GradientBoosting.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import scipy as sp\n",
    "import pandas as pd\n",
    "import sklearn\n",
    "\n",
    "# from sklearn.model_selection import KFold\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "#from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score, recall_score, precision_score, f1_score\n",
    "from sklearn.metrics import roc_auc_score, average_precision_score,roc_curve\n",
    "\n",
    "import matplotlib\n",
    "matplotlib.use('nbagg')\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_set = pd.read_csv(\"data/trn.csv\", sep=',', header=0, low_memory=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "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>UF_1</th>\n",
       "      <th>UF_2</th>\n",
       "      <th>UF_3</th>\n",
       "      <th>UF_4</th>\n",
       "      <th>UF_5</th>\n",
       "      <th>UF_6</th>\n",
       "      <th>UF_7</th>\n",
       "      <th>IDADE</th>\n",
       "      <th>SEXO_1</th>\n",
       "      <th>NIVEL_RELACIONAMENTO_CREDITO01</th>\n",
       "      <th>...</th>\n",
       "      <th>CEP4_6</th>\n",
       "      <th>CEP4_7</th>\n",
       "      <th>CEP4_8</th>\n",
       "      <th>CEP4_9</th>\n",
       "      <th>CEP4_10</th>\n",
       "      <th>CEP4_11</th>\n",
       "      <th>CEP4_12</th>\n",
       "      <th>CEP4_13</th>\n",
       "      <th>CEP4_14</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.357559</td>\n",
       "      <td>1</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.382661</td>\n",
       "      <td>1</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.753467</td>\n",
       "      <td>1</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.848142</td>\n",
       "      <td>1</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.650444</td>\n",
       "      <td>1</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 244 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   UF_1  UF_2  UF_3  UF_4  UF_5  UF_6  UF_7     IDADE  SEXO_1  \\\n",
       "0     1     1     0     0     0     0     1  0.357559       1   \n",
       "1     0     1     1     0     1     0     0  0.382661       1   \n",
       "2     1     1     1     0     0     0     0  0.753467       1   \n",
       "3     1     1     0     0     1     0     0  0.848142       1   \n",
       "4     1     1     0     1     0     0     0  0.650444       1   \n",
       "\n",
       "   NIVEL_RELACIONAMENTO_CREDITO01 ...  CEP4_6  CEP4_7  CEP4_8  CEP4_9  \\\n",
       "0                        0.777778 ...       1       0       1       0   \n",
       "1                        0.111111 ...       0       1       0       1   \n",
       "2                        0.111111 ...       1       1       0       0   \n",
       "3                        0.111111 ...       1       1       0       1   \n",
       "4                        0.111111 ...       0       1       0       1   \n",
       "\n",
       "   CEP4_10  CEP4_11  CEP4_12  CEP4_13  CEP4_14  y  \n",
       "0        1        1        0        0        1  1  \n",
       "1        1        1        0        0        0  1  \n",
       "2        1        0        0        1        0  1  \n",
       "3        0        0        0        1        0  1  \n",
       "4        1        1        0        0        1  1  \n",
       "\n",
       "[5 rows x 244 columns]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#IMPORTANTE PARA USAR A ENTRADA DO KAGGLE\n",
    "data_set = data_set.drop(columns=['Unnamed: 0', 'INDEX'])\n",
    "data_set.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy import stats\n",
    "\n",
    "def get_metrics(y, pred, probas):\n",
    "    accuracy = accuracy_score(y.values, pred)\n",
    "    recall = recall_score(y.values, pred)\n",
    "    precision = precision_score(y.values, pred)\n",
    "    f1 = f1_score(y.values, pred)\n",
    "    aupr = average_precision_score(y.values, pred)\n",
    "    fpr, tpr, thresholds = roc_curve(y.values, probas, pos_label=1)\n",
    "    ks = stats.ks_2samp(y.iloc[:,0].values, probas)\n",
    "    return (accuracy, recall, precision, f1, aupr, fpr, tpr, thresholds, ks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "# encontra as variavies categoricas\n",
    "for col in data_set.columns:\n",
    "    if ((data_set[col].quantile() == 1.0) | (data_set[col].quantile() == 0.0)):\n",
    "        data_set[col] = data_set[col].astype('category')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cross_validation import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data_set.loc[:, data_set.columns !='y']\n",
    "y = data_set.iloc[:, data_set.columns == 'y']\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=1/4, random_state=42, stratify=y)\n",
