DesafioBB.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Teste BB\n",
    "### Dataset Census.csv\n",
    "1. Coluna age: idade da pessoa que participou da avaliação de quanto ganha anualmente;\n",
    "\n",
    "2. Coluna worclass: é qual o setor que a pessoa trabalha;\n",
    "\n",
    "3. Coluna education_level: qual é o nível de instrução do usuário, segue abaixo os tipos que serão encontrados e válidos:\n",
    "\n",
    "a. Bachelors;\n",
    "\n",
    "b. Masters;\n",
    "\n",
    "c. HS-grad;\n",
    "\n",
    "d. Some-college;\n",
    "\n",
    "e. Doctorate.\n",
    "\n",
    "4. Coluna education_num: número da educação, é o processo de quantificar o nível educacional e dar um peso para seu identificador;\n",
    "\n",
    "5. Coluna marital-status: o status de matrimônio do usuário que foi entrevistado;\n",
    "\n",
    "6. Coluna occupation: ocupação do usuário que foi entrevistado, vale lembrar que essa variável pertence ao estado irregular;\n",
    "\n",
    "7. Coluna relationship: é a coluna de relacionamento, onde detalha seu grau de relacionamento atual, segue as variáveis válidas:\n",
    "\n",
    "a. Not-in-family;\n",
    "\n",
    "b. Husband;\n",
    "\n",
    "c. Wife;\n",
    "\n",
    "d. Own-child;\n",
    "\n",
    "e. Unmarried.\n",
    "\n",
    "8. Coluna race: coluna que demonstra a tonalidade da pele do usuário, as variáveis válidas são:\n",
    "\n",
    "a. White;\n",
    "\n",
    "b. Black;\n",
    "\n",
    "c. Asian-Pac-Islander;\n",
    "\n",
    "d. Amer-Indian-Eskimo;\n",
    "\n",
    "e. Other.\n",
    "\n",
    "9. Coluna sex: o sexo da pessoa (Masculino ou Feminino);\n",
    "\n",
    "10. Coluna capital-gain: capital ganho ou adquirido;\n",
    "\n",
    "11. Coluna capital-loss: capital perdido ou não conquistado;\n",
    "\n",
    "12. Coluna hours-per-week: horas por semana trabalhadas;\n",
    "\n",
    "13. Coluna native-country: País nativo do usuário que foi entrevistado;\n",
    "\n",
    "14. Coluna income: a renda anual do usuário que foi entrevistado, saídas:\n",
    "a. Menos que 50k;\n",
    "b. Menor e igual a 50k;\n",
    "c. Maior que 50k.\n",
    "\n",
    "\n",
    "## Saídas esperadas\n",
    "\n",
    "Para este desafio esperamos algumas saídas necessárias, conforme abaixo (via TXT ou Markdown):\n",
    "\n",
    "● Explicar as variáveis que foram exploradas e os motivos;\n",
    "\n",
    "● Caso tenha sido necessário novas variáveis, explicar o motivo da criação;\n",
    "\n",
    "● Explicar o processo de preparação;\n",
    "\n",
    "● Detalhar se houve algum processo que deve ser considerado transformação de recursos contínuos enviesados e o porquê;\n",
    "\n",
    "● Explicar qual foi o processo utilizado para normalização dos dados;\n",
    "\n",
    "● Detalhar qual foi o modelo utilizado para desempenho e predição dos dados;\n",
    "\n",
    "● Se chegou a utilizar Modelos de Aprendizagem Supervisionados, explique quais;\n",
    "\n",
    "● Detalhar o que foi utilizado no treino e de evolução;\n",
    "\n",
    "● Apresentar a acuracidade do modelo e a revalidação do modelo com Log Loss.\n",
    "\n",
    "O arquivo gerado com as saídas solicitadas acima deve ser armazenados em seu repositório pessoal doGitHub."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import and initial parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import missingno as msno\n",
    "import gc\n",
    "\n",
    "import matplotlib.pyplot  as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Parametros de visualização de dados\n",
    "from IPython.display import set_matplotlib_formats\n",
    "set_matplotlib_formats('pdf', 'png')\n",
    "pd.options.display.float_format = '{:.2f}'.format\n",
    "rc={'savefig.dpi': 75, 'figure.autolayout': False, 'figure.figsize': [12, 8], 'axes.labelsize': 18,\\\n",
    "   'axes.titlesize': 18, 'font.size': 18, 'lines.linewidth': 2.0, 'lines.markersize': 8, 'legend.fontsize': 16,\\\n",
    "   'xtick.labelsize': 16, 'ytick.labelsize': 16}\n",
    "\n",
    "sns.set(style='dark',rc=rc)\n",
    "default_color = '#56B4E9'\n",
    "colormap = plt.cm.cool"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading files"
   ]
  },
  {
   "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>age</th>\n",
       "      <th>workclass</th>\n",
       "      <th>education_level</th>\n",
       "      <th>education-num</th>\n",
       "      <th>marital-status</th>\n",
       "      <th>occupation</th>\n",
       "      <th>relationship</th>\n",
       "      <th>race</th>\n",
       "      <th>sex</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>native-country</th>\n",
       "      <th>income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>39</td>\n",
       "      <td>State-gov</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13.00</td>\n",
       "      <td>Never-married</td>\n",
       "      <td>Adm-clerical</td>\n",
       "      <td>Not-in-family</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>2174.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>40.00</td>\n",
       "      <td>United-States</td>\n",
       "      <td>&lt;=50K</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>50</td>\n",
       "      <td>Self-emp-not-inc</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>13.00</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Exec-managerial</td>\n",
       "      <td>Husband</td>\n",
       "      <td>White</td>\n",
       "      <td>Male</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>13.00</td>\n",
       "      <td>United-States</td>\n",
       "      <td>&lt;=50K</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age          workclass education_level  education-num       marital-status  \\\n",
       "0   39          State-gov       Bachelors          13.00        Never-married   \n",
       "1   50   Self-emp-not-inc       Bachelors          13.00   Married-civ-spouse   \n",
       "\n",
       "         occupation    relationship    race    sex  capital-gain  \\\n",
       "0      Adm-clerical   Not-in-family   White   Male       2174.00   \n",
       "1   Exec-managerial         Husband   White   Male          0.00   \n",
       "\n",
       "   capital-loss  hours-per-week  native-country income  \n",
       "0          0.00           40.00   United-States  <=50K  \n",
       "1          0.00           13.00   United-States  <=50K  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('census.csv')\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5533, 14)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>education-num</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>5533.00</td>\n",
       "      <td>5533.00</td>\n",
       "      <td>5532.00</td>\n",
       "      <td>5532.00</td>\n",
       "      <td>5532.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>38.45</td>\n",
       "      <td>10.11</td>\n",
       "      <td>1065.11</td>\n",
       "      <td>92.79</td>\n",
       "      <td>41.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>13.10</td>\n",
       "      <td>2.54</td>\n",
       "      <td>7251.50</td>\n",
       "      <td>409.63</td>\n",
       "      <td>11.62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>17.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>28.00</td>\n",
       "      <td>9.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>40.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>37.00</td>\n",
       "      <td>10.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>40.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>47.00</td>\n",
       "      <td>13.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>45.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>90.00</td>\n",
       "      <td>16.00</td>\n",
       "      <td>99999.00</td>\n",
       "      <td>2824.00</td>\n",
       "      <td>99.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          age  education-num  capital-gain  capital-loss  hours-per-week\n",
       "count 5533.00        5533.00       5532.00       5532.00         5532.00\n",
       "mean    38.45          10.11       1065.11         92.79           41.07\n",
       "std     13.10           2.54       7251.50        409.63           11.62\n",
       "min     17.00           1.00          0.00          0.00            1.00\n",
       "25%     28.00           9.00          0.00          0.00           40.00\n",
       "50%     37.00          10.00          0.00          0.00           40.00\n",
       "75%     47.00          13.00          0.00          0.00           45.00\n",
       "max     90.00          16.00      99999.00       2824.00           99.00"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#analyze outliers\n",
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5533 entries, 0 to 5532\n",
      "Data columns (total 14 columns):\n",
      "age                5533 non-null int64\n",
      "workclass          5533 non-null object\n",
      "education_level    5533 non-null object\n",
      "education-num      5533 non-null float64\n",
      "marital-status     5533 non-null object\n",
      "occupation         5533 non-null object\n",
      "relationship       5533 non-null object\n",
      "race               5533 non-null object\n",
      "sex                5533 non-null object\n",
      "capital-gain       5532 non-null float64\n",
      "capital-loss       5532 non-null float64\n",
      "hours-per-week     5532 non-null float64\n",
      "native-country     5532 non-null object\n",
      "income             5532 non-null object\n",
      "dtypes: float64(4), int64(1), object(9)\n",
      "memory usage: 605.2+ KB\n"
     ]
    }
   ],
   "source": [
    "#analyze columns types and NAN values\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age                0\n",
       "workclass          0\n",
       "education_level    0\n",
       "education-num      0\n",
       "marital-status     0\n",
       "occupation         0\n",
       "relationship       0\n",
       "race               0\n",
       "sex                0\n",
       "capital-gain       1\n",
       "capital-loss       1\n",
       "hours-per-week     1\n",
       "native-country     1\n",
       "income             1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#confirm NAN\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>workclass</th>\n",
