Aprendendo_Scikit_learn_ML.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![logos](http://cdn.wittysparks.com/wp-content/uploads/2017/08/09210754/Machine-Learning.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nesse tutorial teremos o 1º contato com a library de Aprendizado de máquina: <b> Scikit-learn </b>.<p>\n",
    "Vamos aprender e interagir com algo supreendente chamado <b> IA - Inteligência Artificial </b>, para o melhor aproveitamento faz-se necessário conhecer <b> PANDAS </b> (ou biblioteca incrível) e sobre visualização de dados (pois criaremos gráficos incríveis).<p>\n",
    "    Não tenha receio de começar com <b> PYTHON </b>, inicie-se e aventure-se! O conjunto de dados escolhido está no site da <b> IBM </b>, com infos da campanha de vendas de um fornecedor atacadista de peças automotivas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#acessando a URL com os dados\n",
    "#https://www.dataquest.io/blog/sci-kit-learn-tutorial/\n",
    "import pandas as pd\n",
    "url = \"https://community.watsonanalytics.com/wp-content/uploads/2015/04/WA_Fn-UseC_-Sales-Win-Loss.csv\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#converter o csv em um dataframe\n",
    "sales_data = pd.read_csv(url)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Opportunity Number</th>\n",
       "      <th>Supplies Subgroup</th>\n",
       "      <th>Supplies Group</th>\n",
       "      <th>Region</th>\n",
       "      <th>Route To Market</th>\n",
       "      <th>Elapsed Days In Sales Stage</th>\n",
       "      <th>Opportunity Result</th>\n",
       "      <th>Sales Stage Change Count</th>\n",
       "      <th>Total Days Identified Through Closing</th>\n",
       "      <th>Total Days Identified Through Qualified</th>\n",
       "      <th>Opportunity Amount USD</th>\n",
       "      <th>Client Size By Revenue</th>\n",
       "      <th>Client Size By Employee Count</th>\n",
       "      <th>Revenue From Client Past Two Years</th>\n",
       "      <th>Competitor Type</th>\n",
       "      <th>Ratio Days Identified To Total Days</th>\n",
       "      <th>Ratio Days Validated To Total Days</th>\n",
       "      <th>Ratio Days Qualified To Total Days</th>\n",
       "      <th>Deal Size Category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1641984</td>\n",
       "      <td>Exterior Accessories</td>\n",
       "      <td>Car Accessories</td>\n",
       "      <td>Northwest</td>\n",
       "      <td>Fields Sales</td>\n",
       "      <td>76</td>\n",
       "      <td>Won</td>\n",
       "      <td>13</td>\n",
       "      <td>104</td>\n",
       "      <td>101</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0.69636</td>\n",
       "      <td>0.113985</td>\n",
       "      <td>0.154215</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1658010</td>\n",
       "      <td>Exterior Accessories</td>\n",
       "      <td>Car Accessories</td>\n",
       "      <td>Pacific</td>\n",
       "      <td>Reseller</td>\n",
       "      <td>63</td>\n",
       "      <td>Loss</td>\n",
       "      <td>2</td>\n",
       "      <td>163</td>\n",
       "      <td>163</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1674737</td>\n",
       "      <td>Motorcycle Parts</td>\n",
       "      <td>Performance &amp; Non-auto</td>\n",
       "      <td>Pacific</td>\n",
       "      <td>Reseller</td>\n",
       "      <td>24</td>\n",
       "      <td>Won</td>\n",
       "      <td>7</td>\n",
       "      <td>82</td>\n",
       "      <td>82</td>\n",
       "      <td>7750</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1675224</td>\n",
       "      <td>Shelters &amp; RV</td>\n",
       "      <td>Performance &amp; Non-auto</td>\n",
       "      <td>Midwest</td>\n",
       "      <td>Reseller</td>\n",
       "      <td>16</td>\n",
       "      <td>Loss</td>\n",
       "      <td>5</td>\n",
       "      <td>124</td>\n",
       "      <td>124</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Known</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1689785</td>\n",
       "      <td>Exterior Accessories</td>\n",
       "      <td>Car Accessories</td>\n",
       "      <td>Pacific</td>\n",
       "      <td>Reseller</td>\n",
       "      <td>69</td>\n",
       "      <td>Loss</td>\n",
       "      <td>11</td>\n",
       "      <td>91</td>\n",
       "      <td>13</td>\n",
       "      <td>69756</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.141125</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Opportunity Number     Supplies Subgroup          Supplies Group  \\\n",
       "0             1641984  Exterior Accessories         Car Accessories   \n",
       "1             1658010  Exterior Accessories         Car Accessories   \n",
