dirty_woe_iv.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qb9mY_HmN913"
   },
   "source": [
    "# Домашнее задание для Junior DS по разработке PD модели"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "NbqnB_V8N915"
   },
   "source": [
    "<b>Задание:</b> Необходиом разработать модель предсказания дефолта. Даны финансовые данные 32 395 компаний (выручка, активы, пассивы и т.д). Результат будет оцениваться на скрытой выборке из 200 компаний по метрике качества accuracy_score\n",
    "\n",
    "<b>Цель задания</b> научиться / показать знания в разработке моделей оценки риска - probability of default. Задание приближено к реальному, данные синтетические. Лидерборд для оценки собственного решения и геймификации процесса обучения. Первое место в лидерборде не явлется целью. Цель - правильно разарботанная модель, удовлетворяющая требованиям банковской системы. В случае возникновения вопросов, сотрудники блока Риски готовы проконсультировать и дать обратную связь на модель (телеграм чат).\n",
    "\n",
    "<b>Метрика качества</b> accuracy_score вычисляется, как ( True positive + True negative ) / Total. Если сабмит-файл, не правильного формата или неверное количество строк, метрика будет = 0."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "EJ9_SMqyN916"
   },
   "source": [
    "### Импортируем библиотеки"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {
    "id": "Vo4g-4Y5N917"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn.metrics import f1_score\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import accuracy_score\n",
    "pd.set_option('display.max_rows', 500)\n",
    "pd.set_option('display.max_columns', 500)\n",
    "import warnings\n",
    "from pandas.core.common import SettingWithCopyWarning\n",
    "warnings.simplefilter(action=\"ignore\", category=SettingWithCopyWarning)\n",
    "warnings.simplefilter(action='ignore', category=FutureWarning)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4k8ndBMHN92A"
   },
   "source": [
    "### Подгружаем данные"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {
    "id": "XwbL4T6rN92G"
   },
   "outputs": [],
   "source": [
    "train_df = pd.read_csv('https://raw.githubusercontent.com/BKHV/risk_models/master/data/PD-data-train.csv', sep=';')\n",
    "test_df = pd.read_csv('https://raw.githubusercontent.com/BKHV/risk_models/master/data/PD-data-test.csv', sep=';')\n",
    "desc_df = pd.read_csv('https://raw.githubusercontent.com/BKHV/risk_models/master/data/PD-data-desc.csv', sep=';')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "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>record_id</th>\n",
       "      <th>ar_revenue</th>\n",
       "      <th>ar_total_expenses</th>\n",
       "      <th>ar_sale_cost</th>\n",
       "      <th>ar_selling_expenses</th>\n",
       "      <th>ar_management_expenses</th>\n",
       "      <th>ar_sale_profit</th>\n",
       "      <th>ar_balance_of_rvns_and_expns</th>\n",
       "      <th>ar_profit_before_tax</th>\n",
       "      <th>ar_taxes</th>\n",
       "      <th>ar_other_profit_and_losses</th>\n",
       "      <th>ar_net_profit</th>\n",
       "      <th>ab_immobilized_assets</th>\n",
       "      <th>ab_mobile_current_assets</th>\n",
       "      <th>ab_inventory</th>\n",
       "      <th>ab_accounts_receivable</th>\n",
       "      <th>ab_other_current_assets</th>\n",
       "      <th>ab_cash_and_securities</th>\n",
       "      <th>ab_losses</th>\n",
       "      <th>ab_own_capital</th>\n",
       "      <th>ab_borrowed_capital</th>\n",
       "      <th>ab_long_term_liabilities</th>\n",
       "      <th>ab_short_term_borrowing</th>\n",
       "      <th>ab_accounts_payable</th>\n",
       "      <th>ab_other_borrowings</th>\n",
       "      <th>bus_age</th>\n",
       "      <th>ogrn_age</th>\n",
       "      <th>adr_actual_age</th>\n",
       "      <th>head_actual_age</th>\n",
       "      <th>cap_actual_age</th>\n",
       "      <th>ul_staff_range</th>\n",
       "      <th>ul_capital_sum</th>\n",
       "      <th>ul_founders_cnt</th>\n",
       "      <th>ul_branch_cnt</th>\n",
       "      <th>ul_strategic_flg</th>\n",
       "      <th>ul_systematizing_flg</th>\n",
       "      <th>default_12m</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7</td>\n",
       "      <td>52996000.0</td>\n",
       "      <td>48344000.0</td>\n",
       "      <td>48344000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4652000.0</td>\n",
       "      <td>-2608000.0</td>\n",
       "      <td>2044000.0</td>\n",
       "      <td>846000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1198000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7240000.0</td>\n",
       "      <td>4988000.0</td>\n",
       "      <td>2252000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>255000.0</td>\n",
       "      <td>1198000.0</td>\n",
       "      <td>663000.0</td>\n",
       "      <td>15872000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11349000.0</td>\n",
       "      <td>4523000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>198</td>\n",
       "      <td>136</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>50000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>13</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>256</td>\n",
       "      <td>135</td>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>33</td>\n",
       "      <td>33</td>\n",
       "      <td>28</td>\n",
       "      <td>28</td>\n",
       "      <td>28</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22</td>\n",
       "      <td>32049000.0</td>\n",
       "      <td>29568000.0</td>\n",
       "      <td>29568000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2481000.0</td>\n",
       "      <td>-260000.0</td>\n",
       "      <td>2221000.0</td>\n",
       "      <td>302000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1919000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4245000.0</td>\n",
       "      <td>4082000.0</td>\n",
       "      <td>163000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>25000.0</td>\n",
       "      <td>1919000.0</td>\n",
       "      <td>2219000.0</td>\n",
       "      <td>3652000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3652000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>185</td>\n",
