)\n",
      "results = small_view.map(example_globals, [sample_size] * random_num, block=False)\n",
      "sum(results)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "1 loops, best of 3: 12.1 s per loop\n"
       ]
      }
     ],
     "prompt_number": 27
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%%timeit\n",
      "full_view.push({\"nodes\": nodes, \"net\": graph}, block=True)\n",
      "results = full_view.map(example_globals, [sample_size] * random_num, block=False)\n",
      "sum(results)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "1 loops, best of 3: 5.73 s per loop\n"
       ]
      }
     ],
     "prompt_number": 28
    }
   ],
   "metadata": {}
  }
 ]
}