       "      <th>education_level</th>\n",
       "      <th>education-num</th>\n",
       "      <th>marital-status</th>\n",
       "      <th>occupation</th>\n",
       "      <th>relationship</th>\n",
       "      <th>race</th>\n",
       "      <th>sex</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>native-country</th>\n",
       "      <th>income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5532</th>\n",
       "      <td>29</td>\n",
       "      <td>Private</td>\n",
       "      <td>HS-grad</td>\n",
       "      <td>9.00</td>\n",
       "      <td>Married-civ-spouse</td>\n",
       "      <td>Prof-specialty</td>\n",
       "      <td>Husband</td>\n",
       "      <td>White</td>\n",
       "      <td>Mal</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      age workclass education_level  education-num       marital-status  \\\n",
       "5532   29   Private         HS-grad           9.00   Married-civ-spouse   \n",
       "\n",
       "           occupation relationship    race   sex  capital-gain  capital-loss  \\\n",
       "5532   Prof-specialty      Husband   White   Mal           nan           nan   \n",
       "\n",
       "      hours-per-week native-country income  \n",
       "5532             nan            NaN    NaN  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#details NAN\n",
    "df[df['capital-gain'].isnull()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Target Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,5))\n",
    "ax = sns.countplot('income',data=df,color=default_color)\n",
    "for p in ax.patches:\n",
    "    ax.annotate('{:.2f}%'.format(100*p.get_height()/len(df['income'])), (p.get_x()+ 0.3, p.get_height()+100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "columns:age\n",
      "type:int64\n",
      "qtd_unique: 67\n",
      "value_unique: [39 50 38 53 28 37 49 52 31 42 30 23 32 34 25 43 40 54 35 59 56 19 20 45\n",
      " 22 48 21 24 57 44 41 29 47 46 36 79 27 18 33 76 55 61 70 64 71 66 51 58\n",
      " 26 17 60 90 75 65 77 62 63 67 74 72 69 68 73 81 78 88 80]\n",
      "\n",
      "columns:workclass\n",
      "type:object\n",
      "qtd_unique: 7\n",
      "value_unique: [' State-gov' ' Self-emp-not-inc' ' Private' ' Federal-gov' ' Local-gov'\n",
      " ' Self-emp-inc' ' Without-pay']\n",
      "\n",
      "columns:education_level\n",
      "type:object\n",
      "qtd_unique: 16\n",
      "value_unique: [' Bachelors' ' HS-grad' ' 11th' ' Masters' ' 9th' ' Some-college'\n",
      " ' Assoc-acdm' ' 7th-8th' ' Doctorate' ' Assoc-voc' ' Prof-school'\n",
      " ' 5th-6th' ' 10th' ' Preschool' ' 12th' ' 1st-4th']\n",
      "\n",
      "columns:education-num\n",
      "type:float64\n",
      "qtd_unique: 16\n",
      "value_unique: [13.  9.  7. 14.  5. 10. 12.  4. 16. 11. 15.  3.  6.  1.  8.  2.]\n",
      "\n",
      "columns:marital-status\n",
      "type:object\n",
      "qtd_unique: 7\n",
      "value_unique: [' Never-married' ' Married-civ-spouse' ' Divorced'\n",
      " ' Married-spouse-absent' ' Separated' ' Married-AF-spouse' ' Widowed']\n",
      "\n",
      "columns:occupation\n",
      "type:object\n",
      "qtd_unique: 14\n",
      "value_unique: [' Adm-clerical' ' Exec-managerial' ' Handlers-cleaners' ' Prof-specialty'\n",
      " ' Other-service' ' Sales' ' Transport-moving' ' Farming-fishing'\n",
      " ' Machine-op-inspct' ' Tech-support' ' Craft-repair' ' Protective-serv'\n",
      " ' Armed-Forces' ' Priv-house-serv']\n",
      "\n",
      "columns:relationship\n",
      "type:object\n",
      "qtd_unique: 6\n",
      "value_unique: [' Not-in-family' ' Husband' ' Wife' ' Own-child' ' Unmarried'\n",
      " ' Other-relative']\n",
      "\n",
      "columns:race\n",
      "type:object\n",
      "qtd_unique: 5\n",
      "value_unique: [' White' ' Black' ' Asian-Pac-Islander' ' Amer-Indian-Eskimo' ' Other']\n",
      "\n",
      "columns:sex\n",
      "type:object\n",
      "qtd_unique: 3\n",
      "value_unique: [' Male' ' Female' ' Mal']\n",
      "\n",
      "columns:capital-gain\n",
      "type:float64\n",
      "qtd_unique: 79\n",
      "value_unique: [ 2174.     0. 14084.  5178.  5013.  2407. 14344. 15024.  7688.  4064.\n",
      "  4386.  7298.  1409.  3674.  1055.  2050.  2176.   594. 20051.  6849.\n",
      "  4101.  8614.  3411.  2597. 25236.  4650.  2463.  3103. 10605.  2964.\n",
      "  3325.  2580.  3471.  4865. 99999.  6514.  1471.  2329.  2105.  2885.\n",
      " 10520.  2202.  2961. 27828.  6767.  2228.  1506. 13550.  2635.  5556.\n",
      "  4787.  3781.  3137.  3818.   914.   401.  2829.  2977.  4934.  2062.\n",
      "  2354.  3464.  5455. 15020.  1424.  3273. 22040.  4416. 10566.  4931.\n",
      "  7430. 34095.  6497.  3908.   114.  2346.  2907.  1151.  9386.    nan]\n",
      "\n",
      "columns:capital-loss\n",
      "type:float64\n",
      "qtd_unique: 57\n",
      "value_unique: [   0. 2042. 1408. 1902. 1887. 1719. 1762. 1564. 2179. 1816. 1980. 1977.\n",
      " 1876. 1340. 2206. 1741. 1485. 2339. 2415. 1380. 1721. 2051. 2377. 1669.\n",
      " 2352. 1672.  653. 2392. 1504. 1590. 1651. 1628. 1848. 2001. 1740. 2002.\n",
      " 1579. 2258. 1602. 2547. 2205. 1726. 2444. 1138. 2238.  625.  213. 1539.\n",
      "  880. 1668. 1092. 1594. 2231. 1844.  810. 2824. 2559.   nan]\n",
      "\n",
      "columns:hours-per-week\n",
      "type:float64\n",
      "qtd_unique: 74\n",
      "value_unique: [40. 13. 16. 45. 50. 80. 30. 35. 60. 20. 52. 44. 15. 25. 43. 38. 55. 48.\n",
      " 58. 32. 70. 22. 56. 41. 28. 36. 24. 46.  2. 42. 12. 65.  1. 34. 75. 98.\n",
      " 33. 54. 10.  6. 64. 19. 18. 72.  8.  9. 47. 37. 21. 26. 14.  5.  7. 99.\n",
      " 53. 39. 62. 59. 57. 78. 90. 66. 11. 49. 84. 17. 68.  3. 27. 85. 31. 51.\n",
      " 77. 63. nan]\n",
      "\n",
      "columns:native-country\n",
      "type:object\n",
      "qtd_unique: 39\n",
      "value_unique: [' United-States' ' Cuba' ' Jamaica' ' India' ' Mexico' ' Puerto-Rico'\n",
      " ' Honduras' ' England' ' Canada' ' Germany' ' Iran' ' Philippines'\n",
      " ' Poland' ' Columbia' ' Cambodia' ' Thailand' ' Ecuador' ' Laos'\n",
      " ' Taiwan' ' Haiti' ' Portugal' ' Dominican-Republic' ' El-Salvador'\n",
      " ' France' ' Guatemala' ' Italy' ' China' ' South' ' Japan' ' Yugoslavia'\n",
      " ' Peru' ' Outlying-US(Guam-USVI-etc)' ' Scotland' ' Trinadad&Tobago'\n",
      " ' Greece' ' Nicaragua' ' Vietnam' ' Hong' ' Ireland' nan]\n",
      "\n",
      "columns:income\n",
      "type:object\n",
      "qtd_unique: 2\n",
      "value_unique: ['<=50K' '>50K' nan]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#analyze frequency values per columns\n",
    "for col in df.columns:\n",
    "    print('columns:{}\\ntype:{}\\nqtd_unique: {}\\nvalue_unique: {}\\n'.format(col, df[col].dtype, df[col].nunique(),df[col].unique()  ))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Nesta analise verificou-se que existem na coluna 'education_level', 16 frequência de valores, onde no enunciado do teste permite somente cinco variavéis váliddas – a. Bachelors; b. Masters; c. HS-grad; d. Some-college; e. Doctorate –, bem como, ocorre com a coluna relationship tendo uma variavel ('Other-relative'). \n",
    "#### Já a coluna 'sex' possui um possivel erro de digitação na variável('mal'), ocorre que este erro esta na mesma linha que contém 5 NAN,sendo descartado esta linha. Lamento que estas ocorrências tenham um custo alto para o Dataset, retirando 23,98% do total de dados.     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#mask of valid variables and elimination of invalid rows\n",
    "mask1 = df['education_level'] == ' Bachelors'\n",
    "mask2 = df['education_level'] == ' HS-grad'\n",
    "mask3 = df['education_level'] == ' Masters'\n",
    "mask4 = df['education_level'] == ' Some-college'\n",
    "mask5 = df['education_level'] == ' Doctorate'\n",
    "df = df[mask1 | mask2 | mask3 | mask4 | mask5]\n",
    "df = df[df.relationship != ' Other-relative']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### A coluna education_level poderá ser excluida porque possui a coluna education-num que possui os mesmo dados em formato ordinal."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# drop column education_level\n",
    "df.drop('education_level', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#changing the column types according to the variables of the values\n",
    "\n",
    "df['workclass'] = df['workclass'].astype('category')\n",
    "df['marital-status'] = df['marital-status'].astype('category')\n",
    "df['occupation'] = df['occupation'].astype('category')\n",
    "df['relationship'] = df['relationship'].astype('category')\n",
    "df['race'] = df['race'].astype('category')\n",
    "df['sex'] = df['sex'].astype('category')\n",
    "df['native-country'] = df['native-country'].astype('category')\n",
    "df['education-num'] = df['education-num'].astype('int64')\n",
    "\n",
    "#Encode the 'income' data to numerical values\n",
    "df['income'] = df['income'].apply(lambda x: 1 if x == \">50K\" else 0)\n",
    "\n",
    "# Drop NAN lines\n",
    "df.drop(5532, inplace= True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data analisys"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_meta(train):\n",
    "    data = []\n",
    "    for col in train.columns:\n",
    "        # Defining the role\n",
    "        if col == 'income':\n",
    "            role = 'target'\n",
    "        else:\n",
    "            role = 'input'\n",
    "\n",
    "        # Defining the level\n",
    "        if col == 'income':\n",
    "            level = 'binary'\n",
    "        elif train[col].dtype == np.float64:\n",
    "            level = 'interval'\n",
    "        elif train[col].dtype == np.int64:\n",
    "            level = 'ordinal'\n",
    "        elif train[col].dtype == 'category':\n",
    "            level = 'category'\n",
    "\n",
    "        # Initialize keep to True for all variables except for id\n",
    "        keep = True\n",
    "        if col == 'id':\n",
    "            keep = False\n",
    "\n",
    "        # Defining the data type \n",
    "        dtype = train[col].dtype\n",
    "\n",
    "        # Creating a Dict that contains all the metadata for the variable\n",
    "        col_dict = {\n",
    "            'varname': col,\n",
    "            'role'   : role,\n",
    "            'level'  : level,\n",
    "            'keep'   : keep,\n",