       "      <td>139</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>300000.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23</td>\n",
       "      <td>16458000.0</td>\n",
       "      <td>14935000.0</td>\n",
       "      <td>14935000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1523000.0</td>\n",
       "      <td>-274000.0</td>\n",
       "      <td>1249000.0</td>\n",
       "      <td>538000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>711000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8950000.0</td>\n",
       "      <td>8606000.0</td>\n",
       "      <td>344000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>492000.0</td>\n",
       "      <td>711000.0</td>\n",
       "      <td>12439000.0</td>\n",
       "      <td>1279000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1279000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>220</td>\n",
       "      <td>138</td>\n",
       "      <td>27</td>\n",
       "      <td>27</td>\n",
       "      <td>27</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>350000.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>266</td>\n",
       "      <td>27</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>35000.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>37</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>31</td>\n",
       "      <td>31</td>\n",
       "      <td>30</td>\n",
       "      <td>30</td>\n",
       "      <td>30</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>38</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>27</td>\n",
       "      <td>27</td>\n",
       "      <td>26</td>\n",
       "      <td>26</td>\n",
       "      <td>26</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>40</td>\n",
       "      <td>125432000.0</td>\n",
       "      <td>122655000.0</td>\n",
       "      <td>122655000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2777000.0</td>\n",
       "      <td>-63000.0</td>\n",
       "      <td>2714000.0</td>\n",
       "      <td>579000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2135000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24680000.0</td>\n",
       "      <td>8263000.0</td>\n",
       "      <td>16417000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>106000.0</td>\n",
       "      <td>2135000.0</td>\n",
       "      <td>11267000.0</td>\n",
       "      <td>11912000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13000.0</td>\n",
       "      <td>11899000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>170</td>\n",
       "      <td>132</td>\n",
       "      <td>15</td>\n",
       "      <td>15</td>\n",
       "      <td>15</td>\n",
       "      <td>(100-500]</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   record_id   ar_revenue  ar_total_expenses  ar_sale_cost  \\\n",
       "0          7   52996000.0         48344000.0    48344000.0   \n",
       "1         13          NaN                NaN           NaN   \n",
       "2         18          NaN                NaN           NaN   \n",
       "3         22   32049000.0         29568000.0    29568000.0   \n",
       "4         23   16458000.0         14935000.0    14935000.0   \n",
       "5         24          NaN                NaN           NaN   \n",
       "6         30          NaN                NaN           NaN   \n",
       "7         37          NaN                NaN           NaN   \n",
       "8         38          NaN                NaN           NaN   \n",
       "9         40  125432000.0        122655000.0   122655000.0   \n",
       "\n",
       "   ar_selling_expenses  ar_management_expenses  ar_sale_profit  \\\n",
       "0                  0.0                     0.0       4652000.0   \n",
       "1                  NaN                     NaN             NaN   \n",
       "2                  NaN                     NaN             NaN   \n",
       "3                  0.0                     0.0       2481000.0   \n",
       "4                  0.0                     0.0       1523000.0   \n",
       "5                  NaN                     NaN             NaN   \n",
       "6                  NaN                     NaN             NaN   \n",
       "7                  NaN                     NaN             NaN   \n",
       "8                  NaN                     NaN             NaN   \n",
       "9                  0.0                     0.0       2777000.0   \n",
       "\n",
       "   ar_balance_of_rvns_and_expns  ar_profit_before_tax  ar_taxes  \\\n",
       "0                    -2608000.0             2044000.0  846000.0   \n",
       "1                           NaN                   NaN       NaN   \n",
       "2                           NaN                   NaN       NaN   \n",
       "3                     -260000.0             2221000.0  302000.0   \n",
       "4                     -274000.0             1249000.0  538000.0   \n",
       "5                           NaN                   NaN       NaN   \n",
       "6                           NaN                   NaN       NaN   \n",
       "7                           NaN                   NaN       NaN   \n",
       "8                           NaN                   NaN       NaN   \n",
       "9                      -63000.0             2714000.0  579000.0   \n",
       "\n",
       "   ar_other_profit_and_losses  ar_net_profit  ab_immobilized_assets  \\\n",
       "0                         0.0      1198000.0                    0.0   \n",
       "1                         NaN            NaN                    NaN   \n",
       "2                         NaN            NaN                    NaN   \n",
       "3                         0.0      1919000.0                    0.0   \n",
       "4                         0.0       711000.0                    0.0   \n",
       "5                         NaN            NaN                    NaN   \n",
       "6                         NaN            NaN                    NaN   \n",
       "7                         NaN            NaN                    NaN   \n",
       "8                         NaN            NaN                    NaN   \n",
       "9                         0.0      2135000.0                    0.0   \n",
