analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Live EstaTiDados\n",
    "\n",
    "## Técnicas de Pré-processamento\n",
    "\n",
    "### Prof. Dr. Neylson Crepalde\n",
    "\n",
    "Primeiro, vamos carregar as bibliotecas necessárias para o exercício de hoje:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already up-to-date: xgboost in /home/neylson/anaconda3/lib/python3.7/site-packages (1.1.1)\n",
      "Requirement already up-to-date: category_encoders in /home/neylson/anaconda3/lib/python3.7/site-packages (2.2.2)\n",
      "Collecting pmdarima\n",
      "  Downloading pmdarima-1.6.1-cp37-cp37m-manylinux1_x86_64.whl (1.5 MB)\n",
      "\u001b[K     |████████████████████████████████| 1.5 MB 1.2 MB/s eta 0:00:01\n",
      "\u001b[?25hRequirement already satisfied, skipping upgrade: numpy in /home/neylson/anaconda3/lib/python3.7/site-packages (from xgboost) (1.18.3)\n",
      "Requirement already satisfied, skipping upgrade: scipy in /home/neylson/anaconda3/lib/python3.7/site-packages (from xgboost) (1.4.1)\n",
      "Requirement already satisfied, skipping upgrade: scikit-learn>=0.20.0 in /home/neylson/anaconda3/lib/python3.7/site-packages (from category_encoders) (0.22.2)\n",
      "Requirement already satisfied, skipping upgrade: pandas>=0.21.1 in /home/neylson/anaconda3/lib/python3.7/site-packages (from category_encoders) (0.25.3)\n",
      "Requirement already satisfied, skipping upgrade: statsmodels>=0.9.0 in /home/neylson/anaconda3/lib/python3.7/site-packages (from category_encoders) (0.11.0)\n",
      "Requirement already satisfied, skipping upgrade: patsy>=0.5.1 in /home/neylson/anaconda3/lib/python3.7/site-packages (from category_encoders) (0.5.1)\n",
      "Requirement already satisfied, skipping upgrade: urllib3 in /home/neylson/anaconda3/lib/python3.7/site-packages (from pmdarima) (1.25.8)\n",
      "Requirement already satisfied, skipping upgrade: Cython<0.29.18,>=0.29 in /home/neylson/anaconda3/lib/python3.7/site-packages (from pmdarima) (0.29.15)\n",
      "Requirement already satisfied, skipping upgrade: joblib>=0.11 in /home/neylson/anaconda3/lib/python3.7/site-packages (from pmdarima) (0.14.1)\n",
      "Requirement already satisfied, skipping upgrade: pytz>=2017.2 in /home/neylson/anaconda3/lib/python3.7/site-packages (from pandas>=0.21.1->category_encoders) (2019.3)\n",
      "Requirement already satisfied, skipping upgrade: python-dateutil>=2.6.1 in /home/neylson/anaconda3/lib/python3.7/site-packages (from pandas>=0.21.1->category_encoders) (2.8.1)\n",
      "Requirement already satisfied, skipping upgrade: six in /home/neylson/anaconda3/lib/python3.7/site-packages (from patsy>=0.5.1->category_encoders) (1.14.0)\n",
      "Installing collected packages: pmdarima\n",
      "  Attempting uninstall: pmdarima\n",
      "    Found existing installation: pmdarima 1.6.0\n",
      "    Uninstalling pmdarima-1.6.0:\n",
      "      Successfully uninstalled pmdarima-1.6.0\n",
      "Successfully installed pmdarima-1.6.1\n"
     ]
    }
   ],
   "source": [
    "!pip install -U xgboost category_encoders pmdarima"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import re\n",
    "from sklearn.model_selection import train_test_split\n",
    "import category_encoders as ce\n",
    "from xgboost import XGBRegressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Vamos trabalhar com dados de *Supermarket sales*. Estes dados estão disponíveis em https://www.kaggle.com/aungpyaeap/supermarket-sales."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "bd = pd.read_csv('datasets_205965_451952_supermarket_sales - Sheet1.csv', parse_dates=['Date'])"
   ]
  },
  {
   "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>Invoice ID</th>\n",
       "      <th>Branch</th>\n",
       "      <th>City</th>\n",
       "      <th>Customer type</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Product line</th>\n",
       "      <th>Unit price</th>\n",
       "      <th>Quantity</th>\n",
       "      <th>Tax 5%</th>\n",
       "      <th>Total</th>\n",
       "      <th>Date</th>\n",
       "      <th>Time</th>\n",
       "      <th>Payment</th>\n",
       "      <th>cogs</th>\n",
       "      <th>gross margin percentage</th>\n",
       "      <th>gross income</th>\n",
       "      <th>Rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>750-67-8428</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Member</td>\n",
       "      <td>Female</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>74.69</td>\n",
       "      <td>7</td>\n",
       "      <td>26.1415</td>\n",
       "      <td>548.9715</td>\n",
       "      <td>2019-01-05</td>\n",
       "      <td>13:08</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>522.83</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>26.1415</td>\n",
       "      <td>9.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>226-31-3081</td>\n",
       "      <td>C</td>\n",
       "      <td>Naypyitaw</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Female</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>15.28</td>\n",
       "      <td>5</td>\n",
       "      <td>3.8200</td>\n",
       "      <td>80.2200</td>\n",
       "      <td>2019-03-08</td>\n",
       "      <td>10:29</td>\n",
       "      <td>Cash</td>\n",
       "      <td>76.40</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>3.8200</td>\n",
       "      <td>9.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>631-41-3108</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Male</td>\n",
       "      <td>Home and lifestyle</td>\n",
       "      <td>46.33</td>\n",
       "      <td>7</td>\n",
       "      <td>16.2155</td>\n",
       "      <td>340.5255</td>\n",
       "      <td>2019-03-03</td>\n",
       "      <td>13:23</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>324.31</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>16.2155</td>\n",
       "      <td>7.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>123-19-1176</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Member</td>\n",
       "      <td>Male</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>58.22</td>\n",
       "      <td>8</td>\n",
       "      <td>23.2880</td>\n",
       "      <td>489.0480</td>\n",
       "      <td>2019-01-27</td>\n",
       "      <td>20:33</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>465.76</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>23.2880</td>\n",
       "      <td>8.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>373-73-7910</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Male</td>\n",
       "      <td>Sports and travel</td>\n",
       "      <td>86.31</td>\n",
       "      <td>7</td>\n",
       "      <td>30.2085</td>\n",
       "      <td>634.3785</td>\n",
       "      <td>2019-02-08</td>\n",
       "      <td>10:37</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>604.17</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>30.2085</td>\n",
       "      <td>5.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>233-67-5758</td>\n",
       "      <td>C</td>\n",
       "      <td>Naypyitaw</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Male</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>40.35</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0175</td>\n",
       "      <td>42.3675</td>\n",
       "      <td>2019-01-29</td>\n",
       "      <td>13:46</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>40.35</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>2.0175</td>\n",
       "      <td>6.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>303-96-2227</td>\n",
       "      <td>B</td>\n",
       "      <td>Mandalay</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Female</td>\n",
       "      <td>Home and lifestyle</td>\n",
       "      <td>97.38</td>\n",
       "      <td>10</td>\n",
       "      <td>48.6900</td>\n",
       "      <td>1022.4900</td>\n",
       "      <td>2019-03-02</td>\n",
       "      <td>17:16</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>973.80</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>48.6900</td>\n",
       "      <td>4.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>727-02-1313</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Member</td>\n",
       "      <td>Male</td>\n",
       "      <td>Food and beverages</td>\n",
       "      <td>31.84</td>\n",
       "      <td>1</td>\n",
       "      <td>1.5920</td>\n",
       "      <td>33.4320</td>\n",
       "      <td>2019-02-09</td>\n",