    "            'dtype'  : dtype\n",
    "        }\n",
    "        data.append(col_dict)\n",
    "    meta = pd.DataFrame(data, columns=['varname', 'role', 'level', 'keep', 'dtype'])\n",
    "    meta.set_index('varname', inplace=True)\n",
    "    return meta"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>role</th>\n",
       "      <th>level</th>\n",
       "      <th>keep</th>\n",
       "      <th>dtype</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>varname</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>input</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>True</td>\n",
       "      <td>int64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>workclass</th>\n",
       "      <td>input</td>\n",
       "      <td>category</td>\n",
       "      <td>True</td>\n",
       "      <td>category</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>education-num</th>\n",
       "      <td>input</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>True</td>\n",
       "      <td>int64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>marital-status</th>\n",
       "      <td>input</td>\n",
       "      <td>category</td>\n",
       "      <td>True</td>\n",
       "      <td>category</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>occupation</th>\n",
       "      <td>input</td>\n",
       "      <td>category</td>\n",
       "      <td>True</td>\n",
       "      <td>category</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relationship</th>\n",
       "      <td>input</td>\n",
       "      <td>category</td>\n",
       "      <td>True</td>\n",
       "      <td>category</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>race</th>\n",
       "      <td>input</td>\n",
       "      <td>category</td>\n",
       "      <td>True</td>\n",
       "      <td>category</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sex</th>\n",
       "      <td>input</td>\n",
       "      <td>category</td>\n",
       "      <td>True</td>\n",
       "      <td>category</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>capital-gain</th>\n",
       "      <td>input</td>\n",
       "      <td>interval</td>\n",
       "      <td>True</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>capital-loss</th>\n",
       "      <td>input</td>\n",
       "      <td>interval</td>\n",
       "      <td>True</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hours-per-week</th>\n",
       "      <td>input</td>\n",
       "      <td>interval</td>\n",
       "      <td>True</td>\n",
       "      <td>float64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>native-country</th>\n",
       "      <td>input</td>\n",
       "      <td>category</td>\n",
       "      <td>True</td>\n",
       "      <td>category</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>income</th>\n",
       "      <td>target</td>\n",
       "      <td>binary</td>\n",
       "      <td>True</td>\n",
       "      <td>int64</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  role     level  keep     dtype\n",
       "varname                                         \n",
       "age              input   ordinal  True     int64\n",
       "workclass        input  category  True  category\n",
       "education-num    input   ordinal  True     int64\n",
       "marital-status   input  category  True  category\n",
       "occupation       input  category  True  category\n",
       "relationship     input  category  True  category\n",
       "race             input  category  True  category\n",
       "sex              input  category  True  category\n",
       "capital-gain     input  interval  True   float64\n",
       "capital-loss     input  interval  True   float64\n",
       "hours-per-week   input  interval  True   float64\n",
       "native-country   input  category  True  category\n",
       "income          target    binary  True     int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "meta_data = get_meta(df)\n",
    "meta_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>role</th>\n",
       "      <th>level</th>\n",
       "      <th>dtype</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>input</td>\n",
       "      <td>category</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>input</td>\n",
       "      <td>interval</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>input</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>target</td>\n",
       "      <td>binary</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     role     level  dtype\n",
       "0   input  category      7\n",
       "1   input  interval      3\n",
       "2   input   ordinal      2\n",
       "3  target    binary      1"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Group By DataType\n",
    "meta_counts = meta_data.groupby(['role', 'level']).agg({'dtype': lambda x: x.count()}).reset_index()\n",
    "meta_counts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0,0.5,'Count'),\n",
       " Text(0.5,0,'Variable Type'),\n",
       " Text(0.5,1,'Variables Count Across Datatype')]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot Variables Count Across Datatype\n",
    "fig,ax = plt.subplots()\n",
    "fig.set_size_inches(20,5)\n",
    "sns.barplot(data=meta_counts[(meta_counts.role != 'target')],x=\"level\",y=\"dtype\",ax=ax,color=default_color)\n",
    "ax.set(xlabel='Variable Type', ylabel='Count',title=\"Variables Count Across Datatype\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "col_ordinal   = meta_data[(meta_data.level == 'ordinal') & (meta_data.keep)].index\n",
    "col_category   = meta_data[(meta_data.level == 'category') & (meta_data.keep)].index\n",
    "col_interval = meta_data[(meta_data.level == 'interval') & (meta_data.keep)].index"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Continuous features analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a154fa630>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x1152 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Checking if there is continuous data that has a strong correlation with Person correlation\n",
    "plt.figure(figsize=(18,16))\n",
    "plt.title('Pearson correlation of continuous features', y=1.05, size=15)\n",
    "sns.heatmap(df[col_interval].corr(),linewidths=0.1,vmax=1.0, square=True, cmap=colormap, linecolor='white', annot=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Split the data into features and target\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "target = df['income']\n",
    "previsores = df.drop('income', axis = 1)\n",
    "\n",
    "\n",
    "#I created the variable features to test the importance of features with RandonForest\n",
    "features = df.drop('income', axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# Import sklearn.preprocessing.StandardScaler\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "scaler = MinMaxScaler()\n",
    "#previsores[col_ordinal] = scaler.fit_transform(df[col_ordinal])\n",
    "previsores[col_interval] = scaler.fit_transform(df[col_interval])\n",
    "features[col_interval] = scaler.fit_transform(df[col_interval])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "#LabelEncoder to prepare the cathegorical data in formats accepted by the algorithms\n",
    "from sklearn.preprocessing import LabelEncoder, OneHotEncoder\n",
    "labelencoder_previsores = LabelEncoder()\n",
    "for col in col_category:\n",
    "    features[col] = labelencoder_previsores.fit_transform(features[col])\n",
    "    previsores[col] = labelencoder_previsores.fit_transform(previsores[col])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>workclass</th>\n",
       "      <th>education-num</th>\n",
       "      <th>marital-status</th>\n",
       "      <th>occupation</th>\n",
       "      <th>relationship</th>\n",
       "      <th>race</th>\n",
       "      <th>sex</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>native-country</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>39</td>\n",
       "      <td>5</td>\n",
       "      <td>13</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.40</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  workclass  education-num  marital-status  occupation  relationship  \\\n",
       "0   39          5             13               4           0             1   \n",
       "\n",
       "   race  sex  capital-gain  capital-loss  hours-per-week  native-country  \n",
       "0     4    1          0.02          0.00            0.40              35  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "previsores.head(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Testing feature importantes with Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "----- Training Done -----\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "rf = RandomForestClassifier(n_estimators=150, max_depth=8, min_samples_leaf=4, max_features=0.2, n_jobs=-1, random_state=0)\n",
    "rf.fit(features, target)\n",
    "features_importantes = features.columns.values\n",
    "print(\"----- Training Done -----\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_feature_importance_df(feature_importances, \n",
    "                              column_names, \n",
    "                              top_n=25):\n",
    "    \"\"\"Get feature importance data frame.\n",
    " \n",
    "    Parameters\n",
    "    ----------\n",
    "    feature_importances : numpy ndarray\n",
    "        Feature importances computed by an ensemble \n",
    "            model like random forest or boosting\n",
    "    column_names : array-like\n",
    "        Names of the columns in the same order as feature \n",
    "            importances\n",
    "    top_n : integer\n",
    "        Number of top features\n",
    " \n",
    "    Returns\n",
    "    -------\n",
    "    df : a Pandas data frame\n",
    " \n",
    "    \"\"\"\n",
    "     \n",
    "    imp_dict = dict(zip(column_names, \n",
    "                        feature_importances))\n",
    "    top_features = sorted(imp_dict, \n",
    "                          key=imp_dict.get, \n",
    "                          reverse=True)[0:top_n]\n",
    "    top_importances = [imp_dict[feature] for feature \n",
    "                          in top_features]\n",
    "    df = pd.DataFrame(data={'feature': top_features, \n",