       "\n",
       "   ab_mobile_current_assets  ab_inventory  ab_accounts_receivable  \\\n",
       "0                 7240000.0     4988000.0               2252000.0   \n",
       "1                       NaN           NaN                     NaN   \n",
       "2                       NaN           NaN                     NaN   \n",
       "3                 4245000.0     4082000.0                163000.0   \n",
       "4                 8950000.0     8606000.0                344000.0   \n",
       "5                       NaN           NaN                     NaN   \n",
       "6                       NaN           NaN                     NaN   \n",
       "7                       NaN           NaN                     NaN   \n",
       "8                       NaN           NaN                     NaN   \n",
       "9                24680000.0     8263000.0              16417000.0   \n",
       "\n",
       "   ab_other_current_assets  ab_cash_and_securities  ab_losses  ab_own_capital  \\\n",
       "0                      0.0                255000.0  1198000.0        663000.0   \n",
       "1                      NaN                     NaN        NaN             NaN   \n",
       "2                      NaN                     NaN        NaN             NaN   \n",
       "3                      0.0                 25000.0  1919000.0       2219000.0   \n",
       "4                      0.0                492000.0   711000.0      12439000.0   \n",
       "5                      NaN                     NaN        NaN             NaN   \n",
       "6                      NaN                     NaN        NaN             NaN   \n",
       "7                      NaN                     NaN        NaN             NaN   \n",
       "8                      NaN                     NaN        NaN             NaN   \n",
       "9                      0.0                106000.0  2135000.0      11267000.0   \n",
       "\n",
       "   ab_borrowed_capital  ab_long_term_liabilities  ab_short_term_borrowing  \\\n",
       "0           15872000.0                       0.0               11349000.0   \n",
       "1                  NaN                       NaN                      NaN   \n",
       "2                  NaN                       NaN                      NaN   \n",
       "3            3652000.0                       0.0                3652000.0   \n",
       "4            1279000.0                       0.0                      0.0   \n",
       "5                  NaN                       NaN                      NaN   \n",
       "6                  NaN                       NaN                      NaN   \n",
       "7                  NaN                       NaN                      NaN   \n",
       "8                  NaN                       NaN                      NaN   \n",
       "9           11912000.0                       0.0                  13000.0   \n",
       "\n",
       "   ab_accounts_payable  ab_other_borrowings  bus_age  ogrn_age  \\\n",
       "0            4523000.0                  0.0      198       136   \n",
       "1                  NaN                  NaN      256       135   \n",
       "2                  NaN                  NaN       33        33   \n",
       "3                  0.0                  0.0      185       139   \n",
       "4            1279000.0                  0.0      220       138   \n",
       "5                  NaN                  NaN      100       100   \n",
       "6                  NaN                  NaN      266        27   \n",
       "7                  NaN                  NaN       31        31   \n",
       "8                  NaN                  NaN       27        27   \n",
       "9           11899000.0                  0.0      170       132   \n",
       "\n",
       "   adr_actual_age  head_actual_age  cap_actual_age ul_staff_range  \\\n",
       "0               3                3               3        [1-100]   \n",
       "1              12               12              12        [1-100]   \n",
       "2              28               28              28        [1-100]   \n",
       "3               8                8               8        [1-100]   \n",
       "4              27               27              27        [1-100]   \n",
       "5               5                5               5        [1-100]   \n",
       "6               7                7               7        [1-100]   \n",
       "7              30               30              30        [1-100]   \n",
       "8              26               26              26        [1-100]   \n",
       "9              15               15              15      (100-500]   \n",
       "\n",
       "   ul_capital_sum  ul_founders_cnt  ul_branch_cnt  ul_strategic_flg  \\\n",
       "0         50000.0                2              0                 0   \n",
       "1         10000.0                5              0                 0   \n",
       "2         10000.0                2              0                 0   \n",
       "3        300000.0                1              0                 0   \n",
       "4         10000.0                2              0                 0   \n",
       "5        350000.0                1              0                 0   \n",
       "6         35000.0                3              0                 0   \n",
       "7         10000.0                2              0                 0   \n",
       "8         10000.0                1              0                 0   \n",
       "9         10000.0                2              0                 0   \n",
       "\n",
       "   ul_systematizing_flg  default_12m  \n",
       "0                     0            0  \n",
       "1                     0            0  \n",
       "2                     0            0  \n",
       "3                     0            0  \n",
       "4                     0            0  \n",
       "5                     0            0  \n",
       "6                     0            0  \n",
       "7                     0            0  \n",
       "8                     0            0  \n",
       "9                     0            0  "
      ]
     },
     "execution_count": 236,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "PQgDXmx-QPs_"