       "      <td>13:22</td>\n",
       "      <td>Cash</td>\n",
       "      <td>31.84</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>1.5920</td>\n",
       "      <td>7.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>347-56-2442</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Male</td>\n",
       "      <td>Home and lifestyle</td>\n",
       "      <td>65.82</td>\n",
       "      <td>1</td>\n",
       "      <td>3.2910</td>\n",
       "      <td>69.1110</td>\n",
       "      <td>2019-02-22</td>\n",
       "      <td>15:33</td>\n",
       "      <td>Cash</td>\n",
       "      <td>65.82</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>3.2910</td>\n",
       "      <td>4.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>849-09-3807</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Member</td>\n",
       "      <td>Female</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>88.34</td>\n",
       "      <td>7</td>\n",
       "      <td>30.9190</td>\n",
       "      <td>649.2990</td>\n",
       "      <td>2019-02-18</td>\n",
       "      <td>13:28</td>\n",
       "      <td>Cash</td>\n",
       "      <td>618.38</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>30.9190</td>\n",
       "      <td>6.6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Invoice ID Branch       City Customer type  Gender  \\\n",
       "0    750-67-8428      A     Yangon        Member  Female   \n",
       "1    226-31-3081      C  Naypyitaw        Normal  Female   \n",
       "2    631-41-3108      A     Yangon        Normal    Male   \n",
       "3    123-19-1176      A     Yangon        Member    Male   \n",
       "4    373-73-7910      A     Yangon        Normal    Male   \n",
       "..           ...    ...        ...           ...     ...   \n",
       "995  233-67-5758      C  Naypyitaw        Normal    Male   \n",
       "996  303-96-2227      B   Mandalay        Normal  Female   \n",
       "997  727-02-1313      A     Yangon        Member    Male   \n",
       "998  347-56-2442      A     Yangon        Normal    Male   \n",
       "999  849-09-3807      A     Yangon        Member  Female   \n",
       "\n",
       "               Product line  Unit price  Quantity   Tax 5%      Total  \\\n",
       "0         Health and beauty       74.69         7  26.1415   548.9715   \n",
       "1    Electronic accessories       15.28         5   3.8200    80.2200   \n",
       "2        Home and lifestyle       46.33         7  16.2155   340.5255   \n",
       "3         Health and beauty       58.22         8  23.2880   489.0480   \n",
       "4         Sports and travel       86.31         7  30.2085   634.3785   \n",
       "..                      ...         ...       ...      ...        ...   \n",
       "995       Health and beauty       40.35         1   2.0175    42.3675   \n",
       "996      Home and lifestyle       97.38        10  48.6900  1022.4900   \n",
       "997      Food and beverages       31.84         1   1.5920    33.4320   \n",
       "998      Home and lifestyle       65.82         1   3.2910    69.1110   \n",
       "999     Fashion accessories       88.34         7  30.9190   649.2990   \n",
       "\n",
       "          Date   Time      Payment    cogs  gross margin percentage  \\\n",
       "0   2019-01-05  13:08      Ewallet  522.83                 4.761905   \n",
       "1   2019-03-08  10:29         Cash   76.40                 4.761905   \n",
       "2   2019-03-03  13:23  Credit card  324.31                 4.761905   \n",
       "3   2019-01-27  20:33      Ewallet  465.76                 4.761905   \n",
       "4   2019-02-08  10:37      Ewallet  604.17                 4.761905   \n",
       "..         ...    ...          ...     ...                      ...   \n",
       "995 2019-01-29  13:46      Ewallet   40.35                 4.761905   \n",
       "996 2019-03-02  17:16      Ewallet  973.80                 4.761905   \n",
       "997 2019-02-09  13:22         Cash   31.84                 4.761905   \n",
       "998 2019-02-22  15:33         Cash   65.82                 4.761905   \n",
       "999 2019-02-18  13:28         Cash  618.38                 4.761905   \n",
       "\n",
       "     gross income  Rating  \n",
       "0         26.1415     9.1  \n",
       "1          3.8200     9.6  \n",
       "2         16.2155     7.4  \n",
       "3         23.2880     8.4  \n",
       "4         30.2085     5.3  \n",
       "..            ...     ...  \n",
       "995        2.0175     6.2  \n",
       "996       48.6900     4.4  \n",
       "997        1.5920     7.7  \n",
       "998        3.2910     4.1  \n",
       "999       30.9190     6.6  \n",
       "\n",
       "[1000 rows x 17 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Primeiro, vamos consertar os nomes das colunas do dataset. Nomes de colunas com espaços, letras maiúsculas e caracteres especiais podem gerar erros em algumas situações."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fix_column_name(columns):\n",
    "    res = [col.lower() for col in columns]\n",
    "    res = [re.sub(' ', '_', col) for col in res]\n",
    "    res = [re.sub('%', 'pc', col) for col in res]\n",
    "    return res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "bd.columns = fix_column_name(bd.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Rápida análise exploratória\n",
    "\n",
    "Sabemos que\n",
    "\n",
    "- São 3 supermercados;\n",
    "- cada um com vários tipos de produto"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>branch</th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>product_line</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Electronic accessories</th>\n",
       "      <td>60</td>\n",
       "      <td>55</td>\n",
       "      <td>55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fashion accessories</th>\n",
       "      <td>51</td>\n",
       "      <td>62</td>\n",
       "      <td>65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Food and beverages</th>\n",
       "      <td>58</td>\n",
       "      <td>50</td>\n",
       "      <td>66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Health and beauty</th>\n",
       "      <td>47</td>\n",
       "      <td>53</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Home and lifestyle</th>\n",
       "      <td>65</td>\n",
       "      <td>50</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sports and travel</th>\n",
       "      <td>59</td>\n",
       "      <td>62</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "branch                   A   B   C\n",
       "product_line                      \n",
       "Electronic accessories  60  55  55\n",
       "Fashion accessories     51  62  65\n",
       "Food and beverages      58  50  66\n",
       "Health and beauty       47  53  52\n",
       "Home and lifestyle      65  50  45\n",
       "Sports and travel       59  62  45"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(bd.product_line, bd.branch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>invoice_id</th>\n",
       "      <th>branch</th>\n",
       "      <th>city</th>\n",
       "      <th>customer_type</th>\n",
       "      <th>gender</th>\n",
       "      <th>product_line</th>\n",
       "      <th>unit_price</th>\n",
       "      <th>quantity</th>\n",
       "      <th>tax_5pc</th>\n",
       "      <th>total</th>\n",
       "      <th>date</th>\n",
       "      <th>time</th>\n",
       "      <th>payment</th>\n",
       "      <th>cogs</th>\n",
       "      <th>gross_margin_percentage</th>\n",
       "      <th>gross_income</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>750-67-8428</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Member</td>\n",
       "      <td>Female</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>74.69</td>\n",
       "      <td>7</td>\n",
       "      <td>26.1415</td>\n",
       "      <td>548.9715</td>\n",
       "      <td>2019-01-05</td>\n",
       "      <td>13:08</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>522.83</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>26.1415</td>\n",
       "      <td>9.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>631-41-3108</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Male</td>\n",
       "      <td>Home and lifestyle</td>\n",
       "      <td>46.33</td>\n",
       "      <td>7</td>\n",
       "      <td>16.2155</td>\n",
       "      <td>340.5255</td>\n",
       "      <td>2019-03-03</td>\n",