       "2             1674737      Motorcycle Parts  Performance & Non-auto   \n",
       "3             1675224         Shelters & RV  Performance & Non-auto   \n",
       "4             1689785  Exterior Accessories         Car Accessories   \n",
       "\n",
       "      Region Route To Market  Elapsed Days In Sales Stage Opportunity Result  \\\n",
       "0  Northwest    Fields Sales                           76                Won   \n",
       "1    Pacific        Reseller                           63               Loss   \n",
       "2    Pacific        Reseller                           24                Won   \n",
       "3    Midwest        Reseller                           16               Loss   \n",
       "4    Pacific        Reseller                           69               Loss   \n",
       "\n",
       "   Sales Stage Change Count  Total Days Identified Through Closing  \\\n",
       "0                        13                                    104   \n",
       "1                         2                                    163   \n",
       "2                         7                                     82   \n",
       "3                         5                                    124   \n",
       "4                        11                                     91   \n",
       "\n",
       "   Total Days Identified Through Qualified  Opportunity Amount USD  \\\n",
       "0                                      101                       0   \n",
       "1                                      163                       0   \n",
       "2                                       82                    7750   \n",
       "3                                      124                       0   \n",
       "4                                       13                   69756   \n",
       "\n",
       "   Client Size By Revenue  Client Size By Employee Count  \\\n",
       "0                       5                              5   \n",
       "1                       3                              5   \n",
       "2                       1                              1   \n",
       "3                       1                              1   \n",
       "4                       1                              1   \n",
       "\n",
       "   Revenue From Client Past Two Years Competitor Type  \\\n",
       "0                                   0         Unknown   \n",
       "1                                   0         Unknown   \n",
       "2                                   0         Unknown   \n",
       "3                                   0           Known   \n",
       "4                                   0         Unknown   \n",
       "\n",
       "   Ratio Days Identified To Total Days  Ratio Days Validated To Total Days  \\\n",
       "0                              0.69636                            0.113985   \n",
       "1                              0.00000                            1.000000   \n",
       "2                              1.00000                            0.000000   \n",
       "3                              1.00000                            0.000000   \n",
       "4                              0.00000                            0.141125   \n",
       "\n",
       "   Ratio Days Qualified To Total Days  Deal Size Category  \n",
       "0                            0.154215                   1  \n",
       "1                            0.000000                   1  \n",
       "2                            0.000000                   1  \n",
       "3                            0.000000                   1  \n",
       "4                            0.000000                   4  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#com tudo devidamente importado vamos verificar algumas coisas do dataframe\n",
    "#5 primeiras linhas ... caso queira algo difeente tente por ex: 'sales_data.head(50)'\n",
    "#advinha o que acontecerá?!?!? Por default o '.head()' mostra as 5 primeira linhas\n",
    "sales_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Exploração de Dados<p> <b> </b>\n",
    "    A exploração de dados é um passo muito importante em qualquer projeto de <b> Ciência de Dados </b>ou <b>Aprendizado de Máquina</b>. Mesmo uma exploração rápida do conjunto de dados pode nos dar informações importantes que, de outra forma, poderíamos perder, e essas informações podem sugerir questões importantes que podemos tentar responder por meio de nosso projeto.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Agora que visualizamos os registros iniciais de nosso <b>dataframe</b>, vamos tentar visualizar os últimos registros no conjunto de dados. Isso pode ser feito usando o tail()método, que tem sintaxe semelhante ao head()."