   },
   "source": [
    "Описание данных"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "15_jhk2bQPs_",
    "outputId": "06d6ffe6-50c0-4b54-a762-01d107203536"
   },
   "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>field</th>\n",
       "      <th>desc_eng</th>\n",
       "      <th>desc_rus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>Sales revenue (at the end of the last year)</td>\n",
       "      <td>Выручка от реализации (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>Total expenses for the last year</td>\n",
       "      <td>Общие расходы за последний год</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>Cost of sales (at the end of the last year)</td>\n",
       "      <td>Себестоимость продаж (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ar_selling_expenses</td>\n",
       "      <td>Selling expenses (at the end of the last year)</td>\n",
       "      <td>Коммерческие расходы (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ar_management_expenses</td>\n",
       "      <td>Administrative expenses (at the end of the las...</td>\n",
       "      <td>Управленческие расходы (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>ar_sale_profit</td>\n",
       "      <td>Profit from sales (at the end of the last year)</td>\n",
       "      <td>Прибыль от реализации (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ar_balance_of_rvns_and_expns</td>\n",
       "      <td>Balance of income and expenses (at the end of ...</td>\n",
       "      <td>Баланс доходов и расходов (на конец последнего...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>ar_profit_before_tax</td>\n",
       "      <td>Gross Profit (end of last year)</td>\n",
       "      <td>Валовая прибыль (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>ar_taxes</td>\n",
       "      <td>Current income tax (at the end of the last year)</td>\n",
       "      <td>Текущий на налог на прибыль (на конец последне...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ar_other_profit_and_losses</td>\n",
       "      <td>Other income and expenses (at the end of the l...</td>\n",
       "      <td>Прочие доходы и расходы (на конец последнего г...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>ar_net_profit</td>\n",
       "      <td>Net profit (at the end of the last year)</td>\n",
       "      <td>Чистая прибыль (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>ab_immobilized_assets</td>\n",
       "      <td>Fixed assets for the last year</td>\n",
       "      <td>Основные внеоборотные средства за последний год</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>Current assets (at the end of the last year)</td>\n",
       "      <td>Оборотные активы (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>ab_inventory</td>\n",
       "      <td>Stocks (at the end of the last year)</td>\n",
       "      <td>Запасы (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>Accounts receivable (at the end of the last year)</td>\n",
       "      <td>Дебиторская задолженность (на конец последнего...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>ab_other_current_assets</td>\n",
       "      <td>Other current assets (at the end of the last y...</td>\n",
       "      <td>Прочие оборотные активы (на конец последнего г...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>ab_cash_and_securities</td>\n",
       "      <td>Cash and cash equivalents (at the end of the l...</td>\n",
       "      <td>Денежные средства и денежные эквиваленты (на к...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>ab_losses</td>\n",
       "      <td>Retained earnings (at the end of the last year)</td>\n",
       "      <td>Нераспределенная прибыль (на конец последнего ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>ab_own_capital</td>\n",
       "      <td>Capital (at the end of the last year)</td>\n",
       "      <td>Капитал (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>Total debt (end of last year)</td>\n",
       "      <td>Общие долговые обязательства (на конец последн...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>ab_long_term_liabilities</td>\n",
       "      <td>Long-term liabilities (at the end of the last ...</td>\n",
       "      <td>Долгосрочный обязательства (на конец последнег...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>ab_short_term_borrowing</td>\n",
       "      <td>Short-term borrowed funds (at the end of the l...</td>\n",
       "      <td>Краткосрочные заемные средства (на конец после...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>Accounts payable for the last year</td>\n",
       "      <td>Кредиторская задодженность за последний год</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>ab_other_borrowings</td>\n",
       "      <td>Other liabilities (at the end of the last year)</td>\n",
       "      <td>Прочие обязательства (на конец последнего года)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>bus_age</td>\n",
       "      <td>Business duration</td>\n",
       "      <td>Срок ведения бизнеса</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>ogrn_age</td>\n",
       "      <td>Term from the moment the PSRN was assigned</td>\n",
       "      <td>Срок с момента присваивания ОГРН</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>adr_actual_age</td>\n",
       "      <td>Term from the date of registration of the lega...</td>\n",
       "      <td>Срок с момента регистрации юридического адреса</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>head_actual_age</td>\n",
       "      <td>Term since the appointment of the head</td>\n",
       "      <td>Срок с момента назначения руководителя</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>cap_actual_age</td>\n",
       "      <td>Term from the moment of capital installation</td>\n",
       "      <td>Срок с момента установки капитала</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>ul_staff_range</td>\n",
       "      <td>The number of employees</td>\n",
       "      <td>Количество сотрудников</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>ul_capital_sum</td>\n",
       "      <td>Capital amount</td>\n",