    "                            'importance': top_importances})\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_importance = get_feature_importance_df(rf.feature_importances_, features_importantes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0,0.5,'Importance'),\n",
       " Text(0.5,0,'Variable name'),\n",
       " Text(0.5,1,'Variable importances')]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjEwIDAgb2JqCjw8IC9Bbm5vdHMgWyBdIC9Db250ZW50cyA5IDAgUgovR3JvdXAgPDwgL0NTIC9EZXZpY2VSR0IgL1MgL1RyYW5zcGFyZW5jeSAvVHlwZSAvR3JvdXAgPj4KL01lZGlhQm94IFsgMCAwIDExODkuMDQwNjI1IDYxOS4xODc1IF0gL1BhcmVudCAyIDAgUiAvUmVzb3VyY2VzIDggMCBSCi9UeXBlIC9QYWdlID4+CmVuZG9iago5IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTEgMCBSID4+CnN0cmVhbQp4nO1Yy3IbNxC84ytwTA6EZgbvo12JWc7NMSs5pHLYUGuJMl/hI/LnZ7BLkVgBlEq3PUhVkogmuoEegL0EUD6Imw8o7/aS/0iQD/z7KP+Sf/P/W8amCRcT4OZKIIaowIAjy+3loO0wKgzeMsy9s9a9EN8EqIjeGQ82BPm8YSJgdOCD3KUpTIsO54Z41lsIZ1XoJ2CiMtGlMbuJYokvB7iNWsHTGxeZAdzN/eYDpfqAQsslQi4RZiXq4OnpXfFxJm8+oUQnZ99EANXpaOQiaX4xuxU/7dplc1hs1vv7xfZnOXuQv86GLIykqKDNm+3i0Cwnd81iXaVRdMpH/Zy3anYdb39oDsd9lWlAK+NjOdX29jjv5jpZH1dVqrWGLSJ1RNLpRSI2d221uyOjQJ+6X8a53xx3+8m23U0e2/Z7lektKizdPVVludnXvQVnFGK/sgPmZj4/bjtz9UUAQBWLiuzbH1e6R96ZvCcLb4+b3ff5simnF7r68b7rFw1B+Z6eWH/wqjX/LFu5blb1UpJXrlsxo5kI59KDAniZgE4rcsTVJIP4RLIvkyiQ8taA1zEQnUj4ykgGeCcbG63xpM+kV0ayhMprH7Q1zugTiYqRuvIBfxInyB9DYgexW2TyTpE9VXIlUm4lic+r7WZ3aNbzczX/lbXoQOTpWK6pk7tW/inX8pSOoAxEjdqT5ezjTc9jmNTmSAoGXffj+wAbZmgfaFfYMmeLFGje8wqWgebYVijz7AJb5xWl6SPjF5EB3MXZFzlW6xiCcqV3Yjeh9D6AvbIuBOtSTXKVDB+5eQ1pmxfmNT9hTWk+h9HxKPx4TCZzlRwfuXkOIhVK85ZDGErzGayNSbkRjWE8UxngIzdviZ+/pXmHxLTCfAZrXmHrwXlKRbmoDPCRm3fa1vLO8zeFSt5lMEFQloVjMpmpDPCRm/fG1QKPn321wMtgZNy66FwymakM8JGbDzbUAi8aXwu8DEaugyFtOpOZygAfufnoYi3wEGysJV6OI38GLH9ltD6V5aIzwEduH4HPYZXMQ/RQC70cj3wSACTCri4XnRx/s33qT8C/yYQ+8oTJyam4+aX9bzFvf59+lPO96M/A4uu78Lvwu/BrwsR9a9dJQ0r9+ubKbYxYX7nYWV2/2GHO266IBoRM6uUxILl9ywVRf/zXUVF/VrUxPgkOLgAW54Pr+Qbhi/gf6nU1kAplbmRzdHJlYW0KZW5kb2JqCjExIDAgb2JqCjkwMgplbmRvYmoKMTYgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA0OCA+PgpzdHJlYW0KeJwzMrdQMFCwNAYShhYWCuaGZgophlxgfi6IAgnkcMGkICwDIA1WkcOVBgCHCQxTCmVuZHN0cmVhbQplbmRvYmoKMTcgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCA5OCA+PgpzdHJlYW0KeJw1jbENxDAMA3tNwRFM05KiefL4Kr9/+4qRFAKIE3iMSAysvvBCMvChzWM2+Nna4DKOfNIsgXUnNaMvaA6QDonIwmlabJBQP+kBRe3KaR56NO9Wa7ozbvGevOz7B8FBG4sKZW5kc3RyZWFtCmVuZG9iagoxOCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDQzNyA+PgpzdHJlYW0KeJw1kklyJTEIRPd1Ci7gCDFoOs/v8Mp9/22/pNyLKpAQkJkw57JhuezL3aaXzRz2x58aZavs75PbFc4a5hgfNu3zxBn2NS1qd2J4tv08Pt9S7mFhJ4xyn2dS+6jMTf09N5dyVljx+Ez6WozF9aJsbKNBVNm9FlOv3bfFuuQei307NY4SnFNcng8yb5GGTx4dAJJj05K25Ofli47Io/Nrz2tn/I8cbs4FGnk7reIoMoeV3qJDTaGItqgByb4ZsggF+MrGtvAChoV2dzbznPeVRNL+PJwKjCpGEB61JJmPY4V+nmlzSPzNfIQwBmrGy1PTilZPOeImL9FQLxK5NdPPIwyTkRac6/JN/K1JFnVLGDasqFiHqAt7Hd6IESq3CrLZ1fACPX/a85zEmFh16SWMBVfBGwxpNIbRKAJLFjwcekOi2O+qvdIH5Fm69e6WhhYIGdqO0BqobUjQq61DUGDHuC01NyPNNQCIe6lJ7ySgfR2AEoF42+wcearCUl2YsLynxd8NSfOcQlDWOxgU0fkeRROF9/1dDPYut4phj5r3PC4QICRizj41wXeXfqn+PN//ABlPplMKZW5kc3RyZWFtCmVuZG9iagoxOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI1MSA+PgpzdHJlYW0KeJxNkUtuxEAIRPd9Ci4wEv9uzuPRrJz7b1PgiZSFRYmigYcjkpgy6CVCIUFbkt6yxDfyPwsO092hjXtJ2D/l2aYUU3CS2qYwJq2YeC3TJ2OBqEZ2EkMEjouT1yE3fIfJsYejCs6GY+h9ipwDVUW2FU4wlNWsaewTNftNK9noVEaCKveinDFoIJgbOoBaMvFaqjVKDj6nxgXj9eUGk4MnDRcYxagKlGFL2dB6uhabHQN29jA9/sw01RYWuNkGMlb6IvcowCl2qidq9pla9am6wuYNsIGBXgEoR1b7PAdDzkQAAHEy+BldIe3wrIWetpsp590fyrU+v/AiXAYKZW5kc3RyZWFtCmVuZG9iagoyMCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI0NiA+PgpzdHJlYW0KeJw1UUlyBDEIu/sV+kCqDMLGfk+nckr+f41gZg5dwsZoodecmODFlxniGhYnvm3EDNid+Bt1aXnxO+KosotYiXSEnRp8BtVdIK1JPFfjM3yyK4sNc5iO6+h+T9VRs5at7SIUzQWNCLFux06Uh2echSiCamXCVvksGjuSlR2X43JdwoVi4isH9X6Z5pu2NCLKvr63/zgutd3qCS4qJsVLvWZGT3IJac0rHjFwalJRPG+jojK6MjmL8A4WVl5MJ6Y6rjl/oe/uqKoV1wurZWx9s5PdhdwdbNdCo0DyaqogtX6BSK7X9WFvh9KuVX9+3TN+/gHOaljNCmVuZHN0cmVhbQplbmRvYmoKMjEgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNTcgPj4Kc3RyZWFtCnicNVAxkgMxCOv9Cj6QGYMA2+9J5qrN/9sTeFJJi1cgKSJlClxequLbZWnKR4dP4/zbmCHPQB5RF5j2rxar8T1Mo5muRYU6xMCxzi0eslU4TxPVzSlfNXbdsxkNcTgmgU5xE3Bv0tCpMZXu3DwhvlwU/D5Zy5dKcKFRFJjcgka6YYRiUJOgEVATWi9IBjxLsCtonUga7OtkFfsZvwIum4XdwzPUor1+m+lhIGymJWYyXF3Q4xXWjBHYEOdZBWF6EYBXUpCsYO4+y7pwxPmuezValKIYjGXwDzB4afxqEF0JaMtZpOVFBrfLlBpvWy5+bdEoFl9oHPT2i/Ief//jlV6CCmVuZHN0cmVhbQplbmRvYmoKMjIgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNTUgPj4Kc3RyZWFtCnicNVDJbQQxDPu7CjYQQKdl17NBXpv+v6E0G4wHImyRlJi5IfDClypSE+mCb11hCt2F35VqUHa9V9yCiiFKsBXhBp7X8uvgFzEidp76WiZnkAZ5FBFHHt7nJY421Rpvy2yZooaBr6EyHTHtGgcpGyY101ndqWT0C1FITkcEueS/OKpTxWYjjz3VdnMGZfAmYBxsKq3pYzXovZSaShclU51/JefZs1KgOEpMAr3q7k1dd4OOYF84czvd7ec+gUkHwNk+odKrs5PLeMMexHj1wNOn2w/nJrsxdTrtoL49mdiRTzbm97lhAkF3rcO9xyEZ7eUeTiXu++/4Wj9/SRdcugplbmRzdHJlYW0KZW5kb2JqCjIzIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMjQ2ID4+CnN0cmVhbQp4nEVRO27FMAzbfQpeoID1tX2eFJ1e77+WTAJ0SMTYEkkxVY2JdHyZoayxZuPbBk9sb/wOi4SdwmdYTljpaRg7bRtO4hrOkWXwaPSEE7JcIywEIg9WI9aBzc3z5Ftc6UEOv6tH6UZoF9QRnojN8QpxlfOLjxXiLIrMu1KcTu8TOopDdyeopRtTT9O9ZvPRDJeDWojYWsmfhT8jSf6P2l23pH1RbWRMIp+G1JbZyFVwyqnKYt1I5pOms9hpCiKZUnTf8cYbs6fWErLFfioZlbKY1Z0EszO6JId+jDJVfbIWMipwxjgjIq7KfFMsC++/u8bPHx5KWPYKZW5kc3RyZWFtCmVuZG9iagoyNCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDM1MyA+PgpzdHJlYW0KeJw9UsuNJSEQu3cUTqAl6gvE81Zz2s3/OjbM7KkswO1PdVVjIByvj41cGxUDf+ypgTcd/x6LwluFv0Qb7zJYctoYsOaVmePz2JqEvuDWBDHhlRd8Ht8/MLzvq8j1y4xJpd2IxbsZCD3O1M3mIzJSUnZGL523wR1hQet+AshAcppvnmAPTAkXzUrNYfuk9D6SIvODhqiJ2tTcPCYleWYjkDQiw0lj/PY0lCmbnXpcrJBq8FQoaKxG0eNCljhCZZu1nVL/I6a9gA2yb2ZvJuy8jZJ3molEMuJrPW4xto6ctmI0G2PejTk3cwBd7rjQMvSqmET7aTXRWk9owyxbXs92SfJgKVt7FhqMHw1Tcc0K/FTL7hbTe2LqdWnQzDggYuk2GMNsn7WRSIGkR3nV9L6d6BdQ3aSx7Wg1djhxuw91X9TqO/nvUFXINhdGy8aq0o4tGo3JG5cJw2+Uz/P1DSHihlYKZW5kc3RyZWFtCmVuZG9iagoyNSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE3MyA+PgpzdHJlYW0KeJxNULsRQzEI6z2FRuBr8DzJpXrZvw3wcrkUtoR0BmH3DcLuyzIRvPHkNfW7oYVrsdsfMz/NjsFpQ2zDlaB0Bh9LLW4lGS4Ca4eiHNOAnVI8YcF1aqx5O1kq54TQODfKGScrzLWM6IftfFko9FQX8uLVUbqbukxfrcrcIYcHH0v2V7GenBCuRGrl9D4mBI5KQQLed5JyKq3WP/XuKjYoOW+Kdab5r2u9PvyNPrMKZW5kc3RyZWFtCmVuZG9iagoyNiAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDUwID4+CnN0cmVhbQp4nDM2NlYwUDA2UjAyNFUwNjACYmOFFEMuqEguiAESyuGCSUJYIMkcmKocrjQA4qMNkwplbmRzdHJlYW0KZW5kb2JqCjI3IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNjcgPj4Kc3RyZWFtCnicMzIyUjBQMDMDEoamJgrmhmYKKYZcQL6ZoalCLogBEsrhgklCWCDJHJgqMMMAotjU0BKqBMEygKnI4UoDAJV6FUwKZW5kc3RyZWFtCmVuZG9iagoyOCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDg3ID4+CnN0cmVhbQp4nE2MsRHAMAgDe0/BCICNwPvkUiX7twHHuUuDHpBkzMQE5BgT5AI6pK39LqnD1cTGjzpHUoeSyUwajk3qQV3LV2VcroilmmmF7Qblr6t+eK3nA1GLGuQKZW5kc3RyZWFtCmVuZG9iagoyOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDQ4ID4+CnN0cmVhbQp4nDMyMlIwUDAzARKGpkYK5oZmCimGXGB+LogCCeRwwaQgLAMgDVaRw5UGAIAODCUKZW5kc3RyZWFtCmVuZG9iagozMCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI3NCA+PgpzdHJlYW0KeJxNUkt2xDAI2/sUHMH88Xmmr6vp/bcVONPXRSJigZCSlCptisAt45Drpi9e8/zT4Hzovdj8X2UWXQWTJROXkB0m2RjfSa8lijkRksgR1C2Dr6W27wnmukMPOneAMdlQwaVFhjkzoJ5m6oCBtl2V2PWoRVy1aJadUjBX7SAdLv1MJGsGqN57utpwH8V/qJh7KmSKLNIKIHLBee8JqBQFFKz3GZLbOBCc2IEn7HEjr4uv5S73BCmt+w/ycHuzNORWpER+dAxubgZVe1J0fFA1P1UG/GHfdsxBVbzfaL8vA8s2+6TkcSCe90TBWJIwdmozfHoaXy5zEnIgcbY39hpt9laMQYkzzPU0f8Z7ff8CZ7BqnwplbmRzdHJlYW0KZW5kb2JqCjMxIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTgxID4+CnN0cmVhbQp4nE1QQRICMQi79xU8oRDownt0POn/rwYcHQ/bBFLSsBFHtpw+PC8JbLnrmvrVEFryXOrxx5wfWUJiqxhyxqB78Lbg+ulc7JgLqn1Axc04Y3Swec6DbqdaOclKxS92rajyxvZWMgSZcx9RH9SZIdtMgqofQuPL6IbiLB2RNZzZ2pdZOptbO0KcG1BBb5bj4OFiZYO3ZTynYzrJtVhrz+ihAyulCq9By960WWeaP/lcjzeeU0O7CmVuZHN0cmVhbQplbmRvYmoKMzIgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMzIgPj4Kc3RyZWFtCnicLVBBksRACLrnFX5gq1pFO/2e2Zrb/v+6YHKCxEbAqrZlmfbjbuXHKpf9+sU/Ucf+RLLKyBFt7mnYaZ/La/O9W3iMJnYPfq7EHoZF2WpDuaE1weEXN8gncQajNyfD1uL7Y049biI5NX1sc0EyAGHRcUw6lTt8gstc+LliPVUcMCZz7bxlUORQUee2tx1bBN6eYn44zptiInO5y8pP2d4WGdaPVcspmYMkeUBO8673ORyzAMEKB4PRoQlZhk7AIBujwVI6XRislzwDmFcmmNxyFVMIvVCsR6OguenK4BkPPqW+/1TOVsIKZW5kc3RyZWFtCmVuZG9iagozMyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDExNCA+PgpzdHJlYW0KeJw1TssNQzEMumcKRvDf8Tyv6ind/1rHai8GYUC4BwhM1VdTkVx48bqU8FmyvfEMegwLhRtBtJU2CzGsCs/iSFgWWAMWNqXmdj/NXKvT7Lt7ZFJet2UjRNsjaQh3KBFiJ5RjxjzrP+v8Vp31/gItliJeCmVuZHN0cmVhbQplbmRvYmoKMzQgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNzYgPj4Kc3RyZWFtCnicTZFLcsMwDEP3PgUvkBnxK+k86XSV3H/bR7npdGGThigAhDNLhlTJQ/eS1JL0IV96faB3d6lbXpfG/y5Su6uQmFN0gewppoOZIc/LPCTNxcoOp+2b+3l5jNP53MwuCXXuFicREza+pkmEgjK1Nyc5pnjO49DVTrXyPumuVUeJohULN9Y6UUuwFsgFLkeIWcsDQ4uBhyq27orh+kUw/kg4VSawNt+GegkHmmwVDfM+Ab3+orpzMRJ9n04X15IHA52PjtUybDsZY6AQW9EFV0RF49zGswPriTFYVoNIDIIdp1q1g+56i57oKH3l6eFKQmVlZyKOyDoV8Rw3op2LH4txbGn1DwHBl5vJZ5Xn9f0DZepl8gplbmRzdHJlYW0KZW5kb2JqCjM1IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggNDIgPj4Kc3RyZWFtCnicMzK3UDBQsDQEEoZA0tDAQCHFkAvMz+WCCuRwGaKwQDSUSgMAfswMEgplbmRzdHJlYW0KZW5kb2JqCjM2IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTQ0ID4+CnN0cmVhbQp4nE2PMbIDIQxD+z2Fj2DJNrDnyZ9fbe7fRoYUKUACzZPHEWFuo3RFTqtw+8O13++Wwm3Phcwfl+nthtwchkUrd6NOgfa6CBUxjTF2IeuoknVcQMrvSG+mf5Ldy1u9QeMM6S2aW0XHcURJY09MNo0lBm4QE+IxpsXMTkouaSglvrZyYie9+XN2fa7/D8M0MY8KZW5kc3RyZWFtCmVuZG9iagozNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDQwOCA+PgpzdHJlYW0KeJwtkzlyA0EIRfM5BRdQVbP0dh65HNn3T/0+cqCBoaf5C2iOYcPS7eVutdxmDvvyh4rPab+P+zFfx34ejzDP5LfMB7957Bx7P36G7WHhYTSI2eH9xO0kubauZYWdSTmp+aC+ifEfK3WibKXlAmgfrkzzO3Qi2EtlpMU4YEyL0IlfEEWoFtXRHYP77+dS2GV7W9BjFSLFaoKYYbROmigclUlqlNGhYop2lchumTLtLKsj2mV1S7C4MnmTFxPdvgbxthl0CbDHbS9FThE/9idLvp8BH0TrroQLmZO7wYEJVcFCi05iUnQpMAsX82PjXFiXzCUxPK8cFvnRhlf2N9hVSApMKKYVlR3hMqIz36po1vTPHibPCimjGxoU83qfkO3bO5BM1AsrVw+hGI3Mn7dV+cbW3O0G32a2GylXxNXloJ6hkYoXnuWJnhea2a1gDaQ9Cuboitg9JO6M3XvqJz5xysHOpJu3O1v+0jTaMyhcKIDRqyfwuPYSfvTei9Wrt0PTUyaNdNLGUWWfYAkNufz5Y7yf7z+F45iZCmVuZHN0cmVhbQplbmRvYmoKMzggMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxNyA+PgpzdHJlYW0KeJwzMrdQMIDDFEMuABrjAvEKZW5kc3RyZWFtCmVuZG9iagozOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDE3MCA+PgpzdHJlYW0KeJxFUDkOwzAM2/0KfiCARR+y3pOiU/v/tZRTJItJkLJIm75QYQvHBN3gteJlhWPBA9+SmuFT2AeOBrLtydoTzmLOJNYdhwZbxUrVmCtNu5ohGnqqa2B2LCIiTxtMkeijKkDzNxkWIrJuMhUga8YueLHLzKYP+6+Q+zC77xrV0fXcOoQdscu6I6QrRQ1tqZylHBNyWAUDVILgLOQm7ITrH65vOsv7BzKGPYkKZW5kc3RyZWFtCmVuZG9iago0MCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI2NyA+PgpzdHJlYW0KeJw1UUlywzAMu/sVeIK4i+9Jp6fk/9eCzHTGMmhzA6CIxIE2X3EMJY0feSa8js8GB+/HzgLrVGAGl3lS8HrC0GxUiDr6Qjjx9cyH3IKkQZVHeDKY0eYEvTA3WBFrZk2Psdtjhiv83sVQZWYjzrVuxCWWc/mZHm+kOUwK6QmtL3KPxffPIVFSlkrkucMtKPaSsBXC64tn9zDgqveIimpMC6UL6WWuLJIoDlSR9UqniDhEaiPnoCRNd+Ia5FyVtGBWBCcu6pCfyGmHd8JplNNzt1gizJxaO8YkV4r2uyb1irVwbg+MnbomqdF81uqh9ayV25Q2GaFdo0GSog/1hM71vv7v+f38/gErHWDYCmVuZHN0cmVhbQplbmRvYmoKNDEgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAxODkgPj4Kc3RyZWFtCnicTVDBbQAxCPtnCi9QKUAgME+rvq77f2uSq9QHMjJgA+6BiVj4EMHKBZfCl4w1m/85uAPPsHBIwmSeVl1y8HPoy0iSYY87grRoQTZkFkxRAZ9k0xCJvZCFYIM4yVZmD5cQrwO1m77LPENc/2Vq8maSbWeMnqSXZRuHHV2hC3WkFDzr7rknx4+TXifSFGFi3JNVM7vdxr9w2rYeMUuiVReKp4bCeJIwGvsZXYl3zb8/3mw2nnc+4/sX9s1EjAplbmRzdHJlYW0KZW5kb2JqCjQyIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTE5ID4+CnN0cmVhbQp4nDWOSxLDMAhD9z4FR7D42OY86XSV3H9boGTB8EZCjDYzTZKYDSWD0wcDK8VnSAn3cG3AXsRFPDdhnqJDgIQeijqxaLDRNQSrcxrXTSfdzBlm5FIznf11sffd2yZcyzKRXKgt+5AASRbk1cakvH/xe3x/HxopCQplbmRzdHJlYW0KZW5kb2JqCjQzIDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTIwID4+CnN0cmVhbQp4nDWOOxIDMQhDe59CR+BrlvMkk2pz/zbCSQpARm9kUgSCYkU7UhtPXfN8L+2CleBepse5l/pfmSqcAJUlvBzmBc+ABac3Hss2mS0MGaJhF53edLzilxPcf5XLkDE7XjLf+pVnWl1QZlElGcm5RON4dfrrA6GAJdIKZW5kc3RyZWFtCmVuZG9iago0NCAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI3MiA+PgpzdHJlYW0KeJw1UUtuBTEI288pfIFK/EnOM1V3vf+2JumTZgQJ2BgnsyAIw5cqUhZaN7714Y2n43eS8GaJX6IWMhvvs5jLhhJVwRg89xS0N5qdZn64rPPE93G9Nx7NqPAu1E5WQoLoTRkLRfpgRzFnpQq5WVlUV4HYhjRjJYXClhzNwVkTR/FUFqyIIc5E2WXUtw9bYpPeN5IoqnQZYa3gutbHhBE88X1MbqbJ37mrURXvyaKmY5rpDP+fq/7xbDLzPK4o99Ee9DqUAi5qzoXljKqjQE/isaY6xtz2MWYIgqchnHiHTRbUPR0ZF5NrMENSVnDljCgOuZHD3e8NTSnjo/HB8jyA0vA8W9LUFnxWeZ+fP/SWZUsKZW5kc3RyZWFtCmVuZG9iagoxNCAwIG9iago8PCAvQmFzZUZvbnQgL0FyaWFsTVQgL0NoYXJQcm9jcyAxNSAwIFIKL0VuY29kaW5nIDw8Ci9EaWZmZXJlbmNlcyBbIDMyIC9zcGFjZSA0NSAvaHlwaGVuIC9wZXJpb2QgNDggL3plcm8gL29uZSAvdHdvIDUzIC9maXZlIDczIC9JIDg2IC9WIDk3Ci9hIC9iIC9jIC9kIC9lIDEwMyAvZyAvaCAvaSAxMDcgL2sgL2wgL20gL24gL28gL3AgMTE0IC9yIC9zIC90IC91IDExOSAvdyAveApdCi9UeXBlIC9FbmNvZGluZyA+PgovRmlyc3RDaGFyIDAgL0ZvbnRCQm94IFsgLTY2NSAtMzI1IDIwMDAgMTAwNiBdIC9Gb250RGVzY3JpcHRvciAxMyAwIFIKL0ZvbnRNYXRyaXggWyAwLjAwMSAwIDAgMC4wMDEgMCAwIF0gL0xhc3RDaGFyIDI1NSAvTmFtZSAvQXJpYWxNVAovU3VidHlwZSAvVHlwZTMgL1R5cGUgL0ZvbnQgL1dpZHRocyAxMiAwIFIgPj4KZW5kb2JqCjEzIDAgb2JqCjw8IC9Bc2NlbnQgOTA2IC9DYXBIZWlnaHQgNzE2IC9EZXNjZW50IC0yMTIgL0ZsYWdzIDMyCi9Gb250QkJveCBbIC02NjUgLTMyNSAyMDAwIDEwMDYgXSAvRm9udE5hbWUgL0FyaWFsTVQgL0l0YWxpY0FuZ2xlIDAKL01heFdpZHRoIDEwMTUgL1N0ZW1WIDAgL1R5cGUgL0ZvbnREZXNjcmlwdG9yIC9YSGVpZ2h0IDUxOSA+PgplbmRvYmoKMTIgMCBvYmoKWyA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MAo3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDc1MCA3NTAgNzUwIDI3OCAyNzggMzU1IDU1NiA1NTYKODg5IDY2NyAxOTEgMzMzIDMzMyAzODkgNTg0IDI3OCAzMzMgMjc4IDI3OCA1NTYgNTU2IDU1NiA1NTYgNTU2IDU1NiA1NTYgNTU2CjU1NiA1NTYgMjc4IDI3OCA1ODQgNTg0IDU4NCA1NTYgMTAxNSA2NjcgNjY3IDcyMiA3MjIgNjY3IDYxMSA3NzggNzIyIDI3OAo1MDAgNjY3IDU1NiA4MzMgNzIyIDc3OCA2NjcgNzc4IDcyMiA2NjcgNjExIDcyMiA2NjcgOTQ0IDY2NyA2NjcgNjExIDI3OCAyNzgKMjc4IDQ2OSA1NTYgMzMzIDU1NiA1NTYgNTAwIDU1NiA1NTYgMjc4IDU1NiA1NTYgMjIyIDIyMiA1MDAgMjIyIDgzMyA1NTYgNTU2CjU1NiA1NTYgMzMzIDUwMCAyNzggNTU2IDUwMCA3MjIgNTAwIDUwMCA1MDAgMzM0IDI2MCAzMzQgNTg0IDc1MCA1NTYgNzUwIDIyMgo1NTYgMzMzIDEwMDAgNTU2IDU1NiAzMzMgMTAwMCA2NjcgMzMzIDEwMDAgNzUwIDYxMSA3NTAgNzUwIDIyMiAyMjIgMzMzIDMzMwozNTAgNTU2IDEwMDAgMzMzIDEwMDAgNTAwIDMzMyA5NDQgNzUwIDUwMCA2NjcgMjc4IDMzMyA1NTYgNTU2IDU1NiA1NTYgMjYwCjU1NiAzMzMgNzM3IDM3MCA1NTYgNTg0IDMzMyA3MzcgNTUyIDQwMCA1NDkgMzMzIDMzMyAzMzMgNTc2IDUzNyAzMzMgMzMzIDMzMwozNjUgNTU2IDgzNCA4MzQgODM0IDYxMSA2NjcgNjY3IDY2NyA2NjcgNjY3IDY2NyAxMDAwIDcyMiA2NjcgNjY3IDY2NyA2NjcKMjc4IDI3OCAyNzggMjc4IDcyMiA3MjIgNzc4IDc3OCA3NzggNzc4IDc3OCA1ODQgNzc4IDcyMiA3MjIgNzIyIDcyMiA2NjcgNjY3CjYxMSA1NTYgNTU2IDU1NiA1NTYgNTU2IDU1NiA4ODkgNTAwIDU1NiA1NTYgNTU2IDU1NiAyNzggMjc4IDI3OCAyNzggNTU2IDU1Ngo1NTYgNTU2IDU1NiA1NTYgNTU2IDU0OSA2MTEgNTU2IDU1NiA1NTYgNTU2IDUwMCA1NTYgNTAwIF0KZW5kb2JqCjE1IDAgb2JqCjw8IC9JIDE2IDAgUiAvViAxNyAwIFIgL2EgMTggMCBSIC9iIDE5IDAgUiAvYyAyMCAwIFIgL2QgMjEgMCBSIC9lIDIyIDAgUgovZml2ZSAyMyAwIFIgL2cgMjQgMCBSIC9oIDI1IDAgUiAvaHlwaGVuIDI2IDAgUiAvaSAyNyAwIFIgL2sgMjggMCBSCi9sIDI5IDAgUiAvbSAzMCAwIFIgL24gMzEgMCBSIC9vIDMyIDAgUiAvb25lIDMzIDAgUiAvcCAzNCAwIFIKL3BlcmlvZCAzNSAwIFIgL3IgMzYgMCBSIC9zIDM3IDAgUiAvc3BhY2UgMzggMCBSIC90IDM5IDAgUiAvdHdvIDQwIDAgUgovdSA0MSAwIFIgL3cgNDIgMCBSIC94IDQzIDAgUiAvemVybyA0NCAwIFIgPj4KZW5kb2JqCjMgMCBvYmoKPDwgL0YxIDE0IDAgUiA+PgplbmRvYmoKNCAwIG9iago8PCAvQTEgPDwgL0NBIDAgL1R5cGUgL0V4dEdTdGF0ZSAvY2EgMSA+PgovQTIgPDwgL0NBIDEgL1R5cGUgL0V4dEdTdGF0ZSAvY2EgMSA+PiA+PgplbmRvYmoKNSAwIG9iago8PCA+PgplbmRvYmoKNiAwIG9iago8PCA+PgplbmRvYmoKNyAwIG9iago8PCA+PgplbmRvYmoKMiAwIG9iago8PCAvQ291bnQgMSAvS2lkcyBbIDEwIDAgUiBdIC9UeXBlIC9QYWdlcyA+PgplbmRvYmoKNDUgMCBvYmoKPDwgL0NyZWF0aW9uRGF0ZSAoRDoyMDE4MDgyODE1Mzg0Mi0wMicwMCcpCi9DcmVhdG9yIChtYXRwbG90bGliIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcpCi9Qcm9kdWNlciAobWF0cGxvdGxpYiBwZGYgYmFja2VuZCAyLjIuMikgPj4KZW5kb2JqCnhyZWYKMCA0NgowMDAwMDAwMDAwIDY1NTM1IGYgCjAwMDAwMDAwMTYgMDAwMDAgbiAKMDAwMDAxMTE2MCAwMDAwMCBuIAowMDAwMDEwOTY2IDAwMDAwIG4gCjAwMDAwMTA5OTggMDAwMDAgbiAKMDAwMDAxMTA5NyAwMDAwMCBuIAowMDAwMDExMTE4IDAwMDAwIG4gCjAwMDAwMTExMzkgMDAwMDAgbiAKMDAwMDAwMDA2NSAwMDAwMCBuIAowMDAwMDAwMzk4IDAwMDAwIG4gCjAwMDAwMDAyMDggMDAwMDAgbiAKMDAwMDAwMTM3NSAwMDAwMCBuIAowMDAwMDA5NTc5IDAwMDAwIG4gCjAwMDAwMDkzNzkgMDAwMDAgbiAKMDAwMDAwODk0NSAwMDAwMCBuIAowMDAwMDEwNjMwIDAwMDAwIG4gCjAwMDAwMDEzOTUgMDAwMDAgbiAKMDAwMDAwMTUxNSAwMDAwMCBuIAowMDAwMDAxNjg1IDAwMDAwIG4gCjAwMDAwMDIxOTUgMDAwMDAgbiAKMDAwMDAwMjUxOSAwMDAwMCBuIAowMDAwMDAyODM4IDAwMDAwIG4gCjAwMDAwMDMxNjggMDAwMDAgbiAKMDAwMDAwMzQ5NiAwMDAwMCBuIAowMDAwMDAzODE1IDAwMDAwIG4gCjAwMDAwMDQyNDEgMDAwMDAgbiAKMDAwMDAwNDQ4NyAwMDAwMCBuIAowMDAwMDA0NjA5IDAwMDAwIG4gCjAwMDAwMDQ3NDggMDAwMDAgbiAKMDAwMDAwNDkwNyAwMDAwMCBuIAowMDAwMDA1MDI3IDAwMDAwIG4gCjAwMDAwMDUzNzQgMDAwMDAgbiAKMDAwMDAwNTYyOCAwMDAwMCBuIAowMDAwMDA1OTMzIDAwMDAwIG4gCjAwMDAwMDYxMjAgMDAwMDAgbiAKMDAwMDAwNjQ2OSAwMDAwMCBuIAowMDAwMDA2NTgzIDAwMDAwIG4gCjAwMDAwMDY4MDAgMDAwMDAgbiAKMDAwMDAwNzI4MSAwMDAwMCBuIAowMDAwMDA3MzcwIDAwMDAwIG4gCjAwMDAwMDc2MTMgMDAwMDAgbiAKMDAwMDAwNzk1MyAwMDAwMCBuIAowMDAwMDA4MjE1IDAwMDAwIG4gCjAwMDAwMDg0MDcgMDAwMDAgbiAKMDAwMDAwODYwMCAwMDAwMCBuIAowMDAwMDExMjIwIDAwMDAwIG4gCnRyYWlsZXIKPDwgL0luZm8gNDUgMCBSIC9Sb290IDEgMCBSIC9TaXplIDQ2ID4+CnN0YXJ0eHJlZgoxMTM3NAolJUVPRgo=\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig,ax = plt.subplots()\n",
    "fig.set_size_inches(20,10)\n",
    "sns.barplot(data=feature_importance[:10],x=\"feature\",y=\"importance\",ax=ax,color=default_color,)\n",
    "ax.set(xlabel='Variable name', ylabel='Importance',title=\"Variable importances\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train: 3364\n",
      "Test:842\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(previsores, target, test_size = 0.2, random_state = 0)\n",
    "X_train.shape\n",
    "X_test.shape\n",
    "print('Train: {}\\nTest:{}'.format(len(X_train), len(X_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import  accuracy_score, log_loss\n",
    "\n",
    "def avaliacao(y_test, previ , pred_proba):\n",
    "    \n",
    "    precisao = accuracy_score(y_test, previ)\n",
    "    print('accuracy: {}\\n'.format(precisao))\n",
    "    print('Log_loss: {}\\n\\n'.format(log_loss(y_test,pred_proba)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/eduardoquadros/anaconda3/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras_tqdm import TQDMNotebookCallback\n",
    "from sklearn.svm import SVC\n",
    "from xgboost import XGBClassifier\n",
    "\n",
    "def treinamento(X_train, X_test, y_train, y_test, alg):\n",
    "    \n",
    "    \n",
    "    def a_naivebayes(X_train, X_test, y_train, y_test):\n",
    "        classificador = GaussianNB()\n",
    "        classificador.fit(X_train, y_train)\n",
    "        previsoes = classificador.predict(X_test)\n",
    "        pred_proba = classificador.predict_proba(X_test)\n",
    "        print('Naive Bayes')\n",
    "        avaliacao(y_test, previsoes, pred_proba)\n",
    "        \n",
    "    def a_knn(X_train, X_test, y_train, y_test):\n",
    "        classificador = KNeighborsClassifier(n_neighbors=5, metric='minkowski', p=2)\n",
    "        classificador.fit(X_train, y_train)\n",
    "        previsoes = classificador.predict(X_test)\n",
    "        pred_proba = classificador.predict_proba(X_test)\n",
    "        print('KNN')\n",
    "        avaliacao(y_test, previsoes, pred_proba)\n",
    "        \n",
    "    def a_rf(X_train, X_test, y_train, y_test):\n",
    "        \n",
    "        rf_params = {}\n",
    "        rf_params['n_estimators'] = 200\n",
    "        rf_params['random_state'] = 0\n",
    "        rf_params['criterion'] = 'entropy'\n",
    "        rf_params['max_depth'] = 6\n",
    "        rf_params['min_samples_split'] = 70\n",
    "        rf_params['min_samples_leaf'] = 30\n",
    "        classificador = RandomForestClassifier(**rf_params)\n",
    "        classificador.fit(X_train, y_train)\n",
    "        previsoes = classificador.predict(X_test)\n",
    "        pred_proba = classificador.predict_proba(X_test)\n",
    "        print('RandomForest')\n",
    "        avaliacao(y_test, previsoes, pred_proba)\n",
    "        \n",
    "    def a_lr(X_train, X_test, y_train, y_test):\n",
    "        classificador = LogisticRegression()\n",
    "        classificador.fit(X_train, y_train)\n",
    "        previsoes = classificador.predict(X_test)\n",
    "        pred_proba = classificador.predict_proba(X_test)\n",
    "        print('LogisticRegression')\n",
    "        avaliacao(y_test, previsoes, pred_proba)\n",
    "    \n",
    "    def a_svm(X_train, X_test, y_train, y_test):\n",
    "        classificador = SVC(kernel = 'linear', random_state = 1, probability= True)\n",
    "        classificador.fit(X_train, y_train)\n",
    "        previsoes = classificador.predict(X_test)\n",
    "        pred_proba = classificador.predict_proba(X_test)\n",
    "        print('SVM')\n",
    "        avaliacao(y_test, previsoes, pred_proba)\n",
    "        \n",
    "    def a_keras(X_train, X_test, y_train, y_test):\n",
    "        classificador = Sequential()\n",
    "        classificador.add(Dense(units = round((X_train.shape[1]+1)/2), activation = 'relu', input_dim = X_train.shape[1]))\n",
    "        classificador.add(Dense(units = round((X_train.shape[1]+1)/2), activation = 'relu'))\n",
    "        classificador.add(Dense(units = 21, activation = 'relu'))\n",
    "        classificador.add(Dense(units = 1, activation = 'sigmoid'))\n",
