ch_04/2-dataframe_operations.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DataFrame Operations\n",
    "\n",
    "## About the Data\n",
    "In this notebook, we will be working with 2 datasets:\n",
    "- Facebook's stock price throughout 2018 (obtained using the [`stock_analysis` package](https://github.com/stefmolin/stock-analysis)).\n",
    "- Daily weather data for NYC from the [National Centers for Environmental Information (NCEI) API](https://www.ncdc.noaa.gov/cdo-web/webservices/v2).\n",
    "\n",
    "*Note: The NCEI is part of the National Oceanic and Atmospheric Administration (NOAA) and, as you can see from the URL for the API, this resource was created when the NCEI was called the NCDC. Should the URL for this resource change in the future, you can search for \"NCEI weather API\" to find the updated one.*\n",
    "\n",
    "## Background on the weather data\n",
    "\n",
    "Data meanings:\n",
    "- `AWND`: average wind speed\n",
    "- `PRCP`: precipitation in millimeters\n",
    "- `SNOW`: snowfall in millimeters\n",
    "- `SNWD`: snow depth in millimeters\n",
    "- `TMAX`: maximum daily temperature in Celsius\n",
    "- `TMIN`: minimum daily temperature in Celsius\n",
    "\n",
    "## Setup"
   ]
  },
  {
   "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>date</th>\n",
       "      <th>datatype</th>\n",
       "      <th>station</th>\n",
       "      <th>attributes</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2018-01-01</td>\n",
       "      <td>PRCP</td>\n",
       "      <td>GHCND:US1CTFR0039</td>\n",
       "      <td>,,N,</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2018-01-01</td>\n",
       "      <td>PRCP</td>\n",
       "      <td>GHCND:US1NJBG0015</td>\n",
       "      <td>,,N,</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2018-01-01</td>\n",
       "      <td>SNOW</td>\n",
       "      <td>GHCND:US1NJBG0015</td>\n",
       "      <td>,,N,</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2018-01-01</td>\n",
       "      <td>PRCP</td>\n",
       "      <td>GHCND:US1NJBG0017</td>\n",
       "      <td>,,N,</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2018-01-01</td>\n",
       "      <td>SNOW</td>\n",
       "      <td>GHCND:US1NJBG0017</td>\n",
       "      <td>,,N,</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date datatype            station attributes  value\n",
       "0 2018-01-01     PRCP  GHCND:US1CTFR0039       ,,N,    0.0\n",
       "1 2018-01-01     PRCP  GHCND:US1NJBG0015       ,,N,    0.0\n",
       "2 2018-01-01     SNOW  GHCND:US1NJBG0015       ,,N,    0.0\n",
       "3 2018-01-01     PRCP  GHCND:US1NJBG0017       ,,N,    0.0\n",
       "4 2018-01-01     SNOW  GHCND:US1NJBG0017       ,,N,    0.0"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "weather = pd.read_csv('data/nyc_weather_2018.csv', parse_dates=['date'])\n",
    "weather.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-01-02</th>\n",
       "      <td>177.68</td>\n",
       "      <td>181.58</td>\n",
       "      <td>177.5500</td>\n",
       "      <td>181.42</td>\n",
       "      <td>18151903</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-03</th>\n",
       "      <td>181.88</td>\n",
       "      <td>184.78</td>\n",
       "      <td>181.3300</td>\n",
       "      <td>184.67</td>\n",
       "      <td>16886563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-04</th>\n",
       "      <td>184.90</td>\n",
       "      <td>186.21</td>\n",
       "      <td>184.0996</td>\n",
       "      <td>184.33</td>\n",
       "      <td>13880896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-05</th>\n",
       "      <td>185.59</td>\n",
       "      <td>186.90</td>\n",
       "      <td>184.9300</td>\n",
       "      <td>186.85</td>\n",
       "      <td>13574535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-08</th>\n",
       "      <td>187.20</td>\n",
       "      <td>188.90</td>\n",
       "      <td>186.3300</td>\n",
       "      <td>188.28</td>\n",
       "      <td>17994726</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              open    high       low   close    volume\n",
       "date                                                  \n",
       "2018-01-02  177.68  181.58  177.5500  181.42  18151903\n",
       "2018-01-03  181.88  184.78  181.3300  184.67  16886563\n",
       "2018-01-04  184.90  186.21  184.0996  184.33  13880896\n",
       "2018-01-05  185.59  186.90  184.9300  186.85  13574535\n",
       "2018-01-08  187.20  188.90  186.3300  188.28  17994726"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb = pd.read_csv('data/fb_2018.csv', index_col='date', parse_dates=True)\n",
    "fb.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Arithmetic and statistics\n",
    "We already saw that we can use mathematical operators like `+` and `/` with dataframes directly. However, we can also use methods, which allow us to specify the axis to perform the calculation over. By default, this is per column. Let's find the Z-scores for the volume traded and look at the days where this was more than 3 standard deviations from the mean:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>abs_z_score_volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-03-19</th>\n",
       "      <td>177.01</td>\n",
       "      <td>177.17</td>\n",
       "      <td>170.06</td>\n",
       "      <td>172.56</td>\n",
       "      <td>88140060</td>\n",
       "      <td>3.145078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-20</th>\n",
       "      <td>167.47</td>\n",
       "      <td>170.20</td>\n",
       "      <td>161.95</td>\n",
       "      <td>168.15</td>\n",
       "      <td>129851768</td>\n",
       "      <td>5.315169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-21</th>\n",
       "      <td>164.80</td>\n",
       "      <td>173.40</td>\n",
       "      <td>163.30</td>\n",
       "      <td>169.39</td>\n",
       "      <td>106598834</td>\n",
       "      <td>4.105413</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-26</th>\n",
       "      <td>160.82</td>\n",
       "      <td>161.10</td>\n",
       "      <td>149.02</td>\n",
       "      <td>160.06</td>\n",
       "      <td>126116634</td>\n",
       "      <td>5.120845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-07-26</th>\n",
       "      <td>174.89</td>\n",
       "      <td>180.13</td>\n",
       "      <td>173.75</td>\n",
       "      <td>176.26</td>\n",
       "      <td>169803668</td>\n",
       "      <td>7.393705</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              open    high     low   close     volume  abs_z_score_volume\n",
       "date                                                                     \n",
       "2018-03-19  177.01  177.17  170.06  172.56   88140060            3.145078\n",
       "2018-03-20  167.47  170.20  161.95  168.15  129851768            5.315169\n",
       "2018-03-21  164.80  173.40  163.30  169.39  106598834            4.105413\n",
       "2018-03-26  160.82  161.10  149.02  160.06  126116634            5.120845\n",
       "2018-07-26  174.89  180.13  173.75  176.26  169803668            7.393705"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb.assign(\n",
    "    abs_z_score_volume=lambda x: \\\n",
    "        x.volume.sub(x.volume.mean()).div(x.volume.std()).abs()\n",
    ").query('abs_z_score_volume > 3')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use `rank()` and `pct_change()` to see which days had the largest change in volume traded from the day before:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>volume_pct_change</th>\n",
       "      <th>pct_change_rank</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-01-12</th>\n",
       "      <td>178.06</td>\n",
       "      <td>181.48</td>\n",
       "      <td>177.40</td>\n",
       "      <td>179.37</td>\n",
       "      <td>77551299</td>\n",
       "      <td>7.087876</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-19</th>\n",
       "      <td>177.01</td>\n",
       "      <td>177.17</td>\n",
       "      <td>170.06</td>\n",
       "      <td>172.56</td>\n",
       "      <td>88140060</td>\n",
       "      <td>2.611789</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-07-26</th>\n",
       "      <td>174.89</td>\n",
       "      <td>180.13</td>\n",
       "      <td>173.75</td>\n",
       "      <td>176.26</td>\n",
       "      <td>169803668</td>\n",
       "      <td>1.628841</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-09-21</th>\n",
