Benchmarks.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:e45c80857ca77785d5d8f0d017a7793902a6597f374e0cd88e145456b35d93e6"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%pylab inline\n",
      "import fgzip\n",
      "import os\n",
      "import timeit\n",
      "\n",
      "import numpy as np\n",
      "import matplotlib.pyplot as plt\n",
      "import seaborn as sb"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Populating the interactive namespace from numpy and matplotlib\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING: pylab import has clobbered these variables: ['test']\n",
        "`%matplotlib` prevents importing * from pylab and numpy\n"
       ]
      }
     ],
     "prompt_number": 21
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "test_file = \"/Users/legaultmarc/reference/GRCh37/human_g1k_v37.fasta.gz\"\n",
      "\n",
      "chunk_sizes = {\n",
      "    \"1kb\": 1 * 1024,\n",
      "    \"32kb\": 32 * 1024,\n",
      "    \"64kb\": 64 * 1024,\n",
      "    \"128kb\": 128 * 1024,\n",
      "    \"256kb\": 256 * 1024,\n",
      "    \"512kb\": 512 * 1024,\n",
      "}\n",
      "\n",
      "data = []\n",
      "for k, chunk_size in chunk_sizes.iteritems():\n",
      "    \n",
      "    def test():\n",
      "        i = 0\n",
      "        with fgzip.open(test_file, chunk_size) as f:\n",
      "            for line in f:\n",
      "                i += 1\n",
      "        assert i == 51696831\n",
      "    \n",
      "    t = timeit.timeit(\"test()\", \"from __main__ import test\", number=3)\n",
      "    print k, t\n",
      "    data.append((k, t))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "256kb 81.6056780815\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "1kb 553.92594409\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "512kb 79.5930261612\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "128kb 82.7271800041\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "32kb 94.2784428596\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "64kb 77.2360720634\n"
       ]
      }
     ],
     "prompt_number": 7
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print data"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "[[   1.          553.92594409]\n",
        " [  32.           94.27844286]\n",
        " [  64.           77.23607206]\n",
        " [ 128.           82.72718   ]\n",
        " [ 256.           81.60567808]\n",
        " [ 512.           79.59302616]]\n"
       ]
      }
     ],
     "prompt_number": 28
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "data = [(int(i.rstrip(\"kb\")), j) for (i, j) in data]\n",
      "data = np.array(sorted(data, key=lambda x: x[0]))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 19
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "fig = plt.figure(figsize=(15, 5))\n",
      "plt.plot(data[:, 0], data[:, 1])\n",
      "plt.xlabel(\"Chunk Size (kb)\")\n",
      "plt.ylabel(\"Execution time (s)\")\n",
      "plt.ylim((60, 110))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 27,
       "text": [
        "(60, 110)"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAA4AAAAFICAYAAAABEJCnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJEmSlBIGQEmSJElKCQOgJEmSJKXE/w+w2XpnKhi13wAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x11754f750>"
       ]
      }
     ],
     "prompt_number": 27
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}