    "\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=1/3, random_state=42, stratify=y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib.ticker import FormatStrFormatter\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "ax.yaxis.set_major_formatter(FormatStrFormatter('%.2f'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingClassifier as gb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "gb_gs = gb(n_estimators=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "gb_grid_params = {'learning_rate': [0.1,0.05,0.02,0.01],\n",
    "                 'max_depth': [4,6,8],\n",
    "                  #'min_sample_leaf': [20,50,100,150],\n",
    "                 }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.05, loss='deviance', max_depth=6,\n",
       "              max_features=None, max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=1, min_samples_split=2,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=200,\n",
       "              presort='auto', random_state=None, subsample=1.0, verbose=0,\n",
       "              warm_start=False)"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gb_clf = gb(n_estimators=200,learning_rate= 0.05, max_depth = 6)\n",
    "gb_clf.fit(X_train.values, y_train.values.ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean accuracy score: 0.655\n"
     ]
    }
   ],
   "source": [
    "predicted = gb_clf.predict(X_test.values)\n",
    "accuracy = accuracy_score(y_test.values, predicted)\n",
    "print(\"Mean accuracy score: {:.3}\".format(accuracy))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>34.565765</td>\n",
       "      <td>15.434980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>19.027875</td>\n",
       "      <td>30.971379</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           0          1\n",
       "0  34.565765  15.434980\n",
       "1  19.027875  30.971379"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "cm = pd.DataFrame(confusion_matrix(y_test, predicted))\n",
    "tot = cm.sum().sum()\n",
    "cm = cm.applymap(lambda x: x*100/tot)\n",
    "fig = plt.figure(figsize=(10,7))\n",
    "heatmap = sns.heatmap(cm, annot=True)\n",
    "plt.ylabel('True label(%)')\n",
    "plt.xlabel('Predicted label(%)')\n",
    "cm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_clf = gb_clf.predict_proba(X_test)[:, 1]\n",
    "fpr, tpr, thresholds = roc_curve(y_test.values, y_pred_clf, pos_label=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "from sklearn.metrics import auc\n",
    "\n",
    "roc_auc = auc(fpr, tpr)\n",
    "\n",
    "plt.figure(figsize=(10,10))\n",
    "plt.plot(fpr, tpr, color='darkorange', lw=1, label='ROC curve (area = %0.2f)' % roc_auc)\n",
    "#plt.plot(tpr, fpr, color='darkblue', lw=1, label='ROC curve (area = %0.2f)' % roc_auc)\n",
    "plt.plot([0, 1], [0, 1], color='navy', linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver Operating Oharacteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy score: 0.655\n",
      "Recall score: 0.619\n",
      "Precision score: 0.667\n",
      "f1 score: 0.643\n",
      "Mean precison score: 0.604\n"
     ]
    }
   ],
   "source": [
    "accuracy, recall, precision, f1, aupr, fpr, tpr, thresholds, ks = get_metrics(y_test, predicted, y_pred_clf)\n",
    "print(\"Accuracy score: {:.3}\".format(accuracy))\n",
    "print(\"Recall score: {:.3}\".format(recall))\n",
    "print(\"Precision score: {:.3}\".format(precision))\n",
    "print(\"f1 score: {:.3}\".format(f1))\n",
    "print(\"Mean precison score: {:.3}\".format(aupr))\n",
    "#print(\"KS: {:.3}\".format(ks[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\eduar\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:1: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n",
      "C:\\Users\\eduar\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:2: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "y_train_pred = gb_clf.predict(X_train.as_matrix()).ravel()\n",
    "y_test_pred = gb_clf.predict(X_test.as_matrix()).ravel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\eduar\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:1: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "y_test_pred_prob = gb_clf.predict_proba(X_test.as_matrix())[:, 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "33525 33524\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Ks_2sampResult(statistic=0.17821782178217827, pvalue=0.07148265768062076)"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import math\n",
    "from scipy import stats\n",
    "\n",
    "position = 0\n",
    "class_0 = 0\n",
    "class_1 = 0\n",
    "counting_class_0 = [0] * 101\n",
    "counting_class_1 = [0] * 101\n",
    "grouped_counting_class_0 = [0] * 101\n",
    "grouped_counting_class_1 = [0] * 101\n",
    "percentage_counting_class_0 = [0.0] * 101\n",
