modelo.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Case Aquisição"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "O escopo deste projeto está dividido nos seguinte tópicos:\n",
    "- Avaliação e limpeza de dados\n",
    "- Análise exploratória\n",
    "- Pré-processamento\n",
    "- Modelagem\n",
    "- Tuning\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importações e configurações"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 353,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.patches as mpatches\n",
    "from time import time\n",
    "from datetime import datetime\n",
    "from datetime import date\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import f1_score, accuracy_score\n",
    "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import make_scorer\n",
    "from sklearn import tree\n",
    "import warnings\n",
    "\n",
    "# Oculta os warnings dos códigos\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "# Aumenta a fonte nos gráficos\n",
    "plt.rcParams.update({'font.size': 16})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Leitura de dados"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 354,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('dataset/Planilha_1.csv', sep = ';')\n",
    "df2 = pd.read_csv('dataset/Planilha_2.csv', sep = ';')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Limpeza de dados"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Avaliação"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Planilha_1.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 355,
   "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>ID</th>\n",
       "      <th>QT_TOTAL_HIT_PAYWALL</th>\n",
       "      <th>DIASNAVEGADOS</th>\n",
       "      <th>NOTICIASLIDAS</th>\n",
       "      <th>VISITAS_CAPA</th>\n",
       "      <th>USOU_APP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5395</th>\n",
       "      <td>51596</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5013</th>\n",
       "      <td>362499</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3771</th>\n",
       "      <td>194923</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>NAO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4614</th>\n",
       "      <td>244164</td>\n",
       "      <td>27</td>\n",
       "      <td>28</td>\n",
       "      <td>99</td>\n",
       "      <td>149</td>\n",
       "      <td>SIM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>259</th>\n",
       "      <td>636</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4844</th>\n",
       "      <td>112756</td>\n",
       "      <td>24</td>\n",
       "      <td>17</td>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>945</th>\n",
       "      <td>103568</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5165</th>\n",
       "      <td>93876</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>785</th>\n",
       "      <td>264374</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>23</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2929</th>\n",
       "      <td>305332</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          ID  QT_TOTAL_HIT_PAYWALL  DIASNAVEGADOS  NOTICIASLIDAS  \\\n",
       "5395   51596                     0             20              8   \n",
       "5013  362499                     0              1              6   \n",
       "3771  194923                     0              3              3   \n",
       "4614  244164                    27             28             99   \n",
       "259      636                     0              0              0   \n",
       "4844  112756                    24             17             15   \n",
       "945   103568                     0              2              1   \n",
       "5165   93876                     0              3              4   \n",
       "785   264374                     0             21             23   \n",
       "2929  305332                     2              4              5   \n",
       "\n",
       "      VISITAS_CAPA USOU_APP  \n",
       "5395             0      NAO  \n",
       "5013             0      NAO  \n",
       "3771             2      NAO  \n",
       "4614           149      SIM  \n",
       "259              0      NAO  \n",
       "4844             0      NAO  \n",
       "945              0      NAO  \n",
       "5165             0      NAO  \n",
       "785              0      NAO  \n",
       "2929             0      NAO  "
      ]
     },
     "execution_count": 355,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 356,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5600 entries, 0 to 5599\n",
      "Data columns (total 6 columns):\n",
      "ID                      5600 non-null int64\n",
      "QT_TOTAL_HIT_PAYWALL    5600 non-null int64\n",
      "DIASNAVEGADOS           5600 non-null int64\n",
      "NOTICIASLIDAS           5600 non-null int64\n",
      "VISITAS_CAPA            5600 non-null int64\n",
      "USOU_APP                5600 non-null object\n",
      "dtypes: int64(5), object(1)\n",
      "memory usage: 262.6+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 357,
   "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>ID</th>\n",
       "      <th>QT_TOTAL_HIT_PAYWALL</th>\n",
       "      <th>DIASNAVEGADOS</th>\n",
       "      <th>NOTICIASLIDAS</th>\n",
       "      <th>VISITAS_CAPA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>5600.000000</td>\n",
       "      <td>5600.000000</td>\n",
       "      <td>5600.000000</td>\n",
       "      <td>5600.000000</td>\n",
       "      <td>5600.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>165460.766786</td>\n",
       "      <td>6.795000</td>\n",
       "      <td>11.600714</td>\n",
       "      <td>17.768036</td>\n",
       "      <td>34.611250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>116184.700778</td>\n",
       "      <td>25.313927</td>\n",
       "      <td>14.711436</td>\n",
       "      <td>72.011940</td>\n",
       "      <td>126.670657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>58354.250000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>163883.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>265199.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>7.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>371187.000000</td>\n",
       "      <td>894.000000</td>\n",
       "      <td>60.000000</td>\n",
       "      <td>1773.000000</td>\n",
       "      <td>2963.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  ID  QT_TOTAL_HIT_PAYWALL  DIASNAVEGADOS  NOTICIASLIDAS  \\\n",
       "count    5600.000000           5600.000000    5600.000000    5600.000000   \n",
       "mean   165460.766786              6.795000      11.600714      17.768036   \n",
       "std    116184.700778             25.313927      14.711436      72.011940   \n",
       "min         3.000000              0.000000       0.000000       0.000000   \n",
       "25%     58354.250000              0.000000       2.000000       1.000000   \n",
       "50%    163883.000000              0.000000       5.000000       3.000000   \n",
       "75%    265199.000000              3.000000      16.000000       8.000000   \n",
       "max    371187.000000            894.000000      60.000000    1773.000000   \n",
       "\n",
       "       VISITAS_CAPA  \n",
       "count   5600.000000  \n",
       "mean      34.611250  \n",
       "std      126.670657  \n",
       "min        0.000000  \n",
       "25%        0.000000  \n",
       "50%        0.000000  \n",
       "75%        7.000000  \n",
       "max     2963.000000  "
      ]
     },
     "execution_count": 357,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 358,
   "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>ID</th>\n",
       "      <th>QT_TOTAL_HIT_PAYWALL</th>\n",
       "      <th>DIASNAVEGADOS</th>\n",
       "      <th>NOTICIASLIDAS</th>\n",
       "      <th>VISITAS_CAPA</th>\n",