       "      <td>166.64</td>\n",
       "      <td>167.25</td>\n",
       "      <td>162.81</td>\n",
       "      <td>162.93</td>\n",
       "      <td>45994800</td>\n",
       "      <td>1.428956</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-26</th>\n",
       "      <td>160.82</td>\n",
       "      <td>161.10</td>\n",
       "      <td>149.02</td>\n",
       "      <td>160.06</td>\n",
       "      <td>126116634</td>\n",
       "      <td>1.352496</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              open    high     low   close     volume  volume_pct_change  \\\n",
       "date                                                                       \n",
       "2018-01-12  178.06  181.48  177.40  179.37   77551299           7.087876   \n",
       "2018-03-19  177.01  177.17  170.06  172.56   88140060           2.611789   \n",
       "2018-07-26  174.89  180.13  173.75  176.26  169803668           1.628841   \n",
       "2018-09-21  166.64  167.25  162.81  162.93   45994800           1.428956   \n",
       "2018-03-26  160.82  161.10  149.02  160.06  126116634           1.352496   \n",
       "\n",
       "            pct_change_rank  \n",
       "date                         \n",
       "2018-01-12              1.0  \n",
       "2018-03-19              2.0  \n",
       "2018-07-26              3.0  \n",
       "2018-09-21              4.0  \n",
       "2018-03-26              5.0  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb.assign(\n",
    "    volume_pct_change=fb.volume.pct_change(),\n",
    "    pct_change_rank=lambda x: \\\n",
    "        x.volume_pct_change.abs().rank(ascending=False)\n",
    ").nsmallest(5, 'pct_change_rank')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "January 12th was when the news that Facebook changed its news feed product to focus more on content from a users' friends over the brands they follow. Given that Facebook's advertising is a key component of its business ([nearly 89% in 2017](https://www.investopedia.com/ask/answers/120114/how-does-facebook-fb-make-money.asp)), many shares were sold and the price dropped in panic:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-01-11</th>\n",
       "      <td>188.40</td>\n",
       "      <td>188.40</td>\n",
       "      <td>187.38</td>\n",
       "      <td>187.77</td>\n",
       "      <td>9588587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-12</th>\n",
       "      <td>178.06</td>\n",
       "      <td>181.48</td>\n",
       "      <td>177.40</td>\n",
       "      <td>179.37</td>\n",
       "      <td>77551299</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              open    high     low   close    volume\n",
       "date                                                \n",
       "2018-01-11  188.40  188.40  187.38  187.77   9588587\n",
       "2018-01-12  178.06  181.48  177.40  179.37  77551299"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb['2018-01-11':'2018-01-12']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Throughout 2018, Facebook's stock price never had a low above $215:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "open       True\n",
       "high       True\n",
       "low       False\n",
       "close      True\n",
       "volume     True\n",
       "dtype: bool"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(fb > 215).any()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Facebook's OHLC (open, high, low, and close) prices all had at least one day they were at $215 or less:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "open      False\n",
       "high      False\n",
       "low       False\n",
       "close     False\n",
       "volume     True\n",
       "dtype: bool"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(fb > 215).all()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Binning\n",
    "When working with volume traded, we may be interested in ranges of volume rather than the exact values. No two days have the same volume traded:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(fb.volume.value_counts() > 1).sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use `pd.cut()` to create 3 bins of even range in volume traded and name them. Then we can work with low, medium, and high volume traded categories:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "low     240\n",
       "med       8\n",
       "high      3\n",
       "Name: volume, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "volume_binned = pd.cut(fb.volume, bins=3, labels=['low', 'med', 'high'])\n",
    "volume_binned.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's look at the days with high trading volume:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-07-26</th>\n",
       "      <td>174.89</td>\n",
       "      <td>180.13</td>\n",
       "      <td>173.75</td>\n",
       "      <td>176.26</td>\n",
       "      <td>169803668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-20</th>\n",
       "      <td>167.47</td>\n",
       "      <td>170.20</td>\n",
       "      <td>161.95</td>\n",
       "      <td>168.15</td>\n",
       "      <td>129851768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-26</th>\n",
       "      <td>160.82</td>\n",
       "      <td>161.10</td>\n",
       "      <td>149.02</td>\n",
       "      <td>160.06</td>\n",
       "      <td>126116634</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              open    high     low   close     volume\n",
       "date                                                 \n",
       "2018-07-26  174.89  180.13  173.75  176.26  169803668\n",
       "2018-03-20  167.47  170.20  161.95  168.15  129851768\n",
       "2018-03-26  160.82  161.10  149.02  160.06  126116634"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb[volume_binned == 'high'].sort_values('volume', ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "July 25th Facebook announced disappointing user growth and the stock tanked in the after hours:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-07-25</th>\n",
       "      <td>215.715</td>\n",
       "      <td>218.62</td>\n",
       "      <td>214.27</td>\n",
       "      <td>217.50</td>\n",
       "      <td>64592585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-07-26</th>\n",
       "      <td>174.890</td>\n",
       "      <td>180.13</td>\n",
       "      <td>173.75</td>\n",
       "      <td>176.26</td>\n",
       "      <td>169803668</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               open    high     low   close     volume\n",
       "date                                                  \n",
       "2018-07-25  215.715  218.62  214.27  217.50   64592585\n",
       "2018-07-26  174.890  180.13  173.75  176.26  169803668"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb['2018-07-25':'2018-07-26']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cambridge Analytica scandal broke on Saturday, March 17th, so we look at the Monday after for the numbers:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-03-16</th>\n",
       "      <td>184.49</td>\n",
       "      <td>185.33</td>\n",
       "      <td>183.41</td>\n",
       "      <td>185.09</td>\n",
       "      <td>24403438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-19</th>\n",
       "      <td>177.01</td>\n",
       "      <td>177.17</td>\n",
       "      <td>170.06</td>\n",
       "      <td>172.56</td>\n",
       "      <td>88140060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-03-20</th>\n",
       "      <td>167.47</td>\n",
       "      <td>170.20</td>\n",
       "      <td>161.95</td>\n",
       "      <td>168.15</td>\n",
       "      <td>129851768</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              open    high     low   close     volume\n",
       "date                                                 \n",
       "2018-03-16  184.49  185.33  183.41  185.09   24403438\n",
       "2018-03-19  177.01  177.17  170.06  172.56   88140060\n",
       "2018-03-20  167.47  170.20  161.95  168.15  129851768"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb['2018-03-16':'2018-03-20']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since most days have similar volume, but a few are very large, we have very wide bins. Most of the data is in the low bin. \n",
    "\n",
    "*Note: visualizations will be covered in chapters 5 and 6.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:title={'center':'Daily Volume Traded of Facebook Stock in 2018 (with bins)'}, ylabel='volume traded'>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from visual_aids.misc_viz import low_med_high_bins_viz\n",