   ]
  },
  {
   "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>Opportunity Number</th>\n",
       "      <th>Supplies Subgroup</th>\n",
       "      <th>Supplies Group</th>\n",
       "      <th>Region</th>\n",
       "      <th>Route To Market</th>\n",
       "      <th>Elapsed Days In Sales Stage</th>\n",
       "      <th>Opportunity Result</th>\n",
       "      <th>Sales Stage Change Count</th>\n",
       "      <th>Total Days Identified Through Closing</th>\n",
       "      <th>Total Days Identified Through Qualified</th>\n",
       "      <th>Opportunity Amount USD</th>\n",
       "      <th>Client Size By Revenue</th>\n",
       "      <th>Client Size By Employee Count</th>\n",
       "      <th>Revenue From Client Past Two Years</th>\n",
       "      <th>Competitor Type</th>\n",
       "      <th>Ratio Days Identified To Total Days</th>\n",
       "      <th>Ratio Days Validated To Total Days</th>\n",
       "      <th>Ratio Days Qualified To Total Days</th>\n",
       "      <th>Deal Size Category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>78020</th>\n",
       "      <td>10089932</td>\n",
       "      <td>Batteries &amp; Accessories</td>\n",
       "      <td>Car Accessories</td>\n",
       "      <td>Southeast</td>\n",
       "      <td>Reseller</td>\n",
       "      <td>0</td>\n",
       "      <td>Loss</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>250000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78021</th>\n",
       "      <td>10089961</td>\n",
       "      <td>Shelters &amp; RV</td>\n",
       "      <td>Performance &amp; Non-auto</td>\n",
       "      <td>Northeast</td>\n",
       "      <td>Reseller</td>\n",
       "      <td>0</td>\n",
       "      <td>Won</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>180000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78022</th>\n",
       "      <td>10090145</td>\n",
       "      <td>Exterior Accessories</td>\n",
       "      <td>Car Accessories</td>\n",
       "      <td>Southeast</td>\n",
       "      <td>Reseller</td>\n",
       "      <td>0</td>\n",
       "      <td>Loss</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>90000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78023</th>\n",
       "      <td>10090430</td>\n",
       "      <td>Exterior Accessories</td>\n",
       "      <td>Car Accessories</td>\n",
       "      <td>Southeast</td>\n",
       "      <td>Fields Sales</td>\n",
       "      <td>0</td>\n",
       "      <td>Loss</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>120000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78024</th>\n",
       "      <td>10094255</td>\n",
       "      <td>Interior Accessories</td>\n",
       "      <td>Car Accessories</td>\n",
       "      <td>Mid-Atlantic</td>\n",
       "      <td>Reseller</td>\n",
       "      <td>0</td>\n",
       "      <td>Loss</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>90000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Opportunity Number        Supplies Subgroup          Supplies Group  \\\n",
       "78020            10089932  Batteries & Accessories         Car Accessories   \n",
       "78021            10089961            Shelters & RV  Performance & Non-auto   \n",
       "78022            10090145     Exterior Accessories         Car Accessories   \n",
       "78023            10090430     Exterior Accessories         Car Accessories   \n",
       "78024            10094255     Interior Accessories         Car Accessories   \n",
       "\n",
       "             Region Route To Market  Elapsed Days In Sales Stage  \\\n",
       "78020     Southeast        Reseller                            0   \n",
       "78021     Northeast        Reseller                            0   \n",
       "78022     Southeast        Reseller                            0   \n",
       "78023     Southeast    Fields Sales                            0   \n",
       "78024  Mid-Atlantic        Reseller                            0   \n",
       "\n",
       "      Opportunity Result  Sales Stage Change Count  \\\n",
       "78020               Loss                         2   \n",
       "78021                Won                         1   \n",
       "78022               Loss                         2   \n",
       "78023               Loss                         2   \n",
       "78024               Loss                         1   \n",
       "\n",
       "       Total Days Identified Through Closing  \\\n",
       "78020                                      0   \n",
       "78021                                      0   \n",
       "78022                                      0   \n",