       "      <th>USOU_APP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [ID, QT_TOTAL_HIT_PAYWALL, DIASNAVEGADOS, NOTICIASLIDAS, VISITAS_CAPA, USOU_APP]\n",
       "Index: []"
      ]
     },
     "execution_count": 358,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.duplicated(keep=False)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Planilha_2.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 359,
   "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>ID</th>\n",
       "      <th>PERFIL</th>\n",
       "      <th>PES_GENERO</th>\n",
       "      <th>PES_NASCIMENTO_DATA</th>\n",
       "      <th>ATR_PF_GEO_RENDA_FAM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2036</th>\n",
       "      <td>354780</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1060</th>\n",
       "      <td>230911</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4715</th>\n",
       "      <td>258219</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DE 4SM ATE 8SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2619</th>\n",
       "      <td>310029</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>09.09.1976 00:00:00</td>\n",
       "      <td>ACIMA DE 25SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1339</th>\n",
       "      <td>72282</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>05.01.1986 00:00:00</td>\n",
       "      <td>DE 3SM ATE 4SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1498</th>\n",
       "      <td>213404</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3721</th>\n",
       "      <td>300070</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DE 3SM ATE 4SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5151</th>\n",
       "      <td>55080</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>593</th>\n",
       "      <td>97</td>\n",
       "      <td>ASSINANTE</td>\n",
       "      <td>M</td>\n",
       "      <td>01.01.1900 00:00:00</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1484</th>\n",
       "      <td>304640</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DE 4SM ATE 8SM</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          ID     PERFIL PES_GENERO  PES_NASCIMENTO_DATA ATR_PF_GEO_RENDA_FAM\n",
       "2036  354780  PROSPECT           F                  NaN                  NaN\n",
       "1060  230911  PROSPECT           F                  NaN                  NaN\n",
       "4715  258219  PROSPECT           M                  NaN       DE 4SM ATE 8SM\n",
       "2619  310029  PROSPECT           F  09.09.1976 00:00:00        ACIMA DE 25SM\n",
       "1339   72282  PROSPECT           M  05.01.1986 00:00:00       DE 3SM ATE 4SM\n",
       "1498  213404  PROSPECT           F                  NaN                  NaN\n",
       "3721  300070  PROSPECT           F                  NaN       DE 3SM ATE 4SM\n",
       "5151   55080  PROSPECT           F                  NaN                  NaN\n",
       "593       97  ASSINANTE          M  01.01.1900 00:00:00                  NaN\n",
       "1484  304640  PROSPECT           M                  NaN       DE 4SM ATE 8SM"
      ]
     },
     "execution_count": 359,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.sample(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 360,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5600 entries, 0 to 5599\n",
      "Data columns (total 5 columns):\n",
      "ID                      5600 non-null int64\n",
      "PERFIL                  5600 non-null object\n",
      "PES_GENERO              5600 non-null object\n",
      "PES_NASCIMENTO_DATA     1935 non-null object\n",
      "ATR_PF_GEO_RENDA_FAM    2648 non-null object\n",
      "dtypes: int64(1), object(4)\n",
      "memory usage: 218.8+ KB\n"
     ]
    }
   ],
   "source": [
    "df2.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 361,
   "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>ID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>5600.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>165460.766786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>116184.700778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>58354.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>163883.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>265199.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>371187.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  ID\n",
       "count    5600.000000\n",
       "mean   165460.766786\n",
       "std    116184.700778\n",
       "min         3.000000\n",
       "25%     58354.250000\n",
       "50%    163883.000000\n",
       "75%    265199.000000\n",
       "max    371187.000000"
      ]
     },
     "execution_count": 361,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 362,
   "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>ID</th>\n",
       "      <th>PERFIL</th>\n",
       "      <th>PES_GENERO</th>\n",
       "      <th>PES_NASCIMENTO_DATA</th>\n",
       "      <th>ATR_PF_GEO_RENDA_FAM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [ID, PERFIL, PES_GENERO, PES_NASCIMENTO_DATA, ATR_PF_GEO_RENDA_FAM]\n",
       "Index: []"
      ]
     },
     "execution_count": 362,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2[df2.duplicated(keep=False)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problemas detectados"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Os dados estão em dois datasets diferentes.\n",
    "- O tipo de dado da coluna PES_NASCIMENTO_DATA está incorreto.\n",
    "- PES_NASCIMENTO_DATA com datas inválidas.\n",
    "- PES_NASCIMENTO_DATA e ATR_PF_GEO_RENDA_FAM com dados ausentes."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Limpeza"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Backup dos datasets originais"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 363,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_copy = df.copy()\n",
    "df2_copy = df2.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Definição"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Unir os dois datasets."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Código"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 364,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.merge(left = df, right = df2, left_on = 'ID', right_on = 'ID', how = 'left')\n",
    "df = df.drop('ID', axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Teste"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 365,
   "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>QT_TOTAL_HIT_PAYWALL</th>\n",
       "      <th>DIASNAVEGADOS</th>\n",
       "      <th>NOTICIASLIDAS</th>\n",
       "      <th>VISITAS_CAPA</th>\n",
       "      <th>USOU_APP</th>\n",
       "      <th>PERFIL</th>\n",
       "      <th>PES_GENERO</th>\n",
       "      <th>PES_NASCIMENTO_DATA</th>\n",
       "      <th>ATR_PF_GEO_RENDA_FAM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3747</th>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>64</td>\n",
       "      <td>SIM</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DE 3SM ATE 4SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2636</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>12.12.1958 00:00:00</td>\n",
       "      <td>DE 8SM ATE 14SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3060</th>\n",
       "      <td>0</td>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3384</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>23.02.1989 00:00:00</td>\n",
       "      <td>DE 14SM ATE 25SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5129</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DE 4SM ATE 8SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>271</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>ASSINANTE</td>\n",
       "      <td>F</td>\n",
       "      <td>01.01.1900 00:00:00</td>\n",
       "      <td>DE 8SM ATE 14SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4210</th>\n",
       "      <td>22</td>\n",
       "      <td>34</td>\n",
       "      <td>0</td>\n",
       "      <td>99</td>\n",