    "\n",
    "low_med_high_bins_viz(\n",
    "    fb, 'volume', ylabel='volume traded',\n",
    "    title='Daily Volume Traded of Facebook Stock in 2018 (with bins)'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we split using quantiles, the bins will have roughly the same number of observations. For this, we use `qcut()`. We will make 4 quartiles:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "q1    63\n",
       "q2    63\n",
       "q4    63\n",
       "q3    62\n",
       "Name: volume, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "volume_qbinned = pd.qcut(fb.volume, q=4, labels=['q1', 'q2', 'q3', 'q4'])\n",
    "volume_qbinned.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Notice the bins don't cover ranges of the same size anymore:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:title={'center':'Daily Volume Traded of Facebook Stock in 2018 (with quartile bins)'}, ylabel='volume traded'>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from visual_aids.misc_viz import quartile_bins_viz\n",
    "\n",
    "quartile_bins_viz(\n",
    "    fb, 'volume', ylabel='volume traded', \n",
    "    title='Daily Volume Traded of Facebook Stock in 2018 (with quartile bins)'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Applying Functions\n",
    "We can use the `apply()` method to run the same operation on all columns (or rows) of the dataframe. First, let's isolate the weather observations from the Central Park station and pivot the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "central_park_weather = weather\\\n",
    "    .query('station == \"GHCND:USW00094728\"')\\\n",
    "    .pivot(index='date', columns='datatype', values='value')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's calculate the Z-scores of the TMIN, TMAX, and PRCP observations in Central Park in October 2018:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datatype</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>TMIN</th>\n",
       "      <td>31.0</td>\n",
       "      <td>-1.790682e-16</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.339112</td>\n",
       "      <td>-0.751019</td>\n",
       "      <td>-0.474269</td>\n",
       "      <td>1.065152</td>\n",
       "      <td>1.843511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TMAX</th>\n",
       "      <td>31.0</td>\n",
       "      <td>1.951844e-16</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.305582</td>\n",
       "      <td>-0.870013</td>\n",
       "      <td>-0.138258</td>\n",
       "      <td>1.011643</td>\n",
       "      <td>1.604016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PRCP</th>\n",
       "      <td>31.0</td>\n",
       "      <td>4.655774e-17</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.394438</td>\n",
       "      <td>-0.394438</td>\n",
       "      <td>-0.394438</td>\n",
       "      <td>-0.240253</td>\n",
       "      <td>3.936167</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          count          mean  std       min       25%       50%       75%  \\\n",
       "datatype                                                                     \n",
       "TMIN       31.0 -1.790682e-16  1.0 -1.339112 -0.751019 -0.474269  1.065152   \n",
       "TMAX       31.0  1.951844e-16  1.0 -1.305582 -0.870013 -0.138258  1.011643   \n",
       "PRCP       31.0  4.655774e-17  1.0 -0.394438 -0.394438 -0.394438 -0.240253   \n",
       "\n",
       "               max  \n",
       "datatype            \n",
       "TMIN      1.843511  \n",
       "TMAX      1.604016  \n",
       "PRCP      3.936167  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "oct_weather_z_scores = central_park_weather\\\n",
    "    .loc['2018-10', ['TMIN', 'TMAX', 'PRCP']]\\\n",
    "    .apply(lambda x: x.sub(x.mean()).div(x.std()))\n",
    "oct_weather_z_scores.describe().T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "October 27th rained much more than the rest of the days:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date\n",
       "2018-10-27    3.936167\n",
       "Name: PRCP, dtype: float64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "oct_weather_z_scores.query('PRCP > 3').PRCP"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Indeed, this day was much higher than the rest:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    31.000000\n",
       "mean      2.941935\n",
       "std       7.458542\n",
       "min       0.000000\n",
       "25%       0.000000\n",
       "50%       0.000000\n",
       "75%       1.150000\n",
       "max      32.300000\n",
       "Name: PRCP, dtype: float64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "central_park_weather.loc['2018-10', 'PRCP'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When the function we want to apply isn't vectorized, we can:\n",
    "- use `np.vectorize()` to vectorize it (similar to how `map()` works) and then use it with `apply()`\n",
    "- use `applymap()` and pass it the non-vectorized function directly\n",
    "\n",
    "Say we wanted to count the digits of the whole numbers for the Facebook data; `len()` is not vectorized, so we can use `np.vectorize()` or `applymap()`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb.apply(\n",
    "    lambda x: np.vectorize(lambda y: len(str(np.ceil(y))))(x)\n",
    ").astype('int64').equals(\n",
    "    fb.applymap(lambda x: len(str(np.ceil(x))))\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A simple operation of addition to each element in a series grows linearly in time complexity when using `iteritems()`, but stays near 0 when using vectorized operations. `iteritems()` and related methods should only be used if there is no vectorized solution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "np.random.seed(0)\n",
    "\n",
    "vectorized_results = {}\n",
    "iteritems_results = {}\n",
    "\n",
    "for size in [10, 100, 1000, 10000, 100000, 500000, 1000000, 5000000, 10000000]:\n",
    "    # set of numbers to use\n",
    "    test = pd.Series(np.random.uniform(size=size))\n",
    "    \n",
    "    # time the vectorized operation\n",
    "    start = time.time()\n",
    "    x = test + 10\n",
    "    end = time.time()\n",
    "    vectorized_results[size] = end - start\n",
    "    \n",
    "    # time the operation with `iteritems()`\n",
    "    start = time.time()\n",
    "    x = []\n",
    "    for i, v in test.iteritems():\n",
    "        x.append(v + 10)\n",
    "    x = pd.Series(x)\n",
    "    end = time.time()\n",
    "    iteritems_results[size] = end - start\n",
    "\n",
    "results = pd.DataFrame(\n",
    "    [pd.Series(vectorized_results, name='vectorized'), pd.Series(iteritems_results, name='iteritems')]\n",
    ").T    \n",
    "\n",
    "# plotting\n",
    "ax = results.plot(title='Time Complexity', color=['blue', 'red'], legend=False)\n",
    "\n",
    "# formatting\n",
    "ax.set(xlabel='item size (rows)', ylabel='time (s)')\n",
    "ax.text(0.5e7, iteritems_results[0.5e7] * .9, 'iteritems()', rotation=34, color='red', fontsize=12, ha='center', va='bottom')\n",
    "ax.text(0.5e7, vectorized_results[0.5e7], 'vectorized', color='blue', fontsize=12, ha='center', va='bottom')\n",
    "for spine in ['top', 'right']:\n",
    "    ax.spines[spine].set_visible(False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Window Calculations\n",
    "*Consult the [`understanding_window_calculations.ipynb`](./understanding_window_calculations.ipynb) notebook for interactive visualizations using widgets to help understand window calculations.*\n",
    "\n",
    "The `rolling()` method allows us to perform rolling window calculations. We simply specify the window size (3 days here) and follow it with a call to an aggregation function (sum here):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>date</th>\n",
       "      <th>2018-10-01</th>\n",
       "      <th>2018-10-02</th>\n",
       "      <th>2018-10-03</th>\n",
       "      <th>2018-10-04</th>\n",
       "      <th>2018-10-05</th>\n",
       "      <th>2018-10-06</th>\n",
       "      <th>2018-10-07</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datatype</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>PRCP</th>\n",
       "      <td>0.0</td>\n",
       "      <td>17.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rolling_PRCP</th>\n",
       "      <td>0.0</td>\n",
       "      <td>17.5</td>\n",