       "      <td>Величина капитала</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>ul_founders_cnt</td>\n",
       "      <td>Number of shareholders</td>\n",
       "      <td>Количество акционеров</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>ul_branch_cnt</td>\n",
       "      <td>Number of branches</td>\n",
       "      <td>Количество филиалов</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>ul_strategic_flg</td>\n",
       "      <td>A company of strategic importance</td>\n",
       "      <td>Компания стратегического значения</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>ul_systematizing_flg</td>\n",
       "      <td>Backbone company</td>\n",
       "      <td>Системообразующая компания</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           field  \\\n",
       "0                     ar_revenue   \n",
       "1              ar_total_expenses   \n",
       "2                   ar_sale_cost   \n",
       "3            ar_selling_expenses   \n",
       "4         ar_management_expenses   \n",
       "5                 ar_sale_profit   \n",
       "6   ar_balance_of_rvns_and_expns   \n",
       "7           ar_profit_before_tax   \n",
       "8                       ar_taxes   \n",
       "9     ar_other_profit_and_losses   \n",
       "10                 ar_net_profit   \n",
       "11         ab_immobilized_assets   \n",
       "12      ab_mobile_current_assets   \n",
       "13                  ab_inventory   \n",
       "14        ab_accounts_receivable   \n",
       "15       ab_other_current_assets   \n",
       "16        ab_cash_and_securities   \n",
       "17                     ab_losses   \n",
       "18                ab_own_capital   \n",
       "19           ab_borrowed_capital   \n",
       "20      ab_long_term_liabilities   \n",
       "21       ab_short_term_borrowing   \n",
       "22           ab_accounts_payable   \n",
       "23           ab_other_borrowings   \n",
       "24                       bus_age   \n",
       "25                      ogrn_age   \n",
       "26                adr_actual_age   \n",
       "27               head_actual_age   \n",
       "28                cap_actual_age   \n",
       "29                ul_staff_range   \n",
       "30                ul_capital_sum   \n",
       "31               ul_founders_cnt   \n",
       "32                 ul_branch_cnt   \n",
       "33              ul_strategic_flg   \n",
       "34          ul_systematizing_flg   \n",
       "\n",
       "                                             desc_eng  \\\n",
       "0         Sales revenue (at the end of the last year)   \n",
       "1                    Total expenses for the last year   \n",
       "2         Cost of sales (at the end of the last year)   \n",
       "3      Selling expenses (at the end of the last year)   \n",
       "4   Administrative expenses (at the end of the las...   \n",
       "5     Profit from sales (at the end of the last year)   \n",
       "6   Balance of income and expenses (at the end of ...   \n",
       "7                     Gross Profit (end of last year)   \n",
       "8    Current income tax (at the end of the last year)   \n",
       "9   Other income and expenses (at the end of the l...   \n",
       "10           Net profit (at the end of the last year)   \n",
       "11                     Fixed assets for the last year   \n",
       "12       Current assets (at the end of the last year)   \n",
       "13               Stocks (at the end of the last year)   \n",
       "14  Accounts receivable (at the end of the last year)   \n",
       "15  Other current assets (at the end of the last y...   \n",
       "16  Cash and cash equivalents (at the end of the l...   \n",
       "17    Retained earnings (at the end of the last year)   \n",
       "18              Capital (at the end of the last year)   \n",
       "19                      Total debt (end of last year)   \n",
       "20  Long-term liabilities (at the end of the last ...   \n",
       "21  Short-term borrowed funds (at the end of the l...   \n",
       "22                 Accounts payable for the last year   \n",
       "23    Other liabilities (at the end of the last year)   \n",
       "24                                  Business duration   \n",
       "25         Term from the moment the PSRN was assigned   \n",
       "26  Term from the date of registration of the lega...   \n",
       "27             Term since the appointment of the head   \n",
       "28       Term from the moment of capital installation   \n",
       "29                            The number of employees   \n",
       "30                                     Capital amount   \n",
       "31                             Number of shareholders   \n",
       "32                                 Number of branches   \n",
       "33                  A company of strategic importance   \n",
       "34                                   Backbone company   \n",
       "\n",
       "                                             desc_rus  \n",
       "0    Выручка от реализации (на конец последнего года)  \n",
       "1                      Общие расходы за последний год  \n",
       "2     Себестоимость продаж (на конец последнего года)  \n",
       "3     Коммерческие расходы (на конец последнего года)  \n",
       "4   Управленческие расходы (на конец последнего года)  \n",
       "5    Прибыль от реализации (на конец последнего года)  \n",
       "6   Баланс доходов и расходов (на конец последнего...  \n",
       "7          Валовая прибыль (на конец последнего года)  \n",
       "8   Текущий на налог на прибыль (на конец последне...  \n",
       "9   Прочие доходы и расходы (на конец последнего г...  \n",
       "10          Чистая прибыль (на конец последнего года)  \n",
       "11    Основные внеоборотные средства за последний год  \n",
       "12        Оборотные активы (на конец последнего года)  \n",
       "13                  Запасы (на конец последнего года)  \n",
       "14  Дебиторская задолженность (на конец последнего...  \n",
       "15  Прочие оборотные активы (на конец последнего г...  \n",
       "16  Денежные средства и денежные эквиваленты (на к...  \n",
       "17  Нераспределенная прибыль (на конец последнего ...  \n",
       "18                 Капитал (на конец последнего года)  \n",