       "      <td>SIM</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1432</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>SIM</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DE 2SM ATE 3SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375</th>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>ASSINANTE</td>\n",
       "      <td>M</td>\n",
       "      <td>01.01.1900 00:00:00</td>\n",
       "      <td>DE 3SM ATE 4SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3850</th>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DE 4SM ATE 8SM</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      QT_TOTAL_HIT_PAYWALL  DIASNAVEGADOS  NOTICIASLIDAS  VISITAS_CAPA  \\\n",
       "3747                    10             11              0            64   \n",
       "2636                     0              1              1             0   \n",
       "3060                     0             16              0            20   \n",
       "3384                     0              2              2             0   \n",
       "5129                     0              1              1             0   \n",
       "271                      0              0              0             0   \n",
       "4210                    22             34              0            99   \n",
       "1432                     0              2              0             2   \n",
       "375                     16              0              0             0   \n",
       "3850                     0              7              4             1   \n",
       "\n",
       "     USOU_APP     PERFIL PES_GENERO  PES_NASCIMENTO_DATA ATR_PF_GEO_RENDA_FAM  \n",
       "3747      SIM  PROSPECT           M                  NaN       DE 3SM ATE 4SM  \n",
       "2636      NAO  PROSPECT           F  12.12.1958 00:00:00      DE 8SM ATE 14SM  \n",
       "3060      NAO  PROSPECT           M                  NaN                  NaN  \n",
       "3384      NAO  PROSPECT           M  23.02.1989 00:00:00     DE 14SM ATE 25SM  \n",
       "5129      NAO  PROSPECT           M                  NaN       DE 4SM ATE 8SM  \n",
       "271       NAO  ASSINANTE          F  01.01.1900 00:00:00      DE 8SM ATE 14SM  \n",
       "4210      SIM  PROSPECT           M                  NaN                  NaN  \n",
       "1432      SIM  PROSPECT           M                  NaN       DE 2SM ATE 3SM  \n",
       "375       NAO  ASSINANTE          M  01.01.1900 00:00:00       DE 3SM ATE 4SM  \n",
       "3850      NAO  PROSPECT           M                  NaN       DE 4SM ATE 8SM  "
      ]
     },
     "execution_count": 365,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Definição"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Trocar o tipo de dado da coluna PES_NASCIMENTO_DATA para timestamp."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Código"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 366,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['PES_NASCIMENTO_DATA'] = pd.to_datetime(df['PES_NASCIMENTO_DATA'], errors = 'coerce')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Teste"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 367,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 5600 entries, 0 to 5599\n",
      "Data columns (total 9 columns):\n",
      "QT_TOTAL_HIT_PAYWALL    5600 non-null int64\n",
      "DIASNAVEGADOS           5600 non-null int64\n",
      "NOTICIASLIDAS           5600 non-null int64\n",
      "VISITAS_CAPA            5600 non-null int64\n",
      "USOU_APP                5600 non-null object\n",
      "PERFIL                  5600 non-null object\n",
      "PES_GENERO              5600 non-null object\n",
      "PES_NASCIMENTO_DATA     1925 non-null datetime64[ns]\n",
      "ATR_PF_GEO_RENDA_FAM    2648 non-null object\n",
      "dtypes: datetime64[ns](1), int64(4), object(4)\n",
      "memory usage: 437.5+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Definição"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Corrigir datas inválidas da coluna PES_NASCIMENTO_DATA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Código"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 368,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['PES_NASCIMENTO_DATA'][(df.PES_NASCIMENTO_DATA > '2001-06-15') | (df.PES_NASCIMENTO_DATA < '1919-06-15')] = np.nan"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Teste"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 369,
   "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>QT_TOTAL_HIT_PAYWALL</th>\n",
       "      <th>DIASNAVEGADOS</th>\n",
       "      <th>NOTICIASLIDAS</th>\n",
       "      <th>VISITAS_CAPA</th>\n",
       "      <th>USOU_APP</th>\n",
       "      <th>PERFIL</th>\n",
       "      <th>PES_GENERO</th>\n",
       "      <th>PES_NASCIMENTO_DATA</th>\n",
       "      <th>ATR_PF_GEO_RENDA_FAM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [QT_TOTAL_HIT_PAYWALL, DIASNAVEGADOS, NOTICIASLIDAS, VISITAS_CAPA, USOU_APP, PERFIL, PES_GENERO, PES_NASCIMENTO_DATA, ATR_PF_GEO_RENDA_FAM]\n",
       "Index: []"
      ]
     },
     "execution_count": 369,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[(df.PES_NASCIMENTO_DATA > '2001-06-15') | (df.PES_NASCIMENTO_DATA < '1919-06-15')]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Definição"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Preencher dados ausentes com `None`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Código"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 370,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.astype(object).replace(np.nan, 'None')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Teste"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 371,
   "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>QT_TOTAL_HIT_PAYWALL</th>\n",
       "      <th>DIASNAVEGADOS</th>\n",
       "      <th>NOTICIASLIDAS</th>\n",
       "      <th>VISITAS_CAPA</th>\n",
       "      <th>USOU_APP</th>\n",
       "      <th>PERFIL</th>\n",
       "      <th>PES_GENERO</th>\n",
       "      <th>PES_NASCIMENTO_DATA</th>\n",
       "      <th>ATR_PF_GEO_RENDA_FAM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>405</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>ASSINANTE</td>\n",
       "      <td>F</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2222</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4357</th>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>25</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>1987-05-08 00:00:00</td>\n",
       "      <td>DE 8SM ATE 14SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2152</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>None</td>\n",
       "      <td>DE 3SM ATE 4SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3166</th>\n",
       "      <td>46</td>\n",
       "      <td>29</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>1980-08-30 00:00:00</td>\n",
       "      <td>ACIMA DE 25SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4673</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>1979-11-23 00:00:00</td>\n",
       "      <td>DE 8SM ATE 14SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1330</th>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>48</td>\n",
       "      <td>SIM</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3027</th>\n",
       "      <td>2</td>\n",
       "      <td>27</td>\n",
       "      <td>17</td>\n",
       "      <td>42</td>\n",
       "      <td>SIM</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1430</th>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>58</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>1975-09-08 00:00:00</td>\n",
       "      <td>DE 14SM ATE 25SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4180</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1291</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>None</td>\n",