       "      <td>17.5</td>\n",
       "      <td>18.5</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "date          2018-10-01  2018-10-02  2018-10-03  2018-10-04  2018-10-05  \\\n",
       "datatype                                                                   \n",
       "PRCP                 0.0        17.5         0.0         1.0         0.0   \n",
       "rolling_PRCP         0.0        17.5        17.5        18.5         1.0   \n",
       "\n",
       "date          2018-10-06  2018-10-07  \n",
       "datatype                              \n",
       "PRCP                 0.0         0.0  \n",
       "rolling_PRCP         1.0         0.0  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "central_park_weather.loc['2018-10'].assign(\n",
    "    rolling_PRCP=lambda x: x.PRCP.rolling('3D').sum()\n",
    ")[['PRCP', 'rolling_PRCP']].head(7).T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also perform the rolling calculations on the entire dataframe at once. This will apply the same aggregation function to each column:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>datatype</th>\n",
       "      <th>AWND</th>\n",
       "      <th>PRCP</th>\n",
       "      <th>SNOW</th>\n",
       "      <th>SNWD</th>\n",
       "      <th>TMAX</th>\n",
       "      <th>TMIN</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-10-01</th>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24.400000</td>\n",
       "      <td>17.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-02</th>\n",
       "      <td>0.900000</td>\n",
       "      <td>8.750000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24.700000</td>\n",
       "      <td>17.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-03</th>\n",
       "      <td>0.966667</td>\n",
       "      <td>5.833333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24.233333</td>\n",
       "      <td>17.566667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-04</th>\n",
       "      <td>0.800000</td>\n",
       "      <td>6.166667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24.233333</td>\n",
       "      <td>17.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-05</th>\n",
       "      <td>1.033333</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>23.133333</td>\n",
       "      <td>16.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-06</th>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>22.033333</td>\n",
       "      <td>16.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-07</th>\n",
       "      <td>1.066667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>22.600000</td>\n",
       "      <td>17.400000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "datatype        AWND      PRCP  SNOW  SNWD       TMAX       TMIN\n",
       "date                                                            \n",
       "2018-10-01  0.900000  0.000000   0.0   0.0  24.400000  17.200000\n",
       "2018-10-02  0.900000  8.750000   0.0   0.0  24.700000  17.750000\n",
       "2018-10-03  0.966667  5.833333   0.0   0.0  24.233333  17.566667\n",
       "2018-10-04  0.800000  6.166667   0.0   0.0  24.233333  17.200000\n",
       "2018-10-05  1.033333  0.333333   0.0   0.0  23.133333  16.300000\n",
       "2018-10-06  0.833333  0.333333   0.0   0.0  22.033333  16.300000\n",
       "2018-10-07  1.066667  0.000000   0.0   0.0  22.600000  17.400000"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "central_park_weather.loc['2018-10'].rolling('3D').mean().head(7).iloc[:,:6]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use different aggregation functions per column if we use `agg()` instead. We pass in a dictionary mapping the column to the aggregation to perform on it. Here, we join the result to the original data to see what is happening:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>AWND</th>\n",
       "      <th>AWND_rolling</th>\n",
       "      <th>PRCP</th>\n",
       "      <th>PRCP_rolling</th>\n",
       "      <th>TMAX</th>\n",
       "      <th>TMAX_rolling</th>\n",
       "      <th>TMIN</th>\n",
       "      <th>TMIN_rolling</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-10-01</th>\n",
       "      <td>0.9</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24.4</td>\n",
       "      <td>24.4</td>\n",
       "      <td>17.2</td>\n",
       "      <td>17.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-02</th>\n",
       "      <td>0.9</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>17.5</td>\n",
       "      <td>17.5</td>\n",
       "      <td>25.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>18.3</td>\n",
       "      <td>17.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-03</th>\n",
       "      <td>1.1</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17.5</td>\n",
       "      <td>23.3</td>\n",
       "      <td>25.0</td>\n",
       "      <td>17.2</td>\n",
       "      <td>17.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-04</th>\n",
       "      <td>0.4</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>18.5</td>\n",
       "      <td>24.4</td>\n",
       "      <td>25.0</td>\n",
       "      <td>16.1</td>\n",
       "      <td>16.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-05</th>\n",
       "      <td>1.6</td>\n",
       "      <td>1.033333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>21.7</td>\n",
       "      <td>24.4</td>\n",
       "      <td>15.6</td>\n",
       "      <td>15.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-06</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>24.4</td>\n",
       "      <td>17.2</td>\n",
       "      <td>15.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-07</th>\n",
       "      <td>1.1</td>\n",
       "      <td>1.066667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>26.1</td>\n",
       "      <td>26.1</td>\n",
       "      <td>19.4</td>\n",
       "      <td>15.6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            AWND  AWND_rolling  PRCP  PRCP_rolling  TMAX  TMAX_rolling  TMIN  \\\n",
       "date                                                                           \n",
       "2018-10-01   0.9      0.900000   0.0           0.0  24.4          24.4  17.2   \n",
       "2018-10-02   0.9      0.900000  17.5          17.5  25.0          25.0  18.3   \n",
       "2018-10-03   1.1      0.966667   0.0          17.5  23.3          25.0  17.2   \n",
       "2018-10-04   0.4      0.800000   1.0          18.5  24.4          25.0  16.1   \n",
       "2018-10-05   1.6      1.033333   0.0           1.0  21.7          24.4  15.6   \n",
       "2018-10-06   0.5      0.833333   0.0           1.0  20.0          24.4  17.2   \n",
       "2018-10-07   1.1      1.066667   0.0           0.0  26.1          26.1  19.4   \n",
       "\n",
       "            TMIN_rolling  \n",
       "date                      \n",
       "2018-10-01          17.2  \n",
       "2018-10-02          17.2  \n",
       "2018-10-03          17.2  \n",
       "2018-10-04          16.1  \n",
       "2018-10-05          15.6  \n",
       "2018-10-06          15.6  \n",
       "2018-10-07          15.6  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "central_park_weather['2018-10-01':'2018-10-07'].rolling('3D').agg(\n",
    "    {'TMAX': 'max', 'TMIN': 'min', 'AWND': 'mean', 'PRCP': 'sum'}\n",
    ").join( # join with original data for comparison\n",
    "    central_park_weather[['TMAX', 'TMIN', 'AWND', 'PRCP']], \n",
    "    lsuffix='_rolling'\n",
    ").sort_index(axis=1) # sort columns so rolling calcs are next to originals"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Suppose we reindexed the Facebook stock data as we did with the S&P 500 data in chapter 3. If we were to use rolling calculations on this data, we would be including the values when the market was closed:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>day</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-01-01</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Monday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-02</th>\n",
       "      <td>177.68</td>\n",
       "      <td>181.58</td>\n",
       "      <td>177.5500</td>\n",
       "      <td>181.42</td>\n",
       "      <td>18151903.0</td>\n",
       "      <td>Tuesday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-03</th>\n",
       "      <td>181.88</td>\n",
       "      <td>184.78</td>\n",
       "      <td>181.3300</td>\n",
       "      <td>184.67</td>\n",
       "      <td>16886563.0</td>\n",