    "        classificador.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy'])\n",
    "        #print(classificador.summary())\n",
    "\n",
    "\n",
    "        classificador.fit(X_train, y_train, batch_size = 10, epochs = 100, verbose=0, callbacks=[TQDMNotebookCallback()])\n",
    "\n",
    "        previsoes = classificador.predict(X_test)\n",
    "        previsoes = (previsoes > 0.5)\n",
    "        pred_proba = classificador.predict_proba(X_test)\n",
    "        pred_proba = pred_proba.astype(np.float64)\n",
    "\n",
    "\n",
    "        print('NeuralNetwork Keras')\n",
    "        avaliacao(y_test, previsoes, pred_proba)\n",
    "        \n",
    "    def a_xgb(X_train, X_test, y_train, y_test):\n",
    "        \n",
    "        # XGBoost params# XGBoos \n",
    "        xgb_params = {}\n",
    "        xgb_params['learning_rate'] = 0.02\n",
    "        xgb_params['n_estimators'] = 1000\n",
    "        xgb_params['max_depth'] = 4\n",
    "        xgb_params['subsample'] = 0.9\n",
    "        xgb_params['colsample_bytree'] = 0.9\n",
    "        classificador = XGBClassifier(**xgb_params)\n",
    "        classificador.fit(X_train, y_train)\n",
    "        previsoes = classificador.predict(X_test)\n",
    "        pred_proba = classificador.predict_proba(X_test)\n",
    "        print('XGBoost')\n",
    "        avaliacao(y_test, previsoes, pred_proba)\n",
    "        \n",
    "\n",
    "    if alg == 'naive_bayes':\n",
    "        a_naivebayes(X_train, X_test, y_train, y_test)\n",
    "    elif alg == 'knn':\n",
    "        a_knn(X_train, X_test, y_train, y_test)\n",
    "    elif alg == 'rf':\n",
    "        a_rf(X_train, X_test, y_train, y_test)        \n",
    "    elif alg == 'lr':\n",
    "        a_lr(X_train, X_test, y_train, y_test)\n",
    "    elif alg == 'svm':\n",
    "        a_svm(X_train, X_test, y_train, y_test)\n",
    "    elif alg == 'keras':\n",
    "        a_keras(X_train, X_test, y_train, y_test)\n",
    "    elif alg == 'xgb':\n",
    "        a_xgb(X_train, X_test, y_train, y_test)\n",
    "    elif alg == 'all':\n",
    "        a_naivebayes(X_train, X_test, y_train, y_test)\n",
    "        a_knn(X_train, X_test, y_train, y_test)\n",
    "        a_rf(X_train, X_test, y_train, y_test)\n",
    "        a_lr(X_train, X_test, y_train, y_test)\n",
    "        a_keras(X_train, X_test, y_train, y_test)\n",
    "        a_xgb(X_train, X_test, y_train, y_test)\n",
    "        a_svm(X_train, X_test, y_train, y_test)\n",
    "        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Resultados realizando a limpeza de dados, escalonando as variaveis no dataset, com excessão das target e das colunas com types ordinais."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive Bayes\n",
      "accuracy: 0.7755344418052257\n",
      "\n",
      "Log_loss: 0.8448387043647315\n",
      "\n",
      "\n",
      "KNN\n",
      "accuracy: 0.7351543942992874\n",
      "\n",
      "Log_loss: 1.6499625579148174\n",
      "\n",
      "\n",
      "RandomForest\n",
      "accuracy: 0.7992874109263658\n",
      "\n",
      "Log_loss: 0.39656537762251526\n",
      "\n",
      "\n",
      "LogisticRegression\n",
      "accuracy: 0.7648456057007126\n",
      "\n",
      "Log_loss: 0.4944409665293562\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5f52a97208b046c1aed17522a4c49d04",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Training'), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 0', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 1', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 2', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 3', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 4', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 5', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 6', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 7', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 8', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 9', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 10', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 11', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 12', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 13', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 14', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 15', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 16', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 17', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 18', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 19', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 20', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 21', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 22', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 23', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 24', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 25', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 26', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 27', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 28', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 29', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 30', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 31', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 32', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 33', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 34', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 35', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 36', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 37', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 38', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 39', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 40', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 41', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 42', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 43', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 44', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 45', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 46', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 47', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 48', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 49', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 50', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 51', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 52', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 53', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 54', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 55', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 56', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 57', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 58', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 59', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 60', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 61', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 62', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 63', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 64', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 65', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 66', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 67', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 68', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 69', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 70', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 71', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 72', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 73', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 74', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 75', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 76', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 77', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 78', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 79', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 80', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 81', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 82', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 83', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 84', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 85', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 86', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 87', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 88', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 89', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 90', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 91', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 92', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 93', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 94', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 95', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 96', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 97', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 98', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 99', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "NeuralNetwork Keras\n",
      "accuracy: 0.7945368171021377\n",
      "\n",
      "Log_loss: 0.4278727345390083\n",
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/eduardoquadros/anaconda3/lib/python3.6/site-packages/sklearn/preprocessing/label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBoost\n",
      "accuracy: 0.832541567695962\n",
      "\n",
      "Log_loss: 0.34077230518316504\n",
      "\n",
      "\n",
      "SVM\n",
      "accuracy: 0.7731591448931117\n",
      "\n",
      "Log_loss: 0.48146906115596944\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "treinamento(X_train, X_test, y_train, y_test, 'all')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Engineering\n",
    "\n",
    "\n",
    "#### One-hot Encoding effect\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total features after one-hot: 83\n",
      "Train: 3364\n",
      "Test:842\n"
     ]
    },
    {
     "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>age</th>\n",
       "      <th>education-num</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>workclass_ Federal-gov</th>\n",
       "      <th>workclass_ Local-gov</th>\n",