       "      <td>13:23</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>324.31</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>16.2155</td>\n",
       "      <td>7.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>123-19-1176</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Member</td>\n",
       "      <td>Male</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>58.22</td>\n",
       "      <td>8</td>\n",
       "      <td>23.2880</td>\n",
       "      <td>489.0480</td>\n",
       "      <td>2019-01-27</td>\n",
       "      <td>20:33</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>465.76</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>23.2880</td>\n",
       "      <td>8.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>373-73-7910</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Male</td>\n",
       "      <td>Sports and travel</td>\n",
       "      <td>86.31</td>\n",
       "      <td>7</td>\n",
       "      <td>30.2085</td>\n",
       "      <td>634.3785</td>\n",
       "      <td>2019-02-08</td>\n",
       "      <td>10:37</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>604.17</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>30.2085</td>\n",
       "      <td>5.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>355-53-5943</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Member</td>\n",
       "      <td>Female</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>68.84</td>\n",
       "      <td>6</td>\n",
       "      <td>20.6520</td>\n",
       "      <td>433.6920</td>\n",
       "      <td>2019-02-25</td>\n",
       "      <td>14:36</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>413.04</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>20.6520</td>\n",
       "      <td>5.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>990</th>\n",
       "      <td>886-18-2897</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Female</td>\n",
       "      <td>Food and beverages</td>\n",
       "      <td>56.56</td>\n",
       "      <td>5</td>\n",
       "      <td>14.1400</td>\n",
       "      <td>296.9400</td>\n",
       "      <td>2019-03-22</td>\n",
       "      <td>19:06</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>282.80</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>14.1400</td>\n",
       "      <td>4.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>992</th>\n",
       "      <td>745-74-0715</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Male</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>58.03</td>\n",
       "      <td>2</td>\n",
       "      <td>5.8030</td>\n",
       "      <td>121.8630</td>\n",
       "      <td>2019-03-10</td>\n",
       "      <td>20:46</td>\n",
       "      <td>Ewallet</td>\n",
       "      <td>116.06</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>5.8030</td>\n",
       "      <td>8.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>727-02-1313</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Member</td>\n",
       "      <td>Male</td>\n",
       "      <td>Food and beverages</td>\n",
       "      <td>31.84</td>\n",
       "      <td>1</td>\n",
       "      <td>1.5920</td>\n",
       "      <td>33.4320</td>\n",
       "      <td>2019-02-09</td>\n",
       "      <td>13:22</td>\n",
       "      <td>Cash</td>\n",
       "      <td>31.84</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>1.5920</td>\n",
       "      <td>7.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>347-56-2442</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Male</td>\n",
       "      <td>Home and lifestyle</td>\n",
       "      <td>65.82</td>\n",
       "      <td>1</td>\n",
       "      <td>3.2910</td>\n",
       "      <td>69.1110</td>\n",
       "      <td>2019-02-22</td>\n",
       "      <td>15:33</td>\n",
       "      <td>Cash</td>\n",
       "      <td>65.82</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>3.2910</td>\n",
       "      <td>4.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>849-09-3807</td>\n",
       "      <td>A</td>\n",
       "      <td>Yangon</td>\n",
       "      <td>Member</td>\n",
       "      <td>Female</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>88.34</td>\n",
       "      <td>7</td>\n",
       "      <td>30.9190</td>\n",
       "      <td>649.2990</td>\n",
       "      <td>2019-02-18</td>\n",
       "      <td>13:28</td>\n",
       "      <td>Cash</td>\n",
       "      <td>618.38</td>\n",
       "      <td>4.761905</td>\n",
       "      <td>30.9190</td>\n",
       "      <td>6.6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>340 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      invoice_id branch    city customer_type  gender            product_line  \\\n",
       "0    750-67-8428      A  Yangon        Member  Female       Health and beauty   \n",
       "2    631-41-3108      A  Yangon        Normal    Male      Home and lifestyle   \n",
       "3    123-19-1176      A  Yangon        Member    Male       Health and beauty   \n",
       "4    373-73-7910      A  Yangon        Normal    Male       Sports and travel   \n",
       "6    355-53-5943      A  Yangon        Member  Female  Electronic accessories   \n",
       "..           ...    ...     ...           ...     ...                     ...   \n",
       "990  886-18-2897      A  Yangon        Normal  Female      Food and beverages   \n",
       "992  745-74-0715      A  Yangon        Normal    Male  Electronic accessories   \n",
       "997  727-02-1313      A  Yangon        Member    Male      Food and beverages   \n",
       "998  347-56-2442      A  Yangon        Normal    Male      Home and lifestyle   \n",
       "999  849-09-3807      A  Yangon        Member  Female     Fashion accessories   \n",
       "\n",
       "     unit_price  quantity  tax_5pc     total       date   time      payment  \\\n",
       "0         74.69         7  26.1415  548.9715 2019-01-05  13:08      Ewallet   \n",
       "2         46.33         7  16.2155  340.5255 2019-03-03  13:23  Credit card   \n",
       "3         58.22         8  23.2880  489.0480 2019-01-27  20:33      Ewallet   \n",
       "4         86.31         7  30.2085  634.3785 2019-02-08  10:37      Ewallet   \n",
       "6         68.84         6  20.6520  433.6920 2019-02-25  14:36      Ewallet   \n",
       "..          ...       ...      ...       ...        ...    ...          ...   \n",
       "990       56.56         5  14.1400  296.9400 2019-03-22  19:06  Credit card   \n",
       "992       58.03         2   5.8030  121.8630 2019-03-10  20:46      Ewallet   \n",
       "997       31.84         1   1.5920   33.4320 2019-02-09  13:22         Cash   \n",
       "998       65.82         1   3.2910   69.1110 2019-02-22  15:33         Cash   \n",
       "999       88.34         7  30.9190  649.2990 2019-02-18  13:28         Cash   \n",
       "\n",
       "       cogs  gross_margin_percentage  gross_income  rating  \n",
       "0    522.83                 4.761905       26.1415     9.1  \n",
       "2    324.31                 4.761905       16.2155     7.4  \n",
       "3    465.76                 4.761905       23.2880     8.4  \n",
       "4    604.17                 4.761905       30.2085     5.3  \n",
       "6    413.04                 4.761905       20.6520     5.8  \n",
       "..      ...                      ...           ...     ...  \n",
       "990  282.80                 4.761905       14.1400     4.5  \n",
       "992  116.06                 4.761905        5.8030     8.8  \n",
       "997   31.84                 4.761905        1.5920     7.7  \n",
       "998   65.82                 4.761905        3.2910     4.1  \n",
       "999  618.38                 4.761905       30.9190     6.6  \n",
       "\n",
       "[340 rows x 17 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bd.loc[bd.branch == 'A']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Home and lifestyle        65\n",
       "Electronic accessories    60\n",
       "Sports and travel         59\n",
       "Food and beverages        58\n",
       "Fashion accessories       51\n",
       "Health and beauty         47\n",
       "Name: product_line, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bd.loc[bd.branch == 'A', 'product_line'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Construindo a série de receita consolidada para cada supermercado e tipo de produto"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "cons = bd.groupby(['branch', 'product_line', 'date'])\\\n",
    "    .agg({'gross_income': 'sum'})\\\n",
