project-competition/prediction.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# necessary libraries for prediciton\n",
    "import utils\n",
    "import os\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import pprint\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn import metrics\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "\n",
    "from imblearn.pipeline import Pipeline\n",
    "from imblearn.over_sampling import SMOTE\n",
    "from imblearn.under_sampling import NeighbourhoodCleaningRule\n",
    "from imblearn.combine import SMOTEENN\n",
    "\n",
    "from sklearn.tree import export_graphviz\n",
    "from subprocess import call\n",
    "from IPython.display import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.set_option('display.max_columns', 200)\n",
    "os.makedirs('images', exist_ok=True)\n",
    "SEED = 42"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Useful Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def pretty_matrix(matrix, row_label, col_label):\n",
    "    \"\"\"Pretty print of the given matrix \"\"\"\n",
    "\n",
    "    # Restraining labels that are too big\n",
    "    row_label = [el[:10] + '..' if len(el) > 10 else el\n",
    "                for el in row_label]\n",
    "    col_label = [el[:10] + '..' if len(el) > 10 else el\n",
    "                for el in col_label]\n",
    "\n",
    "    # Stringfying everything & Joining top label\n",
    "    s_matrix = [list([\" \"] + (col_label))] + \\\n",
    "               [[row_label[row_idx]] + \\\n",
    "                [str(e) for e in row] for row_idx, row in enumerate(matrix)]\n",
    "\n",
    "    # Length of each matrix column\n",
    "    len_s = [max(map(len, col)) for col in zip(*s_matrix)]\n",
    "\n",
    "    # Cell formatation\n",
    "    formatation = '\\t'.join('{{:{}}}'.format(x) for x in len_s)\n",
    "\n",
    "    # Apply cell formation to each matrix element\n",
    "    pretty_mat = [formatation.format(*row) for row in s_matrix]\n",
    "\n",
    "    # Print Pretty Matrix\n",
    "    print('\\n'.join(pretty_mat))\n",
    "\n",
    "\n",
    "def display_confusion_matrix(values):\n",
    "    '''Display the given array as a confusion matrix'''\n",
    "    pretty_matrix([values[0:2], values[2:4]],\n",
    "                  ['Actual NO', 'Actual YES'],\n",
    "                  ['Predic NO', 'Predic YES'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_roc_auc(fpr, tpr, roc_auc):\n",
    "    '''Plot the ROC-AUC curve'''\n",
    "    fig, ax = plt.subplots()\n",
    "    plt.title('Receiver Operating Characteristic (ROC)')\n",
    "    plt.plot(fpr, tpr, 'b', label = 'AUC = %0.2f' % roc_auc)\n",
    "    plt.legend(loc = 'lower right')\n",
    "    plt.plot([0, 1], [0, 1],'w--')\n",
    "    plt.xlim([-0.02, 1.02])\n",
    "    plt.ylim([0, 1])\n",
    "    ax.fill_between(fpr, 0, tpr)\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def visualize_tree(classifier, name):\n",
    "    '''Visualize the given classifier,\n",
    "    by creating a .png with a given name.'''\n",
    "    dot_name = name + '.dot'\n",
    "    png_name = name + '.png'\n",
    "\n",
    "    export_graphviz(classifier,\n",
    "                    out_file=dot_name, \n",
    "                    feature_names = list(X.columns),\n",
    "                    class_names = 'status',\n",
    "                    rounded = True, proportion = False, \n",
    "                    precision = 2, filled = True)\n",
    "\n",
    "    # Convert to png using system command (requires Graphviz installed)\n",
    "    call(['dot', '-Tpng', dot_name, '-o', png_name, '-Gdpi=600'])\n",
    "    call(['rm', dot_name])\n",
    "\n",
    "    # Display \n",
    "    Image('tree.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Necessary functions for the prediction process"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def normalize_columns(df, columns):\n",
    "    '''Normalize the given columns for range between [0, 1]'''\n",
    "    for col in columns:\n",
    "        col_min = df[col].min()\n",
    "        col_max =  df[col].max()\n",
    "        \n",
    "        df[col] = (df[col] - col_min)/\\\n",
    "                    (col_max - col_min)\n",
    "        \n",
    "def normalize_df(df):\n",
    "    '''Normalize all columns of the given df'''\n",
    "    normalize_columns(df, df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def apply_PCA(df, variance_val=0.95, debug=True):\n",
    "    '''Apply the PCA algorithm to given dataframe,\n",
    "    using the given variance val to trim the df'''\n",
    "    # Necessary to normalize all data to use PCA\n",
    "    scaler=StandardScaler()\n",
    "    X_scaled=scaler.fit_transform(df)\n",
    "\n",
    "    # PCA - keep, by default mode, 90% variance\n",
    "    pca = PCA(variance_val)    \n",
    "    pca.fit(X_scaled)\n",
    "    X_pca = pca.transform(X_scaled)\n",
    "\n",
    "    if debug:\n",
    "        ex_variance=np.var(X_pca,axis=0)\n",
    "        ex_variance_ratio = ex_variance/np.sum(ex_variance)\n",
    "        print(' > Impact in total variance of each generated feature by PCA:')\n",
    "        print(ex_variance_ratio)\n",
    "\n",
    "    principal_df = pd.DataFrame(data = X_pca, index = df.reset_index()['loan_id'])\n",
    "    \n",
    "    return (principal_df, pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def apply_sampling(algorithm, oversample, undersample):\n",
    "    '''Apply sampling according to the control flags'''\n",
    "    # Applying sampling techniques\n",
    "    pipeline = Pipeline([\n",
    "        ('classification', algorithm)\n",
    "    ])\n",
    "    \n",
    "    if oversample:\n",
    "        if undersample:\n",
    "            pipeline = Pipeline([\n",
    "                ('sampling', SMOTEENN(random_state = SEED)),\n",
    "                ('classification', algorithm)\n",
    "            ])\n",
    "        else:\n",
    "            pipeline = Pipeline([\n",
    "                ('sampling', SMOTE(random_state = SEED)),\n",
    "                ('classification', algorithm)\n",
    "            ])\n",
    "    elif undersample:\n",
    "        pipeline = Pipeline([\n",
    "        ('sampling', NeighbourhoodCleaningRule(random_state = SEED)),\n",
    "        ('classification', algorithm)\n",
    "    ])\n",
    "        \n",
    "    return pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def auc_scorer(y_true, y_pred):\n",
    "    '''Scorer of Area Under Curve value'''\n",
    "    fpr, tpr, _ = metrics.roc_curve(y_true, y_pred)\n",
    "    return metrics.auc(fpr, tpr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction Algorithms\n",
    "\n",
    "* Logistic Regression\n",
    "* Decision Tree\n",
    "* Random Forest\n",
    "* Gradient Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_LR():\n",
    "    '''Create a Logistic Regression model'''\n",
    "    return LogisticRegression(random_state=SEED)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_DT():\n",
    "    '''Create a new Decision Tree'''\n",
    "    return DecisionTreeClassifier(random_state=SEED)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_RF():\n",
    "    '''Create a new Ranfom Forest model'''\n",
    "    return RandomForestClassifier(random_state=SEED)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_GB():\n",
    "    '''Create a new Gradient Boosting model'''\n",
    "    return GradientBoostingClassifier(random_state=SEED)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Useful Macros\n",
    "K_FOLD_NUM_SPLITS = 5\n",
    "USE_PCA = False\n",
    "UNDERSAMPLE = False\n",
    "OVERSAMPLE = True\n",
    "\n",
    "# Pretty printer\n",
    "pp = pprint.PrettyPrinter(indent=4)"
   ]
  },
  {
   "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>payments</th>\n",
       "      <th>balance_mean</th>\n",
       "      <th>balance_min</th>\n",
       "      <th>reached_negative_balance</th>\n",
       "      <th>unknown_sum</th>\n",
       "      <th>sanctions_mean</th>\n",
       "      <th>sanctions_sum</th>\n",
       "      <th>sanctions_count</th>\n",
       "      <th>sanctions_std</th>\n",
       "      <th>CC_max</th>\n",
       "      <th>CC_std</th>\n",
       "      <th>CAB_mean</th>\n",
       "      <th>CAB_sum</th>\n",
       "      <th>CAB_max</th>\n",
       "      <th>CAB_std</th>\n",
       "      <th>CAB_min</th>\n",
       "      <th>IC_mean</th>\n",
       "      <th>IC_min</th>\n",
       "      <th>RAB</th>\n",
       "      <th>RAB_mean</th>\n",
       "      <th>RAB_sum</th>\n",
       "      <th>RAB_max</th>\n",
       "      <th>RAB_min</th>\n",
       "      <th>mean_trans_profit</th>\n",
       "      <th>credit_ratio</th>\n",
       "      <th>withdrawal_ratio</th>\n",
       "      <th>ratio_CC</th>\n",
       "      <th>ratio_RAB</th>\n",
       "      <th>owner_count</th>\n",
       "      <th>disponent_count</th>\n",
       "      <th>salary_from_month_records</th>\n",
       "      <th>withdrawal_to_credit_month_ratio</th>\n",
       "      <th>actual_month_income</th>\n",
       "      <th>actual_income_to_payments_ratio</th>\n",
       "      <th>log_sanctions_std</th>\n",
       "      <th>log_max_value_in_account_to_loan_ratio</th>\n",
       "      <th>status</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loan_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>5314</td>\n",
       "      <td>8033</td>\n",
       "      <td>12250.000000</td>\n",
       "      <td>1100.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>20100.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9900.0</td>\n",
       "      <td>3774.806838</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5025.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5025.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5025.000000</td>\n",
       "      <td>0.625545</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.113178</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5316</td>\n",
       "      <td>4610</td>\n",
       "      <td>52083.859459</td>\n",
       "      <td>700.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>62612.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36574.0</td>\n",
       "      <td>13044.711945</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>168.540000</td>\n",
       "      <td>2.9</td>\n",
       "      <td>8.0</td>\n",
       "      <td>-2682.375000</td>\n",
       "      <td>-21459.0</td>\n",
       "      <td>-956.0</td>\n",
       "      <td>-4167.0</td>\n",
       "      <td>4638.918824</td>\n",
       "      <td>0.459459</td>\n",
       "      <td>0.540541</td>\n",
       "      <td>0.324324</td>\n",
       "      <td>0.216216</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4638.918824</td>\n",
       "      <td>-0.656965</td>\n",
       "      <td>4291.668824</td>\n",
       "      <td>0.930948</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493154</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6863</td>\n",
       "      <td>2118</td>\n",
       "      <td>30060.954167</td>\n",
       "      <td>800.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>27534.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3700.0</td>\n",
       "      <td>2050.609665</td>\n",
       "      <td>13981.000000</td>\n",
       "      <td>69905.0</td>\n",
       "      <td>19065.0</td>\n",
       "      <td>2842.042399</td>\n",
       "      <td>12710.0</td>\n",
       "      <td>92.625000</td>\n",
       "      <td>48.6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>-2824.333333</td>\n",
       "      <td>-8473.0</td>\n",
       "      <td>-66.0</td>\n",
       "      <td>-7936.0</td>\n",
       "      <td>-1087.266667</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.375000</td>\n",
       "      <td>0.083333</td>\n",
       "      <td>0.125000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1087.266667</td>\n",
       "      <td>-1.217031</td>\n",
       "      <td>-1748.600000</td>\n",
       "      <td>-0.825590</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.263553</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5325</td>\n",
       "      <td>2939</td>\n",
       "      <td>41297.480000</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>47508.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3900.0</td>\n",
       "      <td>2050.609665</td>\n",
       "      <td>19101.333333</td>\n",
       "      <td>114608.0</td>\n",
       "      <td>26448.0</td>\n",
       "      <td>3599.116929</td>\n",
       "      <td>17632.0</td>\n",
       "      <td>160.360000</td>\n",
       "      <td>132.8</td>\n",
       "      <td>2.0</td>\n",
       "      <td>-6994.000000</td>\n",
       "      <td>-13988.0</td>\n",
       "      <td>-6994.0</td>\n",
       "      <td>-6994.0</td>\n",
       "      <td>2086.500000</td>\n",
       "      <td>0.520000</td>\n",
       "      <td>0.480000</td>\n",
       "      <td>0.080000</td>\n",
       "      <td>0.080000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2086.500000</td>\n",
       "      <td>-0.774545</td>\n",
       "      <td>1503.666667</td>\n",
       "      <td>0.511625</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.427978</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7240</td>\n",
       "      <td>4579</td>\n",
       "      <td>57188.211111</td>\n",
       "      <td>600.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>40788.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>600.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>45764.333333</td>\n",
       "      <td>274586.0</td>\n",
       "      <td>63366.0</td>\n",
       "      <td>8623.020391</td>\n",
       "      <td>42244.0</td>\n",
       "      <td>190.183333</td>\n",
       "      <td>77.1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>-6584.666667</td>\n",
       "      <td>-19754.0</td>\n",
       "      <td>-756.0</td>\n",
       "      <td>-9499.0</td>\n",
       "      <td>4454.930769</td>\n",
       "      <td>0.481481</td>\n",
       "      <td>0.518519</td>\n",
       "      <td>0.037037</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4454.930769</td>\n",
       "      <td>-0.790415</td>\n",
       "      <td>4391.930769</td>\n",