       "      <th>workclass_ Private</th>\n",
       "      <th>workclass_ Self-emp-inc</th>\n",
       "      <th>workclass_ Self-emp-not-inc</th>\n",
       "      <th>...</th>\n",
       "      <th>native-country_ Portugal</th>\n",
       "      <th>native-country_ Puerto-Rico</th>\n",
       "      <th>native-country_ Scotland</th>\n",
       "      <th>native-country_ South</th>\n",
       "      <th>native-country_ Taiwan</th>\n",
       "      <th>native-country_ Thailand</th>\n",
       "      <th>native-country_ Trinadad&amp;Tobago</th>\n",
       "      <th>native-country_ United-States</th>\n",
       "      <th>native-country_ Vietnam</th>\n",
       "      <th>native-country_ Yugoslavia</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>39</td>\n",
       "      <td>13</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 83 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  education-num  capital-gain  capital-loss  hours-per-week  \\\n",
       "0   39             13          0.02          0.00            0.40   \n",
       "\n",
       "   workclass_ Federal-gov  workclass_ Local-gov  workclass_ Private  \\\n",
       "0                       0                     0                   0   \n",
       "\n",
       "   workclass_ Self-emp-inc  workclass_ Self-emp-not-inc  \\\n",
       "0                        0                            0   \n",
       "\n",
       "              ...              native-country_ Portugal  \\\n",
       "0             ...                                     0   \n",
       "\n",
       "   native-country_ Puerto-Rico  native-country_ Scotland  \\\n",
       "0                            0                         0   \n",
       "\n",
       "   native-country_ South  native-country_ Taiwan  native-country_ Thailand  \\\n",
       "0                      0                       0                         0   \n",
       "\n",
       "   native-country_ Trinadad&Tobago  native-country_ United-States  \\\n",
       "0                                0                              1   \n",
       "\n",
       "   native-country_ Vietnam  native-country_ Yugoslavia  \n",
       "0                        0                           0  \n",
       "\n",
       "[1 rows x 83 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gc.collect()\n",
    "\n",
    "target = df['income']\n",
    "previsores = df.drop('income', axis = 1)\n",
    "previsores[col_interval] = scaler.fit_transform(df[col_interval])\n",
    "previsores = pd.get_dummies(previsores)\n",
    "print(\"total features after one-hot: {}\".format(len(list(previsores.columns))))\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(previsores, target, test_size = 0.2, random_state = 0)\n",
    "X_train.shape\n",
    "X_test.shape\n",
    "print('Train: {}\\nTest:{}'.format(len(X_train), len(X_test)))\n",
    "previsores.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive Bayes\n",
      "accuracy: 0.4524940617577197\n",
      "\n",
      "Log_loss: 16.86875696669395\n",
      "\n",
      "\n",
      "KNN\n",
      "accuracy: 0.7541567695961995\n",
      "\n",
      "Log_loss: 1.7190415455559125\n",
      "\n",
      "\n",
      "RandomForest\n",
      "accuracy: 0.7945368171021377\n",
      "\n",
      "Log_loss: 0.4191325186618585\n",
      "\n",
      "\n",
      "LogisticRegression\n",
      "accuracy: 0.7921615201900237\n",
      "\n",
      "Log_loss: 0.4061965704033903\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "983d6a63635149ab8a4b56a7fd6f63dd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Training'), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 0', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 1', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 2', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 3', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 4', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 5', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 6', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 7', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 8', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 9', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 10', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 11', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 12', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 13', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 14', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 15', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 16', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 17', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 18', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 19', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 20', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 21', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 22', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 23', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 24', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 25', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 26', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 27', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 28', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 29', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 30', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 31', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 32', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 33', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 34', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 35', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 36', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 37', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 38', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 39', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 40', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 41', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 42', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 43', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 44', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 45', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 46', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 47', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 48', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 49', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 50', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 51', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 52', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 53', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 54', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 55', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 56', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 57', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 58', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 59', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 60', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 61', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 62', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 63', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 64', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 65', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 66', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 67', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 68', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 69', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 70', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 71', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 72', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 73', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 74', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 75', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 76', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 77', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 78', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 79', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 80', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 81', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 82', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 83', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 84', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 85', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 86', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 87', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 88', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 89', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 90', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 91', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 92', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 93', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 94', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 95', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 96', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 97', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 98', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 99', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "NeuralNetwork Keras\n",
      "accuracy: 0.7969121140142518\n",
      "\n",
      "Log_loss: 0.4870076240902428\n",
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/eduardoquadros/anaconda3/lib/python3.6/site-packages/sklearn/preprocessing/label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBoost\n",
      "accuracy: 0.831353919239905\n",
      "\n",
      "Log_loss: 0.3445993785078701\n",
      "\n",
      "\n",
      "SVM\n",
      "accuracy: 0.7921615201900237\n",
      "\n",
      "Log_loss: 0.4142309507480343\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "treinamento(X_train, X_test, y_train, y_test, 'all')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Extra features effect\n",
    "\n",
    "\n",
    "### Extra Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total features after one-hot: 85\n",
      "Train: 3364\n",
      "Test:842\n"
     ]
    },
    {
     "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>age</th>\n",