    "    .reset_index()"
   ]
  },
  {
   "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>branch</th>\n",
       "      <th>product_line</th>\n",
       "      <th>date</th>\n",
       "      <th>gross_income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-01</td>\n",
       "      <td>37.1100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-05</td>\n",
       "      <td>32.8580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-07</td>\n",
       "      <td>15.5440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-08</td>\n",
       "      <td>21.9845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-12</td>\n",
       "      <td>44.3350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-13</td>\n",
       "      <td>7.0380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-15</td>\n",
       "      <td>27.3960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-18</td>\n",
       "      <td>6.5775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-19</td>\n",
       "      <td>1.7910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-23</td>\n",
       "      <td>19.8180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-24</td>\n",
       "      <td>23.6720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-25</td>\n",
       "      <td>11.8035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-26</td>\n",
       "      <td>18.0915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-27</td>\n",
       "      <td>3.6630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-28</td>\n",
       "      <td>9.6720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-29</td>\n",
       "      <td>1.9300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-30</td>\n",
       "      <td>14.0670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-01-31</td>\n",
       "      <td>7.4715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-04</td>\n",
       "      <td>38.5275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-05</td>\n",
       "      <td>2.0890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-07</td>\n",
       "      <td>1.4480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-12</td>\n",
       "      <td>11.7375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-13</td>\n",
       "      <td>10.2910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-14</td>\n",
       "      <td>39.8240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-17</td>\n",
       "      <td>8.6400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-18</td>\n",
       "      <td>3.1240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-19</td>\n",
       "      <td>31.3110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-21</td>\n",
       "      <td>14.1960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-22</td>\n",
       "      <td>8.7100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-24</td>\n",
       "      <td>20.1300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-25</td>\n",
       "      <td>20.6520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-26</td>\n",
       "      <td>4.6610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-02-27</td>\n",
       "      <td>32.4100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-01</td>\n",
       "      <td>22.7400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-05</td>\n",
       "      <td>11.7465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-06</td>\n",
       "      <td>9.6750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-08</td>\n",
       "      <td>4.8580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-10</td>\n",
       "      <td>5.8030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-14</td>\n",
       "      <td>37.1960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-18</td>\n",
       "      <td>10.2380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-19</td>\n",
       "      <td>41.9545</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-20</td>\n",
       "      <td>22.3530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-21</td>\n",
       "      <td>47.0925</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-22</td>\n",
       "      <td>4.7575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-23</td>\n",
       "      <td>36.7350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-25</td>\n",
       "      <td>14.0340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-26</td>\n",
       "      <td>25.2700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-28</td>\n",
       "      <td>9.1070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "      <td>2019-03-30</td>\n",
       "      <td>16.1100</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   branch            product_line       date  gross_income\n",
       "0       A  Electronic accessories 2019-01-01       37.1100\n",
       "1       A  Electronic accessories 2019-01-05       32.8580\n",
       "2       A  Electronic accessories 2019-01-07       15.5440\n",
       "3       A  Electronic accessories 2019-01-08       21.9845\n",
       "4       A  Electronic accessories 2019-01-12       44.3350\n",
       "5       A  Electronic accessories 2019-01-13        7.0380\n",
       "6       A  Electronic accessories 2019-01-15       27.3960\n",
       "7       A  Electronic accessories 2019-01-18        6.5775\n",
       "8       A  Electronic accessories 2019-01-19        1.7910\n",
       "9       A  Electronic accessories 2019-01-23       19.8180\n",
       "10      A  Electronic accessories 2019-01-24       23.6720\n",
       "11      A  Electronic accessories 2019-01-25       11.8035\n",
       "12      A  Electronic accessories 2019-01-26       18.0915\n",
       "13      A  Electronic accessories 2019-01-27        3.6630\n",
       "14      A  Electronic accessories 2019-01-28        9.6720\n",
       "15      A  Electronic accessories 2019-01-29        1.9300\n",
       "16      A  Electronic accessories 2019-01-30       14.0670\n",
       "17      A  Electronic accessories 2019-01-31        7.4715\n",
       "18      A  Electronic accessories 2019-02-04       38.5275\n",
       "19      A  Electronic accessories 2019-02-05        2.0890\n",
       "20      A  Electronic accessories 2019-02-07        1.4480\n",
       "21      A  Electronic accessories 2019-02-12       11.7375\n",
       "22      A  Electronic accessories 2019-02-13       10.2910\n",
       "23      A  Electronic accessories 2019-02-14       39.8240\n",
       "24      A  Electronic accessories 2019-02-17        8.6400\n",
       "25      A  Electronic accessories 2019-02-18        3.1240\n",
       "26      A  Electronic accessories 2019-02-19       31.3110\n",
       "27      A  Electronic accessories 2019-02-21       14.1960\n",
       "28      A  Electronic accessories 2019-02-22        8.7100\n",
       "29      A  Electronic accessories 2019-02-24       20.1300\n",
       "30      A  Electronic accessories 2019-02-25       20.6520\n",
       "31      A  Electronic accessories 2019-02-26        4.6610\n",
       "32      A  Electronic accessories 2019-02-27       32.4100\n",
       "33      A  Electronic accessories 2019-03-01       22.7400\n",
       "34      A  Electronic accessories 2019-03-05       11.7465\n",
       "35      A  Electronic accessories 2019-03-06        9.6750\n",
       "36      A  Electronic accessories 2019-03-08        4.8580\n",
       "37      A  Electronic accessories 2019-03-10        5.8030\n",
       "38      A  Electronic accessories 2019-03-14       37.1960\n",
       "39      A  Electronic accessories 2019-03-18       10.2380\n",
       "40      A  Electronic accessories 2019-03-19       41.9545\n",
       "41      A  Electronic accessories 2019-03-20       22.3530\n",
       "42      A  Electronic accessories 2019-03-21       47.0925\n",
       "43      A  Electronic accessories 2019-03-22        4.7575\n",
       "44      A  Electronic accessories 2019-03-23       36.7350\n",
       "45      A  Electronic accessories 2019-03-25       14.0340\n",
       "46      A  Electronic accessories 2019-03-26       25.2700\n",
       "47      A  Electronic accessories 2019-03-28        9.1070\n",
       "48      A  Electronic accessories 2019-03-30       16.1100"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "br = 'A'\n",
    "pl = 'Electronic accessories'\n",
    "\n",
    "cons.loc[(cons.branch == br) & (cons.product_line == pl)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = cons.loc[(cons.branch == br) & (cons.product_line == pl)]\n",