       "      <td>0.959146</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.306472</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6818</td>\n",
       "      <td>3242</td>\n",
       "      <td>44197.509884</td>\n",
       "      <td>200.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>101122.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>26388.0</td>\n",
       "      <td>6517.394689</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>145.472727</td>\n",
       "      <td>23.7</td>\n",
       "      <td>36.0</td>\n",
       "      <td>-1709.000000</td>\n",
       "      <td>-61524.0</td>\n",
       "      <td>-831.0</td>\n",
       "      <td>-2587.0</td>\n",
       "      <td>2618.124034</td>\n",
       "      <td>0.406977</td>\n",
       "      <td>0.593023</td>\n",
       "      <td>0.151163</td>\n",
       "      <td>0.209302</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2618.124034</td>\n",
       "      <td>-0.594000</td>\n",
       "      <td>2402.540700</td>\n",
       "      <td>0.741067</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.333010</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5625</td>\n",
       "      <td>3703</td>\n",
       "      <td>55230.444068</td>\n",
       "      <td>800.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>56990.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>44352.0</td>\n",
       "      <td>13832.527973</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>223.050000</td>\n",
       "      <td>103.2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3220.376498</td>\n",
       "      <td>0.474576</td>\n",
       "      <td>0.525424</td>\n",
       "      <td>0.271186</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3220.376498</td>\n",
       "      <td>-0.759988</td>\n",
       "      <td>3220.376498</td>\n",
       "      <td>0.869667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.405061</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6805</td>\n",
       "      <td>938</td>\n",
       "      <td>41994.907692</td>\n",
       "      <td>800.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>51200.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>31636.5</td>\n",
       "      <td>9517.351233</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>114.216667</td>\n",
       "      <td>41.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>-3387.500000</td>\n",
       "      <td>-20325.0</td>\n",
       "      <td>-2056.0</td>\n",
       "      <td>-4719.0</td>\n",
       "      <td>1570.498421</td>\n",
       "      <td>0.512821</td>\n",
       "      <td>0.487179</td>\n",
       "      <td>0.205128</td>\n",
       "      <td>0.153846</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1570.498421</td>\n",
       "      <td>-0.816207</td>\n",
       "      <td>1177.248421</td>\n",
       "      <td>1.255062</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.843812</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7233</td>\n",
       "      <td>3217</td>\n",
       "      <td>56646.516129</td>\n",
       "      <td>1100.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>38211.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>49887.0</td>\n",
       "      <td>14644.196076</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>206.068421</td>\n",
       "      <td>51.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5562.846548</td>\n",
       "      <td>0.411290</td>\n",
       "      <td>0.588710</td>\n",
       "      <td>0.258065</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5562.846548</td>\n",
       "      <td>-0.663978</td>\n",
       "      <td>5562.846548</td>\n",
       "      <td>1.729203</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.664437</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7308</td>\n",
       "      <td>5392</td>\n",
       "      <td>30869.781308</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>73977.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24212.0</td>\n",
       "      <td>6318.812124</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>111.846154</td>\n",
       "      <td>50.1</td>\n",
       "      <td>40.0</td>\n",
       "      <td>-1323.750000</td>\n",
       "      <td>-52950.0</td>\n",
       "      <td>-56.0</td>\n",
       "      <td>-4780.0</td>\n",
       "      <td>5845.843236</td>\n",
       "      <td>0.271028</td>\n",
       "      <td>0.728972</td>\n",
       "      <td>0.149533</td>\n",
       "      <td>0.373832</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5845.843236</td>\n",
       "      <td>-0.336586</td>\n",
       "      <td>5447.509903</td>\n",
       "      <td>1.010295</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.296012</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>328 rows × 37 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         payments  balance_mean  balance_min  reached_negative_balance  \\\n",
       "loan_id                                                                  \n",
       "5314         8033  12250.000000       1100.0                       1.0   \n",
       "5316         4610  52083.859459        700.0                       1.0   \n",
       "6863         2118  30060.954167        800.0                       1.0   \n",
       "5325         2939  41297.480000       1000.0                       1.0   \n",
       "7240         4579  57188.211111        600.0                       1.0   \n",
       "...           ...           ...          ...                       ...   \n",
       "6818         3242  44197.509884        200.0                       1.0   \n",
       "5625         3703  55230.444068        800.0                       1.0   \n",
       "6805          938  41994.907692        800.0                       1.0   \n",
       "7233         3217  56646.516129       1100.0                       1.0   \n",
       "7308         5392  30869.781308       1000.0                       1.0   \n",
       "\n",
       "         unknown_sum  sanctions_mean  sanctions_sum  sanctions_count  \\\n",
       "loan_id                                                                \n",
       "5314         20100.0             0.0            0.0              0.0   \n",
       "5316         62612.0             0.0            0.0              0.0   \n",
       "6863         27534.0             0.0            0.0              0.0   \n",
       "5325         47508.0             0.0            0.0              0.0   \n",
       "7240         40788.0             0.0            0.0              0.0   \n",
       "...              ...             ...            ...              ...   \n",
       "6818        101122.0             0.0            0.0              0.0   \n",
       "5625         56990.0             0.0            0.0              0.0   \n",
       "6805         51200.0             0.0            0.0              0.0   \n",
       "7233         38211.0             0.0            0.0              0.0   \n",
       "7308         73977.0             0.0            0.0              0.0   \n",
       "\n",
       "         sanctions_std   CC_max        CC_std      CAB_mean   CAB_sum  \\\n",
       "loan_id                                                                 \n",
       "5314               0.0   9900.0   3774.806838      0.000000       0.0   \n",
       "5316               0.0  36574.0  13044.711945      0.000000       0.0   \n",
       "6863               0.0   3700.0   2050.609665  13981.000000   69905.0   \n",
       "5325               0.0   3900.0   2050.609665  19101.333333  114608.0   \n",
       "7240               0.0    600.0      0.000000  45764.333333  274586.0   \n",
       "...                ...      ...           ...           ...       ...   \n",
       "6818               0.0  26388.0   6517.394689      0.000000       0.0   \n",
       "5625               0.0  44352.0  13832.527973      0.000000       0.0   \n",
       "6805               0.0  31636.5   9517.351233      0.000000       0.0   \n",
       "7233               0.0  49887.0  14644.196076      0.000000       0.0   \n",
       "7308               0.0  24212.0   6318.812124      0.000000       0.0   \n",
       "\n",
       "         CAB_max      CAB_std  CAB_min     IC_mean  IC_min   RAB     RAB_mean  \\\n",
       "loan_id                                                                         \n",
       "5314         0.0     0.000000      0.0    0.000000     0.0   0.0     0.000000   \n",
       "5316         0.0     0.000000      0.0  168.540000     2.9   8.0 -2682.375000   \n",
       "6863     19065.0  2842.042399  12710.0   92.625000    48.6   3.0 -2824.333333   \n",
       "5325     26448.0  3599.116929  17632.0  160.360000   132.8   2.0 -6994.000000   \n",
       "7240     63366.0  8623.020391  42244.0  190.183333    77.1   3.0 -6584.666667   \n",
       "...          ...          ...      ...         ...     ...   ...          ...   \n",
       "6818         0.0     0.000000      0.0  145.472727    23.7  36.0 -1709.000000   \n",
       "5625         0.0     0.000000      0.0  223.050000   103.2   0.0     0.000000   \n",
       "6805         0.0     0.000000      0.0  114.216667    41.0   6.0 -3387.500000   \n",
       "7233         0.0     0.000000      0.0  206.068421    51.1   0.0     0.000000   \n",
       "7308         0.0     0.000000      0.0  111.846154    50.1  40.0 -1323.750000   \n",
       "\n",
       "         RAB_sum  RAB_max  RAB_min  mean_trans_profit  credit_ratio  \\\n",
       "loan_id                                                               \n",
       "5314         0.0      0.0      0.0        5025.000000      1.000000   \n",
       "5316    -21459.0   -956.0  -4167.0        4638.918824      0.459459   \n",
       "6863     -8473.0    -66.0  -7936.0       -1087.266667      0.625000   \n",
       "5325    -13988.0  -6994.0  -6994.0        2086.500000      0.520000   \n",
       "7240    -19754.0   -756.0  -9499.0        4454.930769      0.481481   \n",
       "...          ...      ...      ...                ...           ...   \n",
       "6818    -61524.0   -831.0  -2587.0        2618.124034      0.406977   \n",
       "5625         0.0      0.0      0.0        3220.376498      0.474576   \n",
       "6805    -20325.0  -2056.0  -4719.0        1570.498421      0.512821   \n",
       "7233         0.0      0.0      0.0        5562.846548      0.411290   \n",
       "7308    -52950.0    -56.0  -4780.0        5845.843236      0.271028   \n",
       "\n",
       "         withdrawal_ratio  ratio_CC  ratio_RAB  owner_count  disponent_count  \\\n",
       "loan_id                                                                        \n",
       "5314             0.000000  1.000000   0.000000            1                1   \n",
       "5316             0.540541  0.324324   0.216216            1                1   \n",
       "6863             0.375000  0.083333   0.125000            1                1   \n",
       "5325             0.480000  0.080000   0.080000            1                1   \n",
       "7240             0.518519  0.037037   0.111111            1                1   \n",
       "...                   ...       ...        ...          ...              ...   \n",
       "6818             0.593023  0.151163   0.209302            2                2   \n",
       "5625             0.525424  0.271186   0.000000            1                1   \n",
       "6805             0.487179  0.205128   0.153846            1                1   \n",
       "7233             0.588710  0.258065   0.000000            1                1   \n",
       "7308             0.728972  0.149533   0.373832            1                1   \n",
       "\n",
       "         salary_from_month_records  withdrawal_to_credit_month_ratio  \\\n",
       "loan_id                                                                \n",
       "5314                   5025.000000                          0.000000   \n",
       "5316                   4638.918824                         -0.656965   \n",
       "6863                  -1087.266667                         -1.217031   \n",
       "5325                   2086.500000                         -0.774545   \n",
       "7240                   4454.930769                         -0.790415   \n",
       "...                            ...                               ...   \n",
       "6818                   2618.124034                         -0.594000   \n",
       "5625                   3220.376498                         -0.759988   \n",
       "6805                   1570.498421                         -0.816207   \n",
       "7233                   5562.846548                         -0.663978   \n",
       "7308                   5845.843236                         -0.336586   \n",
       "\n",
       "         actual_month_income  actual_income_to_payments_ratio  \\\n",
       "loan_id                                                         \n",
       "5314             5025.000000                         0.625545   \n",
       "5316             4291.668824                         0.930948   \n",
       "6863            -1748.600000                        -0.825590   \n",
       "5325             1503.666667                         0.511625   \n",
       "7240             4391.930769                         0.959146   \n",
       "...                      ...                              ...   \n",
       "6818             2402.540700                         0.741067   \n",
       "5625             3220.376498                         0.869667   \n",
       "6805             1177.248421                         1.255062   \n",
       "7233             5562.846548                         1.729203   \n",
       "7308             5447.509903                         1.010295   \n",
       "\n",
       "         log_sanctions_std  log_max_value_in_account_to_loan_ratio  status  \n",
       "loan_id                                                                     \n",
       "5314                   0.0                                0.113178      -1  \n",
       "5316                   0.0                                0.493154       1  \n",
       "6863                   0.0                                0.263553       1  \n",
       "5325                   0.0                                0.427978       1  \n",
