dict_to_2d_array.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from collections import OrderedDict\n",
    "from operator import itemgetter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 309 ms\n"
     ]
    }
   ],
   "source": [
    "%time key_vars = OrderedDict({str(key) : np.array([key, key, key], dtype=np.float64) for key in np.random.choice(np.arange(0,1_000_000), 100_000, replace=False)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_array(gen, shape, dtype):\n",
    "    arr = np.empty(shape, dtype=dtype)\n",
    "    for i, val in enumerate(gen):\n",
    "        arr[i] = val\n",
    "    return arr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43.7 ms ± 1.28 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)\n",
      "56.3 ms ± 3.01 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)\n",
      "45.1 ms ± 1.46 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)\n"
     ]
    }
   ],
   "source": [
    "# select a subset of keys (this case, use all)\n",
    "keys = list(key_vars.keys())\n",
    "# naive method\n",
    "%timeit arr0 = np.array([key_vars[key] for key in keys], dtype=np.float64)\n",
    "\n",
    "# iter over generator\n",
    "%timeit arr2 = create_array((key_vars[key] for key in keys), (len(keys), 3), np.float64)\n",
    "\n",
    "# use itemgetter\n",
    "%timeit arr1 = np.array(itemgetter(*keys)(key_vars), dtype=np.float64)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}