    "dt_inicio = a.date.min()\n",
    "dt_final = a.date.max()\n",
    "ids = pd.date_range(dt_inicio, dt_final)\n",
    "a.index = pd.DatetimeIndex(a.date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = a[['gross_income']].reindex(ids, fill_value=0)\n",
    "a['branch'] = br\n",
    "a['product_line'] = pl\n",
    "a = a.reset_index().rename(columns={'index':'date'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>gross_income</th>\n",
       "      <th>branch</th>\n",
       "      <th>product_line</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-01-01</td>\n",
       "      <td>37.110</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-01-02</td>\n",
       "      <td>0.000</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-01-03</td>\n",
       "      <td>0.000</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-01-04</td>\n",
       "      <td>0.000</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-01-05</td>\n",
       "      <td>32.858</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>2019-03-26</td>\n",
       "      <td>25.270</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>2019-03-27</td>\n",
       "      <td>0.000</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>2019-03-28</td>\n",
       "      <td>9.107</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>2019-03-29</td>\n",
       "      <td>0.000</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88</th>\n",
       "      <td>2019-03-30</td>\n",
       "      <td>16.110</td>\n",
       "      <td>A</td>\n",
       "      <td>Electronic accessories</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>89 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         date  gross_income branch            product_line\n",
       "0  2019-01-01        37.110      A  Electronic accessories\n",
       "1  2019-01-02         0.000      A  Electronic accessories\n",
       "2  2019-01-03         0.000      A  Electronic accessories\n",
       "3  2019-01-04         0.000      A  Electronic accessories\n",
       "4  2019-01-05        32.858      A  Electronic accessories\n",
       "..        ...           ...    ...                     ...\n",
       "84 2019-03-26        25.270      A  Electronic accessories\n",
       "85 2019-03-27         0.000      A  Electronic accessories\n",
       "86 2019-03-28         9.107      A  Electronic accessories\n",
       "87 2019-03-29         0.000      A  Electronic accessories\n",
       "88 2019-03-30        16.110      A  Electronic accessories\n",
       "\n",
       "[89 rows x 4 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encapsulando tudo isso numa função e \"arrumando\" todas as séries:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_br_pl(df):\n",
    "    brs = df.branch.unique()\n",
    "    pls = df.product_line.unique()\n",
    "    \n",
    "    cons = df.groupby(['branch', 'product_line', 'date'])\\\n",
    "        .agg({'gross_income': 'sum'})\\\n",
    "        .reset_index()\n",
    "    \n",
    "    lista_dfs=[]\n",
    "    \n",
    "    for br in brs:\n",
    "        for pl in pls:\n",
    "            a = cons.loc[(cons.branch == br) & (cons.product_line == pl)]\n",
    "            dt_inicio = a.date.min()\n",
    "            dt_final = a.date.max()\n",
    "            ids = pd.date_range(dt_inicio, dt_final)\n",
    "            a.index = pd.DatetimeIndex(a.date)\n",
    "            a = a[['gross_income']].reindex(ids, fill_value=0)\n",
    "            a['branch'] = br\n",
    "            a['product_line'] = pl\n",
    "            a = a.reset_index().rename(columns={'index':'date'})\n",
    "            lista_dfs.append(a)\n",
    "    \n",
    "    res = pd.concat(lista_dfs, ignore_index=True)\n",
    "    return res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>gross_income</th>\n",
       "      <th>branch</th>\n",
       "      <th>product_line</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-01-03</td>\n",
       "      <td>27.8660</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-01-04</td>\n",
       "      <td>5.2425</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-01-05</td>\n",
       "      <td>26.1415</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-01-06</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-01-07</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1548</th>\n",
       "      <td>2019-03-24</td>\n",
       "      <td>17.6250</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1549</th>\n",
       "      <td>2019-03-25</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1550</th>\n",
       "      <td>2019-03-26</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1551</th>\n",
       "      <td>2019-03-27</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1552</th>\n",
       "      <td>2019-03-28</td>\n",
       "      <td>4.5555</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1553 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           date  gross_income branch         product_line\n",
       "0    2019-01-03       27.8660      A    Health and beauty\n",
       "1    2019-01-04        5.2425      A    Health and beauty\n",
       "2    2019-01-05       26.1415      A    Health and beauty\n",
       "3    2019-01-06        0.0000      A    Health and beauty\n",
       "4    2019-01-07        0.0000      A    Health and beauty\n",
       "...         ...           ...    ...                  ...\n",
       "1548 2019-03-24       17.6250      B  Fashion accessories\n",
       "1549 2019-03-25        0.0000      B  Fashion accessories\n",
       "1550 2019-03-26        0.0000      B  Fashion accessories\n",
       "1551 2019-03-27        0.0000      B  Fashion accessories\n",
       "1552 2019-03-28        4.5555      B  Fashion accessories\n",
       "\n",
       "[1553 rows x 4 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "consolidado_br_pl = make_br_pl(bd)\n",
    "consolidado_br_pl"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A partir daqui, há dois caminhos para modelagem. Um seria modelar cada série separadamente. O próximo passo do pré-processamento seria separar cada série e fazer a modelagem."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Outro caminho seria modelar tudo junto entrando as colunas de `branch` e `product_line` como features do modelo. O que precisaríamos fazer caso optássemos pela segunda opção?\n",
    "\n",
    "## Trabalhando com variáveis categóricas"
   ]
  },
  {
   "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>date</th>\n",
       "      <th>gross_income</th>\n",
       "      <th>branch_A</th>\n",
       "      <th>branch_B</th>\n",
       "      <th>branch_C</th>\n",
       "      <th>product_line_Electronic accessories</th>\n",
       "      <th>product_line_Fashion accessories</th>\n",
       "      <th>product_line_Food and beverages</th>\n",
       "      <th>product_line_Health and beauty</th>\n",
       "      <th>product_line_Home and lifestyle</th>\n",
       "      <th>product_line_Sports and travel</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-01-03</td>\n",
       "      <td>27.8660</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-01-04</td>\n",
       "      <td>5.2425</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-01-05</td>\n",
       "      <td>26.1415</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-01-06</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-01-07</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1548</th>\n",
       "      <td>2019-03-24</td>\n",
       "      <td>17.6250</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1549</th>\n",
       "      <td>2019-03-25</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1550</th>\n",
       "      <td>2019-03-26</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1551</th>\n",
       "      <td>2019-03-27</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1552</th>\n",
       "      <td>2019-03-28</td>\n",
       "      <td>4.5555</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1553 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           date  gross_income  branch_A  branch_B  branch_C  \\\n",
       "0    2019-01-03       27.8660         1         0         0   \n",
       "1    2019-01-04        5.2425         1         0         0   \n",
       "2    2019-01-05       26.1415         1         0         0   \n",
       "3    2019-01-06        0.0000         1         0         0   \n",
       "4    2019-01-07        0.0000         1         0         0   \n",