       "7240                   0.0                                0.306472       1  \n",
       "...                    ...                                     ...     ...  \n",
       "6818                   0.0                                0.333010       1  \n",
       "5625                   0.0                                0.405061      -1  \n",
       "6805                   0.0                                0.843812       1  \n",
       "7233                   0.0                                0.664437       1  \n",
       "7308                   0.0                                0.296012       1  \n",
       "\n",
       "[328 rows x 37 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " > Dataset after normalization\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>payments</th>\n",
       "      <th>balance_mean</th>\n",
       "      <th>balance_min</th>\n",
       "      <th>reached_negative_balance</th>\n",
       "      <th>unknown_sum</th>\n",
       "      <th>sanctions_mean</th>\n",
       "      <th>sanctions_sum</th>\n",
       "      <th>sanctions_count</th>\n",
       "      <th>sanctions_std</th>\n",
       "      <th>CC_max</th>\n",
       "      <th>CC_std</th>\n",
       "      <th>CAB_mean</th>\n",
       "      <th>CAB_sum</th>\n",
       "      <th>CAB_max</th>\n",
       "      <th>CAB_std</th>\n",
       "      <th>CAB_min</th>\n",
       "      <th>IC_mean</th>\n",
       "      <th>IC_min</th>\n",
       "      <th>RAB</th>\n",
       "      <th>RAB_mean</th>\n",
       "      <th>RAB_sum</th>\n",
       "      <th>RAB_max</th>\n",
       "      <th>RAB_min</th>\n",
       "      <th>mean_trans_profit</th>\n",
       "      <th>credit_ratio</th>\n",
       "      <th>withdrawal_ratio</th>\n",
       "      <th>ratio_CC</th>\n",
       "      <th>ratio_RAB</th>\n",
       "      <th>owner_count</th>\n",
       "      <th>disponent_count</th>\n",
       "      <th>salary_from_month_records</th>\n",
       "      <th>withdrawal_to_credit_month_ratio</th>\n",
       "      <th>actual_month_income</th>\n",
       "      <th>actual_income_to_payments_ratio</th>\n",
       "      <th>log_sanctions_std</th>\n",
       "      <th>log_max_value_in_account_to_loan_ratio</th>\n",
       "      <th>status</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loan_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>5314</td>\n",
       "      <td>0.823266</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.273848</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.094611</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.194857</td>\n",
       "      <td>0.145173</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.624874</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.624874</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.641549</td>\n",
       "      <td>0.299329</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.048220</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5316</td>\n",
       "      <td>0.457951</td>\n",
       "      <td>0.622572</td>\n",
       "      <td>0.249638</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.250351</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.730695</td>\n",
       "      <td>0.501677</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.503220</td>\n",
       "      <td>0.013339</td>\n",
       "      <td>0.117647</td>\n",
       "      <td>0.817811</td>\n",
       "      <td>0.918812</td>\n",
       "      <td>0.935068</td>\n",
       "      <td>0.716973</td>\n",
       "      <td>0.610923</td>\n",
       "      <td>0.286947</td>\n",
       "      <td>0.713053</td>\n",
       "      <td>0.319696</td>\n",
       "      <td>0.578378</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.610923</td>\n",
       "      <td>0.754242</td>\n",
       "      <td>0.614343</td>\n",
       "      <td>0.314285</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.210111</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6863</td>\n",
       "      <td>0.191996</td>\n",
       "      <td>0.278371</td>\n",
       "      <td>0.255691</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.121845</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.070309</td>\n",
       "      <td>0.078863</td>\n",
       "      <td>0.259110</td>\n",
       "      <td>0.065120</td>\n",
       "      <td>0.255830</td>\n",
       "      <td>0.277983</td>\n",
       "      <td>0.255832</td>\n",
       "      <td>0.276556</td>\n",
       "      <td>0.223551</td>\n",
       "      <td>0.044118</td>\n",
       "      <td>0.808169</td>\n",
       "      <td>0.967943</td>\n",
       "      <td>0.995517</td>\n",
       "      <td>0.460979</td>\n",
       "      <td>0.404010</td>\n",
       "      <td>0.505319</td>\n",
       "      <td>0.494681</td>\n",
       "      <td>0.077055</td>\n",
       "      <td>0.334375</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.404010</td>\n",
       "      <td>0.544732</td>\n",
       "      <td>0.390257</td>\n",
       "      <td>0.228264</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.112288</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5325</td>\n",
       "      <td>0.279616</td>\n",
       "      <td>0.453990</td>\n",
       "      <td>0.267796</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.195018</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.074327</td>\n",
       "      <td>0.078863</td>\n",
       "      <td>0.354005</td>\n",
       "      <td>0.106764</td>\n",
       "      <td>0.354902</td>\n",
       "      <td>0.352033</td>\n",
       "      <td>0.354904</td>\n",
       "      <td>0.478797</td>\n",
       "      <td>0.610856</td>\n",
       "      <td>0.029412</td>\n",
       "      <td>0.524961</td>\n",
       "      <td>0.947078</td>\n",
       "      <td>0.524961</td>\n",
       "      <td>0.524961</td>\n",
       "      <td>0.518693</td>\n",
       "      <td>0.366809</td>\n",
       "      <td>0.633191</td>\n",
       "      <td>0.073699</td>\n",
       "      <td>0.214000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.518693</td>\n",
       "      <td>0.710258</td>\n",
       "      <td>0.510912</td>\n",
       "      <td>0.293750</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.182342</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7240</td>\n",
       "      <td>0.454642</td>\n",
       "      <td>0.702349</td>\n",
       "      <td>0.243586</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.170400</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.008035</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.848151</td>\n",
       "      <td>0.255792</td>\n",
       "      <td>0.850299</td>\n",
       "      <td>0.843426</td>\n",
       "      <td>0.850305</td>\n",
       "      <td>0.567842</td>\n",
       "      <td>0.354646</td>\n",
       "      <td>0.044118</td>\n",
       "      <td>0.552763</td>\n",
       "      <td>0.925263</td>\n",
       "      <td>0.948652</td>\n",
       "      <td>0.354819</td>\n",
       "      <td>0.604275</td>\n",
       "      <td>0.315997</td>\n",
       "      <td>0.684003</td>\n",
       "      <td>0.030441</td>\n",
       "      <td>0.297222</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.604275</td>\n",
       "      <td>0.704321</td>\n",
       "      <td>0.618063</td>\n",
       "      <td>0.315666</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.130574</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6818</td>\n",
       "      <td>0.311953</td>\n",
       "      <td>0.499315</td>\n",
       "      <td>0.219376</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.391429</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.526075</td>\n",
       "      <td>0.250648</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.434347</td>\n",
       "      <td>0.109016</td>\n",
       "      <td>0.529412</td>\n",
       "      <td>0.883923</td>\n",
       "      <td>0.767230</td>\n",
       "      <td>0.943558</td>\n",
       "      <td>0.824289</td>\n",
       "      <td>0.537903</td>\n",
       "      <td>0.217714</td>\n",
       "      <td>0.782286</td>\n",
       "      <td>0.145349</td>\n",
       "      <td>0.559884</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.537903</td>\n",
       "      <td>0.777796</td>\n",
       "      <td>0.544259</td>\n",
       "      <td>0.304986</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.141881</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5625</td>\n",
       "      <td>0.361153</td>\n",
       "      <td>0.671751</td>\n",
       "      <td>0.255691</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.229755</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.886943</td>\n",
       "      <td>0.531975</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.665974</td>\n",
       "      <td>0.474701</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.559665</td>\n",
       "      <td>0.306888</td>\n",
       "      <td>0.693112</td>\n",
       "      <td>0.266195</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.559665</td>\n",
       "      <td>0.715703</td>\n",
       "      <td>0.574599</td>\n",
       "      <td>0.311284</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.172578</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6805</td>\n",
       "      <td>0.066062</td>\n",
       "      <td>0.464890</td>\n",
       "      <td>0.255691</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.208544</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.631509</td>\n",
       "      <td>0.366021</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.341024</td>\n",
       "      <td>0.188592</td>\n",
       "      <td>0.088235</td>\n",
       "      <td>0.769918</td>\n",
       "      <td>0.923102</td>\n",
       "      <td>0.860355</td>\n",
       "      <td>0.679481</td>\n",
       "      <td>0.500047</td>\n",
       "      <td>0.357338</td>\n",
       "      <td>0.642662</td>\n",
       "      <td>0.199684</td>\n",
       "      <td>0.411538</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.500047</td>\n",
       "      <td>0.694673</td>\n",
       "      <td>0.498802</td>\n",
       "      <td>0.330157</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.359510</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7233</td>\n",
       "      <td>0.309285</td>\n",
       "      <td>0.693883</td>\n",
       "      <td>0.273848</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.160960</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.998132</td>\n",
       "      <td>0.563191</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.615271</td>\n",
       "      <td>0.235051</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.644309</td>\n",
       "      <td>0.223404</td>\n",
       "      <td>0.776596</td>\n",
       "      <td>0.252983</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.644309</td>\n",
       "      <td>0.751619</td>\n",
       "      <td>0.661502</td>\n",
       "      <td>0.353376</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.283087</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7308</td>\n",
       "      <td>0.541409</td>\n",
       "      <td>0.291013</td>\n",
       "      <td>0.267796</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.291986</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.482362</td>\n",
       "      <td>0.243011</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.333946</td>\n",
       "      <td>0.230451</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>0.910090</td>\n",
       "      <td>0.799669</td>\n",
       "      <td>0.996196</td>\n",
       "      <td>0.675338</td>\n",
       "      <td>0.654535</td>\n",
       "      <td>0.038377</td>\n",
       "      <td>0.961623</td>\n",
       "      <td>0.143708</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.654535</td>\n",
       "      <td>0.874090</td>\n",
       "      <td>0.657223</td>\n",
       "      <td>0.318170</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.126117</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>328 rows × 37 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         payments  balance_mean  balance_min  reached_negative_balance  \\\n",
       "loan_id                                                                  \n",
       "5314     0.823266      0.000000     0.273848                       1.0   \n",
       "5316     0.457951      0.622572     0.249638                       1.0   \n",
       "6863     0.191996      0.278371     0.255691                       1.0   \n",
       "5325     0.279616      0.453990     0.267796                       1.0   \n",
       "7240     0.454642      0.702349     0.243586                       1.0   \n",
       "...           ...           ...          ...                       ...   \n",
       "6818     0.311953      0.499315     0.219376                       1.0   \n",
       "5625     0.361153      0.671751     0.255691                       1.0   \n",
       "6805     0.066062      0.464890     0.255691                       1.0   \n",
       "7233     0.309285      0.693883     0.273848                       1.0   \n",
       "7308     0.541409      0.291013     0.267796                       1.0   \n",
       "\n",
       "         unknown_sum  sanctions_mean  sanctions_sum  sanctions_count  \\\n",
       "loan_id                                                                \n",
       "5314        0.094611             1.0            1.0              0.0   \n",
       "5316        0.250351             1.0            1.0              0.0   \n",
       "6863        0.121845             1.0            1.0              0.0   \n",
       "5325        0.195018             1.0            1.0              0.0   \n",
       "7240        0.170400             1.0            1.0              0.0   \n",
       "...              ...             ...            ...              ...   \n",
       "6818        0.391429             1.0            1.0              0.0   \n",
       "5625        0.229755             1.0            1.0              0.0   \n",
       "6805        0.208544             1.0            1.0              0.0   \n",
       "7233        0.160960             1.0            1.0              0.0   \n",
