pandas-crosstab-groupby-pivot.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Runtime comparison of pandas crosstab, groupby and pivot_table\n",
    "\n",
    "The [pandas library](https://pandas.pydata.org/) is very powerful and offers several ways to group and summarize data. Typically, I use the `groupby` method but find `pivot_table` to be more readable. There is also `crosstab` as another alternative. In this notebook I'll do a short comparison of the runtime of `groupby`, `pivot_table` and `crosstab`. Let's first look at a DataFrame consisting of 10 columns and 100,000 rows of random integers between 1 and 100. The DataFrame will be grouped by the columns `A` and `B` counting the values in column `C`. The three operations result in data frames containing the same values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>E</th>\n",
       "      <th>F</th>\n",
       "      <th>G</th>\n",
       "      <th>H</th>\n",
       "      <th>I</th>\n",
       "      <th>J</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>52</td>\n",
       "      <td>93</td>\n",
       "      <td>15</td>\n",
       "      <td>72</td>\n",
       "      <td>61</td>\n",
       "      <td>21</td>\n",
       "      <td>83</td>\n",
       "      <td>87</td>\n",
       "      <td>75</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>88</td>\n",
       "      <td>24</td>\n",
       "      <td>3</td>\n",
       "      <td>22</td>\n",
       "      <td>53</td>\n",
       "      <td>2</td>\n",
       "      <td>88</td>\n",
       "      <td>30</td>\n",
       "      <td>38</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>64</td>\n",
       "      <td>60</td>\n",
       "      <td>21</td>\n",
       "      <td>33</td>\n",
       "      <td>76</td>\n",
       "      <td>58</td>\n",
       "      <td>22</td>\n",
       "      <td>89</td>\n",
       "      <td>49</td>\n",
       "      <td>91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>59</td>\n",
       "      <td>42</td>\n",
       "      <td>92</td>\n",
       "      <td>60</td>\n",
       "      <td>80</td>\n",
       "      <td>15</td>\n",
       "      <td>62</td>\n",
       "      <td>62</td>\n",
       "      <td>47</td>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>51</td>\n",
       "      <td>55</td>\n",
       "      <td>64</td>\n",
       "      <td>3</td>\n",
       "      <td>51</td>\n",
       "      <td>7</td>\n",
       "      <td>21</td>\n",
       "      <td>73</td>\n",
       "      <td>39</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    A   B   C   D   E   F   G   H   I   J\n",
       "0  52  93  15  72  61  21  83  87  75  75\n",
       "1  88  24   3  22  53   2  88  30  38   2\n",
       "2  64  60  21  33  76  58  22  89  49  91\n",
       "3  59  42  92  60  80  15  62  62  47  62\n",
       "4  51  55  64   3  51   7  21  73  39  18"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "%load_ext signature\n",
    "from string import ascii_uppercase\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "cols = list(ascii_uppercase[:10])\n",
    "np.random.seed(42)\n",
    "data = np.random.randint(1, 100, size=(100_000, 10))\n",
    "df = pd.DataFrame(data, columns=cols)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8.21 ms ± 245 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "df.groupby(['A', 'B'])['C'].count().unstack(fill_value=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "34.2 ms ± 424 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "df.pivot_table(values='C', index='A', columns='B', aggfunc='count', fill_value=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "71.4 ms ± 771 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "pd.crosstab(index=df.A, columns=df.B)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the first test case the `groupby` DataFrame method followed by calls to `count` and `unstack` is clearly the fastest way to do the calculation. I think the `pivot_table` option makes it more obvious what the result will look like, when spelling out the keyword arguments, but it is much slower. The `crosstab` option saves you some typing at the cost of being slow and in my opinion the hardest to understand just by looking at the code.\n",
    "\n",
    "Next, let's compare data frames of different sizes. Since the `groupby` variant involves multiple function calls, we'll create wrapper functions that can be easily passed to the `timeit` function. In the following loop the runtimes are measured for data frames ranging from 10,000 rows to 10 million."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import timeit\n",
    "from collections import defaultdict\n",
    "\n",
    "\n",
    "def crosstab(df):\n",
    "    pd.crosstab(index=df.A, columns=df.B)\n",
    "\n",
    "\n",
    "def groupby(df):\n",
    "    df.groupby(['A', 'B'])['C'].count().unstack(fill_value=0)\n",
    "\n",
    "\n",
    "def pivot_table(df):\n",
    "    df.pivot_table(values='C', index='A', columns='B', aggfunc='count', fill_value=0)\n",
    "\n",
    "\n",
    "funcs = [crosstab, groupby, pivot_table]\n",
    "measurements = []\n",
    "repetitions = 5\n",
    "\n",
    "# Use a seed distinct from above to prevent caching\n",
    "np.random.seed(420)\n",
    "\n",
    "for size in np.logspace(start=4, stop=7, num=4):\n",
    "    size = int(size)\n",
    "    data = np.random.randint(1, 100, size=(size, 10))\n",
    "    df = pd.DataFrame(data, columns=cols)\n",
    "    \n",
    "    for func in funcs:\n",
    "        duration = timeit.timeit('func(df)', number=repetitions, globals=globals()) / repetitions\n",
    "        measurements.append({'Function': func.__name__, 'Row count': size, 'duration': duration})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, plot the measurements as bar charts to visualize the differences in duration."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.style.use('ramiro')\n",
    "fig, ax = plt.subplots(figsize=(10, 6))\n",
    "pd.DataFrame(measurements).groupby(['Row count', 'Function']).duration.mean().unstack().plot(ax=ax, kind='bar')\n",
    "\n",
    "ax.set_xlabel('DataFrame row count')\n",
    "ax.set_ylabel('Duration in seconds', labelpad=25, va='top')\n",
    "\n",
    "fig.suptitle('Runtime Comparison of pandas crosstab, groupby and pivot_table', fontsize=22)\n",
    "fig.text(0.01, -0.15, 'Author: Ramiro Gómez • Source: ramiro.org/notebook/pandas-crosstab-groupby-pivot/', ha='left', va='center', fontsize=14);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For smaller data frames the `groupby` variant is clearly the fastest. For larger record counts the runtime differences between `groupby` and `pivot_table` become less pronounced, both of them are significantly faster than `crosstab`. In this particular case, where all three variants produce equivalent results, `groupby` looks like the best choice. Obviously, runtime may not be your main concern and other factors such as data types and missing values could impact the runtime in favor of the other variants. Better measure than guess and you should end up with a sane choice."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "Author: <a href=\"https://ramiro.org/\">Ramiro Gómez</a>\n",
       "            • Last edited: June 20, 2019<br>Linux 5.0.0-16-generic - CPython 3.7.3 - IPython 7.5.0 - matplotlib 3.1.0 - numpy 1.16.4 - pandas 0.24.2"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%signature"
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}