jupyter_magic.ipynb

jupyter_magic.ipynb

{
  "cells": [
    {
      "metadata": {
        "trusted": true,
        "collapsed": true
      },
      "cell_type": "code",
      "source": "%load_ext cython",
      "execution_count": 1,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%%cython\n\ncpdef int fact1(int n):\n    \"\"\"Doc\n    \n    Parameters\n    ----------\n    n : int\n    \n    \"\"\"\n    if (n <= 1):\n        return 1\n    \n    else:\n        return n * fact1(n - 1)\n    \ndef fact2(n):\n    if (n <= 1):\n        return 1\n    else:\n        return n * fact2(n - 1)\n    \ndef fact3(int n):\n    if (n <= 1):\n        return 1\n    else:\n        return n * fact3(n - 1)\n    \ncpdef int fact4(int n):\n    if n <= 1:\n        return 1\n    else:\n        return n * fact4(n - 1)",
      "execution_count": 2,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 1000 fact1(5)",
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": "51.1 ns ± 1.57 ns per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 1000 fact2(5)",
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": "338 ns ± 50.8 ns per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 1000 fact3(5)",
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": "235 ns ± 3.75 ns per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 1000 fact4(5)",
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": "57.1 ns ± 0.471 ns per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "collapsed": true
      },
      "cell_type": "code",
      "source": "def fact_python(n):\n    if (n <= 1):\n        return 1\n    else:\n        return n * fact_python(n - 1)",
      "execution_count": 7,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 1000 fact_python(5)",
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": "895 ns ± 71.2 ns per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "collapsed": true
      },
      "cell_type": "code",
      "source": "%quickref",
      "execution_count": 9,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%%writefile hello.py\nprint(\"Hello World\")",
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": "Overwriting hello.py\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "# %load hello.py\nprint(\"Hello World\")",
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": "Hello World\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%save save_test.py 2",
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": "File `save_test.py` exists. Overwrite (y/[N])?  y\nThe following commands were written to file `save_test.py`:\nget_ipython().run_cell_magic('cython', '', '\\ncpdef int fact1(int n):\\n    \"\"\"Doc\\n    \\n    Parameters\\n    ----------\\n    n : int\\n    \\n    \"\"\"\\n    if (n <= 1):\\n        return 1\\n    \\n    else:\\n        return n * fact1(n - 1)\\n    \\ndef fact2(n):\\n    if (n <= 1):\\n        return 1\\n    else:\\n        return n * fact2(n - 1)\\n    \\ndef fact3(int n):\\n    if (n <= 1):\\n        return 1\\n    else:\\n        return n * fact3(n - 1)\\n    \\ncpdef int fact4(int n):\\n    if n <= 1:\\n        return 1\\n    else:\\n        return n * fact4(n - 1)')\n"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "# %load save_test.py\nget_ipython().run_cell_magic('cython', '', '\\ncpdef int fact1(int n):\\n    \"\"\"Doc\\n    \\n    Parameters\\n    ----------\\n    n : int\\n    \\n    \"\"\"\\n    if (n <= 1):\\n        return 1\\n    \\n    else:\\n        return n * fact1(n - 1)\\n    \\ndef fact2(n):\\n    if (n <= 1):\\n        return 1\\n    else:\\n        return n * fact2(n - 1)\\n    \\ndef fact3(int n):\\n    if (n <= 1):\\n        return 1\\n    else:\\n        return n * fact3(n - 1)\\n    \\ncpdef int fact4(int n):\\n    if n <= 1:\\n        return 1\\n    else:\\n        return n * fact4(n - 1)')\n",
      "execution_count": 13,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%save save_test.py 7",
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": "File `save_test.py` exists. Overwrite (y/[N])?  y\nThe following commands were written to file `save_test.py`:\ndef fact_python(n):\n    if (n <= 1):\n        return 1\n    else:\n        return n * fact_python(n - 1)\n"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "collapsed": true
      },
      "cell_type": "code",
      "source": "# %load save_test.py\ndef fact_python(n):\n    if (n <= 1):\n        return 1\n    else:\n        return n * fact_python(n - 1)",
      "execution_count": 15,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true,
        "collapsed": true
      },
      "cell_type": "code",
      "source": "import numpy as np",
      "execution_count": 16,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true,
        "collapsed": true
      },
      "cell_type": "code",
      "source": "%%cython\nimport numpy as np\ncimport numpy as np\n\nDTYPE = np.float\nctypedef np.float_t DTYPE_t\n\ndef sum_by_row(np.ndarray[DTYPE_t, ndim=2] f):\n    cdef int n_row = f.shape[0]\n    cdef int n_col = f.shape[1]\n    \n    cdef np.ndarray[DTYPE_t, ndim=1] result = np.zeros([n_col], dtype=DTYPE)\n    \n    cdef int i\n    cdef int j\n    \n    for i in range(n_row):\n        for c in range(n_col):\n            result[c] += f[i, c]\n\n    return result",
      "execution_count": 17,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%%cython\nimport numpy as np\ncimport numpy as np\n\nDTYPE = np.float\nctypedef np.float_t DTYPE_t\n\ndef sum_by_row2(np.ndarray[DTYPE_t, ndim=2] f):\n    cdef int n_row = f.shape[0]\n    cdef int n_col = f.shape[1]\n    \n    cdef np.ndarray[DTYPE_t, ndim=1] result = np.zeros([n_col], dtype=DTYPE)\n    \n    cdef int i\n    cdef int j\n    \n    for i in range(n_row):\n        for c in range(n_col):\n            result[c] += f[i, c]\n\n    return result",
      "execution_count": 18,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%%cython\nimport numpy as np\ncimport numpy as np\n\ndef sum_by_row3(np.ndarray f):\n    cdef int n_row = f.shape[0]\n    cdef int n_col = f.shape[1]\n    \n    cdef np.ndarray result = np.zeros([n_col])\n    \n    cdef int i\n    cdef int j\n    \n    for i in range(n_row):\n        for c in range(n_col):\n            result[c] += f[i, c]\n\n    return result",
      "execution_count": 19,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "A = np.random.randn(32, 32)",
      "execution_count": 20,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 10000 sum_by_row(A)",
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": "3.06 µs ± 180 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 10000 A.sum(0)",
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": "3.41 µs ± 190 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 10000 sum_by_row2(A)",
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "text": "3.53 µs ± 222 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 10000 sum_by_row3(A)",
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "text": "220 µs ± 8.94 µs per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "collapsed": true
      },
      "cell_type": "code",
      "source": "def sum_by_row_python(f):\n    n_row = f.shape[0]\n    n_col = f.shape[1]\n    \n    result = np.zeros([n_col])\n    \n        \n    for i in range(n_row):\n        for c in range(n_col):\n            result[c] += f[i, c]\n\n    return result",
      "execution_count": 25,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "%timeit -n 1000 sum_by_row_python(A)",
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": "355 µs ± 14.8 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
          "name": "stdout"
        }
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3",
      "language": "python"
    },
    "language_info": {
      "name": "python",
      "version": "3.6.1",
      "mimetype": "text/x-python",
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "pygments_lexer": "ipython3",
      "nbconvert_exporter": "python",
      "file_extension": ".py"
    },
    "gist": {
      "id": "0993e87adb7d0332a843c32f35dd8145",
      "data": {
        "description": "jupyter_magic.ipynb",
        "public": true
      }
    },
    "_draft": {
      "nbviewer_url": "https://gist.github.com/0993e87adb7d0332a843c32f35dd8145"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 2
}