       "7308        0.291986             1.0            1.0              0.0   \n",
       "\n",
       "         sanctions_std    CC_max    CC_std  CAB_mean   CAB_sum   CAB_max  \\\n",
       "loan_id                                                                    \n",
       "5314               0.0  0.194857  0.145173  0.000000  0.000000  0.000000   \n",
       "5316               0.0  0.730695  0.501677  0.000000  0.000000  0.000000   \n",
       "6863               0.0  0.070309  0.078863  0.259110  0.065120  0.255830   \n",
       "5325               0.0  0.074327  0.078863  0.354005  0.106764  0.354902   \n",
       "7240               0.0  0.008035  0.000000  0.848151  0.255792  0.850299   \n",
       "...                ...       ...       ...       ...       ...       ...   \n",
       "6818               0.0  0.526075  0.250648  0.000000  0.000000  0.000000   \n",
       "5625               0.0  0.886943  0.531975  0.000000  0.000000  0.000000   \n",
       "6805               0.0  0.631509  0.366021  0.000000  0.000000  0.000000   \n",
       "7233               0.0  0.998132  0.563191  0.000000  0.000000  0.000000   \n",
       "7308               0.0  0.482362  0.243011  0.000000  0.000000  0.000000   \n",
       "\n",
       "          CAB_std   CAB_min   IC_mean    IC_min       RAB  RAB_mean   RAB_sum  \\\n",
       "loan_id                                                                         \n",
       "5314     0.000000  0.000000  0.000000  0.000000  0.000000  1.000000  1.000000   \n",
       "5316     0.000000  0.000000  0.503220  0.013339  0.117647  0.817811  0.918812   \n",
       "6863     0.277983  0.255832  0.276556  0.223551  0.044118  0.808169  0.967943   \n",
       "5325     0.352033  0.354904  0.478797  0.610856  0.029412  0.524961  0.947078   \n",
       "7240     0.843426  0.850305  0.567842  0.354646  0.044118  0.552763  0.925263   \n",
       "...           ...       ...       ...       ...       ...       ...       ...   \n",
       "6818     0.000000  0.000000  0.434347  0.109016  0.529412  0.883923  0.767230   \n",
       "5625     0.000000  0.000000  0.665974  0.474701  0.000000  1.000000  1.000000   \n",
       "6805     0.000000  0.000000  0.341024  0.188592  0.088235  0.769918  0.923102   \n",
       "7233     0.000000  0.000000  0.615271  0.235051  0.000000  1.000000  1.000000   \n",
       "7308     0.000000  0.000000  0.333946  0.230451  0.588235  0.910090  0.799669   \n",
       "\n",
       "          RAB_max   RAB_min  mean_trans_profit  credit_ratio  \\\n",
       "loan_id                                                        \n",
       "5314     1.000000  1.000000           0.624874      1.000000   \n",
       "5316     0.935068  0.716973           0.610923      0.286947   \n",
       "6863     0.995517  0.460979           0.404010      0.505319   \n",
       "5325     0.524961  0.524961           0.518693      0.366809   \n",
       "7240     0.948652  0.354819           0.604275      0.315997   \n",
       "...           ...       ...                ...           ...   \n",
       "6818     0.943558  0.824289           0.537903      0.217714   \n",
       "5625     1.000000  1.000000           0.559665      0.306888   \n",
       "6805     0.860355  0.679481           0.500047      0.357338   \n",
       "7233     1.000000  1.000000           0.644309      0.223404   \n",
       "7308     0.996196  0.675338           0.654535      0.038377   \n",
       "\n",
       "         withdrawal_ratio  ratio_CC  ratio_RAB  owner_count  disponent_count  \\\n",
       "loan_id                                                                        \n",
       "5314             0.000000  1.000000   0.000000          0.0              0.0   \n",
       "5316             0.713053  0.319696   0.578378          0.0              0.0   \n",
       "6863             0.494681  0.077055   0.334375          0.0              0.0   \n",
       "5325             0.633191  0.073699   0.214000          0.0              0.0   \n",
       "7240             0.684003  0.030441   0.297222          0.0              0.0   \n",
       "...                   ...       ...        ...          ...              ...   \n",
       "6818             0.782286  0.145349   0.559884          1.0              1.0   \n",
       "5625             0.693112  0.266195   0.000000          0.0              0.0   \n",
       "6805             0.642662  0.199684   0.411538          0.0              0.0   \n",
       "7233             0.776596  0.252983   0.000000          0.0              0.0   \n",
       "7308             0.961623  0.143708   1.000000          0.0              0.0   \n",
       "\n",
       "         salary_from_month_records  withdrawal_to_credit_month_ratio  \\\n",
       "loan_id                                                                \n",
       "5314                      0.624874                          1.000000   \n",
       "5316                      0.610923                          0.754242   \n",
       "6863                      0.404010                          0.544732   \n",
       "5325                      0.518693                          0.710258   \n",
       "7240                      0.604275                          0.704321   \n",
       "...                            ...                               ...   \n",
       "6818                      0.537903                          0.777796   \n",
       "5625                      0.559665                          0.715703   \n",
       "6805                      0.500047                          0.694673   \n",
       "7233                      0.644309                          0.751619   \n",
       "7308                      0.654535                          0.874090   \n",
       "\n",
       "         actual_month_income  actual_income_to_payments_ratio  \\\n",
       "loan_id                                                         \n",
       "5314                0.641549                         0.299329   \n",
       "5316                0.614343                         0.314285   \n",
       "6863                0.390257                         0.228264   \n",
       "5325                0.510912                         0.293750   \n",
       "7240                0.618063                         0.315666   \n",
       "...                      ...                              ...   \n",
       "6818                0.544259                         0.304986   \n",
       "5625                0.574599                         0.311284   \n",
       "6805                0.498802                         0.330157   \n",
       "7233                0.661502                         0.353376   \n",
       "7308                0.657223                         0.318170   \n",
       "\n",
       "         log_sanctions_std  log_max_value_in_account_to_loan_ratio  status  \n",
       "loan_id                                                                     \n",
       "5314                   0.0                                0.048220     0.0  \n",
       "5316                   0.0                                0.210111     1.0  \n",
       "6863                   0.0                                0.112288     1.0  \n",
       "5325                   0.0                                0.182342     1.0  \n",
       "7240                   0.0                                0.130574     1.0  \n",
       "...                    ...                                     ...     ...  \n",
       "6818                   0.0                                0.141881     1.0  \n",
       "5625                   0.0                                0.172578     0.0  \n",
       "6805                   0.0                                0.359510     1.0  \n",
       "7233                   0.0                                0.283087     1.0  \n",
       "7308                   0.0                                0.126117     1.0  \n",
       "\n",
       "[328 rows x 37 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dataset =  utils.read_csv_to_df('dataset/preprocessed_data.csv')\n",
    "\n",
    "dataset = dataset.set_index('loan_id')\n",
    "display(dataset)\n",
    "# Normalizing dataset\n",
    "print(' > Dataset after normalization')\n",
    "normalize_df(dataset)\n",
    "display(dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "STATUS_COL = dataset.columns.get_loc(\"status\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": false
   },
   "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>payments</th>\n",
       "      <th>balance_mean</th>\n",
       "      <th>balance_min</th>\n",
       "      <th>reached_negative_balance</th>\n",
       "      <th>unknown_sum</th>\n",
       "      <th>sanctions_mean</th>\n",
       "      <th>sanctions_sum</th>\n",
       "      <th>sanctions_count</th>\n",
       "      <th>sanctions_std</th>\n",
       "      <th>CC_max</th>\n",
       "      <th>CC_std</th>\n",
       "      <th>CAB_mean</th>\n",
       "      <th>CAB_sum</th>\n",
       "      <th>CAB_max</th>\n",
       "      <th>CAB_std</th>\n",
       "      <th>CAB_min</th>\n",
       "      <th>IC_mean</th>\n",
       "      <th>IC_min</th>\n",
       "      <th>RAB</th>\n",
       "      <th>RAB_mean</th>\n",
       "      <th>RAB_sum</th>\n",
       "      <th>RAB_max</th>\n",
       "      <th>RAB_min</th>\n",
       "      <th>mean_trans_profit</th>\n",
       "      <th>credit_ratio</th>\n",
       "      <th>withdrawal_ratio</th>\n",
       "      <th>ratio_CC</th>\n",
       "      <th>ratio_RAB</th>\n",
       "      <th>owner_count</th>\n",
       "      <th>disponent_count</th>\n",
       "      <th>salary_from_month_records</th>\n",
       "      <th>withdrawal_to_credit_month_ratio</th>\n",
       "      <th>actual_month_income</th>\n",
       "      <th>actual_income_to_payments_ratio</th>\n",
       "      <th>log_sanctions_std</th>\n",
       "      <th>log_max_value_in_account_to_loan_ratio</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loan_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>5314</td>\n",
       "      <td>0.823266</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.273848</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.094611</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.194857</td>\n",
       "      <td>0.145173</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.624874</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.624874</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.641549</td>\n",
       "      <td>0.299329</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.048220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5316</td>\n",
       "      <td>0.457951</td>\n",
       "      <td>0.622572</td>\n",
       "      <td>0.249638</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.250351</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.730695</td>\n",
       "      <td>0.501677</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.503220</td>\n",
       "      <td>0.013339</td>\n",
       "      <td>0.117647</td>\n",
       "      <td>0.817811</td>\n",
       "      <td>0.918812</td>\n",
       "      <td>0.935068</td>\n",
       "      <td>0.716973</td>\n",
       "      <td>0.610923</td>\n",
       "      <td>0.286947</td>\n",
       "      <td>0.713053</td>\n",
       "      <td>0.319696</td>\n",
       "      <td>0.578378</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.610923</td>\n",
       "      <td>0.754242</td>\n",
       "      <td>0.614343</td>\n",
       "      <td>0.314285</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.210111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6863</td>\n",
       "      <td>0.191996</td>\n",
       "      <td>0.278371</td>\n",
       "      <td>0.255691</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.121845</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.070309</td>\n",
       "      <td>0.078863</td>\n",
       "      <td>0.259110</td>\n",
       "      <td>0.065120</td>\n",
       "      <td>0.255830</td>\n",
       "      <td>0.277983</td>\n",
       "      <td>0.255832</td>\n",
       "      <td>0.276556</td>\n",
       "      <td>0.223551</td>\n",
       "      <td>0.044118</td>\n",
       "      <td>0.808169</td>\n",
       "      <td>0.967943</td>\n",
       "      <td>0.995517</td>\n",
       "      <td>0.460979</td>\n",
       "      <td>0.404010</td>\n",
       "      <td>0.505319</td>\n",
       "      <td>0.494681</td>\n",
       "      <td>0.077055</td>\n",
       "      <td>0.334375</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.404010</td>\n",
       "      <td>0.544732</td>\n",
       "      <td>0.390257</td>\n",
       "      <td>0.228264</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.112288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5325</td>\n",
       "      <td>0.279616</td>\n",
       "      <td>0.453990</td>\n",
       "      <td>0.267796</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.195018</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.074327</td>\n",
       "      <td>0.078863</td>\n",
       "      <td>0.354005</td>\n",
       "      <td>0.106764</td>\n",
       "      <td>0.354902</td>\n",
       "      <td>0.352033</td>\n",
       "      <td>0.354904</td>\n",
       "      <td>0.478797</td>\n",
       "      <td>0.610856</td>\n",
       "      <td>0.029412</td>\n",
       "      <td>0.524961</td>\n",
       "      <td>0.947078</td>\n",
       "      <td>0.524961</td>\n",
       "      <td>0.524961</td>\n",
       "      <td>0.518693</td>\n",
       "      <td>0.366809</td>\n",
       "      <td>0.633191</td>\n",
       "      <td>0.073699</td>\n",
       "      <td>0.214000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.518693</td>\n",
       "      <td>0.710258</td>\n",
       "      <td>0.510912</td>\n",
       "      <td>0.293750</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.182342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7240</td>\n",
       "      <td>0.454642</td>\n",
       "      <td>0.702349</td>\n",
       "      <td>0.243586</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.170400</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.008035</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.848151</td>\n",
       "      <td>0.255792</td>\n",