       "      <td>Wednesday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-04</th>\n",
       "      <td>184.90</td>\n",
       "      <td>186.21</td>\n",
       "      <td>184.0996</td>\n",
       "      <td>184.33</td>\n",
       "      <td>13880896.0</td>\n",
       "      <td>Thursday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-05</th>\n",
       "      <td>185.59</td>\n",
       "      <td>186.90</td>\n",
       "      <td>184.9300</td>\n",
       "      <td>186.85</td>\n",
       "      <td>13574535.0</td>\n",
       "      <td>Friday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-06</th>\n",
       "      <td>186.85</td>\n",
       "      <td>186.85</td>\n",
       "      <td>186.8500</td>\n",
       "      <td>186.85</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Saturday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-07</th>\n",
       "      <td>186.85</td>\n",
       "      <td>186.85</td>\n",
       "      <td>186.8500</td>\n",
       "      <td>186.85</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sunday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-08</th>\n",
       "      <td>187.20</td>\n",
       "      <td>188.90</td>\n",
       "      <td>186.3300</td>\n",
       "      <td>188.28</td>\n",
       "      <td>17994726.0</td>\n",
       "      <td>Monday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-09</th>\n",
       "      <td>188.70</td>\n",
       "      <td>188.80</td>\n",
       "      <td>187.1000</td>\n",
       "      <td>187.87</td>\n",
       "      <td>12393057.0</td>\n",
       "      <td>Tuesday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-10</th>\n",
       "      <td>186.94</td>\n",
       "      <td>187.89</td>\n",
       "      <td>185.6300</td>\n",
       "      <td>187.84</td>\n",
       "      <td>10529894.0</td>\n",
       "      <td>Wednesday</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              open    high       low   close      volume        day\n",
       "2018-01-01     NaN     NaN       NaN     NaN         0.0     Monday\n",
       "2018-01-02  177.68  181.58  177.5500  181.42  18151903.0    Tuesday\n",
       "2018-01-03  181.88  184.78  181.3300  184.67  16886563.0  Wednesday\n",
       "2018-01-04  184.90  186.21  184.0996  184.33  13880896.0   Thursday\n",
       "2018-01-05  185.59  186.90  184.9300  186.85  13574535.0     Friday\n",
       "2018-01-06  186.85  186.85  186.8500  186.85         0.0   Saturday\n",
       "2018-01-07  186.85  186.85  186.8500  186.85         0.0     Sunday\n",
       "2018-01-08  187.20  188.90  186.3300  188.28  17994726.0     Monday\n",
       "2018-01-09  188.70  188.80  187.1000  187.87  12393057.0    Tuesday\n",
       "2018-01-10  186.94  187.89  185.6300  187.84  10529894.0  Wednesday"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb_reindexed = fb\\\n",
    "    .reindex(pd.date_range('2018-01-01', '2018-12-31', freq='D'))\\\n",
    "    .assign(\n",
    "        volume=lambda x: x.volume.fillna(0),\n",
    "        close=lambda x: x.close.fillna(method='ffill'),\n",
    "        open=lambda x: x.open.combine_first(x.close),\n",
    "        high=lambda x: x.high.combine_first(x.close),\n",
    "        low=lambda x: x.low.combine_first(x.close)\n",
    "    )\n",
    "fb_reindexed.assign(day=lambda x: x.index.day_name()).head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As of version 1.0, `pandas` supports defining custom windows for rolling calculations, which makes it possible for us to perform rolling calculations on the days the market was open. One way is to make a new class that inherits from `BaseIndexer` and provide the logic for determining the window bounds in the `get_window_bounds()` method (more info [here](https://pandas.pydata.org/pandas-docs/stable/user_guide/computation.html#custom-window-rolling)). For our use case, we can use the `VariableOffsetWindowIndexer` class, which was introduced in version 1.1, to perform rolling calculations over non-fixed time offsets (like business days). Let's perform a three business day rolling calculation on the reindexed Facebook stock data and join it with the reindexed data for comparison:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>close</th>\n",
       "      <th>close_rolling</th>\n",
       "      <th>high</th>\n",
       "      <th>high_rolling</th>\n",
       "      <th>low</th>\n",
       "      <th>low_rolling</th>\n",
       "      <th>open</th>\n",
       "      <th>open_rolling</th>\n",
       "      <th>volume</th>\n",
       "      <th>volume_rolling</th>\n",
       "      <th>window_start_day</th>\n",
       "      <th>day</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-01-01</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2018-01-01</td>\n",
       "      <td>Monday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-02</th>\n",
       "      <td>181.42</td>\n",
       "      <td>181.420000</td>\n",
       "      <td>181.58</td>\n",
       "      <td>181.58</td>\n",
       "      <td>177.5500</td>\n",
       "      <td>177.5500</td>\n",
       "      <td>177.68</td>\n",
       "      <td>177.680000</td>\n",
       "      <td>18151903.0</td>\n",
       "      <td>18151903.0</td>\n",
       "      <td>2018-01-01</td>\n",
       "      <td>Tuesday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-03</th>\n",
       "      <td>184.67</td>\n",
       "      <td>183.045000</td>\n",
       "      <td>184.78</td>\n",
       "      <td>184.78</td>\n",
       "      <td>181.3300</td>\n",
       "      <td>177.5500</td>\n",
       "      <td>181.88</td>\n",
       "      <td>179.780000</td>\n",
       "      <td>16886563.0</td>\n",
       "      <td>35038466.0</td>\n",
       "      <td>2018-01-01</td>\n",
       "      <td>Wednesday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-04</th>\n",
       "      <td>184.33</td>\n",
       "      <td>183.473333</td>\n",
       "      <td>186.21</td>\n",
       "      <td>186.21</td>\n",
       "      <td>184.0996</td>\n",
       "      <td>177.5500</td>\n",
       "      <td>184.90</td>\n",
       "      <td>181.486667</td>\n",
       "      <td>13880896.0</td>\n",
       "      <td>48919362.0</td>\n",
       "      <td>2018-01-02</td>\n",
       "      <td>Thursday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-05</th>\n",
       "      <td>186.85</td>\n",
       "      <td>185.283333</td>\n",
       "      <td>186.90</td>\n",
       "      <td>186.90</td>\n",
       "      <td>184.9300</td>\n",
       "      <td>181.3300</td>\n",
       "      <td>185.59</td>\n",
       "      <td>184.123333</td>\n",
       "      <td>13574535.0</td>\n",
       "      <td>44341994.0</td>\n",
       "      <td>2018-01-03</td>\n",
       "      <td>Friday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-06</th>\n",
       "      <td>186.85</td>\n",
       "      <td>186.010000</td>\n",
       "      <td>186.85</td>\n",
       "      <td>186.90</td>\n",
       "      <td>186.8500</td>\n",
       "      <td>184.0996</td>\n",
       "      <td>186.85</td>\n",
       "      <td>185.780000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>27455431.0</td>\n",
       "      <td>2018-01-04</td>\n",
       "      <td>Saturday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-07</th>\n",
       "      <td>186.85</td>\n",
       "      <td>186.220000</td>\n",
       "      <td>186.85</td>\n",
       "      <td>186.90</td>\n",
       "      <td>186.8500</td>\n",
       "      <td>184.0996</td>\n",
       "      <td>186.85</td>\n",
       "      <td>186.047500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>27455431.0</td>\n",
       "      <td>2018-01-04</td>\n",
       "      <td>Sunday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-08</th>\n",
       "      <td>188.28</td>\n",
       "      <td>186.632000</td>\n",
       "      <td>188.90</td>\n",
       "      <td>188.90</td>\n",
       "      <td>186.3300</td>\n",
       "      <td>184.0996</td>\n",
       "      <td>187.20</td>\n",
       "      <td>186.278000</td>\n",
       "      <td>17994726.0</td>\n",
       "      <td>45450157.0</td>\n",
       "      <td>2018-01-04</td>\n",
       "      <td>Monday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-09</th>\n",
       "      <td>187.87</td>\n",
       "      <td>187.340000</td>\n",
       "      <td>188.80</td>\n",
       "      <td>188.90</td>\n",
       "      <td>187.1000</td>\n",
       "      <td>184.9300</td>\n",
       "      <td>188.70</td>\n",
       "      <td>187.038000</td>\n",