       "19  Общие долговые обязательства (на конец последн...  \n",
       "20  Долгосрочный обязательства (на конец последнег...  \n",
       "21  Краткосрочные заемные средства (на конец после...  \n",
       "22        Кредиторская задодженность за последний год  \n",
       "23    Прочие обязательства (на конец последнего года)  \n",
       "24                               Срок ведения бизнеса  \n",
       "25                   Срок с момента присваивания ОГРН  \n",
       "26     Срок с момента регистрации юридического адреса  \n",
       "27             Срок с момента назначения руководителя  \n",
       "28                  Срок с момента установки капитала  \n",
       "29                             Количество сотрудников  \n",
       "30                                  Величина капитала  \n",
       "31                              Количество акционеров  \n",
       "32                                Количество филиалов  \n",
       "33                  Компания стратегического значения  \n",
       "34                         Системообразующая компания  "
      ]
     },
     "execution_count": 237,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "desc_df.head(50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Чтобы разобраться в признаках, давайте углубимся в предметную область. В задании сказано, что мы работаем не с ИП, а с ООО, поэтому будем обращать особое внимание на порядок оформления ООО. Для начала участники ООО названчают руководителя и вместе с этим устанавливает капитал и регистрируют юридический адрес,тот адрес, по которому ведомства будут связываться с компанией: присылать запросы, письма, требования и уведомления. Затем необходимо выбрать коды видов деятельности: название и цифровой код для дальнейшего обращения в налоговую службу. До регистрации ООО нужно определиться с системой налогооблежния и собрать все учредительные документы. В течение 10 дней после передачи документов в налоговую службу ООО должен быть присвоен ОГРН - основной государственный регистрационный номер. \n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Исходя из вышесказанного, посмотрим на наши признаки: \n",
    "\n",
    "В наших данных есть признаки с разными названиями, но с одним и тем же назначением, например: \n",
    "1. adr_actual_age, head_actual_age, cap_actual_age  \n",
    "При регистрации юридического адреса необходимо указать руководителя и установить капитал, поэтому все эти признаки обозначают одну и ту же величину. \n",
    "\n",
    "2. ogrn_age и bus_age. При присваении ОГРН владелец бизнеса может официально начать вести свою дейтельность, поэтому эти признаки обозначают одну и ту же величину, за исключением, редких случаев. \n",
    "\n",
    "3. ar_net_profit и ar_profit_before_tax будут определенно коррелировать. Так как они отличаются только тем, что одна прибыль с учетом налога, а другая нет. Поэтому можем удалить один из этих признаков. \n",
    "4. ar_total_expenses и ar_sale_cost также могут коррелировать, если у нас фиксированы расходы, не включающие себестоимость продаж. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "An6p1hDkN92J"
   },
   "source": [
    "Семплы и размеры выборок"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 211
    },
    "id": "FDqOr3Du8C1Y",
    "outputId": "5fa6016a-6113-459c-a67e-9fd3d2fff623"
   },
   "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>record_id</th>\n",
       "      <th>ar_revenue</th>\n",
       "      <th>ar_total_expenses</th>\n",
       "      <th>ar_sale_cost</th>\n",
       "      <th>ar_selling_expenses</th>\n",
       "      <th>ar_management_expenses</th>\n",
       "      <th>ar_sale_profit</th>\n",
       "      <th>ar_balance_of_rvns_and_expns</th>\n",
       "      <th>ar_profit_before_tax</th>\n",
       "      <th>ar_taxes</th>\n",
       "      <th>ar_other_profit_and_losses</th>\n",
       "      <th>ar_net_profit</th>\n",
       "      <th>ab_immobilized_assets</th>\n",
       "      <th>ab_mobile_current_assets</th>\n",
       "      <th>ab_inventory</th>\n",
       "      <th>ab_accounts_receivable</th>\n",
       "      <th>ab_other_current_assets</th>\n",
       "      <th>ab_cash_and_securities</th>\n",
       "      <th>ab_losses</th>\n",
       "      <th>ab_own_capital</th>\n",
       "      <th>ab_borrowed_capital</th>\n",
       "      <th>ab_long_term_liabilities</th>\n",
       "      <th>ab_short_term_borrowing</th>\n",
       "      <th>ab_accounts_payable</th>\n",
       "      <th>ab_other_borrowings</th>\n",
       "      <th>bus_age</th>\n",
       "      <th>ogrn_age</th>\n",
       "      <th>adr_actual_age</th>\n",
       "      <th>head_actual_age</th>\n",
       "      <th>cap_actual_age</th>\n",
       "      <th>ul_staff_range</th>\n",
       "      <th>ul_capital_sum</th>\n",
       "      <th>ul_founders_cnt</th>\n",
       "      <th>ul_branch_cnt</th>\n",
       "      <th>ul_strategic_flg</th>\n",
       "      <th>ul_systematizing_flg</th>\n",
       "      <th>default_12m</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7</td>\n",
       "      <td>52996000.0</td>\n",
       "      <td>48344000.0</td>\n",
       "      <td>48344000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4652000.0</td>\n",
       "      <td>-2608000.0</td>\n",
       "      <td>2044000.0</td>\n",
       "      <td>846000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1198000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7240000.0</td>\n",
       "      <td>4988000.0</td>\n",
       "      <td>2252000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>255000.0</td>\n",
       "      <td>1198000.0</td>\n",
       "      <td>663000.0</td>\n",
       "      <td>15872000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11349000.0</td>\n",
       "      <td>4523000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>198</td>\n",
       "      <td>136</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>50000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>13</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>256</td>\n",
       "      <td>135</td>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>33</td>\n",
       "      <td>33</td>\n",
       "      <td>28</td>\n",
       "      <td>28</td>\n",