       "      <td>0.850299</td>\n",
       "      <td>0.843426</td>\n",
       "      <td>0.850305</td>\n",
       "      <td>0.567842</td>\n",
       "      <td>0.354646</td>\n",
       "      <td>0.044118</td>\n",
       "      <td>0.552763</td>\n",
       "      <td>0.925263</td>\n",
       "      <td>0.948652</td>\n",
       "      <td>0.354819</td>\n",
       "      <td>0.604275</td>\n",
       "      <td>0.315997</td>\n",
       "      <td>0.684003</td>\n",
       "      <td>0.030441</td>\n",
       "      <td>0.297222</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.604275</td>\n",
       "      <td>0.704321</td>\n",
       "      <td>0.618063</td>\n",
       "      <td>0.315666</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.130574</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         payments  balance_mean  balance_min  reached_negative_balance  \\\n",
       "loan_id                                                                  \n",
       "5314     0.823266      0.000000     0.273848                       1.0   \n",
       "5316     0.457951      0.622572     0.249638                       1.0   \n",
       "6863     0.191996      0.278371     0.255691                       1.0   \n",
       "5325     0.279616      0.453990     0.267796                       1.0   \n",
       "7240     0.454642      0.702349     0.243586                       1.0   \n",
       "\n",
       "         unknown_sum  sanctions_mean  sanctions_sum  sanctions_count  \\\n",
       "loan_id                                                                \n",
       "5314        0.094611             1.0            1.0              0.0   \n",
       "5316        0.250351             1.0            1.0              0.0   \n",
       "6863        0.121845             1.0            1.0              0.0   \n",
       "5325        0.195018             1.0            1.0              0.0   \n",
       "7240        0.170400             1.0            1.0              0.0   \n",
       "\n",
       "         sanctions_std    CC_max    CC_std  CAB_mean   CAB_sum   CAB_max  \\\n",
       "loan_id                                                                    \n",
       "5314               0.0  0.194857  0.145173  0.000000  0.000000  0.000000   \n",
       "5316               0.0  0.730695  0.501677  0.000000  0.000000  0.000000   \n",
       "6863               0.0  0.070309  0.078863  0.259110  0.065120  0.255830   \n",
       "5325               0.0  0.074327  0.078863  0.354005  0.106764  0.354902   \n",
       "7240               0.0  0.008035  0.000000  0.848151  0.255792  0.850299   \n",
       "\n",
       "          CAB_std   CAB_min   IC_mean    IC_min       RAB  RAB_mean   RAB_sum  \\\n",
       "loan_id                                                                         \n",
       "5314     0.000000  0.000000  0.000000  0.000000  0.000000  1.000000  1.000000   \n",
       "5316     0.000000  0.000000  0.503220  0.013339  0.117647  0.817811  0.918812   \n",
       "6863     0.277983  0.255832  0.276556  0.223551  0.044118  0.808169  0.967943   \n",
       "5325     0.352033  0.354904  0.478797  0.610856  0.029412  0.524961  0.947078   \n",
       "7240     0.843426  0.850305  0.567842  0.354646  0.044118  0.552763  0.925263   \n",
       "\n",
       "          RAB_max   RAB_min  mean_trans_profit  credit_ratio  \\\n",
       "loan_id                                                        \n",
       "5314     1.000000  1.000000           0.624874      1.000000   \n",
       "5316     0.935068  0.716973           0.610923      0.286947   \n",
       "6863     0.995517  0.460979           0.404010      0.505319   \n",
       "5325     0.524961  0.524961           0.518693      0.366809   \n",
       "7240     0.948652  0.354819           0.604275      0.315997   \n",
       "\n",
       "         withdrawal_ratio  ratio_CC  ratio_RAB  owner_count  disponent_count  \\\n",
       "loan_id                                                                        \n",
       "5314             0.000000  1.000000   0.000000          0.0              0.0   \n",
       "5316             0.713053  0.319696   0.578378          0.0              0.0   \n",
       "6863             0.494681  0.077055   0.334375          0.0              0.0   \n",
       "5325             0.633191  0.073699   0.214000          0.0              0.0   \n",
       "7240             0.684003  0.030441   0.297222          0.0              0.0   \n",
       "\n",
       "         salary_from_month_records  withdrawal_to_credit_month_ratio  \\\n",
       "loan_id                                                                \n",
       "5314                      0.624874                          1.000000   \n",
       "5316                      0.610923                          0.754242   \n",
       "6863                      0.404010                          0.544732   \n",
       "5325                      0.518693                          0.710258   \n",
       "7240                      0.604275                          0.704321   \n",
       "\n",
       "         actual_month_income  actual_income_to_payments_ratio  \\\n",
       "loan_id                                                         \n",
       "5314                0.641549                         0.299329   \n",
       "5316                0.614343                         0.314285   \n",
       "6863                0.390257                         0.228264   \n",
       "5325                0.510912                         0.293750   \n",
       "7240                0.618063                         0.315666   \n",
       "\n",
       "         log_sanctions_std  log_max_value_in_account_to_loan_ratio  \n",
       "loan_id                                                             \n",
       "5314                   0.0                                0.048220  \n",
       "5316                   0.0                                0.210111  \n",
       "6863                   0.0                                0.112288  \n",
       "5325                   0.0                                0.182342  \n",
       "7240                   0.0                                0.130574  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Setting X and Y\n",
    "X = dataset.iloc[:, 0:STATUS_COL]\n",
    "y = dataset.iloc[:, [STATUS_COL]]\n",
    "display(X.head())\n",
    "\n",
    "if USE_PCA:\n",
    "    print(' > Applying PCA to X_train:')\n",
    "    X, pca = apply_PCA(X, debug=True)\n",
    "    display(X.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hyper Parameter Tunning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using GridSearch to obtain best parameters.\n",
    "\n",
    " > Considered using RandomSearch, but since the search is not that big, opted for GridSearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getLogisticRegressionBest(X, y, debug=True):\n",
    "    '''Get the Logistic Regression Hyper Parameters'''\n",
    "\n",
    "    # Maximum number of levels in tree\n",
    "    max_depth = [int(x) for x in range(2, 20, 4)]\n",
    "    max_depth.append(None)\n",
    "\n",
    "    # Create the random grid\n",
    "    grid = {'classification__penalty': ['l2', 'none'],\n",
    "            'classification__C': [0.01, 0.05, 0.1, 0.2, 0.5, 1.0],\n",
    "            'classification__solver': ['newton-cg', 'lbfgs', 'sag', 'saga'],\n",
    "            'classification__class_weight': [\"balanced\", None]}\n",
    "    \n",
    "    if debug:\n",
    "        pp.pprint(grid)\n",
    "    \n",
    "    # Applying sampling techniques\n",
    "    lr = apply_sampling(create_LR(), OVERSAMPLE, UNDERSAMPLE)\n",
    "\n",
    "    # Using the grid search for best hyperparameters\n",
    "    lr_grid = GridSearchCV(estimator = lr,\n",
    "                           param_grid = grid,\n",
    "                           scoring=metrics.make_scorer(auc_scorer,\n",
    "                                                       greater_is_better=True),\n",
    "                           cv=StratifiedKFold(K_FOLD_NUM_SPLITS,\n",
    "                                             random_state=SEED,\n",
    "                                              shuffle=True),\n",
    "                           verbose=2,\n",
    "                           n_jobs = -1)\n",
    "\n",
    "    # Fit the grid search model\n",
    "    lr_grid = lr_grid.fit(X, y)\n",
    "    \n",
    "    if debug:\n",
    "        print('Best Score: ', lr_grid.best_score_)\n",
    "        print('Best Params: ', lr_grid.best_params_)\n",
    "        \n",
    "    # Return score, method & params tuple\n",
    "    return (lr_grid.best_score_, 'Logistic Regression', lr_grid.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getDecisionTreeBest(X, y, debug=True):\n",
    "    '''Get the Decision Tree Hyper Parameters'''\n",
    "\n",
    "    # Maximum number of levels in tree\n",
    "    max_depth = [int(x) for x in range(2, 16, 2)]\n",
    "    max_depth.append(None)\n",
    "\n",
    "    # Create the random grid\n",
    "    grid = {'classification__criterion': ['gini', 'entropy'],\n",
    "            'classification__splitter': ['best'],\n",
    "            'classification__max_features': ['auto', 'sqrt'],\n",
    "            'classification__max_depth': max_depth,\n",
    "            'classification__min_samples_split':  [2, 4, 6, 8],\n",
    "            'classification__min_samples_leaf':  [1, 2, 4, 6],\n",
    "            'classification__min_impurity_split': [0.05, 0.1, 0.23, 0.3],\n",
    "            'classification__class_weight': [\"balanced\", None]}\n",
    "    \n",
    "    if debug:\n",
    "        pp.pprint(grid)\n",
    "\n",
    "    # Applying sampling techniques\n",
    "    dt = apply_sampling(create_DT(), OVERSAMPLE, UNDERSAMPLE)\n",
    "\n",
    "    # Using the grid search for best hyperparameters\n",
    "    dt_grid = GridSearchCV(estimator = dt,\n",
    "                           param_grid = grid,\n",
    "                           scoring=metrics.make_scorer(auc_scorer,\n",
    "                                                       greater_is_better=True),\n",
    "                           cv=StratifiedKFold(K_FOLD_NUM_SPLITS,\n",
    "                                             random_state=SEED,\n",
    "                                              shuffle=True),\n",
    "                           verbose=2,\n",
    "                           n_jobs = -1)\n",
    "\n",
    "    # Fit the grid search model\n",
    "    dt_grid = dt_grid.fit(X, y)\n",
    "    \n",
    "    if debug:\n",
    "        print('Best Score: ', dt_grid.best_score_)\n",
    "        print('Best Params: ', dt_grid.best_params_)\n",
    "        \n",
    "    # Return score, method & params tuple\n",
    "    return (dt_grid.best_score_, 'Decision Tree', dt_grid.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getRandomForestBest(X, y, debug=True):\n",
    "    '''Get the Random Forest Hyper Parameters'''\n",
    "\n",
    "    # Maximum number of levels in tree\n",
    "    max_depth = [int(x) for x in range(2, 16, 4)]\n",
    "    max_depth.append(None)\n",
    "\n",
    "    # Create the random grid\n",
    "    grid = {'classification__n_estimators': [int(x) for x in range(2, 14, 2)],\n",
    "            'classification__max_features': ['auto', 'sqrt'],\n",
    "            'classification__max_depth': max_depth,\n",
    "            'classification__criterion': ['gini', 'entropy'],\n",
    "            'classification__min_samples_split':  [2, 4, 6, 8],\n",
    "            'classification__min_samples_leaf':  [1, 2, 4, 6],\n",
    "            'classification__class_weight': [\"balanced\", \"balanced_subsample\", None]}\n",
    "    \n",
    "    if debug:\n",
    "        pp.pprint(grid)\n",
    "    \n",
    "    # Applying sampling techniques\n",
    "    rf = apply_sampling(create_RF(), OVERSAMPLE, UNDERSAMPLE)\n",
    "\n",
    "    # Using the grid search for best hyperparameters\n",
    "    rf_grid = GridSearchCV(estimator = rf,\n",
    "                           param_grid = grid,\n",
    "                           scoring=metrics.make_scorer(auc_scorer,\n",
    "                                                       greater_is_better=True),\n",
    "                           cv=StratifiedKFold(K_FOLD_NUM_SPLITS,\n",
    "                                             random_state=SEED,\n",
    "                                              shuffle=True),\n",
    "                           verbose=2,\n",
    "                           n_jobs = -1)\n",
    "\n",
    "    # Fit the grid search model\n",
    "    rf_grid = rf_grid.fit(X, y)\n",
    "    \n",
    "    if debug:\n",
    "        print('Best Score: ', rf_grid.best_score_)\n",
    "        print('Best Params: ', rf_grid.best_params_)\n",
    "        \n",
    "    # Return score, method & params tuple\n",
    "    return (rf_grid.best_score_, 'Random Forest', rf_grid.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getGradientBoostBest(X, y, debug=True):\n",
    "    '''Get the Gradient Boost Hyper Parameters'''\n",
    "\n",
    "    # Create the grid parameters\n",
    "    grid = {'classification__n_estimators': [int(x) for x in range(2, 14, 2)],\n",
    "            'classification__learning_rate': [0.1, 0.3, 0.5, 0.7],\n",
    "            'classification__loss': ['deviance', 'exponential'],\n",
    "            'classification__criterion': ['friedman_mse', 'mse', 'mae'],\n",
    "            'classification__min_samples_split':  [4, 6, 8],\n",
    "            'classification__min_samples_leaf':  [2, 4, 6]}\n",
    "    \n",
    "    if debug:\n",
    "        pp.pprint(grid)\n",
    "\n",
    "    # Applying sampling techniques\n",
    "    gb = apply_sampling(create_GB(), OVERSAMPLE, UNDERSAMPLE)\n",
    "\n",
    "    # Using the grid search for best hyperparameters\n",
    "    gb_grid = GridSearchCV(estimator = gb,\n",
    "                           param_grid = grid,\n",
    "                           scoring=metrics.make_scorer(auc_scorer,\n",
    "                                                       greater_is_better=True),\n",
    "                           cv=StratifiedKFold(K_FOLD_NUM_SPLITS,\n",
    "                                             random_state=SEED,\n",