       "...         ...           ...       ...       ...       ...   \n",
       "1548 2019-03-24       17.6250         0         1         0   \n",
       "1549 2019-03-25        0.0000         0         1         0   \n",
       "1550 2019-03-26        0.0000         0         1         0   \n",
       "1551 2019-03-27        0.0000         0         1         0   \n",
       "1552 2019-03-28        4.5555         0         1         0   \n",
       "\n",
       "      product_line_Electronic accessories  product_line_Fashion accessories  \\\n",
       "0                                       0                                 0   \n",
       "1                                       0                                 0   \n",
       "2                                       0                                 0   \n",
       "3                                       0                                 0   \n",
       "4                                       0                                 0   \n",
       "...                                   ...                               ...   \n",
       "1548                                    0                                 1   \n",
       "1549                                    0                                 1   \n",
       "1550                                    0                                 1   \n",
       "1551                                    0                                 1   \n",
       "1552                                    0                                 1   \n",
       "\n",
       "      product_line_Food and beverages  product_line_Health and beauty  \\\n",
       "0                                   0                               1   \n",
       "1                                   0                               1   \n",
       "2                                   0                               1   \n",
       "3                                   0                               1   \n",
       "4                                   0                               1   \n",
       "...                               ...                             ...   \n",
       "1548                                0                               0   \n",
       "1549                                0                               0   \n",
       "1550                                0                               0   \n",
       "1551                                0                               0   \n",
       "1552                                0                               0   \n",
       "\n",
       "      product_line_Home and lifestyle  product_line_Sports and travel  \n",
       "0                                   0                               0  \n",
       "1                                   0                               0  \n",
       "2                                   0                               0  \n",
       "3                                   0                               0  \n",
       "4                                   0                               0  \n",
       "...                               ...                             ...  \n",
       "1548                                0                               0  \n",
       "1549                                0                               0  \n",
       "1550                                0                               0  \n",
       "1551                                0                               0  \n",
       "1552                                0                               0  \n",
       "\n",
       "[1553 rows x 11 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# One Hot Encoding\n",
    "pd.get_dummies(consolidado_br_pl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Divide o dataset em treino e teste\n",
    "X_train = []\n",
    "X_test = []\n",
    "y_train = []\n",
    "y_test = []\n",
    "\n",
    "for br in consolidado_br_pl.branch.unique():\n",
    "    for pl in consolidado_br_pl.product_line.unique():\n",
    "        intermediario = consolidado_br_pl.loc[(consolidado_br_pl.branch == br) &\n",
    "                                              (consolidado_br_pl.product_line == pl)].sort_values('date')\n",
    "        X_train.append(intermediario.drop(columns='gross_income').iloc[:-15])\n",
    "        X_test.append(intermediario.drop(columns='gross_income').iloc[-15:])\n",
    "        y_train.append(intermediario.gross_income.iloc[:-15])\n",
    "        y_test.append(intermediario.gross_income.iloc[-15:])\n",
    "\n",
    "X_train = pd.concat(X_train, ignore_index=True)\n",
    "X_test = pd.concat(X_test, ignore_index=True)\n",
    "y_train = pd.concat(y_train, ignore_index=True)\n",
    "y_test = pd.concat(y_test, ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>branch</th>\n",
       "      <th>product_line</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-01-03</td>\n",
       "      <td>9.424490</td>\n",
       "      <td>9.083908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-01-04</td>\n",
       "      <td>9.424490</td>\n",
       "      <td>9.083908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-01-05</td>\n",
       "      <td>9.424490</td>\n",
       "      <td>9.083908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-01-06</td>\n",
       "      <td>9.424490</td>\n",
       "      <td>9.083908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-01-07</td>\n",
       "      <td>9.424490</td>\n",
       "      <td>9.083908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1278</th>\n",
       "      <td>2019-03-09</td>\n",
       "      <td>10.137406</td>\n",
       "      <td>10.150051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1279</th>\n",
       "      <td>2019-03-10</td>\n",
       "      <td>10.137406</td>\n",
       "      <td>10.150051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1280</th>\n",
       "      <td>2019-03-11</td>\n",
       "      <td>10.137406</td>\n",
       "      <td>10.150051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1281</th>\n",
       "      <td>2019-03-12</td>\n",
       "      <td>10.137406</td>\n",
       "      <td>10.150051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1282</th>\n",
       "      <td>2019-03-13</td>\n",
       "      <td>10.137406</td>\n",
       "      <td>10.150051</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1283 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           date     branch  product_line\n",
       "0    2019-01-03   9.424490      9.083908\n",
       "1    2019-01-04   9.424490      9.083908\n",
       "2    2019-01-05   9.424490      9.083908\n",
       "3    2019-01-06   9.424490      9.083908\n",
       "4    2019-01-07   9.424490      9.083908\n",
       "...         ...        ...           ...\n",
       "1278 2019-03-09  10.137406     10.150051\n",
       "1279 2019-03-10  10.137406     10.150051\n",
       "1280 2019-03-11  10.137406     10.150051\n",
       "1281 2019-03-12  10.137406     10.150051\n",
       "1282 2019-03-13  10.137406     10.150051\n",
       "\n",
       "[1283 rows x 3 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Target encoder\n",
    "encoder = ce.TargetEncoder()\n",
    "encoder.fit(X_train, y_train)\n",
    "X_train_enc = encoder.transform(X_train)\n",
    "X_train_enc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_enc['year'] = X_train_enc.date.dt.year\n",
    "X_train_enc['month'] = X_train_enc.date.dt.month\n",
    "X_train_enc['day'] = X_train_enc.date.dt.day\n",
    "X_train_enc['weekday'] = X_train_enc.date.dt.weekday\n",
    "X_train_enc = X_train_enc.drop(columns='date')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>branch</th>\n",
       "      <th>product_line</th>\n",
       "      <th>gross_income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-03-15</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>20.8585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-03-16</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-03-17</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-03-18</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-03-19</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>265</th>\n",
       "      <td>2019-03-24</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>17.6250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>266</th>\n",
       "      <td>2019-03-25</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>267</th>\n",