       "      <td>DE 4SM ATE 8SM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1943</th>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2275</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>F</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2911</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>I</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1334</th>\n",
       "      <td>3</td>\n",
       "      <td>17</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>NAO</td>\n",
       "      <td>PROSPECT</td>\n",
       "      <td>M</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      QT_TOTAL_HIT_PAYWALL  DIASNAVEGADOS  NOTICIASLIDAS  VISITAS_CAPA  \\\n",
       "405                      0              0              0             0   \n",
       "2222                     0              1              1             0   \n",
       "4357                     0             21             25             0   \n",
       "2152                     0              2              1             0   \n",
       "3166                    46             29             10             0   \n",
       "4673                     0              1              0             0   \n",
       "1330                     7             10              0            48   \n",
       "3027                     2             27             17            42   \n",
       "1430                     0             22             58             0   \n",
       "4180                     0              1              0             1   \n",
       "1291                     0              1              6             0   \n",
       "1943                     0              5              3             0   \n",
       "2275                     0              1              1             0   \n",
       "2911                     0              1              2             0   \n",
       "1334                     3             17             12             0   \n",
       "\n",
       "     USOU_APP     PERFIL PES_GENERO  PES_NASCIMENTO_DATA ATR_PF_GEO_RENDA_FAM  \n",
       "405       NAO  ASSINANTE          F                 None                 None  \n",
       "2222      NAO  PROSPECT           F                 None                 None  \n",
       "4357      NAO  PROSPECT           F  1987-05-08 00:00:00      DE 8SM ATE 14SM  \n",
       "2152      NAO  PROSPECT           F                 None       DE 3SM ATE 4SM  \n",
       "3166      NAO  PROSPECT           M  1980-08-30 00:00:00        ACIMA DE 25SM  \n",
       "4673      NAO  PROSPECT           M  1979-11-23 00:00:00      DE 8SM ATE 14SM  \n",
       "1330      SIM  PROSPECT           M                 None                 None  \n",
       "3027      SIM  PROSPECT           M                 None                 None  \n",
       "1430      NAO  PROSPECT           M  1975-09-08 00:00:00     DE 14SM ATE 25SM  \n",
       "4180      NAO  PROSPECT           F                 None                 None  \n",
       "1291      NAO  PROSPECT           F                 None       DE 4SM ATE 8SM  \n",
       "1943      NAO  PROSPECT           M                 None                 None  \n",
       "2275      NAO  PROSPECT           F                 None                 None  \n",
       "2911      NAO  PROSPECT           I                 None                 None  \n",
       "1334      NAO  PROSPECT           M                 None                 None  "
      ]
     },
     "execution_count": 371,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Análise exploratória"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Distribuição de variáveis quantitativas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 372,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Distribuição de QT_TOTAL_HIT_PAYWALL\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZAAAAEBCAYAAABWhfMqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAGxlJREFUeJzt3XuUFvWd5/H3h3DzEtcmiZ0NhgElM3vwdHR22dnMiUdashPQUZhz8DIRMxM5kejsGDNekjBmXUdxQrxENzomEl13TmCVMZqlOcl6izSZk4VkDCNCm5AhS3uJuGoa0EZE0O/+UdVLPZXq5ukfD3TD83mdU6d8fvWtql/9lOdDXZ5SEYGZmdlgjRjqDpiZ2aHJAWJmZkkcIGZmlsQBYmZmSRwgZmaWxAFiZmZJHCBmZpbEAWJmZkkcIGZmlmTkUHfgQHr/+98fEydOTFp3x44dHHXUUY3t0CHM41HL41HL41HrUB+Pn/3sZ69FxAf2VXdYB8jEiRN56qmnktbt7Oykvb29sR06hHk8ank8ank8ah3q4yHpuXrqfAnLzMySOEDMzCyJA8TMzJI4QMzMLIkDxMzMkjhAzMwsiQPEzMySOEDMzCyJA8TMzJIc1r9E3x/rf72dz3z5+0nrdi/64wb3xsxs+PEZiJmZJXGAmJlZEgeImZklcYCYmVkSB4iZmSVxgJiZWRIHiJmZJXGAmJlZEgeImZklqStAJLVLioppW6muRdI9kl6TtEPSE5LaKrY3VtLNkrZI2ilptaTTKupGSFogqVvSW5LWSZqTfrhmZtYogz0D+Tzwh4XpP/YtkCSgA5gJXAbMAUYBKyUdX9rOvcDFwLXAWcAW4FFJp5TqbgCuA+4EzgDWAA9KOnOQ/TYzswYb7Luwfh4Ra/pZNgs4FZgeESsBJK0GNgNfJAsfJJ0MXADMi4j78rZVQBdwfb4dJB0HXAUsiohb8n2slDQZWAT8YJB9NzOzBmrkPZBZwEt94QEQEduBFcDsUt1uYFmhbg/wADBD0pi8eQYwGlhS2s8SoE3SpAb23czMBmmwAbJU0juSfiPpf0iaUFh2ErChYp0uYIKkowt1myPizYq60cDkQt0uYFNFHcCUQfbdzMwaqN5LWNuBW4FVwOvA7wN/DayW9PsR8QowDuiuWLcnn7cAvXnd1gHqxhXm2yIi9lFXQ9J8YD5Aa2srnZ2dAx1Xv1qPgCvb9iStm7rP4ay3t/ewPK5UHo9aHo9azTIedQVIRPwz8M+FplWSfgT8lOzexlcAAeUve/L28udG1pX7uhhYDDB16tRob28fqLxfdyxdzq3r0/53Kd1z0/Y5nHV2dpI6locjj0ctj0etZhmP5HsgEbEW+CXw7/OmHqrPClry+dY663oK85b86a6B6szMbAjs70304llCF9l9i7IpwPMR0VuomyTpyIq6t9l7z6MLGAOcWFEH8Ox+9NvMzPZTcoBImgr8LvCTvKkDGC9pWqHmGODsfBmFulHAuYW6kcD5wGMRsStvfoQsUOaWdn0hsCEiNqf23czM9l9dF/klLSX7PcdaYBvZTfQFwK+BO/KyDmA1sETS1WSXrBaQnaXc1LetiHha0jLgdkmj8u1eCkyiEBYR8Yqk24AFkt7I930+MJ3ax4LNzGwI1HuXeAPwKbJfmB8JvAw8DPyXiHgNICLelXQWcAtwFzCWLFBOj4gXStu7CLgRWAgcC6wDZub3VYquIXty63Lgg8BG4LyIWDGYgzQzs8ar9ymsrwJfraOuB5iXTwPV7QSuyKeB6t4hC5mF9fTTzMwOHr+N18zMkjhAzMwsiQPEzMySOEDMzCyJA8TMzJI4QMzMLIkDxMzMkjhAzMwsiQPEzMySOEDMzCyJA8TMzJI4QMzMLIkDxMzMkjhAzMwsiQPEzMySOEDMzCyJA8TMzJI4QMzMLIkDxMzMkjhAzMwsiQPEzMySOEDMzCyJA8TMzJI4QMzMLIkDxMzMkiQFiKRHJIWkhaX2Fkn3SHpN0g5JT0hqq1h/rKSbJW2RtFPSakmnVdSNkLRAUrektyStkzQnpc9mZtZYgw4QSZ8CTq5oF9ABzAQuA+YAo4CVko4vld8LXAxcC5wFbAEelXRKqe4G4DrgTuAMYA3woKQzB9tvMzNrrEEFiKRjgduAKyoWzwJOBT4dEfdHxCN52wjgi4VtnAxcAPxVRHw7In4InAc8D1xfqDsOuApYFBG3RMTKiPgcsBJYNJh+m5lZ4w32DOQmoCsi7q9YNgt4KSJW9jVExHZgBTC7VLcbWFao2wM8AMyQNCZvngGMBpaU9rMEaJM0aZB9NzOzBqo7QCSdCvwZ8Bf9lJwEbKho7wImSDq6ULc5It6sqBsNTC7U7QI2VdQBTKm372Zm1nh1BYikUcDdwC0RsbGfsnHA1or2nnzeUmfduMJ8W0TEPurMzGwIjKyz7kvAEcCNA9QIKH/Z97UfyLrahdJ8YD5Aa2srnZ2dA5X3q/UIuLJtT9K6qfscznp7ew/L40rl8ajl8ajVLOOxzwCRNAG4BvgsMKZwj4L887HAG2RnBlVnBX1nHn1nHT3AhAHqegrzFkkqnYWU62pExGJgMcDUqVOjvb29nyMb2B1Ll3Pr+nrztVb33LR9DmednZ2kjuXhyONRy+NRq1nGo55LWCcAY8luXm8tTJA9JbUVaCO7N3FSxfpTgOcjojf/3AVMknRkRd3b7L3n0QWMAU6sqAN4to6+m5nZAVJPgDwNnF4xQRYqp5N96XcA4yVN61tR0jHA2fmyPh1kvw85t1A3EjgfeCwiduXNj5AFytxSfy4ENkTE5jr6bmZmB8g+r9FExDags9ye/W6Q5yKiM//cAawGlki6muzMZAHZPYubCtt7WtIy4Pb85vxm4FJgEoWwiIhXJN0GLJD0BrCWLGSmU/tYsJmZDYG0i/wVIuJdSWcBtwB3kV32Wg2cHhEvlMovIrshvxA4FlgHzIyItaW6a4Be4HLgg8BG4LyIWNGofpuZWZrkAImI33oaKiJ6gHn5NNC6O8l+zV71i/Zi3TtkIbNwoDozMzv4/DZeMzNL4gAxM7MkDhAzM0viADEzsyQOEDMzS+IAMTOzJA4QMzNL4gAxM7MkDhAzM0viADEzsyQOEDMzS+IAMTOzJA4QMzNL4gAxM7MkDhAzM0viADEzsyQOEDMzS+IAMTOzJA4QMzNL4gAxM7MkDhAzM0viADEzsyQOEDMzS+IAMTOzJA4QMzNLUleASJoh6UlJL0vaJelFSf8gaUqp7sOSvitpu6TXJT0saULF9lok3SPpNUk7JD0hqa2ibqykmyVtkbRT0mpJp6UfrpmZNUq9ZyDjgJ8Bfwl8ElgAnASskfQ7AJKOBJ4E/g3w58CngY8AKyUd1bchSQI6gJnAZcAcYFRed3xpv/cCFwPXAmcBW4BHJZ0y6CM1M7OGGllPUUTcD9xfbJP0U+AXwDnArWRf9CcAvxcRm/KaZ4B/AT4HfD1fdRZwKjA9IlbmdauBzcAXgc/nbScDFwDzIuK+vG0V0AVcn2/HzMyGyP7cA/lNPt+dz2cBa/rCAyAiNgM/BmYX1psFvNQXHnnddmBFRd1uYFmhbg/wADBD0pj96LuZme2nQQWIpPdIGi3pI8DdwMtkX+iQXdLaULFaF1C8VzJQ3QRJRxfqNkfEmxV1o4HJg+m7mZk11mDPQH4C7AJ+CXyU7DLUK/myccDWinV6gJbC54HqKNTuq25c/d02M7NGq+seSMGngWPI7nVcBTwu6dSI6M6XR8U6qvjcyLrahdJ8YD5Aa2srnZ2dA5X3q/UIuLJtT9K6qfscznp7ew/L40rl8ajl8ajVLOMxqACJiJ/n//gTSf8L6Aa+DFxCdrZQdVbQQu2ZRM8AdRRqe4DfegS4UNdTsYyIWAwsBpg6dWq0t7dXle3THUuXc+v6weZrpntu2j6Hs87OTlLH8nDk8ajl8ajVLOORfBM9IrYBm9h7L6KL7L5F2RTg2cLngeqej4jeQt2k/PHgct3b+b7NzGyIJAeIpFay33z8Km/qAD4m6YRCzUTg4/kyCnXjJU0r1B0DnF1RNwo4t1A3EjgfeCwidqX23czM9l9d12gkfQ9YCzwDvA78LvBXwB6y34AAfJvsh4bLJX2F7P7FDcALZE9s9ekAVgNLJF1NdslqAdm9jZv6iiLiaUnLgNsljSL7ncilwCRgbsrBmplZ49R7BrIG+BPg74HvA1cAq4BTIuKXABGxA5hO9oTWd4ClZF/60wuXpYiId8l+Vf44cBfwPeAd4PSIeKG034uA+4CF+X4/DMyMiLWDPlIzM2uoen+J/jXga3XUPU/2apJ91fUA8/JpoLqdZGF1RT39NDOzg8dv4zUzsyQOEDMzS+IAMTOzJA4QMzNL4gAxM7MkDhAzM0viADEzsyQOEDMzS+IAMTOzJA4QMzNL4gAxM7MkDhAzM0viADEzsyQOEDMzS+IAMTOzJA4QMzNL4gAxM7MkDhAzM0viADEzsyQOEDMzS+IAMTOzJA4QMzNL4gAxM7MkDhAzM0viADEzsyQOEDMzS7LPAJF0jqSHJD0naaekjZK+Kum9pboWSfdIek3SDklPSGqr2N5YSTdL2pJvb7Wk0yrqRkhaIKlb0luS1kmas3+Ha2ZmjVLPGchVwDvAXwMzgW8ClwKPSxoBIElAR778MmAOMApYKen40vbuBS4GrgXOArYAj0o6pVR3A3AdcCdwBrAGeFDSmYM7RDMzOxBG1lFzdkS8Wvi8SlIP8PdAO/AkMAs4FZgeESsBJK0GNgNfBD6ft50MXADMi4j78rZVQBdwfb4dJB1HFlyLIuKWfL8rJU0GFgE/SD1gMzNrjH2egZTCo88/5fPx+XwW8FJfeOTrbQdWALML680CdgPLCnV7gAeAGZLG5M0zgNHAktJ+lwBtkibtq99mZnZgpd5En5bPf57PTwI2VNR1ARMkHV2o2xwRb1bUjQYmF+p2AZsq6gCmJPbbzMwapJ5LWDUkjSe73PRERDyVN48DuivKe/J5C9Cb120doG5cYb4tImIfdVX9mw/MB2htbaWzs7O/0gG1HgFXtu1JWjd1n8NZb2/vYXlcqTwetTwetZplPAYVIPmZxHJgD3BRcRFQ/rLvay9/bmTdb4mIxcBigKlTp0Z7e/u+Vql0x9Ll3Lp+0PkKQPfctH0OZ52dnaSO5eHI41HL41GrWcaj7ktYksaSPWl1AjAjIl4sLO6h+qygJZ9vrbOupzBvyZ/uGqjOzMyGSF0BImkU8BDwB8CZEbG+VNJFdt+ibArwfET0FuomSTqyou5t9t7z6ALGACdW1AE8W0+/zczswKnnh4QjgKXAJ4DZEbGmoqwDGC9pWmG9Y4Cz82XFulHAuYW6kcD5wGMRsStvfoQsUOaW9nMhsCEiNu+r32ZmdmDVc5H/78i+8G8Edkj6WGHZi/mlrA5gNbBE0tVkl6wWkN2zuKmvOCKelrQMuD0/q9lM9qPESRTCIiJekXQbsEDSG8BaspCZTu1jwWZmNkTqCZAz8vk1+VT0N8B1EfGupLOAW4C7gLFkgXJ6RLxQWucisjBaCBwLrANmRsTaUt01ZE9uXQ58ENgInBcRK+o5MDMzO7D2GSARMbGeDUVEDzAvnwaq2wlckU8D1b1DFjIL69m/mZkdXH4br5mZJXGAmJlZEgeImZklcYCYmVkSB4iZmSVxgJiZWRIHiJmZJXGAmJlZEgeImZklcYCYmVkSB4iZmSVxgJiZWRIHiJmZJXGAmJlZEgeImZklcYCYmVkSB4iZmSVxgJiZWRIHiJmZJXGAmJlZEgeImZklcYCYmVkSB4iZmSVxgJiZWRIHiJmZJakrQCQdL+kOSaslvSkpJE2sqBsr6WZJWyTtzOtPq6gbIWmBpG5Jb0laJ2lOP/u+WNIvJO2StFHSJYM9SDMza7x6z0AmA+cBW4F/HKDuXuBi4FrgLGAL8KikU0p1NwDXAXcCZwBrgAclnVksknQxcDfwEDATeBC4S9KldfbbzMwOkJF11v0oIloBJH0W+GS5QNLJwAXAvIi4L29bBXQB1wOz8rbjgKuARRFxS776SkmTgUXAD/K6kcCNwHci4ppC3YeAGyTdExG7B3vAZmbWGHWdgUTEu3WUzQJ2A8sK6+0BHgBmSBqTN88ARgNLSusvAdokTco//yHwgYq67wDvA06tp+9mZnZgNPIm+knA5oh4s9TeRRYYkwt1u4BNFXUAUwp1ABv2UWdmZkOgkQEyjuweSVlPYXnffFtERB11VGyzXGdmZkOg3nsg9RBQDoW+9tQ6+qntvxPSfGA+QGtrK52dnYNZ/f9rPQKubNuTtG7qPoez3t7ew/K4Unk8ank8ajXLeDQyQHqACRXtLYXlffMWSSqdhVTVQXamsaVQN660vEZELAYWA0ydOjXa29vr7X+NO5Yu59b1acPTPTdtn8NZZ2cnqWN5OPJ41PJ41GqW8WjkJawuYJKkI0vtU4C32XvPowsYA5xYUQfwbKEO9t4L6a/OzMyGQCMDpAMYBZzb15A/ins+8FhE7MqbHyELlLml9S8ENkTE5vzzauC1fup6gB83sO9mZjZIdV+jkXRO/o//Lp+fIelV4NWIWBURT0taBtwuaRSwGbgUmEQhBCLiFUm3AQskvQGsJQuZ6cDsQt1uSf+Z7IeDvwaeyGvmAZdFxNtph2xmZo0wmIv8D5Y+35XPVwHt+T9fRPbjv4XAscA6YGZErC2tew3QC1wOfBDYCJwXESuKRRHxLUkBXAlcDTwP/GVE3IWZmQ2pugMkIspPSVXV7ASuyKeB6t4hC5mFdWzzbrLXmZiZ2TDit/GamVkSB4iZmSVxgJiZWRIHiJmZJXGAmJlZEgeImZklcYCYmVkSB4iZmSVxgJiZWRIHiJmZJXGAmJlZEgeImZklcYCYmVkSB4iZmSVxgJiZWRIHiJmZJXGAmJlZEgeImZklcYCYmVkSB4iZmSVxgJiZWRIHiJmZJXGAmJlZEgeImZklcYCYmVmSYR8gkj4s6buStkt6XdLDkiYMdb/MzJrdyKHuwEAkHQk8CewC/hwIYCGwUtJHI2LHUPavPxO//P3kdbsX/XEDe2JmduAM6wABLgZOAH4vIjYBSHoG+Bfgc8DXh7BvZmZNbbhfwpoFrOkLD4CI2Az8GJg9ZL0yM7NhfwZyErC8or0LOPcg9+Wg2J/LX+BLYGZ28Az3ABkHbK1o7wFaDnJfDgn7G0D9ubJtD58ZYNsOLrPmM9wDBLIb52Xqr1jSfGB+/rFX0sbE/b4feC1x3cPO5/cxHvraQezM8OD/Pmp5PGod6uPxO/UUDfcA2Up2FlLWQvWZCRGxGFi8vzuW9FRETN3f7RwuPB61PB61PB61mmU8hvtN9C6y+yBlU4BnD3JfzMysYLgHSAfwMUkn9DVImgh8PF9mZmZDZLgHyLeBbmC5pNmSZpE9lfUCcPcB3vd+XwY7zHg8ank8ank8ajXFeCii6h718JG/tuQ24I/Ibp7/EPhCRHQPZb/MzJrdsA8QMzMbnob7JayDqhle3CjpHEkPSXpO0k5JGyV9VdJ7S3Utku6R9JqkHZKekNRWsb2xkm6WtCXf3mpJpx28I2o8SY9ICkkLS+1NMyaSzpT0I0m9+Z+FpyRNLyxvprH4uKTHJL2Sj8VaSfNKNXUdp6QRkhZI6pb0lqR1kuYcvKNpsIjwlJ2FHUn2jq0NwJ+QvSplPfAr4Kih7l8Dj3MN8A/AXGAa8AVgW94+Iq8R8I/Ai8CngJnAKrLn2o8vbW9pvv7FwCeAh4GdwClDfayJ4/MpYAv5izsL7U0zJmTvmdvN3kvHM4AvAWc14Vh8NO/vyvw74Y/I7r8GcOlgjxO4kezlsFcBp+fbehc4c6iPNWl8hroDw2UCLgfeASYX2iYBe4Arhrp/DTzOD1S0/Vn+B2J6/nl2/vn0Qs2/InsDwDcKbSfndRcV2kYCG4GOoT7WhLE5Fng5/1IsB0hTjAkwMf/i+8IANU0xFnl//xZ4Gzi61L4GWD2Y4wSOy8Pjb0rb+iHwzFAfa8rkS1h7NcWLGyPi1Yrmf8rn4/P5LOCliFhZWG87sILasZhF9jfVZYW6PcADwAxJYxrY9YPhJqArIu6vWNYsYzKP7G/E3xqgplnGAmA02THsLLVvY+8tgHqPc0a+vSWlbS0B2iRNamzXDzwHyF4nkV2+Kusi++Hi4WxaPv95Ph9oLCZIOrpQtzki3qyoGw1MbnRHDxRJp5Kdif1FPyXNMianAr8A/lTSryTtkbRJ0n8q1DTLWAD893z+DUkfknSspL7LVLfly+o9zpPIzkA2VdTBIfg94wDZqylf3ChpPHA98EREPJU3DzQWsHc89lVX9RqaYUfSKLJr0bdERH/vTmuWMfkQ8BHgZmAR8EngceBOSZfnNc0yFkTEBqCd7Mzq12TH83fAJRHxQF5W73GOA7ZFft1qgLpDxnB/F9bBNqgXNx7q8r8pLie7z3NRcRH1jUW9dcPdl4AjyG5w9qdZxmQE8F7gMxHxcN72ZP4GiAWSvkHzjAWSPgI8RHaWcAnZpazZwLckvRURS2mi8ShzgOw16Bc3HsokjSV7HcwJwLSIeLGwuIf+xwL2jkcPUPWYc0th+bCWP6Z9DfBZYEzpuvwYSccCb9A8Y/IbsjOQx0vtj5E9bfWvaZ6xgOwm+m6yJ9B2520/lPQ+4L9Kup/6j7MHaJGk0lnIoTQeNXwJa6+meXFjfsnmIeAPyB4fXF8qGWgsno+I3kLdpPz/XV+ue5vfvtY7HJ0AjCW7kbm1MEH2qOVWoI3mGZOuftr7/pb8Ls0zFpD9u19XCI8+PwXeR/ZkVb3H2QWMAU6sqIND8HvGAbJXU7y4UdIIsmfWPwHMjog1FWUdwHhJ0wrrHQOcTe1YdACjKPzfISWNBM4HHouIXY0/goZ7mux5/PIEWaicTvYF0Cxj8r18PqPUPgN4MSJepnnGArLHuk+RNLrU/h+At8jOGuo9zkfIAmVuaVsXAhvypz4PLUP9HPFwmYCjyL4o1pNd45wFrAP+D6VnwA/lCfgm+W8cgI+VpuPzmhHA/yZ7aeWfkn15dJL9YflwaXsPkP0t/bNkofRdsj9Y/3aoj3U/x6n8O5CmGBOyM40nyS5lXUJ2E31xPh6faaaxyPt/Tn7sj+bfC58E7szbvj7Y4yR7MOEt4Aqym/PfJDurO3uojzVpfIa6A8NpIruO+RDwOtl17/8JTBzqfjX4GLvz//irpusKdeOA/5Z/KbxJ9mOnkyu2dwTwdbK/qb0F/ARoH+rjbMA41QRIM40JcAzZk0b/l+xvzM8AFzTjWOTHcEYekK/m3wtPkz3u/Z7BHifwHuArwHNkj/Q+A5wz1MeYOvllimZmlsT3QMzMLIkDxMzMkjhAzMwsiQPEzMySOEDMzCyJA8TMzJI4QMzMLIkDxMzMkjhAzMwsyf8DhvUb2b9YhJoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Distribuição de DIASNAVEGADOS\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Distribuição de NOTICIASLIDAS\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Distribuição de VISITAS_CAPA\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAEBCAYAAACwrDhuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAGohJREFUeJzt3X2QVfWd5/H3h/AkOq7NJHZqMQSUPBQUq7PDTiU1ljRkd0BHYarwYSM6M1IB4+wYZ3zIhDHjOooV4kO01qwZiY6VCqwSY7JAJaXGSDNTWXBGGRFaQ8ZZ8CHiGtKANiCCfvePczocjqe7T19+3X3Vz6vq1OH+zvec87v3B/fjeboqIjAzM0th2FB3wMzM3j8cKmZmloxDxczMknGomJlZMg4VMzNLxqFiZmbJOFTMzCwZh4qZmSXjUDEzs2SGD3UHBtKHP/zhmDBhQkPr7t27l2OPPTZthywJj01z8rg0r/6OzVNPPbUzIj7SyL7e16EyYcIEnnzyyYbWbW9vp62tLW2HLAmPTXPyuDSv/o6NpBca3ZdPf5mZWTIOFTMzS8ahYmZmyThUzMwsGYeKmZkl41AxM7NkHCpmZpaMQ8XMzJJxqJiZWTLv6yfqj8bmX+7hT7/yo4bW3b70DxP3xszsvcFHKmZmloxDxczMknGomJlZMrVCRVKbpKiYdpfqWiTdI2mnpL2SHpM0tWJ7oyXdImmHpP2S1ks6o6JumKTFkrZLelPSJknzGn+7ZmY2kPp7pPIl4LOF6T93L5AkYDUwG7gcmAeMANZKOqm0nXuBhcB1wNnADuARSaeV6m4Erge+CZwJbAAelHRWP/ttZmaDoL93fz0XERt6WDYHOB2YGRFrASStB7YBXyYLJCSdClwILIiI+/K2dUAHcEO+HSSdCFwNLI2IW/N9rJU0CVgK/LiffTczswGW8prKHOCV7kABiIg9wBpgbqnuILCyUHcIeACYJWlU3jwLGAksL+1nOTBV0sSEfTczswT6GyorJL0t6deS/pek8YVlU4AtFet0AOMlHVeo2xYR+yrqRgKTCnUHgOcr6gAm97PvZmY2wOqe/toD3AasA14Hfgf4a2C9pN+JiNeAscD2inU783kL0JXX7eqlbmxhvjsioo+6I0haBCwCaG1tpb29vbf31aPWY+CqqYcaWrfRfVo9XV1d/oybkMeleQ3m2NQKlYj4F+BfCk3rJP0D8E9k10q+CggoBwB5e/l1yrpyX5cBywCmTZsWjf4/s+9csYrbNjf2gwPb5ze2T6vH/y/05uRxaV6DOTYNX1OJiI3AL4D/lDd1Un300JLPd9Ws6yzMW/K7ynqrMzOzJnG0F+qLRxMdZNdByiYDL0ZEV6FuoqQxFXVvcfgaSgcwCjilog7g2aPot5mZDYCGQ0XSNOCTwBN502pgnKTphZrjgXPyZRTqRgDnFeqGAxcAj0bEgbz5YbKQmV/a9UXAlojY1mjfzcxsYNS6aCBpBdnzJhuB3WQX6hcDvwTuzMtWA+uB5ZKuITvdtZjsaObm7m1FxNOSVgJ3SBqRb/cyYCKFAImI1yTdDiyW9Ea+7wuAmRx5i7KZmTWJuleitwCfJ3tSfgzwKvAD4L9HxE6AiHhH0tnArcBdwGiykJkRES+VtncJcBOwBDgB2ATMzq/TFF1LdsfYFcBHga3A+RGxpj9v0szMBkfdu7++BnytRl0nsCCfeqvbD1yZT73VvU0WPEvq9NPMzIaWf6XYzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDYWKpIclhaQlpfYWSfdI2ilpr6THJE2tWH+0pFsk7ZC0X9J6SWdU1A2TtFjSdklvStokaV4jfTYzs4HX71CR9Hng1Ip2AauB2cDlwDxgBLBW0kml8nuBhcB1wNnADuARSaeV6m4Erge+CZwJbAAelHRWf/ttZmYDr1+hIukE4HbgyorFc4DTgYsj4v6IeDhvGwZ8ubCNU4ELgb+MiG9HxE+B84EXgRsKdScCVwNLI+LWiFgbEZcCa4Gl/em3mZkNjv4eqdwMdETE/RXL5gCvRMTa7oaI2AOsAeaW6g4CKwt1h4AHgFmSRuXNs4CRwPLSfpYDUyVN7GffzcxsgNUOFUmnA38M/FkPJVOALRXtHcB4SccV6rZFxL6KupHApELdAeD5ijqAyXX7bmZmg6NWqEgaAdwN3BoRW3soGwvsqmjvzOctNevGFua7IyL6qDMzsyYxvGbdXwHHADf1UiOgHADd7QNZd+RCaRGwCKC1tZX29vbeynvUegxcNfVQQ+s2uk+rp6ury59xE/K4NK/BHJs+Q0XSeOBa4AvAqMI1D/LXJwBvkB1BVB09dB+hdB+ddALje6nrLMxbJKl0tFKuO0JELAOWAUybNi3a2tp6eGe9u3PFKm7bXDdzj7R9fmP7tHra29tpdFxt4Hhcmtdgjk2d018nA6PJLpDvKkyQ3Z21C5hKdq1jSsX6k4EXI6Irf90BTJQ0pqLuLQ5fQ+kARgGnVNQBPFuj72ZmNojqhMrTwIyKCbKgmUEWBKuBcZKmd68o6XjgnHxZt9Vkz6+cV6gbDlwAPBoRB/Lmh8lCZn6pPxcBWyJiW42+m5nZIOrz/E5E7Abay+3Zs468EBHt+evVwHpguaRryI5gFpNdA7m5sL2nJa0E7shvANgGXAZMpBAgEfGapNuBxZLeADaSBc9MjrxF2czMmkRjFw0qRMQ7ks4GbgXuIjtlth6YEREvlcovIbvovwQ4AdgEzI6IjaW6a4Eu4Argo8BW4PyIWJOq32Zmlk7DoRIR77oLKyI6gQX