       "78023                                      0   \n",
       "78024                                      0   \n",
       "\n",
       "       Total Days Identified Through Qualified  Opportunity Amount USD  \\\n",
       "78020                                        0                  250000   \n",
       "78021                                        0                  180000   \n",
       "78022                                        0                   90000   \n",
       "78023                                        0                  120000   \n",
       "78024                                        0                   90000   \n",
       "\n",
       "       Client Size By Revenue  Client Size By Employee Count  \\\n",
       "78020                       1                              1   \n",
       "78021                       1                              1   \n",
       "78022                       1                              1   \n",
       "78023                       1                              1   \n",
       "78024                       1                              1   \n",
       "\n",
       "       Revenue From Client Past Two Years Competitor Type  \\\n",
       "78020                                   3         Unknown   \n",
       "78021                                   0         Unknown   \n",
       "78022                                   0         Unknown   \n",
       "78023                                   0         Unknown   \n",
       "78024                                   0         Unknown   \n",
       "\n",
       "       Ratio Days Identified To Total Days  \\\n",
       "78020                                  0.0   \n",
       "78021                                  0.0   \n",
       "78022                                  0.0   \n",
       "78023                                  1.0   \n",
       "78024                                  0.0   \n",
       "\n",
       "       Ratio Days Validated To Total Days  Ratio Days Qualified To Total Days  \\\n",
       "78020                                 0.0                                 0.0   \n",
       "78021                                 0.0                                 0.0   \n",
       "78022                                 0.0                                 0.0   \n",
       "78023                                 0.0                                 0.0   \n",
       "78024                                 0.0                                 0.0   \n",
       "\n",
       "       Deal Size Category  \n",
       "78020                   6  \n",
       "78021                   5  \n",
       "78022                   4  \n",
       "78023                   5  \n",
       "78024                   4  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#os último 5 registros são da linha 78020 até 78024\n",
    "sales_data.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Agora, seria bom se pudéssemos ver os <b>diferentes tipos de dados disponíveis</b> neste conjunto de dados.\n",
    "Esta informação pode ser útil no caso de precisarmos fazer alguma conversão mais tarde. Podemos fazer isso com o dtypes():"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Opportunity Number                           int64\n",
       "Supplies Subgroup                           object\n",
       "Supplies Group                              object\n",
       "Region                                      object\n",
       "Route To Market                             object\n",
       "Elapsed Days In Sales Stage                  int64\n",
       "Opportunity Result                          object\n",
       "Sales Stage Change Count                     int64\n",
       "Total Days Identified Through Closing        int64\n",
       "Total Days Identified Through Qualified      int64\n",
       "Opportunity Amount USD                       int64\n",
       "Client Size By Revenue                       int64\n",
       "Client Size By Employee Count                int64\n",
       "Revenue From Client Past Two Years           int64\n",
       "Competitor Type                             object\n",
       "Ratio Days Identified To Total Days        float64\n",
       "Ratio Days Validated To Total Days         float64\n",
       "Ratio Days Qualified To Total Days         float64\n",
       "Deal Size Category                           int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sales_data.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> </b> <b> Visualização de dados</b> <p>\n",