       "      <td>2019-03-26</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>268</th>\n",
       "      <td>2019-03-27</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269</th>\n",
       "      <td>2019-03-28</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>4.5555</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>270 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          date branch         product_line  gross_income\n",
       "0   2019-03-15      A    Health and beauty       20.8585\n",
       "1   2019-03-16      A    Health and beauty        0.0000\n",
       "2   2019-03-17      A    Health and beauty        0.0000\n",
       "3   2019-03-18      A    Health and beauty        0.0000\n",
       "4   2019-03-19      A    Health and beauty        0.0000\n",
       "..         ...    ...                  ...           ...\n",
       "265 2019-03-24      B  Fashion accessories       17.6250\n",
       "266 2019-03-25      B  Fashion accessories        0.0000\n",
       "267 2019-03-26      B  Fashion accessories        0.0000\n",
       "268 2019-03-27      B  Fashion accessories        0.0000\n",
       "269 2019-03-28      B  Fashion accessories        4.5555\n",
       "\n",
       "[270 rows x 4 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# os dados de treino transformardos\n",
    "para_comparar = pd.concat([X_test, y_test], axis=1)\n",
    "para_comparar"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Um exemplo de modelagem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBRegressor(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "             colsample_bynode=1, colsample_bytree=1, gamma=0, gpu_id=-1,\n",
       "             importance_type='gain', interaction_constraints='',\n",
       "             learning_rate=0.1, max_delta_step=0.699999988, max_depth=10,\n",
       "             min_child_weight=1, missing=nan, monotone_constraints='()',\n",
       "             n_estimators=100, n_jobs=0, num_parallel_tree=1,\n",
       "             objective='count:poisson', random_state=0, reg_alpha=0,\n",
       "             reg_lambda=1, scale_pos_weight=None, subsample=1,\n",
       "             tree_method='exact', validate_parameters=1, verbosity=None)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "modelo = XGBRegressor(objective='count:poisson', max_depth=10, learning_rate=0.1)\n",
    "modelo.fit(X_train_enc, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Para as predições, é necessário aplicar o encoder nos dados de teste e todo o preprocessamento\n",
    "X_test_enc = encoder.transform(X_test)\n",
    "X_test_enc['year'] = X_test_enc.date.dt.year\n",
    "X_test_enc['month'] = X_test_enc.date.dt.month\n",
    "X_test_enc['day'] = X_test_enc.date.dt.day\n",
    "X_test_enc['weekday'] = X_test_enc.date.dt.weekday\n",
    "X_test_enc = X_test_enc.drop(columns='date')\n",
    "\n",
    "yhat = modelo.predict(X_test_enc)"
   ]
  },
  {
   "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>date</th>\n",
       "      <th>branch</th>\n",
       "      <th>product_line</th>\n",
       "      <th>gross_income</th>\n",
       "      <th>obs</th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-03-15</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>20.8585</td>\n",
       "      <td>20.8585</td>\n",
       "      <td>3.013385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-03-16</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>2.185266</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-03-17</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>2.173001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-03-18</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>2.358811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-03-19</td>\n",
       "      <td>A</td>\n",
       "      <td>Health and beauty</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>9.864868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>265</th>\n",
       "      <td>2019-03-24</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>17.6250</td>\n",
       "      <td>17.6250</td>\n",
       "      <td>1.621789</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>266</th>\n",
       "      <td>2019-03-25</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>11.512790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>267</th>\n",
       "      <td>2019-03-26</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>6.740536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>268</th>\n",
       "      <td>2019-03-27</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>10.340576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269</th>\n",
       "      <td>2019-03-28</td>\n",
       "      <td>B</td>\n",
       "      <td>Fashion accessories</td>\n",
       "      <td>4.5555</td>\n",
       "      <td>4.5555</td>\n",
       "      <td>10.104683</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>270 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          date branch         product_line  gross_income      obs       pred\n",
       "0   2019-03-15      A    Health and beauty       20.8585  20.8585   3.013385\n",
       "1   2019-03-16      A    Health and beauty        0.0000   0.0000   2.185266\n",
       "2   2019-03-17      A    Health and beauty        0.0000   0.0000   2.173001\n",
       "3   2019-03-18      A    Health and beauty        0.0000   0.0000   2.358811\n",
       "4   2019-03-19      A    Health and beauty        0.0000   0.0000   9.864868\n",
       "..         ...    ...                  ...           ...      ...        ...\n",
       "265 2019-03-24      B  Fashion accessories       17.6250  17.6250   1.621789\n",
       "266 2019-03-25      B  Fashion accessories        0.0000   0.0000  11.512790\n",
       "267 2019-03-26      B  Fashion accessories        0.0000   0.0000   6.740536\n",
       "268 2019-03-27      B  Fashion accessories        0.0000   0.0000  10.340576\n",
       "269 2019-03-28      B  Fashion accessories        4.5555   4.5555  10.104683\n",
       "\n",
       "[270 rows x 6 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final = pd.concat([para_comparar, pd.DataFrame({'obs':y_test, 'pred':yhat})], axis=1)\n",
    "final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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": [
    "plt.scatter(final.obs, final.pred)\n",
    "#plt.xlim(-1,100)\n",
    "#plt.ylim(-1,100)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    270.000000\n",
       "mean       9.570806\n",
       "std       14.517095\n",
       "min        0.000000\n",
       "25%        0.000000\n",
       "50%        0.000000\n",
       "75%       15.114250\n",
       "max       98.092000\n",
       "Name: obs, dtype: float64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final.obs.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "E se tentarmos fazer uma predição para o agregado dos supermercados?\n",
    "\n",
    "### Prevendo o income total"
   ]
  },
  {
   "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></th>\n",
       "      <th>gross_income</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-01-01</th>\n",
       "      <td>225.9610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-02</th>\n",
       "      <td>92.6430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-03</th>\n",
       "      <td>98.9585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-04</th>\n",
       "      <td>77.3185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-05</th>\n",
       "      <td>168.4135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-26</th>\n",
       "      <td>93.4530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-27</th>\n",
       "      <td>138.2295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-28</th>\n",
       "      <td>106.1620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-29</th>\n",
       "      <td>191.5830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-30</th>\n",
       "      <td>213.6695</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>89 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            gross_income\n",
       "date                    \n",
       "2019-01-01      225.9610\n",
       "2019-01-02       92.6430\n",
       "2019-01-03       98.9585\n",
       "2019-01-04       77.3185\n",
       "2019-01-05      168.4135\n",
       "...                  ...\n",
       "2019-03-26       93.4530\n",
       "2019-03-27      138.2295\n",
       "2019-03-28      106.1620\n",