51Nu6+8meyq96Mr9Y9zZZ8Czprc7MzJqDf6XYzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0umVqhImiXpcUmvSjog6WVJ35M0uVT3MUnfl7RH0uuSfiBpfMX2WiTdI2mnpL2SHpM0taJutKRbJO2QtF/SeklnNP52zcxsINU9UhkLPAX8OfAHwGJgCrBB0scBJI0BHgc+DfwJcDHwCWCtpGO7NyRJwGpgNnA5MA8YkdedVNrvvcBC4DrgbGAH8Iik0/r9Ts3MbMANr1MUEfcD9xfbJP0T8HPgXOA2si//k4FPRcTzec0zwL8ClwLfyFedA5wOzIyItXndemAb8GXgS3nbqcCFwIKIuC9vWwd0ADfk2zEzsyZyNNdUfp3PD+bzOcCG7kABiIhtwM+AuYX15gCvdAdKXrcHWFNRdxBYWag7BDwAzJI06ij6bmZmA6BfoSLpQ5JGSvoEcDfwKtmXPGSnw7ZUrNYBFK+99FY3XtJxhbptEbGvom4kMKk/fTczs4FX6/RXwRPA7+Z/fp7sFNZr+euxwK6KdTqBlsLrscD2HurIa7v62F73dt5F0iJgEUBrayvt7e1VZX1qPQaumnqooXUb3afV09XV5c+4CXlcmtdgjk1/Q+Vi4HiyaydXAz+RdHpEbM+XR8U6qnidsu4IEbEMWAYwbdq0aGtr6628R3euWMVtm/v78WS2z29sn1ZPe3s7jY6rDRyPS/MazLHp1+mviHguIp7IL9x/DjgO+Eq+eBfVRw8tHHnE0dlLHYXavuo6K5aZmdkQavhCfUTsJjsF1n1to4PsOkjZZODZwuve6l6MiK5C3cT8VuVy3Vv5vs3MrIk0HCqSWsmeSfm3vGk18BlJJxdqJgC/ny+jUDdO0vRC3fHAORV1I4DzCnXDgQuARyPiQKN9NzOzgVHrooGkHwIbgWeA14FPAn8JHCJ7RgXg22QPR66S9FWy6yE3Ai+R3SnWbTWwHlgu6Rqy012Lya6V3NxdFBFPS1oJ3CFpBNlzLJcBE4H5jbxZMzMbWHWPVDYAfwR8B/gRcCWwDjgtIn4BEBF7gZnAL4DvAivIgmBm4ZQWEfEO2dPxPwHuAn4IvA3MiIiXSvu9BLgPWJLv92PA7IjY2O93amZmA67uE/VfB75eo+5Fsp9d6auuE1iQT73V7ScLsCvr9NPMzIaWf6XYzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyfYaKpHMlPSTpBUn7JW2V9DVJv1Wqa5F0j6SdkvZKekzS1IrtjZZ0i6Qd+fbWSzqjom6YpMWStkt6U9ImSfOO7u2amdlAqnOkcjXwNvDXwGzgW8BlwE8kDQOQJGB1vvxyYB4wAlgr6aTS9u4FFgLXAWcDO4BHJJ1WqrsRuB74JnAmsAF4UNJZ/XuLZmY2WIbXqDknIn5VeL1OUifwHaANeByYA5wOzIyItQCS1gPbgC8DX8rbTgUuBBZExH152zqgA7gh3w6STiQLs6URcWu+37WSJgFLgR83+obNzGzg9HmkUgqUbv+cz8fl8znAK92Bkq+3B1gDzC2sNwc4CKws1B0CHgBmSRqVN88CRgLLS/tdDkyVNLGvfpuZ2eBr9EL99Hz+XD6fAmypqOsAxks6rlC3LSL2VdSNBCYV6g4Az1fUAUxusN9mZjaA+h0qksaRnap6LCKezJvHArsqyjvzeUvNurGF+e6IiD7qzMysidS5pvIb+RHHKuAQcElxEVAOgO728uuUdVV9XAQsAmhtbaW9vb2vVSq1HgNXTT3U0LqN7tPq6erq8mfchDwuzWswx6Z2qEgaTXaH18nA9Ih4ubC4k+qjh+4jlF2FuvG91HUW5i2SVDpaKde9S0QsA5YBTJs2Ldra2noq7dWdK1Zx2+Z+Ze5vbJ/f2D6tnvb2dhodVxs4HpfmNZhjU+v0l6QRwEPA7wFnRcTmUkkH2XWQssnAixHRVaibKGlMRd1bHL6G0gGMAk6pqAN4tk6/zcxscNV5+HEYsAL4HDA3IjZUlK0GxkmaXljveOCcfFmxbgRwXqFuOHAB8GhEHMibHyYLmfml/VwEbImIbX3128zMBl+d8zv/kywEbgL2SvpMYdnL+Wmw1cB6YLmka8hOdy0muwZyc3dxRDwtaSVwR370s43sQcqJFAIkIl6TdDuwWNIbwEay4JnJkbcom5lZE6kTKmfm82vzqehvgesj4h1JZwO3AncBo8lCZkZEvFRa5xKygFoCnABsAmZHxMZS3bVAF3AF8FFgK3B+RKyp88bMzGzw9RkqETGhzoYiohNYkE+91e0Hrsyn3ureJgueJXX2b2ZmQ8+/UmxmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJeNQMTOzZBwqZmaWjEPFzMyScaiYmVkyDhUzM0vGoWJmZsk4VMzMLBmHipmZJVMrVCSdJOlOSesl7ZMUkiZU1I2WdIukHZL25/VnVNQNk7RY0nZJb0raJGleD/teKOnnkg5I2irpi/19k2ZmNjjqHqlMAs4HdgH/2EvdvcBC4DrgbGAH8Iik00p1NwLXA98EzgQ2AA9KOqtYJGkhcDfwEDAbeBC4S9JlNfttZmaDaHjNun+IiFYASV8A/qBcIOlU4EJgQUTcl7etAzqAG4A5eduJwNXA0oi4NV99raRJwFLgx3ndcOAm4LsRcW2h7t8DN0q6JyIO9vcNm5nZwKl1pBIR79QomwMcBFYW1jsEPADMkjQqb54FjASWl9ZfDkyVNDF//VngIxV13wV+Gzi9Tt/NzGzwpLxQPwXYFhH7Su0dZCEyqVB3AHi+og5gcqEOYEsfdWZm1iTqnv6qYyzZNZeyzsLy7vnuiIgadVRss1x3BEmLgEUAra2ttLe399nxKq3HwFVTDzW0bqP7tHq6urr8GTchj0vzGsyxSRkqAspB0d3eaB091PYoIpYBywCmTZsWbW1t/Vn9N+5csYrbNjf28Wyf39g+rZ729nYaHVcbOB6X5jWYY5Py9Fcn1UcPLYXl3fMWSeUQqaqjYptjS8vNzKxJpAyVDmCipDGl9snAWxy+htIBjAJOqagDeLZQB4evrfRUZ2ZmTSJlqKwGRgDndTfktwVfADwaEQfy5ofJQmZ+af2LgC0RsS1/vR7Y2UNdJ/CzhH03M7MEal80kHRu/sffzednSvoV8KuIWBcRT0taCdwhaQSwDbgMmEghGCLiNUm3A4slvQFsJAuemcDcQt1BSX9D9rDjL4HH8poFwOUR8VZjb9nMzAZKf65EP1h6fVc+Xwe05X++hOyBxSXACcAmYHZEbCytey3QBVwBfBTYCpwfEWuKRRHxd5ICuAq4BngR+POIuAszM2s6tUMlIsoX1qtq9gNX5lNvdW+TBc+SGtu8m+ynWszMrMn5V4rNzCwZh4qZmSXjUDEzs2QcKmZmloxDxczMknGomJlZMg4VMzNLxqFiZmbJOFTMzCwZh4qZmSXjUDEzs2QcKmZmloxDxczMknGomJlZMg4VMzNLxqFiZmbJOFTMzCwZh4qZmSXjUDEzs2QcKmZmloxDxczMknGomJlZMg4VMzNLxqFiZmbJDB/qDrwfTfjKjxped/vSP0zYEzOzweUjFTMzS6bpQ0XSxyR9X9IeSa9L+oGk8UPdLzMze7emDhVJY4DHgU8DfwJcDHwCWCvp2KHsm5mZvVuzX1NZCJwMfCoingeQ9Azwr8ClwDeGsG9mZlbS7KEyB9jQHSgAEbFN0s+AubwPQ+VoLvKDL/Sb2dBq9lCZAqyqaO8Azhvkvrwn+M4zMxtKzR4qY4FdFe2dQMsg9+V972iPkgbLVVMP8aeFvjoMzZpHs4cKQFS0qadiSYuARfnLLklbG9zvh4GdDa5rA+hLpbHR14ewM1bkfzPNq79j8/FGd9TsobKL7GilrIXqIxgiYhmw7Gh3LOnJiJh2tNux9Dw2zcnj0rwGc2ya+pZismsnUyraJwPPDnJfzMysD80eKquBz0g6ubtB0gTg9/NlZmbWRJo9VL4NbAdWSZoraQ7Z3WAvAXcP8L6P+hSaDRiPTXPyuDSvQRsbRVRdB28e+U+y3A78F7IL9D8F/iIitg9lv8zM7N2aPlTMzOy9o9lPfw0q/3jlwJF0kqQ7Ja2XtE9S5NfHynWjJd0iaYek/Xn9GRV1wyQtlrRd0puSNkma18O+F0r6uaQDkrZK+mL6d/jeJOlcSQ9JeiH/vLdK+pqk3yrVtUi6R9JOSXslPSZpasX2ko/fB5WkWZIel/Rq/nf3ZUnfkzS5VFfreyv1GPYoIjxlR2tjyH5TbAvwR2Q/A7MZ+Dfg2KHu33t9AtqA/wf8GHiE7PmjCRV1K4DdZL/79jngB8B+4LRS3U3AAeBqYAbZNbZ3gLNKdQvz9pvyuiX568uG+jNphgnYAHwPmA9MB/4i//w3AMPyGgH/CLwMfB6YDawje+7hpIEcvw/ylH/WtwDn5mNzMdkdsa8DH89ran1vDcQY9tjvof7gmmUCrgDeBiYV2iYCh4Arh7p/7/Wp+wsq//MXqkIFODVvv6TQNhzYCqwutJ2YfyH9bWn9nwLPlNZ9DfhOqe7v839MI4b6cxnqCfhIRdsf5+MwM389N389o1Dz78h+2eJ/DNT4eaocr0/ln/FV+eta31upx7C3yae/Dqv88Uqg+8cr7ShExDs1yuYAB4GVhfUOAQ8AsySNyptnASOB5aX1lwNTJU3MX38W+EhF3XeB3wZO7897eD+KiF9VNP9zPh+Xz+cAr0TE2sJ6e4A1HPlvI/X42bv9Op8fzOd1v7dSj2GPHCqHTSE7hCzrIHvY0gbeFGBbROwrtXeQfQlNKtQdAJ6vqIPD49X94Gx5XMt1dqTp+fy5fN7bv43xko4r1KUcPwMkfUjSSEmfIDtN+CrZlzzU/95KPYY9cqgc5h+vHHq9jUH38u757siPzfuoo2Kb5TrLSRoH3AA8FhFP5s19jUtLzbr+jp9lniAL4V8A/4HstORr+bK631upx7BHDpUj9evHKy05UW8M+lNHD7VWkv/X6iqy8/GXFBeRflz8b62+i4HPABeSXaT/SenOyaYaG4fKYf3+8UpLrpOex6B7efe8RVL5L3pVHRXbHFta/oEnaTTZTx+dDMyKiJcLi/sal1016/o7fgZExHMR8URE3E92N9ZxwFfyxXW/t1KPYY8cKof5xyuHXgcwUdKYUvtk4C0On4PvAEYBp1TUweHx6j5HXx7Xct0HmqQRwEPA75Hd0ru5VNLbv40XI6KrUJdy/KwkInaTfY7d1zbqfm+lHsMeOVQO849XDr3VwAgK/1dPScOBC4BHI+JA3vww2V/w+aX1LwK25He/AKwnu3W4qq6T7A6ZDzRJw8ieS/gcMDciNlSUrQbGSZpeWO944ByO/LeRevysRFIr8Gmy51Cg/vdW6jHs2VDfd90sE3AsWQpvJrvFbg6wCfi/wHFD3b/3w0T2ENe5wLfIzttelr+eXqh5gOxQ/AtkX3TfB94E/mNpW0vz9ivJHqz8FtnDc+eU6r6Yty/J627IX/+3of48mmEqjMUSsvP2xemkvGYY8H/Ifsj1v5LdEtxOFswfK20v6fh9kCfgh8Df5N9HM4BLgZ+TPZj4ybym1vfWQIxhj/0e6g+umSZgPNlpgNeBN4D/TcVT354a/nyjh6m9UHMM8A2y2ybfJLvzpa1iWx8Cvgq8QHZnzDPAuT3s91KyO2cOkD19/GdD/Vk0y0T2K+A9jcv1hbqxZA+NdgL7yB5UPLVie8nH74M6AX8FPJWHyD6yBxDvLn8n1f3eSj2GPU3+QUkzM0vG11TMzCwZh4qZmSXjUDEzs2QcKmZmloxDxczMknGomJlZMg4VMzNLxqFiZmbJOFTMzCyZ/w9oZtvoCeWz0QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Para cada coluna mencionada, gera e exibe a sua distribuição\n",
    "for col in ['QT_TOTAL_HIT_PAYWALL', 'DIASNAVEGADOS', 'NOTICIASLIDAS', 'VISITAS_CAPA']:\n",
    "    hist = df[col].hist(bins=20)\n",
    "    print(\"Distribuição de {}\".format(col))\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tratamento de outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 373,
   "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"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "# Função que encontra os outliers e os substitui pela mediana correspondentes\n",
    "def outliers(col):\n",
    "    q1, q3 = np.percentile(df[col],[2.5,97.5])\n",
    "    iqr = q3 - q1\n",
    "    lower_bound = q1 -(1.5 * iqr) \n",
    "    upper_bound = q3 +(1.5 * iqr)\n",
    "    x = df[(df[col] > lower_bound) | (df[col] < upper_bound)][col].median()\n",
    "    df[col][(df[col] < lower_bound) | (df[col] > upper_bound)] = x\n",
    "    hist = df[col].hist(bins=20)\n",
    "    plt.show()\n",
    "\n",
    "for col in ['QT_TOTAL_HIT_PAYWALL', 'DIASNAVEGADOS', 'NOTICIASLIDAS', 'VISITAS_CAPA']:\n",
    "    outliers(col)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gênero\n",
    "\n",
    "- Qual a porcentagem de pessoas por gênero?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 374,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_temp = df['PES_GENERO'].value_counts(normalize = True)\n",
    "fig,ax = plt.subplots(figsize = (12,6))\n",
    "ax.bar(df_temp.index, df_temp.values*100, width = 0.8)\n",
    "ax.set_ylabel('%')\n",
    "ax.set_xlabel('Gênero')\n",
    "plt.title(\"Porcentagem de pessoas por gênero\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Uso do app\n",
    "\n",
    "- Qual a porcentagem de pessoas que utilizam o app na amostra?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 375,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_temp = df['USOU_APP'].value_counts(normalize = True)\n",
    "fig,ax = plt.subplots(figsize = (12,6))\n",
    "ax.bar(df_temp.index, df_temp.values*100, width = 0.8)\n",
    "ax.set_ylabel('%')\n",