    "Agora que fizemos algumas explorações básicas de dados, vamos tentar criar alguns gráficos interessantes para representar visualmente os dados e descobrir mais histórias ocultas no conjunto de dados. Vamos investigar e entender."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 842.4x595.44 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#usaremos uma biblioteca incrível chamada 'Seaborn', para visualização de dados\n",
    "#antes de tudo precisamos ver se a mesma encontra-se instalada\n",
    "#Dados da coluna 'Opportunity Result'\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "sns.set(style=\"whitegrid\", color_codes=True)\n",
    "sns.set(rc={'figure.figsize':(11.7,8.27)})\n",
    "\n",
    "sns.countplot('Route To Market',data=sales_data,hue = 'Opportunity Result')\n",
    "sns.despine(offset=10, trim=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Curioso e querendo saber + sobre o [Seaborn](https://seaborn.pydata.org/generated/seaborn.countplot.html) !?!?!<p>\n",
    "    Agora veremos outra popular forma de visualização de dados chamada <b>'ViolinPlots'</b> (logo mais você entenderá a razão do nome)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "sns.set(rc={'figure.figsize':(16.7,13.27)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Lucas Barros\\Anaconda3\\lib\\site-packages\\scipy\\stats\\stats.py:1713: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.\n",
      "  return np.add.reduce(sorted[indexer] * weights, axis=axis) / sumval\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1202.4x955.44 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.violinplot(x=\"Opportunity Result\",y=\"Client Size By Revenue\", hue=\"Opportunity Result\", data=sales_data);\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Prestando muito bem atenção, você perceberá que nosso gráfico representa (ou se parece risos) com um violino.<p>\n",
    "    Usamos muito esse tipode gráfico pois eles no traz como os dados estão realmente distríbuidos e quais recursos possuem maior concetração de dados. Para esse exemplo usamos a coluna <b>'Client Size By Revenue'</b>."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Dados de pré-processamento </b> <p>\n",
    "    Agora que temos uma boa compreensão de como são nossos dados, podemos nos preparar para <b> criar modelos de previsão</b> usando o scikit-learn.\n",
    "\n",
    "Vimos em nossa exploração inicial que a maioria das colunas em nosso conjunto de dados são strings, mas os algoritmos no scikit-learn entendem <b>apenas dados numéricos</b>. Felizmente, a biblioteca scikit-learn nos fornece muitos métodos para converter dados de string em dados numéricos. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1 1 2 0]\n"
     ]
    }
   ],
   "source": [
    "#import the necessary module\n",
    "from sklearn import preprocessing\n",
    "le = preprocessing.LabelEncoder()\n",
    "\n",
    "encoded_value = le.fit_transform([\"paris\", \"paris\", \"tokyo\", \"amsterdam\"])\n",
    "print(encoded_value)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Acima usamos o 'fit_transform()', método pertencente ao LabelEncoder() para codificar a coluna 'Route To Market'<p>\n",
    "    4 valores foram utilizados para o nosso exemplo: \"paris\", \"paris\", \"tokyo\", \"amsterdam\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Se estou conseguindo deixar você mais curioso, estou indo bem então! risos<p>\n",
    "    Para saber mais sobre o [scikit-learn](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.LabelEncoder.html) !?!?!<p>\n",
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Supplies Subgroup' :  ['Exterior Accessories' 'Motorcycle Parts' 'Shelters & RV'\n",
      " 'Garage & Car Care' 'Batteries & Accessories' 'Performance Parts'\n",
      " 'Towing & Hitches' 'Replacement Parts' 'Tires & Wheels'\n",
      " 'Interior Accessories' 'Car Electronics']\n",
      "Region :  ['Northwest' 'Pacific' 'Midwest' 'Southwest' 'Mid-Atlantic' 'Northeast'\n",
      " 'Southeast']\n",
      "Route To Market :  ['Fields Sales' 'Reseller' 'Other' 'Telesales' 'Telecoverage']\n",
      "Opportunity Result :  ['Won' 'Loss']\n",
      "Competitor Type :  ['Unknown' 'Known' 'None']\n",
      "'Supplies Group :  ['Car Accessories' 'Performance & Non-auto' 'Tires & Wheels'\n",
      " 'Car Electronics']\n"
     ]
    }
   ],
   "source": [
    "# aprendemos um pouco sobre o 'LabelEncoder()' e como converter dados categóricos em dados numéricos\n",
    "# vamos visualizar como estão as colunas que são formadas por String\n",
    "# lá atrás usamos a sintaze 'sales_data.dtypes' lembra? \n",
    "print(\"Supplies Subgroup' : \",sales_data['Supplies Subgroup'].unique())\n",
    "print(\"Region : \",sales_data['Region'].unique())\n",
    "print(\"Route To Market : \",sales_data['Route To Market'].unique())\n",