       "      <td>12393057.0</td>\n",
       "      <td>43962318.0</td>\n",
       "      <td>2018-01-05</td>\n",
       "      <td>Tuesday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-10</th>\n",
       "      <td>187.84</td>\n",
       "      <td>187.538000</td>\n",
       "      <td>187.89</td>\n",
       "      <td>188.90</td>\n",
       "      <td>185.6300</td>\n",
       "      <td>185.6300</td>\n",
       "      <td>186.94</td>\n",
       "      <td>187.308000</td>\n",
       "      <td>10529894.0</td>\n",
       "      <td>40917677.0</td>\n",
       "      <td>2018-01-06</td>\n",
       "      <td>Wednesday</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             close  close_rolling    high  high_rolling       low  \\\n",
       "2018-01-01     NaN            NaN     NaN           NaN       NaN   \n",
       "2018-01-02  181.42     181.420000  181.58        181.58  177.5500   \n",
       "2018-01-03  184.67     183.045000  184.78        184.78  181.3300   \n",
       "2018-01-04  184.33     183.473333  186.21        186.21  184.0996   \n",
       "2018-01-05  186.85     185.283333  186.90        186.90  184.9300   \n",
       "2018-01-06  186.85     186.010000  186.85        186.90  186.8500   \n",
       "2018-01-07  186.85     186.220000  186.85        186.90  186.8500   \n",
       "2018-01-08  188.28     186.632000  188.90        188.90  186.3300   \n",
       "2018-01-09  187.87     187.340000  188.80        188.90  187.1000   \n",
       "2018-01-10  187.84     187.538000  187.89        188.90  185.6300   \n",
       "\n",
       "            low_rolling    open  open_rolling      volume  volume_rolling  \\\n",
       "2018-01-01          NaN     NaN           NaN         0.0             0.0   \n",
       "2018-01-02     177.5500  177.68    177.680000  18151903.0      18151903.0   \n",
       "2018-01-03     177.5500  181.88    179.780000  16886563.0      35038466.0   \n",
       "2018-01-04     177.5500  184.90    181.486667  13880896.0      48919362.0   \n",
       "2018-01-05     181.3300  185.59    184.123333  13574535.0      44341994.0   \n",
       "2018-01-06     184.0996  186.85    185.780000         0.0      27455431.0   \n",
       "2018-01-07     184.0996  186.85    186.047500         0.0      27455431.0   \n",
       "2018-01-08     184.0996  187.20    186.278000  17994726.0      45450157.0   \n",
       "2018-01-09     184.9300  188.70    187.038000  12393057.0      43962318.0   \n",
       "2018-01-10     185.6300  186.94    187.308000  10529894.0      40917677.0   \n",
       "\n",
       "           window_start_day        day  \n",
       "2018-01-01       2018-01-01     Monday  \n",
       "2018-01-02       2018-01-01    Tuesday  \n",
       "2018-01-03       2018-01-01  Wednesday  \n",
       "2018-01-04       2018-01-02   Thursday  \n",
       "2018-01-05       2018-01-03     Friday  \n",
       "2018-01-06       2018-01-04   Saturday  \n",
       "2018-01-07       2018-01-04     Sunday  \n",
       "2018-01-08       2018-01-04     Monday  \n",
       "2018-01-09       2018-01-05    Tuesday  \n",
       "2018-01-10       2018-01-06  Wednesday  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from pandas.api.indexers import VariableOffsetWindowIndexer\n",
    "\n",
    "indexer = VariableOffsetWindowIndexer(\n",
    "    index=fb_reindexed.index, offset=pd.offsets.BDay(3)\n",
    ")\n",
    "fb_reindexed.assign(window_start_day=0).rolling(indexer).agg({\n",
    "    'window_start_day': lambda x: x.index.min().timestamp(),\n",
    "    'open': 'mean', 'high': 'max', 'low': 'min',\n",
    "    'close': 'mean', 'volume': 'sum'\n",
    "}).join(\n",
    "    fb_reindexed, lsuffix='_rolling'\n",
    ").sort_index(axis=1).assign(\n",
    "    day=lambda x: x.index.day_name(),\n",
    "    window_start_day=lambda x: pd.to_datetime(x.window_start_day, unit='s')\n",
    ").head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Rolling calculations (`rolling()`) use a sliding window. Expanding calculations (`expanding()`), however, grow in size. These are equivalent to cumulative aggregations like `cumsum()`; however, we can specify the minimum number of periods required to start calculating (default is 1), and we aren't limited to predefined aggregations. Therefore, while there is no method for the cumulative mean, we can calculate it using `expanding()`. Let's calculate the month-to-date average precipiation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>date</th>\n",
       "      <th>2018-06-01</th>\n",
       "      <th>2018-06-02</th>\n",
       "      <th>2018-06-03</th>\n",
       "      <th>2018-06-04</th>\n",
       "      <th>2018-06-05</th>\n",
       "      <th>2018-06-06</th>\n",
       "      <th>2018-06-07</th>\n",
       "      <th>2018-06-08</th>\n",
       "      <th>2018-06-09</th>\n",
       "      <th>2018-06-10</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datatype</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>PRCP</th>\n",
       "      <td>6.9</td>\n",
       "      <td>2.00</td>\n",
       "      <td>6.4</td>\n",
       "      <td>4.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TOTAL_PRCP</th>\n",
       "      <td>6.9</td>\n",
       "      <td>8.90</td>\n",
       "      <td>15.3</td>\n",
       "      <td>19.40</td>\n",
       "      <td>19.40</td>\n",
       "      <td>19.400000</td>\n",
       "      <td>19.400000</td>\n",
       "      <td>19.400</td>\n",
       "      <td>19.400000</td>\n",
       "      <td>19.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AVG_PRCP</th>\n",
       "      <td>6.9</td>\n",
       "      <td>4.45</td>\n",
       "      <td>5.1</td>\n",
       "      <td>4.85</td>\n",
       "      <td>3.88</td>\n",
       "      <td>3.233333</td>\n",
       "      <td>2.771429</td>\n",
       "      <td>2.425</td>\n",
       "      <td>2.155556</td>\n",
       "      <td>1.97</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "date        2018-06-01  2018-06-02  2018-06-03  2018-06-04  2018-06-05  \\\n",
       "datatype                                                                 \n",
       "PRCP               6.9        2.00         6.4        4.10        0.00   \n",
       "TOTAL_PRCP         6.9        8.90        15.3       19.40       19.40   \n",
       "AVG_PRCP           6.9        4.45         5.1        4.85        3.88   \n",
       "\n",
       "date        2018-06-06  2018-06-07  2018-06-08  2018-06-09  2018-06-10  \n",
       "datatype                                                                \n",
       "PRCP          0.000000    0.000000       0.000    0.000000        0.30  \n",
       "TOTAL_PRCP   19.400000   19.400000      19.400   19.400000       19.70  \n",
       "AVG_PRCP      3.233333    2.771429       2.425    2.155556        1.97  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "central_park_weather.loc['2018-06'].assign(\n",
    "    TOTAL_PRCP=lambda x: x.PRCP.cumsum(),\n",
    "    AVG_PRCP=lambda x: x.PRCP.expanding().mean()\n",
    ").head(10)[['PRCP', 'TOTAL_PRCP', 'AVG_PRCP']].T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also use `agg()` to specify aggregations per column. Note that this works with NumPy functions as well. Here, we join the expanding calculations with the original results for comparison:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>AWND</th>\n",
       "      <th>AWND_expanding</th>\n",
       "      <th>PRCP</th>\n",
       "      <th>PRCP_expanding</th>\n",
       "      <th>TMAX</th>\n",
       "      <th>TMAX_expanding</th>\n",
       "      <th>TMIN</th>\n",
       "      <th>TMIN_expanding</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-10-01</th>\n",
       "      <td>0.9</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24.4</td>\n",
       "      <td>24.4</td>\n",
       "      <td>17.2</td>\n",
       "      <td>17.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-02</th>\n",
       "      <td>0.9</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>17.5</td>\n",
       "      <td>17.5</td>\n",
       "      <td>25.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>18.3</td>\n",
       "      <td>17.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-03</th>\n",
       "      <td>1.1</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17.5</td>\n",
       "      <td>23.3</td>\n",
       "      <td>25.0</td>\n",
       "      <td>17.2</td>\n",
       "      <td>17.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-04</th>\n",
       "      <td>0.4</td>\n",