    "                                              shuffle=True),\n",
    "                           verbose=2,\n",
    "                           n_jobs = -1)\n",
    "\n",
    "    # Fit the grid search model\n",
    "    gb_grid = gb_grid.fit(X, y)\n",
    "    \n",
    "    if debug:\n",
    "        print('Best Score: ', gb_grid.best_score_)\n",
    "        print('Best Params: ', gb_grid.best_params_)\n",
    "        \n",
    "    # Return score, method & params tuple\n",
    "    return (gb_grid.best_score_, 'Gradient Boosting', gb_grid.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{   'classification__C': [0.01, 0.05, 0.1, 0.2, 0.5, 1.0],\n",
      "    'classification__class_weight': ['balanced', None],\n",
      "    'classification__penalty': ['l2', 'none'],\n",
      "    'classification__solver': ['newton-cg', 'lbfgs', 'sag', 'saga']}\n",
      "Fitting 5 folds for each of 96 candidates, totalling 480 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  40 tasks      | elapsed:    2.4s\n",
      "[Parallel(n_jobs=-1)]: Done 480 out of 480 | elapsed:    8.3s finished\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/model_selection/_search.py:814: DeprecationWarning: The default of the `iid` parameter will change from True to False in version 0.22 and will be removed in 0.24. This will change numeric results when test-set sizes are unequal.\n",
      "  DeprecationWarning)\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/utils/validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/linear_model/logistic.py:1510: UserWarning: Setting penalty='none' will ignore the C and l1_ratio parameters\n",
      "  \"Setting penalty='none' will ignore the C and l1_ratio \"\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/linear_model/sag.py:337: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  \"the coef_ did not converge\", ConvergenceWarning)\n",
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Score:  0.8278157920207021\n",
      "Best Params:  {'classification__C': 0.01, 'classification__class_weight': 'balanced', 'classification__penalty': 'none', 'classification__solver': 'saga'}\n",
      "{   'classification__class_weight': ['balanced', None],\n",
      "    'classification__criterion': ['gini', 'entropy'],\n",
      "    'classification__max_depth': [2, 4, 6, 8, 10, 12, 14, None],\n",
      "    'classification__max_features': ['auto', 'sqrt'],\n",
      "    'classification__min_impurity_split': [0.05, 0.1, 0.23, 0.3],\n",
      "    'classification__min_samples_leaf': [1, 2, 4, 6],\n",
      "    'classification__min_samples_split': [2, 4, 6, 8],\n",
      "    'classification__splitter': ['best']}\n",
      "Fitting 5 folds for each of 4096 candidates, totalling 20480 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done 408 tasks      | elapsed:    5.0s\n",
      "[Parallel(n_jobs=-1)]: Done 2344 tasks      | elapsed:   21.5s\n",
      "[Parallel(n_jobs=-1)]: Done 5592 tasks      | elapsed:   43.1s\n",
      "[Parallel(n_jobs=-1)]: Done 10120 tasks      | elapsed:  1.3min\n",
      "[Parallel(n_jobs=-1)]: Done 15960 tasks      | elapsed:  1.9min\n",
      "[Parallel(n_jobs=-1)]: Done 20480 out of 20480 | elapsed:  2.4min finished\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/model_selection/_search.py:814: DeprecationWarning: The default of the `iid` parameter will change from True to False in version 0.22 and will be removed in 0.24. This will change numeric results when test-set sizes are unequal.\n",
      "  DeprecationWarning)\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/utils/validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/tree/tree.py:297: DeprecationWarning: The min_impurity_split parameter is deprecated. Its default value will change from 1e-7 to 0 in version 0.23, and it will be removed in 0.25. Use the min_impurity_decrease parameter instead.\n",
      "  DeprecationWarning)\n",
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Score:  0.8148510885954725\n",
      "Best Params:  {'classification__class_weight': 'balanced', 'classification__criterion': 'entropy', 'classification__max_depth': 8, 'classification__max_features': 'auto', 'classification__min_impurity_split': 0.3, 'classification__min_samples_leaf': 2, 'classification__min_samples_split': 2, 'classification__splitter': 'best'}\n",
      "{   'classification__class_weight': ['balanced', 'balanced_subsample', None],\n",
      "    'classification__criterion': ['gini', 'entropy'],\n",
      "    'classification__max_depth': [2, 6, 10, 14, None],\n",
      "    'classification__max_features': ['auto', 'sqrt'],\n",
      "    'classification__min_samples_leaf': [1, 2, 4, 6],\n",
      "    'classification__min_samples_split': [2, 4, 6, 8],\n",
      "    'classification__n_estimators': [2, 4, 6, 8, 10, 12]}\n",
      "Fitting 5 folds for each of 5760 candidates, totalling 28800 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done 408 tasks      | elapsed:    4.5s\n",
      "[Parallel(n_jobs=-1)]: Done 2344 tasks      | elapsed:   29.2s\n",
      "[Parallel(n_jobs=-1)]: Done 5592 tasks      | elapsed:  1.2min\n",
      "[Parallel(n_jobs=-1)]: Done 10120 tasks      | elapsed:  2.2min\n",
      "[Parallel(n_jobs=-1)]: Done 15960 tasks      | elapsed:  3.4min\n",
      "[Parallel(n_jobs=-1)]: Done 23080 tasks      | elapsed:  5.0min\n",
      "[Parallel(n_jobs=-1)]: Done 28800 out of 28800 | elapsed:  6.2min finished\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/model_selection/_search.py:814: DeprecationWarning: The default of the `iid` parameter will change from True to False in version 0.22 and will be removed in 0.24. This will change numeric results when test-set sizes are unequal.\n",
      "  DeprecationWarning)\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/utils/validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n",
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Score:  0.8680515121645577\n",
      "Best Params:  {'classification__class_weight': 'balanced', 'classification__criterion': 'gini', 'classification__max_depth': 6, 'classification__max_features': 'auto', 'classification__min_samples_leaf': 6, 'classification__min_samples_split': 2, 'classification__n_estimators': 10}\n",
      "{   'classification__criterion': ['friedman_mse', 'mse', 'mae'],\n",
      "    'classification__learning_rate': [0.1, 0.3, 0.5, 0.7],\n",
      "    'classification__loss': ['deviance', 'exponential'],\n",
      "    'classification__min_samples_leaf': [2, 4, 6],\n",
      "    'classification__min_samples_split': [4, 6, 8],\n",
      "    'classification__n_estimators': [2, 4, 6, 8, 10, 12]}\n",
      "Fitting 5 folds for each of 1296 candidates, totalling 6480 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done 488 tasks      | elapsed:    6.3s\n",
      "[Parallel(n_jobs=-1)]: Done 1698 tasks      | elapsed:   21.2s\n",
      "[Parallel(n_jobs=-1)]: Done 3728 tasks      | elapsed:   45.3s\n",
      "[Parallel(n_jobs=-1)]: Done 4872 tasks      | elapsed:  1.6min\n",
      "[Parallel(n_jobs=-1)]: Done 5359 tasks      | elapsed:  2.4min\n",
      "[Parallel(n_jobs=-1)]: Done 5804 tasks      | elapsed:  3.1min\n",
      "[Parallel(n_jobs=-1)]: Done 6331 tasks      | elapsed:  3.9min\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Score:  0.8424238632658881\n",
      "Best Params:  {'classification__criterion': 'friedman_mse', 'classification__learning_rate': 0.3, 'classification__loss': 'deviance', 'classification__min_samples_leaf': 4, 'classification__min_samples_split': 4, 'classification__n_estimators': 12}\n",
      "1. Random Forest - 0.868052\n",
      "---------\n",
      "2. Gradient Boosting - 0.842424\n",
      "---------\n",
      "3. Logistic Regression - 0.827816\n",
      "---------\n",
      "4. Decision Tree - 0.814851\n",
      "---------\n",
      "Best algorithm: Random Forest\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done 6480 out of 6480 | elapsed:  4.1min finished\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/model_selection/_search.py:814: DeprecationWarning: The default of the `iid` parameter will change from True to False in version 0.22 and will be removed in 0.24. This will change numeric results when test-set sizes are unequal.\n",
      "  DeprecationWarning)\n",
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/utils/validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    }
   ],
   "source": [
    "# Getting the best algorithm\n",
    "algorithms = [getLogisticRegressionBest(X, y),\n",
    "              getDecisionTreeBest(X, y),\n",
    "              getRandomForestBest(X, y),\n",
    "              getGradientBoostBest(X, y)]\n",
    "algorithms.sort(reverse=True, key=lambda el: el[0])\n",
    "\n",
    "for index, entry in enumerate(algorithms):\n",
    "    print('%i. %s - %f\\n---------' % (index + 1, entry[1], entry[0]))\n",
    "    \n",
    "print('Best algorithm: %s' % algorithms[0][1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using method with higher score with our data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/utils/validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/utils/validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/utils/validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/utils/validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/edgarcarneiro/Documents/University/feup-ecac/project-competition/env/lib/python3.7/site-packages/sklearn/utils/validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Cross validation settings\n",
    "auc_scores = []\n",
    "confusion_matrixes = []\n",
    "cv = StratifiedKFold(n_splits=K_FOLD_NUM_SPLITS,\n",
    "                     random_state=SEED,\n",
    "                     shuffle=True)\n",
    "\n",
    "# CHANGE THIS LINE TO CHANGE THE USED CLASSIFICATION METHOD\n",
    "classifier = RandomForestClassifier(bootstrap=False,\n",
    "                                    class_weight='balanced',\n",
    "                                    criterion='gini',\n",
    "                                    max_depth=6,\n",
    "                                    max_features='auto',\n",
    "                                    min_samples_leaf=6,\n",
    "                                    min_samples_split=2,\n",
    "                                    n_estimators=10,\n",
    "                                    random_state=SEED)\n",
    "\n",
    "\n",
    "# Applying sampling techniques\n",
    "classifier = apply_sampling(classifier, OVERSAMPLE, UNDERSAMPLE)\n",
    "\n",
    "# Applying Cross validation\n",
    "for train_index, test_index in cv.split(X, y):\n",
    "    X_train, X_test = X.iloc[train_index], X.iloc[test_index]\n",
    "    y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "\n",
    "    # Training with this fold\n",
    "    classifier.fit(X_train, y_train)\n",
    "    \n",
    "    # Testing & Measuring accuracy\n",
    "    y_pred = classifier.predict(X_test)\n",
    "    \n",
    "    fpr, tpr, _ = metrics.roc_curve(y_test, y_pred)\n",
    "    auc = metrics.auc(fpr, tpr)\n",
    "\n",
    "    auc_scores.append(auc)\n",
    "    confusion_matrixes.append(metrics.confusion_matrix(y_test, y_pred).ravel())\n",
    "\n",
    "    plot_roc_auc(fpr, tpr, auc)\n",
    "    \n",
    "# Visualization of classifier result - if DT or RF\n",
    "algorithm = classifier.get_params().get('classification')\n",
    "\n",
    "if isinstance(algorithm, DecisionTreeClassifier):\n",
    "    visualize_tree(algorithm, 'images/DT')\n",
    "\n",
    "elif isinstance(algorithm, RandomForestClassifier):\n",
    "    for index, tree in enumerate(algorithm.estimators_):\n",
    "        visualize_tree(tree, 'images/RF_' + str(index))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Method used: Pipeline(memory=None,\n",
      "         steps=[('sampling',\n",
      "                 SMOTE(k_neighbors=5, kind='deprecated',\n",
      "                       m_neighbors='deprecated', n_jobs=1,\n",
      "                       out_step='deprecated', random_state=42, ratio=None,\n",
      "                       sampling_strategy='auto', svm_estimator='deprecated')),\n",
      "                ('classification',\n",
      "                 RandomForestClassifier(bootstrap=False,\n",
      "                                        class_weight='balanced',\n",
      "                                        criterion='gini', max_depth=6,\n",
      "                                        max_features='auto',\n",
      "                                        max_leaf_nodes=None,\n",
      "                                        min_impurity_decrease=0.0,\n",
      "                                        min_impurity_split=None,\n",
      "                                        min_samples_leaf=6, min_samples_split=2,\n",
      "                                        min_weight_fraction_leaf=0.0,\n",
      "                                        n_estimators=10, n_jobs=None,\n",
      "                                        oob_score=False, random_state=42,\n",
      "                                        verbose=0, warm_start=False))],\n",
      "         verbose=False) \n",
      "\n",
      "AUC scores: [0.8649122807017544, 0.736842105263158, 0.8640873015873016, 0.7728174603174603, 0.8442460317460316]\n",
      "> Average:  0.8165810359231411\n",
      "          \tPredic NO\tPredic YES\n",
      "Actual NO \t8        \t2         \n",
      "Actual YES\t4        \t53        \n",
      "          \tPredic NO\tPredic YES\n",
      "Actual NO \t6        \t3         \n",
      "Actual YES\t11       \t46        \n",
      "          \tPredic NO\tPredic YES\n",
      "Actual NO \t8        \t1         \n",
      "Actual YES\t9        \t47        \n",
      "          \tPredic NO\tPredic YES\n",
      "Actual NO \t7        \t2         \n",
      "Actual YES\t13       \t43        \n",
      "          \tPredic NO\tPredic YES\n",
      "Actual NO \t7        \t2         \n",
      "Actual YES\t5        \t51        \n"
     ]
    }
   ],
   "source": [
    "# Printing the obtained results\n",
    "print('Classification Method used:', classifier, '\\n')\n",
    "print('AUC scores:', auc_scores)\n",
    "print('> Average: ', sum(auc_scores)/len(auc_scores))\n",