       "      <th>education-num</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>ord_sum</th>\n",
       "      <th>intenval_median</th>\n",
       "      <th>workclass_ Federal-gov</th>\n",
       "      <th>workclass_ Local-gov</th>\n",
       "      <th>workclass_ Private</th>\n",
       "      <th>...</th>\n",
       "      <th>native-country_ Portugal</th>\n",
       "      <th>native-country_ Puerto-Rico</th>\n",
       "      <th>native-country_ Scotland</th>\n",
       "      <th>native-country_ South</th>\n",
       "      <th>native-country_ Taiwan</th>\n",
       "      <th>native-country_ Thailand</th>\n",
       "      <th>native-country_ Trinadad&amp;Tobago</th>\n",
       "      <th>native-country_ United-States</th>\n",
       "      <th>native-country_ Vietnam</th>\n",
       "      <th>native-country_ Yugoslavia</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>39</td>\n",
       "      <td>13</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.40</td>\n",
       "      <td>52</td>\n",
       "      <td>2214.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 85 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  education-num  capital-gain  capital-loss  hours-per-week  ord_sum  \\\n",
       "0   39             13          0.02          0.00            0.40       52   \n",
       "\n",
       "   intenval_median  workclass_ Federal-gov  workclass_ Local-gov  \\\n",
       "0          2214.00                       0                     0   \n",
       "\n",
       "   workclass_ Private             ...              native-country_ Portugal  \\\n",
       "0                   0             ...                                     0   \n",
       "\n",
       "   native-country_ Puerto-Rico  native-country_ Scotland  \\\n",
       "0                            0                         0   \n",
       "\n",
       "   native-country_ South  native-country_ Taiwan  native-country_ Thailand  \\\n",
       "0                      0                       0                         0   \n",
       "\n",
       "   native-country_ Trinadad&Tobago  native-country_ United-States  \\\n",
       "0                                0                              1   \n",
       "\n",
       "   native-country_ Vietnam  native-country_ Yugoslavia  \n",
       "0                        0                           0  \n",
       "\n",
       "[1 rows x 85 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gc.collect()\n",
    "\n",
    "target = df['income']\n",
    "previsores = df.drop('income', axis = 1)\n",
    "#previsores['bin_sum']  = previsores[col_category].sum(axis=1)\n",
    "previsores['ord_sum']  = previsores[col_ordinal].sum(axis=1)\n",
    "previsores['intenval_median']  = previsores[col_interval].sum(axis=1)\n",
    "\n",
    "previsores[col_interval] = scaler.fit_transform(df[col_interval])\n",
    "previsores = pd.get_dummies(previsores)\n",
    "print(\"total features after one-hot: {}\".format(len(list(previsores.columns))))\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(previsores, target, test_size = 0.2, random_state = 0)\n",
    "X_train.shape\n",
    "X_test.shape\n",
    "print('Train: {}\\nTest:{}'.format(len(X_train), len(X_test)))\n",
    "previsores.head(1)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Naive Bayes\n",
      "accuracy: 0.7553444180522565\n",
      "\n",
      "Log_loss: 1.1772505401039122\n",
      "\n",
      "\n",
      "KNN\n",
      "accuracy: 0.7707838479809976\n",
      "\n",
      "Log_loss: 1.4946824028403476\n",
      "\n",
      "\n",
      "RandomForest\n",
      "accuracy: 0.8004750593824228\n",
      "\n",
      "Log_loss: 0.40456072681821387\n",
      "\n",
      "\n",
      "LogisticRegression\n",
      "accuracy: 0.8135391923990499\n",
      "\n",
      "Log_loss: 0.38364689416450554\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e5677b8e0ba64fea8728232dbd39a1dd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Training'), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 0', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 1', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 2', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 3', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 4', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 5', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 6', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 7', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 8', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 9', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 10', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 11', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 12', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 13', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 14', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 15', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 16', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 17', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 18', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 19', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 20', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 21', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 22', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 23', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 24', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 25', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 26', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 27', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 28', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 29', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 30', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 31', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 32', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 33', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 34', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 35', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 36', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 37', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 38', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 39', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 40', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 41', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 42', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 43', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 44', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 45', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 46', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 47', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 48', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 49', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 50', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 51', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 52', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 53', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 54', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 55', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 56', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 57', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 58', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 59', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 60', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 61', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 62', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 63', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 64', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 65', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 66', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 67', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 68', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 69', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 70', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 71', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 72', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 73', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 74', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 75', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 76', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 77', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 78', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 79', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 80', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 81', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 82', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 83', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 84', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 85', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 86', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 87', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 88', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 89', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 90', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 91', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 92', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 93', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 94', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 95', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 96', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 97', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 98', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 99', max=3364), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "NeuralNetwork Keras\n",
      "accuracy: 0.7482185273159145\n",
      "\n",
      "Log_loss: 3.3030618119366015\n",
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/eduardoquadros/anaconda3/lib/python3.6/site-packages/sklearn/preprocessing/label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBoost\n",
      "accuracy: 0.831353919239905\n",
      "\n",
      "Log_loss: 0.3461598818425923\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "treinamento(X_train, X_test, y_train, y_test, 'all')"
   ]
  },
  {
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}