       "      <td>0.825000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>18.5</td>\n",
       "      <td>24.4</td>\n",
       "      <td>25.0</td>\n",
       "      <td>16.1</td>\n",
       "      <td>16.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-05</th>\n",
       "      <td>1.6</td>\n",
       "      <td>0.980000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18.5</td>\n",
       "      <td>21.7</td>\n",
       "      <td>25.0</td>\n",
       "      <td>15.6</td>\n",
       "      <td>15.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-06</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18.5</td>\n",
       "      <td>20.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>17.2</td>\n",
       "      <td>15.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-07</th>\n",
       "      <td>1.1</td>\n",
       "      <td>0.928571</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18.5</td>\n",
       "      <td>26.1</td>\n",
       "      <td>26.1</td>\n",
       "      <td>19.4</td>\n",
       "      <td>15.6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            AWND  AWND_expanding  PRCP  PRCP_expanding  TMAX  TMAX_expanding  \\\n",
       "date                                                                           \n",
       "2018-10-01   0.9        0.900000   0.0             0.0  24.4            24.4   \n",
       "2018-10-02   0.9        0.900000  17.5            17.5  25.0            25.0   \n",
       "2018-10-03   1.1        0.966667   0.0            17.5  23.3            25.0   \n",
       "2018-10-04   0.4        0.825000   1.0            18.5  24.4            25.0   \n",
       "2018-10-05   1.6        0.980000   0.0            18.5  21.7            25.0   \n",
       "2018-10-06   0.5        0.900000   0.0            18.5  20.0            25.0   \n",
       "2018-10-07   1.1        0.928571   0.0            18.5  26.1            26.1   \n",
       "\n",
       "            TMIN  TMIN_expanding  \n",
       "date                              \n",
       "2018-10-01  17.2            17.2  \n",
       "2018-10-02  18.3            17.2  \n",
       "2018-10-03  17.2            17.2  \n",
       "2018-10-04  16.1            16.1  \n",
       "2018-10-05  15.6            15.6  \n",
       "2018-10-06  17.2            15.6  \n",
       "2018-10-07  19.4            15.6  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "central_park_weather['2018-10-01':'2018-10-07'].expanding().agg(\n",
    "    {'TMAX': np.max, 'TMIN': np.min, 'AWND': np.mean, 'PRCP': np.sum}\n",
    ").join(\n",
    "    central_park_weather[['TMAX', 'TMIN', 'AWND', 'PRCP']], \n",
    "    lsuffix='_expanding'\n",
    ").sort_index(axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pandas provides the `ewm()` method for exponentially weighted moving calculations. As we saw in chapter 1, we can use the exponentially weighted moving average to smooth the data. Let's compare the rolling mean to the exponentially weighted moving average with the maximum daily temperature. Note that `span` here is the periods to use:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>date</th>\n",
       "      <th>2018-09-29</th>\n",
       "      <th>2018-09-30</th>\n",
       "      <th>2018-10-01</th>\n",
       "      <th>2018-10-02</th>\n",
       "      <th>2018-10-03</th>\n",
       "      <th>2018-10-04</th>\n",
       "      <th>2018-10-05</th>\n",
       "      <th>2018-10-06</th>\n",
       "      <th>2018-10-07</th>\n",
       "      <th>2018-10-08</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datatype</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>TMAX</th>\n",
       "      <td>22.200000</td>\n",
       "      <td>21.100000</td>\n",
       "      <td>24.400000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>23.300000</td>\n",
       "      <td>24.400000</td>\n",
       "      <td>21.700000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>26.100000</td>\n",
       "      <td>23.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EWMA</th>\n",
       "      <td>24.410887</td>\n",
       "      <td>24.197281</td>\n",
       "      <td>24.210360</td>\n",
       "      <td>24.261304</td>\n",
       "      <td>24.199285</td>\n",
       "      <td>24.212234</td>\n",
       "      <td>24.050154</td>\n",
       "      <td>23.788854</td>\n",
       "      <td>23.937960</td>\n",
       "      <td>23.896802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AVG</th>\n",
       "      <td>24.723333</td>\n",
       "      <td>24.573333</td>\n",
       "      <td>24.533333</td>\n",
       "      <td>24.460000</td>\n",
       "      <td>24.163333</td>\n",
       "      <td>23.866667</td>\n",
       "      <td>23.533333</td>\n",
       "      <td>23.070000</td>\n",
       "      <td>23.143333</td>\n",
       "      <td>23.196667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "date      2018-09-29  2018-09-30  2018-10-01  2018-10-02  2018-10-03  \\\n",
       "datatype                                                               \n",
       "TMAX       22.200000   21.100000   24.400000   25.000000   23.300000   \n",
       "EWMA       24.410887   24.197281   24.210360   24.261304   24.199285   \n",
       "AVG        24.723333   24.573333   24.533333   24.460000   24.163333   \n",
       "\n",
       "date      2018-10-04  2018-10-05  2018-10-06  2018-10-07  2018-10-08  \n",
       "datatype                                                              \n",
       "TMAX       24.400000   21.700000   20.000000   26.100000   23.300000  \n",
       "EWMA       24.212234   24.050154   23.788854   23.937960   23.896802  \n",
       "AVG        23.866667   23.533333   23.070000   23.143333   23.196667  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "central_park_weather.assign(\n",
    "    AVG=lambda x: x.TMAX.rolling('30D').mean(),\n",
    "    EWMA=lambda x: x.TMAX.ewm(span=30).mean()\n",
    ").loc['2018-09-29':'2018-10-08', ['TMAX', 'EWMA', 'AVG']].T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Consult the [`understanding_window_calculations.ipynb`](./understanding_window_calculations.ipynb) notebook for interactive visualizations to help understand window calculations.*\n",
    "\n",
    "## Pipes\n",
    "Pipes are a way to streamline our `pandas` code and make it more readable and flexible. Using pipes, we can take a nested call like \n",
    "\n",
    "```python\n",
    "f(g(h(data), 20), x=True)\n",
    "```\n",
    "\n",
    "and turn it into something more readable:\n",
    "\n",
    "```python\n",
    "data.pipe(h)\\\n",
    "    .pipe(g, 20)\\\n",
    "    .pipe(f, x=True)\\\n",
    "```\n",
    "\n",
    "We can use pipes to apply any function that accepts our data as the first argument and pass in any additional arguments. This makes it easy to chain steps together regardless of whether they are methods or functions:\n",
    "\n",
    "We can pass any function that will accept the caller of `pipe()` as the first argument:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_info(df):\n",
    "    return '%d rows, %d columns and max closing Z-score was %d' % (*df.shape, df.close.max())\n",
    "\n",
    "get_info(fb.loc['2018-Q1'].apply(lambda x: (x - x.mean())/x.std()))\\\n",
    "    == fb.loc['2018-Q1'].apply(lambda x: (x - x.mean())/x.std()).pipe(get_info)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For example, passing `pd.DataFrame.rolling` to `pipe()` is equivalent to calling `rolling()` directly on the dataframe, except we have more flexiblity to change this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fb.pipe(pd.DataFrame.rolling, '20D').mean().equals(fb.rolling('20D').mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The pipe takes the function passed in and calls it with the object that called `pipe()` as the first argument. Positional and keyword arguments are passed down:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame.rolling(fb, '20D').mean().equals(fb.rolling('20D').mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use a pipe to make a function that we can use for all of our window calculation needs:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[0;31mSignature:\u001b[0m \u001b[0mwindow_calc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0magg_dict\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
       "\u001b[0;31mSource:\u001b[0m   \n",
       "\u001b[0;32mdef\u001b[0m \u001b[0mwindow_calc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0magg_dict\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0;34m\"\"\"\u001b[0m\n",