    "print(\"Opportunity Result : \",sales_data['Opportunity Result'].unique())\n",
    "print(\"Competitor Type : \",sales_data['Competitor Type'].unique())\n",
    "print(\"'Supplies Group : \",sales_data['Supplies Group'].unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Opportunity Number</th>\n",
       "      <th>Supplies Subgroup</th>\n",
       "      <th>Supplies Group</th>\n",
       "      <th>Region</th>\n",
       "      <th>Route To Market</th>\n",
       "      <th>Elapsed Days In Sales Stage</th>\n",
       "      <th>Opportunity Result</th>\n",
       "      <th>Sales Stage Change Count</th>\n",
       "      <th>Total Days Identified Through Closing</th>\n",
       "      <th>Total Days Identified Through Qualified</th>\n",
       "      <th>Opportunity Amount USD</th>\n",
       "      <th>Client Size By Revenue</th>\n",
       "      <th>Client Size By Employee Count</th>\n",
       "      <th>Revenue From Client Past Two Years</th>\n",
       "      <th>Competitor Type</th>\n",
       "      <th>Ratio Days Identified To Total Days</th>\n",
       "      <th>Ratio Days Validated To Total Days</th>\n",
       "      <th>Ratio Days Qualified To Total Days</th>\n",
       "      <th>Deal Size Category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1641984</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>76</td>\n",
       "      <td>1</td>\n",
       "      <td>13</td>\n",
       "      <td>104</td>\n",
       "      <td>101</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.69636</td>\n",
       "      <td>0.113985</td>\n",
       "      <td>0.154215</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1658010</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>63</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>163</td>\n",
       "      <td>163</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1674737</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>24</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>82</td>\n",
       "      <td>82</td>\n",
       "      <td>7750</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1675224</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>124</td>\n",
       "      <td>124</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1689785</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>69</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>91</td>\n",
       "      <td>13</td>\n",
       "      <td>69756</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.141125</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Opportunity Number  Supplies Subgroup  Supplies Group  Region  \\\n",
       "0             1641984                  2               0       3   \n",
       "1             1658010                  2               0       4   \n",
       "2             1674737                  5               2       4   \n",
       "3             1675224                  8               2       1   \n",
       "4             1689785                  2               0       4   \n",
       "\n",
       "   Route To Market  Elapsed Days In Sales Stage  Opportunity Result  \\\n",
       "0                0                           76                   1   \n",
       "1                2                           63                   0   \n",
       "2                2                           24                   1   \n",
       "3                2                           16                   0   \n",
       "4                2                           69                   0   \n",
       "\n",
       "   Sales Stage Change Count  Total Days Identified Through Closing  \\\n",
       "0                        13                                    104   \n",
       "1                         2                                    163   \n",
       "2                         7                                     82   \n",
       "3                         5                                    124   \n",
       "4                        11                                     91   \n",
       "\n",
       "   Total Days Identified Through Qualified  Opportunity Amount USD  \\\n",
       "0                                      101                       0   \n",
       "1                                      163                       0   \n",
       "2                                       82                    7750   \n",
       "3                                      124                       0   \n",
       "4                                       13                   69756   \n",
       "\n",
       "   Client Size By Revenue  Client Size By Employee Count  \\\n",