    "percentage_counting_class_1 = [0.0] * 101\n",
    "\n",
    "for x in y_test['y']:\n",
    "    score = math.floor(y_test_pred_prob[position]*100)\n",
    "    if x == 0:\n",
    "        class_0 = class_0 + 1\n",
    "        counting_class_0[score] = counting_class_0[score] + 1\n",
    "    else:\n",
    "        class_1 = class_1 + 1\n",
    "        counting_class_1[score]= counting_class_1[score] + 1\n",
    "    position = position + 1\n",
    "\n",
    "print(class_0, class_1)\n",
    "\n",
    "last_value_class_0 = 0\n",
    "last_value_class_1 = 0\n",
    "\n",
    "for x in range(0, 101):\n",
    "    last_value_class_0 += counting_class_0[x]\n",
    "    grouped_counting_class_0[x] = last_value_class_0\n",
    "    percentage_counting_class_0[x] = last_value_class_0/class_0\n",
    "    \n",
    "    last_value_class_1 += counting_class_1[x]\n",
    "    grouped_counting_class_1[x] = last_value_class_1\n",
    "    percentage_counting_class_1[x] = last_value_class_1/class_1\n",
    "\n",
    "#stats.ks_2samp(percentage_counting_class_0, percentage_counting_class_1)\n",
    "stats.ks_2samp(grouped_counting_class_0, grouped_counting_class_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "kaggle_test_data = pd.read_csv(\"data/real_test_set.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "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>UF_1</th>\n",
       "      <th>UF_2</th>\n",
       "      <th>UF_3</th>\n",
       "      <th>UF_4</th>\n",
       "      <th>UF_5</th>\n",
       "      <th>UF_6</th>\n",
       "      <th>UF_7</th>\n",
       "      <th>IDADE</th>\n",
       "      <th>SEXO_1</th>\n",
       "      <th>NIVEL_RELACIONAMENTO_CREDITO01</th>\n",
       "      <th>...</th>\n",
       "      <th>CEP4_5</th>\n",
       "      <th>CEP4_6</th>\n",
       "      <th>CEP4_7</th>\n",
       "      <th>CEP4_8</th>\n",
       "      <th>CEP4_9</th>\n",
       "      <th>CEP4_10</th>\n",
       "      <th>CEP4_11</th>\n",
       "      <th>CEP4_12</th>\n",
       "      <th>CEP4_13</th>\n",
       "      <th>CEP4_14</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.898745</td>\n",
       "      <td>1</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.847404</td>\n",
       "      <td>1</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.016513</td>\n",
       "      <td>0</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.688950</td>\n",
       "      <td>0</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.230114</td>\n",
       "      <td>1</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</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",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 243 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   UF_1  UF_2  UF_3  UF_4  UF_5  UF_6  UF_7     IDADE  SEXO_1  \\\n",
       "0     1     1     1     0     0     0     0  0.898745       1   \n",
       "1     1     1     0     1     0     0     0  0.847404       1   \n",
       "2     1     1     0     0     0     0     1  0.016513       0   \n",
       "3     1     1     0     1     0     0     0  0.688950       0   \n",
       "4     0     1     1     0     0     1     0  0.230114       1   \n",
       "\n",
       "   NIVEL_RELACIONAMENTO_CREDITO01   ...     CEP4_5  CEP4_6  CEP4_7  CEP4_8  \\\n",
       "0                        0.111111   ...          1       1       0       1   \n",
       "1                        0.111111   ...          1       0       0       0   \n",
       "2                        0.111111   ...          1       1       0       1   \n",
       "3                        0.111111   ...          1       1       0       0   \n",
       "4                        0.111111   ...          0       0       0       1   \n",
       "\n",
       "   CEP4_9  CEP4_10  CEP4_11  CEP4_12  CEP4_13  CEP4_14  \n",
       "0       1        1        0        0        0        0  \n",
       "1       0        1        1        1        0        0  \n",
       "2       0        0        1        0        0        1  \n",
       "3       0        0        0        1        0        1  \n",
       "4       0        0        1        0        1        0  \n",
       "\n",
       "[5 rows x 243 columns]"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "kaggle_test_data.shape\n",
    "features_kaggle = kaggle_test_data.drop([\"id\"], axis=1)\n",
    "features_kaggle.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "gb_pred_test_class = gb_clf.predict(features_kaggle)\n",
    "gb_pred_test_scores = gb_clf.predict_proba(features_kaggle)[:, 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame(data=gb_pred_test_class)\n",
    "df.to_csv('test.csv', mode='a', index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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
}