    "for cf in confusion_matrixes:\n",
    "    display_confusion_matrix(cf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### After having our model trained we shall use the model on the data to be sumitted in the kaggle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true
   },
   "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>payments</th>\n",
       "      <th>balance_mean</th>\n",
       "      <th>balance_min</th>\n",
       "      <th>reached_negative_balance</th>\n",
       "      <th>unknown_sum</th>\n",
       "      <th>sanctions_mean</th>\n",
       "      <th>sanctions_sum</th>\n",
       "      <th>sanctions_count</th>\n",
       "      <th>sanctions_std</th>\n",
       "      <th>CC_max</th>\n",
       "      <th>CC_std</th>\n",
       "      <th>CAB_mean</th>\n",
       "      <th>CAB_sum</th>\n",
       "      <th>CAB_max</th>\n",
       "      <th>CAB_std</th>\n",
       "      <th>CAB_min</th>\n",
       "      <th>IC_mean</th>\n",
       "      <th>IC_min</th>\n",
       "      <th>RAB</th>\n",
       "      <th>RAB_mean</th>\n",
       "      <th>RAB_sum</th>\n",
       "      <th>RAB_max</th>\n",
       "      <th>RAB_min</th>\n",
       "      <th>mean_trans_profit</th>\n",
       "      <th>credit_ratio</th>\n",
       "      <th>withdrawal_ratio</th>\n",
       "      <th>ratio_CC</th>\n",
       "      <th>ratio_RAB</th>\n",
       "      <th>owner_count</th>\n",
       "      <th>disponent_count</th>\n",
       "      <th>salary_from_month_records</th>\n",
       "      <th>withdrawal_to_credit_month_ratio</th>\n",
       "      <th>actual_month_income</th>\n",
       "      <th>actual_income_to_payments_ratio</th>\n",
       "      <th>log_sanctions_std</th>\n",
       "      <th>log_max_value_in_account_to_loan_ratio</th>\n",
       "      <th>status</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loan_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>5895</td>\n",
       "      <td>0.131376</td>\n",
       "      <td>0.663113</td>\n",
       "      <td>0.474157</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.360467</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.731809</td>\n",
       "      <td>0.483870</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.689672</td>\n",
       "      <td>0.373467</td>\n",
       "      <td>0.226190</td>\n",
       "      <td>0.699383</td>\n",
       "      <td>0.613525</td>\n",
       "      <td>0.763137</td>\n",
       "      <td>0.628545</td>\n",
       "      <td>0.660156</td>\n",
       "      <td>0.146538</td>\n",
       "      <td>0.853462</td>\n",
       "      <td>0.175218</td>\n",
       "      <td>0.546415</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.660156</td>\n",
       "      <td>0.845588</td>\n",
       "      <td>0.659778</td>\n",
       "      <td>0.413987</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.246558</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7122</td>\n",
       "      <td>0.722049</td>\n",
       "      <td>0.352906</td>\n",
       "      <td>0.433774</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.147094</td>\n",
       "      <td>0.870568</td>\n",
       "      <td>0.987057</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.836918</td>\n",
       "      <td>0.726426</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.439087</td>\n",
       "      <td>0.070423</td>\n",
       "      <td>0.273810</td>\n",
       "      <td>0.793889</td>\n",
       "      <td>0.679238</td>\n",
       "      <td>0.977758</td>\n",
       "      <td>0.570286</td>\n",
       "      <td>0.572825</td>\n",
       "      <td>0.191233</td>\n",
       "      <td>0.808767</td>\n",
       "      <td>0.174929</td>\n",
       "      <td>0.754731</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.572825</td>\n",
       "      <td>0.802215</td>\n",
       "      <td>0.573668</td>\n",
       "      <td>0.281695</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.103447</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6173</td>\n",
       "      <td>0.472725</td>\n",
       "      <td>0.469661</td>\n",
       "      <td>0.458202</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.341665</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.190798</td>\n",
       "      <td>0.224621</td>\n",
       "      <td>0.574792</td>\n",
       "      <td>0.524729</td>\n",
       "      <td>0.584189</td>\n",
       "      <td>0.482511</td>\n",
       "      <td>0.584192</td>\n",
       "      <td>0.539962</td>\n",
       "      <td>0.207633</td>\n",
       "      <td>0.726190</td>\n",
       "      <td>0.918482</td>\n",
       "      <td>0.663539</td>\n",
       "      <td>0.981118</td>\n",
       "      <td>0.819110</td>\n",
       "      <td>0.764744</td>\n",
       "      <td>0.027116</td>\n",
       "      <td>0.972884</td>\n",
       "      <td>0.011585</td>\n",
       "      <td>0.887108</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.764744</td>\n",
       "      <td>0.895074</td>\n",
       "      <td>0.773518</td>\n",
       "      <td>0.339933</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.103590</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6142</td>\n",
       "      <td>0.353321</td>\n",
       "      <td>0.527183</td>\n",
       "      <td>0.479476</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.161846</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.020084</td>\n",
       "      <td>0.006688</td>\n",
       "      <td>0.438695</td>\n",
       "      <td>0.200243</td>\n",
       "      <td>0.445862</td>\n",
       "      <td>0.378339</td>\n",
       "      <td>0.445871</td>\n",
       "      <td>0.502754</td>\n",
       "      <td>0.246252</td>\n",
       "      <td>0.285714</td>\n",
       "      <td>0.940499</td>\n",
       "      <td>0.903375</td>\n",
       "      <td>0.999866</td>\n",
       "      <td>0.807015</td>\n",
       "      <td>0.674889</td>\n",
       "      <td>0.093363</td>\n",
       "      <td>0.906637</td>\n",
       "      <td>0.022770</td>\n",
       "      <td>0.865191</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.674889</td>\n",
       "      <td>0.876517</td>\n",
       "      <td>0.687726</td>\n",
       "      <td>0.336087</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.101718</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5358</td>\n",
       "      <td>0.302519</td>\n",
       "      <td>0.200691</td>\n",
       "      <td>0.476817</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.237976</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.172722</td>\n",
       "      <td>0.096925</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.249445</td>\n",
       "      <td>0.089959</td>\n",
       "      <td>0.190476</td>\n",
       "      <td>0.790687</td>\n",
       "      <td>0.773395</td>\n",
       "      <td>0.790687</td>\n",
       "      <td>0.790687</td>\n",
       "      <td>0.471897</td>\n",
       "      <td>0.299821</td>\n",
       "      <td>0.700179</td>\n",
       "      <td>0.285526</td>\n",
       "      <td>0.359231</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.471897</td>\n",
       "      <td>0.736241</td>\n",
       "      <td>0.473703</td>\n",
       "      <td>0.267758</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.220466</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         payments  balance_mean  balance_min  reached_negative_balance  \\\n",
       "loan_id                                                                  \n",
       "5895     0.131376      0.663113     0.474157                       1.0   \n",
       "7122     0.722049      0.352906     0.433774                       0.0   \n",
       "6173     0.472725      0.469661     0.458202                       1.0   \n",
       "6142     0.353321      0.527183     0.479476                       1.0   \n",
       "5358     0.302519      0.200691     0.476817                       1.0   \n",
       "\n",
       "         unknown_sum  sanctions_mean  sanctions_sum  sanctions_count  \\\n",
       "loan_id                                                                \n",
       "5895        0.360467        1.000000       1.000000              0.0   \n",
       "7122        0.147094        0.870568       0.987057              0.1   \n",
       "6173        0.341665        1.000000       1.000000              0.0   \n",
       "6142        0.161846        1.000000       1.000000              0.0   \n",
       "5358        0.237976        1.000000       1.000000              0.0   \n",
       "\n",
       "         sanctions_std    CC_max    CC_std  CAB_mean   CAB_sum   CAB_max  \\\n",
       "loan_id                                                                    \n",
       "5895               0.0  0.731809  0.483870  0.000000  0.000000  0.000000   \n",
       "7122               0.0  0.836918  0.726426  0.000000  0.000000  0.000000   \n",
       "6173               0.0  0.190798  0.224621  0.574792  0.524729  0.584189   \n",
       "6142               0.0  0.020084  0.006688  0.438695  0.200243  0.445862   \n",
       "5358               0.0  0.172722  0.096925  0.000000  0.000000  0.000000   \n",
       "\n",
       "          CAB_std   CAB_min   IC_mean    IC_min       RAB  RAB_mean   RAB_sum  \\\n",
       "loan_id                                                                         \n",
       "5895     0.000000  0.000000  0.689672  0.373467  0.226190  0.699383  0.613525   \n",
       "7122     0.000000  0.000000  0.439087  0.070423  0.273810  0.793889  0.679238   \n",
       "6173     0.482511  0.584192  0.539962  0.207633  0.726190  0.918482  0.663539   \n",
       "6142     0.378339  0.445871  0.502754  0.246252  0.285714  0.940499  0.903375   \n",
       "5358     0.000000  0.000000  0.249445  0.089959  0.190476  0.790687  0.773395   \n",
       "\n",
       "          RAB_max   RAB_min  mean_trans_profit  credit_ratio  \\\n",
       "loan_id                                                        \n",
       "5895     0.763137  0.628545           0.660156      0.146538   \n",
       "7122     0.977758  0.570286           0.572825      0.191233   \n",
       "6173     0.981118  0.819110           0.764744      0.027116   \n",
       "6142     0.999866  0.807015           0.674889      0.093363   \n",
       "5358     0.790687  0.790687           0.471897      0.299821   \n",
       "\n",
       "         withdrawal_ratio  ratio_CC  ratio_RAB  owner_count  disponent_count  \\\n",
       "loan_id                                                                        \n",
       "5895             0.853462  0.175218   0.546415          0.0              0.0   \n",
       "7122             0.808767  0.174929   0.754731          0.0              0.0   \n",
       "6173             0.972884  0.011585   0.887108          0.0              0.0   \n",
       "6142             0.906637  0.022770   0.865191          1.0              1.0   \n",
       "5358             0.700179  0.285526   0.359231          0.0              0.0   \n",
       "\n",
       "         salary_from_month_records  withdrawal_to_credit_month_ratio  \\\n",
       "loan_id                                                                \n",
       "5895                      0.660156                          0.845588   \n",
       "7122                      0.572825                          0.802215   \n",
       "6173                      0.764744                          0.895074   \n",
       "6142                      0.674889                          0.876517   \n",
       "5358                      0.471897                          0.736241   \n",
       "\n",
       "         actual_month_income  actual_income_to_payments_ratio  \\\n",
       "loan_id                                                         \n",
       "5895                0.659778                         0.413987   \n",
       "7122                0.573668                         0.281695   \n",
       "6173                0.773518                         0.339933   \n",
       "6142                0.687726                         0.336087   \n",
       "5358                0.473703                         0.267758   \n",
       "\n",
       "         log_sanctions_std  log_max_value_in_account_to_loan_ratio  status  \n",
       "loan_id                                                                     \n",
       "5895                   0.0                                0.246558     NaN  \n",
       "7122                   0.0                                0.103447     NaN  \n",
       "6173                   0.0                                0.103590     NaN  \n",
       "6142                   0.0                                0.101718     NaN  \n",
       "5358                   0.0                                0.220466     NaN  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_dataset =  utils.read_csv_to_df('dataset/test_dataset.csv')\n",
    "test_dataset = test_dataset.set_index('loan_id')\n",
    "normalize_df(test_dataset)\n",
    "display(test_dataset.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>payments</th>\n",
       "      <th>balance_mean</th>\n",
       "      <th>balance_min</th>\n",
       "      <th>reached_negative_balance</th>\n",
       "      <th>unknown_sum</th>\n",
       "      <th>sanctions_mean</th>\n",
       "      <th>sanctions_sum</th>\n",
       "      <th>sanctions_count</th>\n",
       "      <th>sanctions_std</th>\n",
       "      <th>CC_max</th>\n",
       "      <th>CC_std</th>\n",
       "      <th>CAB_mean</th>\n",
       "      <th>CAB_sum</th>\n",
       "      <th>CAB_max</th>\n",
       "      <th>CAB_std</th>\n",
       "      <th>CAB_min</th>\n",
       "      <th>IC_mean</th>\n",
       "      <th>IC_min</th>\n",
       "      <th>RAB</th>\n",
       "      <th>RAB_mean</th>\n",
       "      <th>RAB_sum</th>\n",
       "      <th>RAB_max</th>\n",
       "      <th>RAB_min</th>\n",
       "      <th>mean_trans_profit</th>\n",
       "      <th>credit_ratio</th>\n",
       "      <th>withdrawal_ratio</th>\n",
       "      <th>ratio_CC</th>\n",
       "      <th>ratio_RAB</th>\n",
       "      <th>owner_count</th>\n",
       "      <th>disponent_count</th>\n",
       "      <th>salary_from_month_records</th>\n",
       "      <th>withdrawal_to_credit_month_ratio</th>\n",