    "ax.set_xlabel('Usa o app')\n",
    "plt.title(\"Porcentagem de pessoas que utilizam o app\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Assinantes vs Prospects\n",
    "- Qual a porcentagem de pessoas que são assinantes?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 376,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_temp = df['PERFIL'].value_counts(normalize = True)\n",
    "fig,ax = plt.subplots(figsize = (12,6))\n",
    "ax.bar(df_temp.index, df_temp.values*100, width = 0.8)\n",
    "ax.set_ylabel('%')\n",
    "ax.set_xlabel('Perfil')\n",
    "plt.title(\"Porcentagem de pessoas que são assinantes\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Distribuição de renda dos assinantes\n",
    "- Como é a distribuição de renda dos assinantes da amostra?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 377,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_temp = df['ATR_PF_GEO_RENDA_FAM'][(df.PERFIL == 'ASSINANTE')].value_counts(normalize = True)\n",
    "df_temp = df_temp[['ATE 1SM', 'DE 2SM ATE 3SM', 'DE 3SM ATE 4SM', 'DE 4SM ATE 8SM', 'DE 8SM ATE 14SM', 'DE 14SM ATE 25SM', 'ACIMA DE 25SM']]\n",
    "fig,ax = plt.subplots(figsize = (25,8))\n",
    "ax.bar(df_temp.index, df_temp.values*100, width = 0.8)\n",
    "ax.set_ylabel('%')\n",
    "ax.set_xlabel('Faixa de renda')\n",
    "plt.title(\"Distribuição de renda dos assinantes\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Correlações"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 378,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                      QT_TOTAL_HIT_PAYWALL  DIASNAVEGADOS  NOTICIASLIDAS  \\\n",
      "QT_TOTAL_HIT_PAYWALL              1.000000       0.459818       0.061129   \n",
      "DIASNAVEGADOS                     0.459818       1.000000       0.539521   \n",
      "NOTICIASLIDAS                     0.061129       0.539521       1.000000   \n",
      "VISITAS_CAPA                      0.317927       0.666299       0.343202   \n",
      "\n",
      "                      VISITAS_CAPA  \n",
      "QT_TOTAL_HIT_PAYWALL      0.317927  \n",
      "DIASNAVEGADOS             0.666299  \n",
      "NOTICIASLIDAS             0.343202  \n",
      "VISITAS_CAPA              1.000000  \n"
     ]
    }
   ],
   "source": [
    "print(df.corr())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pré-procesamento"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Divisão de features e target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 379,
   "metadata": {},
   "outputs": [],
   "source": [
    "target = df['PERFIL']\n",
    "features = df.drop('PERFIL', axis = 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Transformação de data de nascimento para idade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 380,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Função para calcular a idade baseada na diferença entre a data do dia atual e a data do nascimento\n",
    "# Retorna a idade em um número inteiro\n",
    "def calcula_idade(nascimento):\n",
    "    nascimento = nascimento.date()\n",
    "    hoje = date.today()\n",
    "    return hoje.year - nascimento.year - ((hoje.month, hoje.day) < (nascimento.month, nascimento.day))\n",
    "\n",
    "# Aplica a função para na variável da data de nascimento\n",
    "features['IDADE'] = pd.to_datetime(features['PES_NASCIMENTO_DATA'], errors = 'coerce').apply(calcula_idade)\n",
    "\n",
    "# Remove a coluna da data de nascimento\n",
    "features = features.drop('PES_NASCIMENTO_DATA', axis = 1)\n",
    "\n",
    "# Preenche os dados ausentes com a média das idades\n",
    "features['IDADE'] = features['IDADE'].fillna(features['IDADE'].mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Distribuição de idade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 381,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x111173358>"
      ]
     },
     "execution_count": 381,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "features['IDADE'].hist()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Normalização"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 382,
   "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>QT_TOTAL_HIT_PAYWALL</th>\n",
       "      <th>DIASNAVEGADOS</th>\n",
       "      <th>NOTICIASLIDAS</th>\n",
       "      <th>VISITAS_CAPA</th>\n",
       "      <th>USOU_APP</th>\n",
       "      <th>PES_GENERO</th>\n",
       "      <th>ATR_PF_GEO_RENDA_FAM</th>\n",
       "      <th>IDADE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NAO</td>\n",
       "      <td>F</td>\n",
       "      <td>DE 4SM ATE 8SM</td>\n",
       "      <td>0.246914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.007576</td>\n",
       "      <td>0.383333</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.428402</td>\n",
       "      <td>SIM</td>\n",
       "      <td>M</td>\n",
       "      <td>DE 3SM ATE 4SM</td>\n",
       "      <td>0.301414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.037879</td>\n",
       "      <td>0.266667</td>\n",
       "      <td>0.026829</td>\n",
       "      <td>0.004734</td>\n",
       "      <td>NAO</td>\n",
       "      <td>M</td>\n",
       "      <td>None</td>\n",
       "      <td>0.301414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.068182</td>\n",
       "      <td>0.216667</td>\n",
       "      <td>0.019512</td>\n",
       "      <td>0.029586</td>\n",
       "      <td>SIM</td>\n",
       "      <td>F</td>\n",
       "      <td>ACIMA DE 25SM</td>\n",
       "      <td>0.301414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.007576</td>\n",
       "      <td>0.016667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.010651</td>\n",
       "      <td>SIM</td>\n",
       "      <td>M</td>\n",
       "      <td>DE 4SM ATE 8SM</td>\n",
       "      <td>0.283951</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   QT_TOTAL_HIT_PAYWALL  DIASNAVEGADOS  NOTICIASLIDAS  VISITAS_CAPA USOU_APP  \\\n",
       "0              0.000000       0.000000       0.000000      0.000000      NAO   \n",
       "1              0.007576       0.383333       0.000000      0.428402      SIM   \n",
       "2              0.037879       0.266667       0.026829      0.004734      NAO   \n",
       "3              0.068182       0.216667       0.019512      0.029586      SIM   \n",
       "4              0.007576       0.016667       0.000000      0.010651      SIM   \n",
       "\n",
       "  PES_GENERO ATR_PF_GEO_RENDA_FAM     IDADE  \n",
       "0          F       DE 4SM ATE 8SM  0.246914  \n",
       "1          M       DE 3SM ATE 4SM  0.301414  \n",
       "2          M                 None  0.301414  \n",
       "3          F        ACIMA DE 25SM  0.301414  \n",
       "4          M       DE 4SM ATE 8SM  0.283951  "
      ]
     },
     "execution_count": 382,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = MinMaxScaler()\n",
    "columns = ['QT_TOTAL_HIT_PAYWALL', 'DIASNAVEGADOS', 'NOTICIASLIDAS', 'VISITAS_CAPA', 'IDADE']\n",
    "\n",
    "# Aplica a normalização nas colunas mencionadas\n",
    "features[columns] = scaler.fit_transform(features[columns])\n",
    "\n",
    "features.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### One hot encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 383,
   "metadata": {},
   "outputs": [],
   "source": [
    "features = pd.get_dummies(features)\n",
    "\n",
    "target = target.replace({'PROSPECT ':0, 'ASSINANTE':1})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Split em dataset de treino e teste"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 384,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset de treino tem 4480 registros.\n",
      "Dataset de teste tem 1120 registros.\n"
     ]
    }
   ],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(features, target, test_size = 0.2, random_state = 0)\n",
    "\n",
    "print(\"Dataset de treino tem {} registros.\".format(X_train.shape[0]))\n",