       "2019-03-29      191.5830\n",
       "2019-03-30      213.6695\n",
       "\n",
       "[89 rows x 1 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total = bd.groupby('date').agg({'gross_income':'sum'})\n",
    "total"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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": [
    "total.plot()\n",
    "plt.show()"
   ]
  },
  {
   "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></th>\n",
       "      <th>gross_income</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-01-01</th>\n",
       "      <td>225.9610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-02</th>\n",
       "      <td>92.6430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-03</th>\n",
       "      <td>98.9585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-04</th>\n",
       "      <td>77.3185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-01-05</th>\n",
       "      <td>168.4135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-26</th>\n",
       "      <td>93.4530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-27</th>\n",
       "      <td>138.2295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-28</th>\n",
       "      <td>106.1620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-29</th>\n",
       "      <td>191.5830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-03-30</th>\n",
       "      <td>213.6695</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>89 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            gross_income\n",
       "date                    \n",
       "2019-01-01      225.9610\n",
       "2019-01-02       92.6430\n",
       "2019-01-03       98.9585\n",
       "2019-01-04       77.3185\n",
       "2019-01-05      168.4135\n",
       "...                  ...\n",
       "2019-03-26       93.4530\n",
       "2019-03-27      138.2295\n",
       "2019-03-28      106.1620\n",
       "2019-03-29      191.5830\n",
       "2019-03-30      213.6695\n",
       "\n",
       "[89 rows x 1 columns]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = total.gross_income.iloc[:-21]\n",
    "test = total.gross_income.iloc[-21:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date\n",
       "2019-03-10    150.6300\n",
       "2019-03-11    141.0120\n",
       "2019-03-12    175.1215\n",
       "2019-03-13     98.2670\n",
       "2019-03-14    343.5540\n",
       "2019-03-15    140.1150\n",
       "2019-03-16    150.2130\n",
       "2019-03-17     94.1090\n",
       "2019-03-18     61.5635\n",
       "2019-03-19    273.3520\n",
       "2019-03-20    259.9145\n",
       "2019-03-21     89.4055\n",
       "2019-03-22    151.3880\n",
       "2019-03-23    195.0020\n",
       "2019-03-24    165.5935\n",
       "2019-03-25    108.2365\n",
       "2019-03-26     93.4530\n",
       "2019-03-27    138.2295\n",
       "2019-03-28    106.1620\n",
       "2019-03-29    191.5830\n",
       "2019-03-30    213.6695\n",
       "Name: gross_income, dtype: float64"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing stepwise search to minimize aic\n",
      "Fit ARIMA(1,1,1)x(0,1,1,7) [intercept=True]; AIC=720.317, BIC=730.788, Time=0.291 seconds\n",
      "Fit ARIMA(0,1,0)x(0,1,0,7) [intercept=True]; AIC=769.936, BIC=774.125, Time=0.013 seconds\n",
      "Fit ARIMA(1,1,0)x(1,1,0,7) [intercept=True]; AIC=741.074, BIC=749.451, Time=0.112 seconds\n",
      "Fit ARIMA(0,1,1)x(0,1,1,7) [intercept=True]; AIC=718.382, BIC=726.759, Time=0.173 seconds\n",
      "Near non-invertible roots for order (0, 1, 1)(0, 1, 1, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 0.999)\n",
      "Fit ARIMA(0,1,0)x(0,1,0,7) [intercept=False]; AIC=767.936, BIC=770.031, Time=0.010 seconds\n",
      "Fit ARIMA(1,1,1)x(0,1,0,7) [intercept=True]; AIC=731.002, BIC=739.379, Time=0.159 seconds\n",
      "Near non-invertible roots for order (1, 1, 1)(0, 1, 0, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 1.000)\n",
      "Fit ARIMA(1,1,1)x(1,1,1,7) [intercept=True]; AIC=719.584, BIC=732.150, Time=0.212 seconds\n",
      "Near non-invertible roots for order (1, 1, 1)(1, 1, 1, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 0.999)\n",
      "Fit ARIMA(1,1,1)x(0,1,2,7) [intercept=True]; AIC=720.274, BIC=732.840, Time=0.466 seconds\n",
      "Near non-invertible roots for order (1, 1, 1)(0, 1, 2, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 1.000)\n",
      "Fit ARIMA(1,1,1)x(1,1,0,7) [intercept=True]; AIC=722.002, BIC=732.474, Time=0.179 seconds\n",
      "Fit ARIMA(1,1,1)x(1,1,2,7) [intercept=True]; AIC=724.160, BIC=738.820, Time=0.602 seconds\n",
      "Near non-invertible roots for order (1, 1, 1)(1, 1, 2, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 1.000)\n",
      "Fit ARIMA(1,1,0)x(0,1,1,7) [intercept=True]; AIC=740.262, BIC=748.639, Time=0.189 seconds\n",
      "Near non-invertible roots for order (1, 1, 0)(0, 1, 1, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 1.000)\n",
      "Fit ARIMA(2,1,1)x(0,1,1,7) [intercept=True]; AIC=721.342, BIC=733.908, Time=0.195 seconds\n",
      "Fit ARIMA(1,1,2)x(0,1,1,7) [intercept=True]; AIC=720.483, BIC=733.049, Time=0.276 seconds\n",
      "Near non-invertible roots for order (1, 1, 2)(0, 1, 1, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 0.997)\n",
      "Fit ARIMA(0,1,0)x(0,1,1,7) [intercept=True]; AIC=751.811, BIC=758.094, Time=0.062 seconds\n",
      "Near non-invertible roots for order (0, 1, 0)(0, 1, 1, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 0.998)\n",
      "Fit ARIMA(0,1,2)x(0,1,1,7) [intercept=True]; AIC=720.284, BIC=730.756, Time=0.140 seconds\n",
      "Fit ARIMA(0,1,2)x(0,1,0,7) [intercept=True]; AIC=731.012, BIC=739.390, Time=0.048 seconds\n",
      "Near non-invertible roots for order (0, 1, 2)(0, 1, 0, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 0.997)\n",
      "Fit ARIMA(0,1,2)x(1,1,1,7) [intercept=True]; AIC=719.574, BIC=732.140, Time=0.365 seconds\n",
      "Near non-invertible roots for order (0, 1, 2)(1, 1, 1, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 0.998)\n",
      "Fit ARIMA(0,1,2)x(0,1,2,7) [intercept=True]; AIC=720.261, BIC=732.827, Time=0.515 seconds\n",
      "Near non-invertible roots for order (0, 1, 2)(0, 1, 2, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 1.000)\n",
      "Fit ARIMA(0,1,2)x(1,1,0,7) [intercept=True]; AIC=722.035, BIC=732.506, Time=0.240 seconds\n",
      "Near non-invertible roots for order (0, 1, 2)(1, 1, 0, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 0.999)\n",
      "Fit ARIMA(0,1,2)x(1,1,2,7) [intercept=True]; AIC=724.105, BIC=738.765, Time=0.640 seconds\n",
      "Near non-invertible roots for order (0, 1, 2)(1, 1, 2, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 1.000)\n",
      "Fit ARIMA(0,1,3)x(0,1,1,7) [intercept=True]; AIC=721.298, BIC=733.864, Time=0.309 seconds\n",
      "Near non-invertible roots for order (0, 1, 3)(0, 1, 1, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 0.994)\n",
      "Fit ARIMA(1,1,3)x(0,1,1,7) [intercept=True]; AIC=723.921, BIC=738.581, Time=0.262 seconds\n",
      "Near non-invertible roots for order (1, 1, 3)(0, 1, 1, 7); setting score to inf (at least one inverse root too close to the border of the unit circle: 0.998)\n",
      "Total fit time: 5.469 seconds\n"
     ]
    }
   ],
   "source": [
    "import pmdarima as pm\n",
    "\n",
    "model = pm.auto_arima(train, start_p=1, start_q=1,\n",
    "                           max_p=3, max_q=3, m=7,\n",
    "                           start_P=0, seasonal=True,\n",
    "                           d=1, D=1, trace=True,\n",
    "                           error_action='ignore',  \n",
    "                           suppress_warnings=True, \n",
    "                           stepwise=True)\n",
    "yhat = model.predict(n_periods=21)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([195.66414349, 170.93309755, 194.15009143, 161.89152186,\n",
       "       132.555256  , 161.54884203, 232.36672938, 192.52694264,\n",
       "       167.65460774, 190.73031267, 158.33045415, 128.85289933,\n",
       "       157.7051964 , 228.38179479, 188.4007191 , 163.38709525,\n",
       "       186.32151122, 153.78036375, 124.16151997, 152.87252809,\n",
       "       223.40783753])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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": [
    "plt.plot(test.index, test, '--', label='obs', alpha=.5)\n",
    "plt.plot(test.index, yhat, '-o', label='pred', alpha=.8)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}