       "      <th>actual_month_income</th>\n",
       "      <th>actual_income_to_payments_ratio</th>\n",
       "      <th>log_sanctions_std</th>\n",
       "      <th>log_max_value_in_account_to_loan_ratio</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loan_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>5895</td>\n",
       "      <td>0.131376</td>\n",
       "      <td>0.663113</td>\n",
       "      <td>0.474157</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.360467</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.731809</td>\n",
       "      <td>0.483870</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.689672</td>\n",
       "      <td>0.373467</td>\n",
       "      <td>0.226190</td>\n",
       "      <td>0.699383</td>\n",
       "      <td>0.613525</td>\n",
       "      <td>0.763137</td>\n",
       "      <td>0.628545</td>\n",
       "      <td>0.660156</td>\n",
       "      <td>0.146538</td>\n",
       "      <td>0.853462</td>\n",
       "      <td>0.175218</td>\n",
       "      <td>0.546415</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.660156</td>\n",
       "      <td>0.845588</td>\n",
       "      <td>0.659778</td>\n",
       "      <td>0.413987</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.246558</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7122</td>\n",
       "      <td>0.722049</td>\n",
       "      <td>0.352906</td>\n",
       "      <td>0.433774</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.147094</td>\n",
       "      <td>0.870568</td>\n",
       "      <td>0.987057</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.836918</td>\n",
       "      <td>0.726426</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.439087</td>\n",
       "      <td>0.070423</td>\n",
       "      <td>0.273810</td>\n",
       "      <td>0.793889</td>\n",
       "      <td>0.679238</td>\n",
       "      <td>0.977758</td>\n",
       "      <td>0.570286</td>\n",
       "      <td>0.572825</td>\n",
       "      <td>0.191233</td>\n",
       "      <td>0.808767</td>\n",
       "      <td>0.174929</td>\n",
       "      <td>0.754731</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.572825</td>\n",
       "      <td>0.802215</td>\n",
       "      <td>0.573668</td>\n",
       "      <td>0.281695</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.103447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6173</td>\n",
       "      <td>0.472725</td>\n",
       "      <td>0.469661</td>\n",
       "      <td>0.458202</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.341665</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.190798</td>\n",
       "      <td>0.224621</td>\n",
       "      <td>0.574792</td>\n",
       "      <td>0.524729</td>\n",
       "      <td>0.584189</td>\n",
       "      <td>0.482511</td>\n",
       "      <td>0.584192</td>\n",
       "      <td>0.539962</td>\n",
       "      <td>0.207633</td>\n",
       "      <td>0.726190</td>\n",
       "      <td>0.918482</td>\n",
       "      <td>0.663539</td>\n",
       "      <td>0.981118</td>\n",
       "      <td>0.819110</td>\n",
       "      <td>0.764744</td>\n",
       "      <td>0.027116</td>\n",
       "      <td>0.972884</td>\n",
       "      <td>0.011585</td>\n",
       "      <td>0.887108</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.764744</td>\n",
       "      <td>0.895074</td>\n",
       "      <td>0.773518</td>\n",
       "      <td>0.339933</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.103590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6142</td>\n",
       "      <td>0.353321</td>\n",
       "      <td>0.527183</td>\n",
       "      <td>0.479476</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.161846</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.020084</td>\n",
       "      <td>0.006688</td>\n",
       "      <td>0.438695</td>\n",
       "      <td>0.200243</td>\n",
       "      <td>0.445862</td>\n",
       "      <td>0.378339</td>\n",
       "      <td>0.445871</td>\n",
       "      <td>0.502754</td>\n",
       "      <td>0.246252</td>\n",
       "      <td>0.285714</td>\n",
       "      <td>0.940499</td>\n",
       "      <td>0.903375</td>\n",
       "      <td>0.999866</td>\n",
       "      <td>0.807015</td>\n",
       "      <td>0.674889</td>\n",
       "      <td>0.093363</td>\n",
       "      <td>0.906637</td>\n",
       "      <td>0.022770</td>\n",
       "      <td>0.865191</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.674889</td>\n",
       "      <td>0.876517</td>\n",
       "      <td>0.687726</td>\n",
       "      <td>0.336087</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.101718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5358</td>\n",
       "      <td>0.302519</td>\n",
       "      <td>0.200691</td>\n",
       "      <td>0.476817</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.237976</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.172722</td>\n",
       "      <td>0.096925</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.249445</td>\n",
       "      <td>0.089959</td>\n",
       "      <td>0.190476</td>\n",
       "      <td>0.790687</td>\n",
       "      <td>0.773395</td>\n",
       "      <td>0.790687</td>\n",
       "      <td>0.790687</td>\n",
       "      <td>0.471897</td>\n",
       "      <td>0.299821</td>\n",
       "      <td>0.700179</td>\n",
       "      <td>0.285526</td>\n",
       "      <td>0.359231</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.471897</td>\n",
       "      <td>0.736241</td>\n",
       "      <td>0.473703</td>\n",
       "      <td>0.267758</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.220466</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         payments  balance_mean  balance_min  reached_negative_balance  \\\n",
       "loan_id                                                                  \n",
       "5895     0.131376      0.663113     0.474157                       1.0   \n",
       "7122     0.722049      0.352906     0.433774                       0.0   \n",
       "6173     0.472725      0.469661     0.458202                       1.0   \n",
       "6142     0.353321      0.527183     0.479476                       1.0   \n",
       "5358     0.302519      0.200691     0.476817                       1.0   \n",
       "\n",
       "         unknown_sum  sanctions_mean  sanctions_sum  sanctions_count  \\\n",
       "loan_id                                                                \n",
       "5895        0.360467        1.000000       1.000000              0.0   \n",
       "7122        0.147094        0.870568       0.987057              0.1   \n",
       "6173        0.341665        1.000000       1.000000              0.0   \n",
       "6142        0.161846        1.000000       1.000000              0.0   \n",
       "5358        0.237976        1.000000       1.000000              0.0   \n",
       "\n",
       "         sanctions_std    CC_max    CC_std  CAB_mean   CAB_sum   CAB_max  \\\n",
       "loan_id                                                                    \n",
       "5895               0.0  0.731809  0.483870  0.000000  0.000000  0.000000   \n",
       "7122               0.0  0.836918  0.726426  0.000000  0.000000  0.000000   \n",
       "6173               0.0  0.190798  0.224621  0.574792  0.524729  0.584189   \n",
       "6142               0.0  0.020084  0.006688  0.438695  0.200243  0.445862   \n",
       "5358               0.0  0.172722  0.096925  0.000000  0.000000  0.000000   \n",
       "\n",
       "          CAB_std   CAB_min   IC_mean    IC_min       RAB  RAB_mean   RAB_sum  \\\n",
       "loan_id                                                                         \n",
       "5895     0.000000  0.000000  0.689672  0.373467  0.226190  0.699383  0.613525   \n",
       "7122     0.000000  0.000000  0.439087  0.070423  0.273810  0.793889  0.679238   \n",
       "6173     0.482511  0.584192  0.539962  0.207633  0.726190  0.918482  0.663539   \n",
       "6142     0.378339  0.445871  0.502754  0.246252  0.285714  0.940499  0.903375   \n",
       "5358     0.000000  0.000000  0.249445  0.089959  0.190476  0.790687  0.773395   \n",
       "\n",
       "          RAB_max   RAB_min  mean_trans_profit  credit_ratio  \\\n",
       "loan_id                                                        \n",
       "5895     0.763137  0.628545           0.660156      0.146538   \n",
       "7122     0.977758  0.570286           0.572825      0.191233   \n",
       "6173     0.981118  0.819110           0.764744      0.027116   \n",
       "6142     0.999866  0.807015           0.674889      0.093363   \n",
       "5358     0.790687  0.790687           0.471897      0.299821   \n",
       "\n",
       "         withdrawal_ratio  ratio_CC  ratio_RAB  owner_count  disponent_count  \\\n",
       "loan_id                                                                        \n",
       "5895             0.853462  0.175218   0.546415          0.0              0.0   \n",
       "7122             0.808767  0.174929   0.754731          0.0              0.0   \n",
       "6173             0.972884  0.011585   0.887108          0.0              0.0   \n",
       "6142             0.906637  0.022770   0.865191          1.0              1.0   \n",
       "5358             0.700179  0.285526   0.359231          0.0              0.0   \n",
       "\n",
       "         salary_from_month_records  withdrawal_to_credit_month_ratio  \\\n",
       "loan_id                                                                \n",
       "5895                      0.660156                          0.845588   \n",
       "7122                      0.572825                          0.802215   \n",
       "6173                      0.764744                          0.895074   \n",
       "6142                      0.674889                          0.876517   \n",
       "5358                      0.471897                          0.736241   \n",
       "\n",
       "         actual_month_income  actual_income_to_payments_ratio  \\\n",
       "loan_id                                                         \n",
       "5895                0.659778                         0.413987   \n",
       "7122                0.573668                         0.281695   \n",
       "6173                0.773518                         0.339933   \n",
       "6142                0.687726                         0.336087   \n",
       "5358                0.473703                         0.267758   \n",
       "\n",
       "         log_sanctions_std  log_max_value_in_account_to_loan_ratio  \n",
       "loan_id                                                             \n",
       "5895                   0.0                                0.246558  \n",
       "7122                   0.0                                0.103447  \n",
       "6173                   0.0                                0.103590  \n",
       "6142                   0.0                                0.101718  \n",
       "5358                   0.0                                0.220466  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# We now remove the Y column with NaNs\n",
    "test_dataset = test_dataset.iloc[:, 0:STATUS_COL]\n",
    "\n",
    "display(test_dataset.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>Predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>5895</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>7122</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>6173</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>6142</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>5358</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>349</td>\n",
       "      <td>4989</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>350</td>\n",
       "      <td>5221</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>351</td>\n",
       "      <td>6402</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>352</td>\n",
       "      <td>5346</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>353</td>\n",
       "      <td>6748</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>354 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Id  Predicted\n",
       "0    5895        1.0\n",
       "1    7122        1.0\n",
       "2    6173        1.0\n",
       "3    6142        1.0\n",
       "4    5358        1.0\n",
       "..    ...        ...\n",
       "349  4989       -1.0\n",
       "350  5221       -1.0\n",
       "351  6402        1.0\n",
       "352  5346       -1.0\n",
       "353  6748        1.0\n",
       "\n",
       "[354 rows x 2 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "final_df = pd.DataFrame()\n",
    "\n",
    "if USE_PCA:\n",
    "    # Using train PCA and classifying\n",
    "    scaler=StandardScaler()\n",
    "    X_test_scaled=scaler.fit_transform(test_dataset)\n",
    "    predictions_df = pd.DataFrame(data = pca.transform(X_test_scaled),\n",
    "                                  index=test_dataset.reset_index()['loan_id'])\n",
    "    display(predictions_df)\n",
    "\n",
    "    predictions_df['Predicted'] = classifier.predict(predictions_df)\n",
    "    final_df = predictions_df.reset_index()\\\n",
    "                    [['loan_id', 'Predicted']]\\\n",
    "                    .rename(columns={\n",
    "                        'loan_id': 'Id'\n",
    "                    })\n",
    "    \n",
    "else:\n",
    "    final_df = test_dataset.copy()\n",
    "    final_df['Predicted'] = classifier.predict(final_df)\n",
    "    final_df = final_df.reset_index()\\\n",
    "                        [['loan_id', 'Predicted']]\\\n",
    "                        .rename(columns={\n",
    "                            'loan_id': 'Id'\n",
    "                        })\\\n",
    "                        .drop_duplicates()\n",
    "\n",
    "final_df.loc[final_df[\"Predicted\"]== 0.0, \"Predicted\"] = -1\n",
    "display(final_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Outputting predictions to .csv\n",
    "# CHANGE FILE NAME TO PRESERVE DIFFERENT INSTANCES\n",
    "utils.write_df_to_csv(final_df, 'predictions', 'prediction.csv')"
   ]
  },
  {
   "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}