       "      <td>28</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22</td>\n",
       "      <td>32049000.0</td>\n",
       "      <td>29568000.0</td>\n",
       "      <td>29568000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2481000.0</td>\n",
       "      <td>-260000.0</td>\n",
       "      <td>2221000.0</td>\n",
       "      <td>302000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1919000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4245000.0</td>\n",
       "      <td>4082000.0</td>\n",
       "      <td>163000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>25000.0</td>\n",
       "      <td>1919000.0</td>\n",
       "      <td>2219000.0</td>\n",
       "      <td>3652000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3652000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>185</td>\n",
       "      <td>139</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>300000.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23</td>\n",
       "      <td>16458000.0</td>\n",
       "      <td>14935000.0</td>\n",
       "      <td>14935000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1523000.0</td>\n",
       "      <td>-274000.0</td>\n",
       "      <td>1249000.0</td>\n",
       "      <td>538000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>711000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8950000.0</td>\n",
       "      <td>8606000.0</td>\n",
       "      <td>344000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>492000.0</td>\n",
       "      <td>711000.0</td>\n",
       "      <td>12439000.0</td>\n",
       "      <td>1279000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1279000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>220</td>\n",
       "      <td>138</td>\n",
       "      <td>27</td>\n",
       "      <td>27</td>\n",
       "      <td>27</td>\n",
       "      <td>[1-100]</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   record_id  ar_revenue  ar_total_expenses  ar_sale_cost  \\\n",
       "0          7  52996000.0         48344000.0    48344000.0   \n",
       "1         13         NaN                NaN           NaN   \n",
       "2         18         NaN                NaN           NaN   \n",
       "3         22  32049000.0         29568000.0    29568000.0   \n",
       "4         23  16458000.0         14935000.0    14935000.0   \n",
       "\n",
       "   ar_selling_expenses  ar_management_expenses  ar_sale_profit  \\\n",
       "0                  0.0                     0.0       4652000.0   \n",
       "1                  NaN                     NaN             NaN   \n",
       "2                  NaN                     NaN             NaN   \n",
       "3                  0.0                     0.0       2481000.0   \n",
       "4                  0.0                     0.0       1523000.0   \n",
       "\n",
       "   ar_balance_of_rvns_and_expns  ar_profit_before_tax  ar_taxes  \\\n",
       "0                    -2608000.0             2044000.0  846000.0   \n",
       "1                           NaN                   NaN       NaN   \n",
       "2                           NaN                   NaN       NaN   \n",
       "3                     -260000.0             2221000.0  302000.0   \n",
       "4                     -274000.0             1249000.0  538000.0   \n",
       "\n",
       "   ar_other_profit_and_losses  ar_net_profit  ab_immobilized_assets  \\\n",
       "0                         0.0      1198000.0                    0.0   \n",
       "1                         NaN            NaN                    NaN   \n",
       "2                         NaN            NaN                    NaN   \n",
       "3                         0.0      1919000.0                    0.0   \n",
       "4                         0.0       711000.0                    0.0   \n",
       "\n",
       "   ab_mobile_current_assets  ab_inventory  ab_accounts_receivable  \\\n",
       "0                 7240000.0     4988000.0               2252000.0   \n",
       "1                       NaN           NaN                     NaN   \n",
       "2                       NaN           NaN                     NaN   \n",
       "3                 4245000.0     4082000.0                163000.0   \n",
       "4                 8950000.0     8606000.0                344000.0   \n",
       "\n",
       "   ab_other_current_assets  ab_cash_and_securities  ab_losses  ab_own_capital  \\\n",
       "0                      0.0                255000.0  1198000.0        663000.0   \n",
       "1                      NaN                     NaN        NaN             NaN   \n",
       "2                      NaN                     NaN        NaN             NaN   \n",
       "3                      0.0                 25000.0  1919000.0       2219000.0   \n",
       "4                      0.0                492000.0   711000.0      12439000.0   \n",
       "\n",
       "   ab_borrowed_capital  ab_long_term_liabilities  ab_short_term_borrowing  \\\n",
       "0           15872000.0                       0.0               11349000.0   \n",
       "1                  NaN                       NaN                      NaN   \n",
       "2                  NaN                       NaN                      NaN   \n",
       "3            3652000.0                       0.0                3652000.0   \n",
       "4            1279000.0                       0.0                      0.0   \n",
       "\n",
       "   ab_accounts_payable  ab_other_borrowings  bus_age  ogrn_age  \\\n",
       "0            4523000.0                  0.0      198       136   \n",
       "1                  NaN                  NaN      256       135   \n",
       "2                  NaN                  NaN       33        33   \n",
       "3                  0.0                  0.0      185       139   \n",
       "4            1279000.0                  0.0      220       138   \n",
       "\n",
       "   adr_actual_age  head_actual_age  cap_actual_age ul_staff_range  \\\n",
       "0               3                3               3        [1-100]   \n",
       "1              12               12              12        [1-100]   \n",
       "2              28               28              28        [1-100]   \n",