       "0                       5                              5   \n",
       "1                       3                              5   \n",
       "2                       1                              1   \n",
       "3                       1                              1   \n",
       "4                       1                              1   \n",
       "\n",
       "   Revenue From Client Past Two Years  Competitor Type  \\\n",
       "0                                   0                2   \n",
       "1                                   0                2   \n",
       "2                                   0                2   \n",
       "3                                   0                0   \n",
       "4                                   0                2   \n",
       "\n",
       "   Ratio Days Identified To Total Days  Ratio Days Validated To Total Days  \\\n",
       "0                              0.69636                            0.113985   \n",
       "1                              0.00000                            1.000000   \n",
       "2                              1.00000                            0.000000   \n",
       "3                              1.00000                            0.000000   \n",
       "4                              0.00000                            0.141125   \n",
       "\n",
       "   Ratio Days Qualified To Total Days  Deal Size Category  \n",
       "0                            0.154215                   1  \n",
       "1                            0.000000                   1  \n",
       "2                            0.000000                   1  \n",
       "3                            0.000000                   1  \n",
       "4                            0.000000                   4  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import preprocessing\n",
    "le = preprocessing.LabelEncoder()\n",
    "\n",
    "#conversão das colunas categóricas em numéricas\n",
    "sales_data['Supplies Subgroup'] = le.fit_transform(sales_data['Supplies Subgroup'])\n",
    "sales_data['Region'] = le.fit_transform(sales_data['Region'])\n",
    "sales_data['Route To Market'] = le.fit_transform(sales_data['Route To Market'])\n",
    "sales_data['Opportunity Result'] = le.fit_transform(sales_data['Opportunity Result'])\n",
    "sales_data['Competitor Type'] = le.fit_transform(sales_data['Competitor Type'])\n",
    "sales_data['Supplies Group'] = le.fit_transform(sales_data['Supplies Group'])\n",
    "\n",
    "sales_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Agora <b> SIM </b>, avançamos bem! Com todos os dados já convertidos podemos em breve a construção do <b> Modelo Preditivo </b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Conjunto de Treinamento e Conjunto de Testes </b><p>\n",
    "Um algoritmo de aprendizado de máquina precisa ser treinado em um conjunto de dados para aprender as relações entre diferentes recursos e como esses recursos afetam a variável de destino. Para isso, precisamos dividir todo o <b>conjunto de dados em dois conjuntos</b>. Um é o conjunto de <b>treinamento</b> no qual vamos treinar nosso algoritmo para construir um modelo.<p> O outro é o conjunto de <b>testes</b> no qual testaremos nosso modelo para ver quão precisas são suas previsões.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>Supplies Subgroup</th>\n",
       "      <th>Supplies Group</th>\n",
       "      <th>Region</th>\n",
       "      <th>Route To Market</th>\n",
       "      <th>Elapsed Days In Sales Stage</th>\n",
       "      <th>Sales Stage Change Count</th>\n",
       "      <th>Total Days Identified Through Closing</th>\n",
       "      <th>Total Days Identified Through Qualified</th>\n",
       "      <th>Opportunity Amount USD</th>\n",
       "      <th>Client Size By Revenue</th>\n",
       "      <th>Client Size By Employee Count</th>\n",
       "      <th>Revenue From Client Past Two Years</th>\n",
       "      <th>Competitor Type</th>\n",
       "      <th>Ratio Days Identified To Total Days</th>\n",
       "      <th>Ratio Days Validated To Total Days</th>\n",
       "      <th>Ratio Days Qualified To Total Days</th>\n",
       "      <th>Deal Size Category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>76</td>\n",
       "      <td>13</td>\n",
       "      <td>104</td>\n",
       "      <td>101</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.69636</td>\n",
       "      <td>0.113985</td>\n",
       "      <td>0.154215</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>63</td>\n",
       "      <td>2</td>\n",
       "      <td>163</td>\n",
       "      <td>163</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Supplies Subgroup  Supplies Group  Region  Route To Market  \\\n",
