{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This code snippet demonstrates the speed of numba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numba\n",
    "import numpy as np\n",
    "import scipy.spatial.distance as distance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "@numba.jit(nopython=True)\n",
    "def ngrams(string, n=3):\n",
    "    res = []\n",
    "    for i in range(len(string) - n):\n",
    "        res.append(string[i:i+n])\n",
    "    return res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "@numba.jit(nopython=True)\n",
    "def cosine_sim(u, v):\n",
    "    ulen, vlen, s = 0, 0, 0\n",
    "    for k in range(len(u)):\n",
    "        i = u[k]\n",
    "        j = v[k]\n",
    "        s += i * j\n",
    "        ulen += i ** 2\n",
    "        vlen += j ** 2\n",
    "    ulen = np.sqrt(ulen)\n",
    "    vlen = np.sqrt(vlen)\n",
    "    return 1 - s / (ulen * vlen)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Show the caluclation is correct"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.01613008990009257"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cosine_sim([1,2], [3,4])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.01613008990009257"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distance.cosine([1, 2], [3, 4])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It's faster than scipy!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11.3 µs ± 993 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%timeit cosine_sim([1,2], [3,4])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "26.8 µs ± 273 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%timeit distance.cosine([1, 2], [3, 4])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}