       "\u001b[0;34m    Run a window calculation of your choice on a `DataFrame` object.\u001b[0m\n",
       "\u001b[0;34m    \u001b[0m\n",
       "\u001b[0;34m    Parameters:\u001b[0m\n",
       "\u001b[0;34m        - df: The `DataFrame` object to run the calculation on.\u001b[0m\n",
       "\u001b[0;34m        - func: The window calculation method that takes `df`\u001b[0m\n",
       "\u001b[0;34m          as the first argument.\u001b[0m\n",
       "\u001b[0;34m        - agg_dict: Information to pass to `agg()`, could be a \u001b[0m\n",
       "\u001b[0;34m          dictionary mapping the columns to the aggregation \u001b[0m\n",
       "\u001b[0;34m          function to use, a string name for the function, \u001b[0m\n",
       "\u001b[0;34m          or the function itself.\u001b[0m\n",
       "\u001b[0;34m        - args: Positional arguments to pass to `func`.\u001b[0m\n",
       "\u001b[0;34m        - kwargs: Keyword arguments to pass to `func`.\u001b[0m\n",
       "\u001b[0;34m    \u001b[0m\n",
       "\u001b[0;34m    Returns:\u001b[0m\n",
       "\u001b[0;34m        A new `DataFrame` object.\u001b[0m\n",
       "\u001b[0;34m    \"\"\"\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0;32mreturn\u001b[0m \u001b[0mdf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpipe\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfunc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0magg\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0magg_dict\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
       "\u001b[0;31mFile:\u001b[0m      ~/Hands-On-Data-Analysis-with-Pandas-2nd-edition/ch_04/window_calc.py\n",
       "\u001b[0;31mType:\u001b[0m      function\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from window_calc import window_calc\n",
    "window_calc??"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use the same interface to calculate various window calculations now. Let's find the expanding median for the Facebook data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-01-02</th>\n",
       "      <td>177.68</td>\n",
       "      <td>181.580</td>\n",
       "      <td>177.5500</td>\n",
       "      <td>181.420</td>\n",
       "      <td>18151903.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-03</th>\n",
       "      <td>179.78</td>\n",
       "      <td>183.180</td>\n",
       "      <td>179.4400</td>\n",
       "      <td>183.045</td>\n",
       "      <td>17519233.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-04</th>\n",
       "      <td>181.88</td>\n",
       "      <td>184.780</td>\n",
       "      <td>181.3300</td>\n",
       "      <td>184.330</td>\n",
       "      <td>16886563.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-05</th>\n",
       "      <td>183.39</td>\n",
       "      <td>185.495</td>\n",
       "      <td>182.7148</td>\n",
       "      <td>184.500</td>\n",
       "      <td>15383729.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-08</th>\n",
       "      <td>184.90</td>\n",
       "      <td>186.210</td>\n",
       "      <td>184.0996</td>\n",
       "      <td>184.670</td>\n",
       "      <td>16886563.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              open     high       low    close      volume\n",
       "date                                                      \n",
       "2018-01-02  177.68  181.580  177.5500  181.420  18151903.0\n",
       "2018-01-03  179.78  183.180  179.4400  183.045  17519233.0\n",
       "2018-01-04  181.88  184.780  181.3300  184.330  16886563.0\n",
       "2018-01-05  183.39  185.495  182.7148  184.500  15383729.5\n",
       "2018-01-08  184.90  186.210  184.0996  184.670  16886563.0"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "window_calc(fb, pd.DataFrame.expanding, np.median).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using the exponentially weighted moving average requires we pass in a keyword argument:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-01-02</th>\n",
       "      <td>177.680000</td>\n",
       "      <td>181.580000</td>\n",
       "      <td>177.550000</td>\n",
       "      <td>181.420000</td>\n",
       "      <td>1.815190e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-03</th>\n",
       "      <td>180.480000</td>\n",
       "      <td>183.713333</td>\n",
       "      <td>180.070000</td>\n",
       "      <td>183.586667</td>\n",
       "      <td>1.730834e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-04</th>\n",
       "      <td>183.005714</td>\n",
       "      <td>185.140000</td>\n",
       "      <td>182.372629</td>\n",
       "      <td>184.011429</td>\n",
       "      <td>1.534980e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-05</th>\n",
       "      <td>184.384000</td>\n",
       "      <td>186.078667</td>\n",
       "      <td>183.736560</td>\n",
       "      <td>185.525333</td>\n",
       "      <td>1.440299e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-08</th>\n",
       "      <td>185.837419</td>\n",
       "      <td>187.534839</td>\n",
       "      <td>185.075110</td>\n",
       "      <td>186.947097</td>\n",
       "      <td>1.625679e+07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  open        high         low       close        volume\n",
       "date                                                                    \n",
       "2018-01-02  177.680000  181.580000  177.550000  181.420000  1.815190e+07\n",
       "2018-01-03  180.480000  183.713333  180.070000  183.586667  1.730834e+07\n",
       "2018-01-04  183.005714  185.140000  182.372629  184.011429  1.534980e+07\n",
       "2018-01-05  184.384000  186.078667  183.736560  185.525333  1.440299e+07\n",
       "2018-01-08  185.837419  187.534839  185.075110  186.947097  1.625679e+07"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "window_calc(fb, pd.DataFrame.ewm, 'mean', span=3).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With rolling calculations, we can pass in a positional argument for the window size:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "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>TMAX</th>\n",
       "      <th>TMIN</th>\n",
       "      <th>AWND</th>\n",
       "      <th>PRCP</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-10-01</th>\n",
       "      <td>24.4</td>\n",
       "      <td>17.2</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-02</th>\n",
       "      <td>25.0</td>\n",
       "      <td>17.2</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>17.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-03</th>\n",
       "      <td>25.0</td>\n",
       "      <td>17.2</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>17.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-04</th>\n",
       "      <td>25.0</td>\n",
       "      <td>16.1</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>18.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-10-05</th>\n",
       "      <td>24.4</td>\n",
       "      <td>15.6</td>\n",
       "      <td>1.033333</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            TMAX  TMIN      AWND  PRCP\n",
       "date                                  \n",
       "2018-10-01  24.4  17.2  0.900000   0.0\n",
       "2018-10-02  25.0  17.2  0.900000  17.5\n",
       "2018-10-03  25.0  17.2  0.966667  17.5\n",
       "2018-10-04  25.0  16.1  0.800000  18.5\n",
       "2018-10-05  24.4  15.6  1.033333   1.0"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "window_calc(\n",
    "    central_park_weather.loc['2018-10'], \n",
    "    pd.DataFrame.rolling, \n",
    "    {'TMAX': 'max', 'TMIN': 'min', 'AWND': 'mean', 'PRCP': 'sum'},\n",
    "    '3D'\n",
    ").head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<hr>\n",
    "<div style=\"overflow: hidden; margin-bottom: 10px;\">\n",
    "    <div style=\"float: left;\">\n",
    "         <a href=\"./1-querying_and_merging.ipynb\">\n",
    "            <button>&#8592; Previous Notebook</button>\n",
    "        </a>\n",
    "        <a href=\"./understanding_window_calculations.ipynb\">\n",
    "            <button>Understanding Window Calculations</button>\n",
    "        </a>\n",
    "    </div>\n",
    "    <div style=\"float: right;\">\n",
    "        <a href=\"./3-aggregations.ipynb\">\n",
    "            <button>Next Notebook &#8594;</button>\n",
    "        </a>\n",
    "    </div>\n",
    "</div>\n",
    "<hr>"
   ]
  }
 ],
 "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": 4
}