       "0                  2               0       3                0   \n",
       "1                  2               0       4                2   \n",
       "\n",
       "   Elapsed Days In Sales Stage  Sales Stage Change Count  \\\n",
       "0                           76                        13   \n",
       "1                           63                         2   \n",
       "\n",
       "   Total Days Identified Through Closing  \\\n",
       "0                                    104   \n",
       "1                                    163   \n",
       "\n",
       "   Total Days Identified Through Qualified  Opportunity Amount USD  \\\n",
       "0                                      101                       0   \n",
       "1                                      163                       0   \n",
       "\n",
       "   Client Size By Revenue  Client Size By Employee Count  \\\n",
       "0                       5                              5   \n",
       "1                       3                              5   \n",
       "\n",
       "   Revenue From Client Past Two Years  Competitor Type  \\\n",
       "0                                   0                2   \n",
       "1                                   0                2   \n",
       "\n",
       "   Ratio Days Identified To Total Days  Ratio Days Validated To Total Days  \\\n",
       "0                              0.69636                            0.113985   \n",
       "1                              0.00000                            1.000000   \n",
       "\n",
       "   Ratio Days Qualified To Total Days  Deal Size Category  \n",
       "0                            0.154215                   1  \n",
       "1                            0.000000                   1  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#selecionando as colunas 'Opportunity Number' e 'Opportunity Result'\n",
    "#não precisaremos das colunas citadas, pois a 1ª refere-se a um identificador excluisivo para cada registros e \n",
    "#e a outra é o nosso alvo! O que buscamos logo, não faz sentido possuir as mesmas\n",
    "cols = [col for col in sales_data.columns if col not in ['Opportunity Number','Opportunity Result']]\n",
    "\n",
    "data = sales_data[cols]\n",
    "target = sales_data['Opportunity Result']\n",
    "data.head(n=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Estamos todos prontos para definir nossos recursos e segmentar em dois quadros de dados separados. Em seguida, vamos dividir as dataframes 'data' e target' em conjuntos de formação e conjuntos de testes. Ao dividir o conjunto de dados, manteremos <b>30% dos dados como dados de teste</b> e os <b> 70% restantes como dados de treinamento</b>. Porém, fique a vontade para usar o modelo mais a sua escolha, usa-se muito o príncipio <b> '80/20'</b>, no qual são mantidos '80%' com dados de treinamento e os outros 20%' como dados de teste."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "#criação do 'data_train' e do 'data_test'\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "data_train, data_test, target_train, target_test = train_test_split(data,target, test_size = 0.30, random_state = 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Para a escolha do melhor algoritmo use o <b> MAP </b> como referência<p>\n",
    "Machine Learn [MAP](https://scikit-learn.org/stable/tutorial/machine_learning_map/) !?!?!<p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![MAP](https://scikit-learn.org/stable/_static/ml_map.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "O algoritmo que vamos usar para os nossos dados de vendas será o <b> [Gaussian Naive Bayes](http://scikit-learn.org/stable/modules/naive_bayes.html)</b> !?!?!<p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive-Bayes accuracy Teste:  0.7590567327409433\n"
     ]
    }
   ],
   "source": [
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "gnb = GaussianNB()\n",
    "\n",
    "pred = gnb.fit(data_train, target_train).predict(data_test)\n",
    "\n",
    "print(\"Naive-Bayes accuracy Teste: \",accuracy_score(target_test, pred, normalize = True))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive-Bayes accuracy Treinamento:  0.7579508211728949\n"
     ]
    }
   ],
   "source": [
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "gnb = GaussianNB()\n",
    "\n",
    "pred = gnb.fit(data_train, target_train).predict(data_train)\n",
    "\n",
    "print(\"Naive-Bayes accuracy Treinamento: \",accuracy_score(target_train, pred, normalize = True))"
   ]
  }
 ],
 "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}