    "print(\"Dataset de teste tem {} registros.\".format(X_test.shape[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Benchmark"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Aplicação dos dados em uma árvore de decisão"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 385,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DecisionTreeClassifier treinado.\n",
      "acc_train: 0.996875\n",
      "acc_test: 0.9089285714285714\n",
      "f_train: 0.9849785407725321\n",
      "f_test: 0.6076923076923078\n"
     ]
    }
   ],
   "source": [
    "# Treinamento do benchmark com uma árvore de decisão\n",
    "learner = tree.DecisionTreeClassifier()\n",
    "learner = learner.fit(X_train, y_train)\n",
    "\n",
    "predictions_test = learner.predict(X_test)\n",
    "predictions_train = learner.predict(X_train)\n",
    "\n",
    "print(\"{} treinado.\".format(learner.__class__.__name__))\n",
    "print(\"acc_train: {}\".format(accuracy_score(y_train, predictions_train)))\n",
    "print(\"acc_test: {}\".format(accuracy_score(y_test, predictions_test)))\n",
    "print(\"f_train: {}\".format(f1_score(y_train, predictions_train)))\n",
    "print(\"f_test: {}\".format(f1_score(y_test, predictions_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modelo"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Pipeline de treinamento e predição"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 386,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Função para testar algoritmos de uma maneira fácil\n",
    "def train_predict(learner, X_train, y_train, X_test, y_test): \n",
    "    \n",
    "    results = {}\n",
    "\n",
    "    learner = learner.fit(X_train, y_train)\n",
    "    \n",
    "    predictions_test = learner.predict(X_test)\n",
    "    predictions_train = learner.predict(X_train)\n",
    "    \n",
    "    results['acc_train'] = accuracy_score(y_train, predictions_train)\n",
    "    results['acc_test'] = accuracy_score(y_test, predictions_test)\n",
    "    results['f_train'] = f1_score(y_train, predictions_train)\n",
    "    results['f_test'] = f1_score(y_test, predictions_test)\n",
    "    \n",
    "    print(\"{} treinado.\".format(learner.__class__.__name__))\n",
    "        \n",
    "    return results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Validação inicial do modelo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 387,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForestClassifier treinado.\n",
      "GradientBoostingClassifier treinado.\n",
      "KNeighborsClassifier treinado.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'RandomForestClassifier': {'acc_train': 0.9897321428571428,\n",
       "  'acc_test': 0.9464285714285714,\n",
       "  'f_train': 0.9491150442477876,\n",
       "  'f_test': 0.7115384615384616},\n",
       " 'GradientBoostingClassifier': {'acc_train': 0.9558035714285714,\n",
       "  'acc_test': 0.9508928571428571,\n",
       "  'f_train': 0.7367021276595745,\n",
       "  'f_test': 0.7236180904522613},\n",
       " 'KNeighborsClassifier': {'acc_train': 0.9455357142857143,\n",
       "  'acc_test': 0.9375,\n",
       "  'f_train': 0.6620498614958448,\n",
       "  'f_test': 0.6568627450980392}}"
      ]
     },
     "execution_count": 387,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_A = RandomForestClassifier()\n",
    "clf_B = GradientBoostingClassifier()\n",
    "clf_C = KNeighborsClassifier()\n",
    "\n",
    "results = {}\n",
    "for clf in [clf_A, clf_B, clf_C]:\n",
    "    clf_name = clf.__class__.__name__\n",
    "    results[clf_name] = {}\n",
    "    results[clf_name] = train_predict(clf, X_train, y_train, X_test, y_test)\n",
    "\n",
    "results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 388,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Função para gerar o gráfico comparativo das métricas entre os algoritmos\n",
    "def evaluate(results):\n",
    "\n",
    "    fig, ax = plt.subplots(2, 2, figsize = (11,6))\n",
    "\n",
    "    bar_width = 0.34\n",
    "    colors = ['#A00000','#00A0A0','#00A000']\n",
    "    \n",
    "    for k, learner in enumerate(results.keys()):\n",
    "        for j, metric in enumerate(['acc_train', 'f_train', 'acc_test', 'f_test']):\n",
    "            ax[j//2, j%2].bar(0.12+1*k*bar_width, results[learner][metric], width = bar_width, color = colors[k])\n",
    "            ax[j//2, j%2].set_xlabel(\" \")\n",
    "            ax[j//2, j%2].set_xlim((-1, 2))\n",
    "\n",
    "    ax[0, 0].set_title(\"Acurácia no dataset de treino\")\n",
    "    ax[0, 1].set_title(\"F-score no dataset de treino\")\n",
    "    ax[1, 0].set_title(\"Acurácia no dataset de teste\")\n",
    "    ax[1, 1].set_title(\"F-score no dataset de teste\")\n",
    "    \n",
    "    ax[0, 0].set_ylim((0, 1))\n",
    "    ax[0, 1].set_ylim((0, 1))\n",
    "    ax[1, 0].set_ylim((0, 1))\n",
    "    ax[1, 1].set_ylim((0, 1))\n",
    "    \n",
    "    ax[0, 0].xaxis.set_major_locator(plt.NullLocator())\n",
    "    ax[0, 1].xaxis.set_major_locator(plt.NullLocator())\n",
    "    ax[1, 0].xaxis.set_major_locator(plt.NullLocator())\n",
    "    ax[1, 1].xaxis.set_major_locator(plt.NullLocator())\n",
    "    \n",
    "    patches = []\n",
    "    for i, learner in enumerate(results.keys()):\n",
    "        patches.append(mpatches.Patch(color = colors[i], label = learner))\n",
    "    plt.legend(handles = patches, bbox_to_anchor = (-0.15, 3.17), \\\n",
    "               loc = 'upper center', borderaxespad = 0., ncol = 3, fontsize = 'large')\n",
    "    \n",
    "    plt.suptitle(\"Métricas de performace nos três modelos\", fontsize = 16, y = 1.3)\n",
    "    plt.show()\n",
    "    \n",
    "evaluate(results)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Tuning do modelo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 390,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Modelo não-otimizado\n",
      "------\n",
      "Acurácia no dataset de teste: 0.9509\n",
      "F-score no dataset de teste: 0.7236\n",
      "\n",
      " Modelo otimizado\n",
      "------\n",
      "Acurácia final no dataset de teste: 0.9500\n",
      "F-score final no dataset de teste: 0.7358\n",
      "\n",
      " Modelo final\n",
      "------\n",
      "GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
      "              learning_rate=0.05, loss='deviance', max_depth=3,\n",
      "              max_features=None, max_leaf_nodes=None,\n",
      "              min_impurity_split=1e-07, min_samples_leaf=1,\n",
      "              min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
      "              n_estimators=500, presort='auto', random_state=None,\n",
      "              subsample=0.5, verbose=0, warm_start=False)\n"
     ]
    }
   ],
   "source": [
    "clf = GradientBoostingClassifier()\n",
    "\n",
    "# Parâmetros e valores para serem testados no GridSearchCV\n",
    "parameters = {\n",
    "    'n_estimators': [100, 200, 500, 1000],\n",
    "    \"learning_rate\": [0.01, 0.05, 0.075, 0.1],\n",
    "    \"subsample\": [0.5, 0.8, 0.9, 1.0],\n",
    "}\n",
    "\n",
    "scorer = make_scorer(f1_score)\n",
    "\n",
    "grid_obj = GridSearchCV(clf, parameters, scoring = scorer)\n",
    "\n",
    "grid_fit = grid_obj.fit(X_train, y_train)\n",
    "\n",
    "best_clf = grid_fit.best_estimator_\n",
    "\n",
    "predictions = (clf.fit(X_train, y_train)).predict(X_test)\n",
    "best_predictions = best_clf.predict(X_test)\n",
    "\n",
    "print(\"Modelo não-otimizado\\n------\")\n",
    "print(\"Acurácia no dataset de teste: {:.4f}\".format(accuracy_score(y_test, predictions)))\n",
    "print(\"F-score no dataset de teste: {:.4f}\".format(f1_score(y_test, predictions)))\n",
    "print(\"\\n Modelo otimizado\\n------\")\n",
    "print(\"Acurácia final no dataset de teste: {:.4f}\".format(accuracy_score(y_test, best_predictions)))\n",
    "print(\"F-score final no dataset de teste: {:.4f}\".format(f1_score(y_test, best_predictions)))\n",
    "print(\"\\n Modelo final\\n------\")\n",
    "print(best_clf)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}