       "3               8                8               8        [1-100]   \n",
       "4              27               27              27        [1-100]   \n",
       "\n",
       "   ul_capital_sum  ul_founders_cnt  ul_branch_cnt  ul_strategic_flg  \\\n",
       "0         50000.0                2              0                 0   \n",
       "1         10000.0                5              0                 0   \n",
       "2         10000.0                2              0                 0   \n",
       "3        300000.0                1              0                 0   \n",
       "4         10000.0                2              0                 0   \n",
       "\n",
       "   ul_systematizing_flg  default_12m  \n",
       "0                     0            0  \n",
       "1                     0            0  \n",
       "2                     0            0  \n",
       "3                     0            0  \n",
       "4                     0            0  "
      ]
     },
     "execution_count": 238,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 33
    },
    "id": "wW_OWNg-8Hfi",
    "outputId": "bf27a3ab-e63d-4f0a-e9c8-8636e885ab01"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(32395, 37)"
      ]
     },
     "execution_count": 239,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "H06Wcg09N92N"
   },
   "source": [
    "Количество разных типов фичей"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 82
    },
    "id": "cYOxzyvLN92O",
    "outputId": "8d0185c7-5342-4f91-c287-1dc422fa1b55"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "float64    25\n",
       "int64      11\n",
       "object      1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 240,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.dtypes.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "f8ZLAgMeN92R"
   },
   "source": [
    "Уровень дефолт-рейта в выборке = 6,7219%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 33
    },
    "id": "LTHNGVxTN92S",
    "outputId": "0235aafa-033e-432e-99ef-08326f1880a1"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.06454699799351751"
      ]
     },
     "execution_count": 241,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df['default_12m'].values.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 33
    },
    "id": "47MrTU_xN92X",
    "outputId": "f759ebbb-4ae4-4732-ab21-136a0c530ef3"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "32395"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.record_id.nunique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "vo_LbDC0N92a"
   },
   "source": [
    "32395 уникальных наблюдений, что совпадает с размерностью нашей выборки. Следовательно, нам не надо бороться с проблемой дублирования строк. И первичным ключом является record_id. Этот признак в модели нам не нужен, поэтому можем его удалить. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [],
   "source": [
    "del train_df['record_id']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "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>ar_revenue</th>\n",
       "      <th>ar_total_expenses</th>\n",
       "      <th>ar_sale_cost</th>\n",
       "      <th>ar_selling_expenses</th>\n",
       "      <th>ar_management_expenses</th>\n",
       "      <th>ar_sale_profit</th>\n",
       "      <th>ar_balance_of_rvns_and_expns</th>\n",
       "      <th>ar_profit_before_tax</th>\n",
       "      <th>ar_taxes</th>\n",
       "      <th>ar_other_profit_and_losses</th>\n",
       "      <th>ar_net_profit</th>\n",
       "      <th>ab_immobilized_assets</th>\n",
       "      <th>ab_mobile_current_assets</th>\n",
       "      <th>ab_inventory</th>\n",
       "      <th>ab_accounts_receivable</th>\n",
       "      <th>ab_other_current_assets</th>\n",
       "      <th>ab_cash_and_securities</th>\n",
       "      <th>ab_losses</th>\n",
       "      <th>ab_own_capital</th>\n",
       "      <th>ab_borrowed_capital</th>\n",
       "      <th>ab_long_term_liabilities</th>\n",
       "      <th>ab_short_term_borrowing</th>\n",
       "      <th>ab_accounts_payable</th>\n",
       "      <th>ab_other_borrowings</th>\n",
       "      <th>bus_age</th>\n",
       "      <th>ogrn_age</th>\n",
       "      <th>adr_actual_age</th>\n",
       "      <th>head_actual_age</th>\n",
       "      <th>cap_actual_age</th>\n",
       "      <th>ul_capital_sum</th>\n",
       "      <th>ul_founders_cnt</th>\n",
       "      <th>ul_branch_cnt</th>\n",
       "      <th>ul_strategic_flg</th>\n",
       "      <th>ul_systematizing_flg</th>\n",
       "      <th>default_12m</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>1.725000e+04</td>\n",
       "      <td>32395.000000</td>\n",
       "      <td>32395.000000</td>\n",
       "      <td>32395.000000</td>\n",
       "      <td>32395.000000</td>\n",
       "      <td>32395.000000</td>\n",
       "      <td>3.239500e+04</td>\n",
       "      <td>32395.000000</td>\n",
       "      <td>32395.000000</td>\n",
       "      <td>32395.000000</td>\n",
       "      <td>32395.0</td>\n",
       "      <td>32395.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>8.153776e+07</td>\n",
       "      <td>7.678474e+07</td>\n",
       "      <td>7.320328e+07</td>\n",
       "      <td>2.499198e+06</td>\n",
       "      <td>1.082263e+06</td>\n",
       "      <td>4.753016e+06</td>\n",
       "      <td>-6.424773e+05</td>\n",
       "      <td>4.110538e+06</td>\n",
       "      <td>4.188071e+05</td>\n",
       "      <td>3.937304e+04</td>\n",
       "      <td>3.731104e+06</td>\n",
       "      <td>9.007101e+06</td>\n",
       "      <td>3.044379e+07</td>\n",
       "      <td>1.627389e+07</td>\n",
       "      <td>1.336024e+07</td>\n",
       "      <td>8.096597e+05</td>\n",
       "      <td>1.610857e+06</td>\n",
       "      <td>3.307750e+06</td>\n",
       "      <td>1.911815e+07</td>\n",
       "      <td>2.419023e+07</td>\n",
       "      <td>3.149311e+06</td>\n",
       "      <td>4.983973e+06</td>\n",
       "      <td>1.508316e+07</td>\n",
       "      <td>9.737877e+05</td>\n",
       "      <td>97.121500</td>\n",
       "      <td>78.644822</td>\n",
       "      <td>15.010341</td>\n",
       "      <td>15.010341</td>\n",
       "      <td>15.010341</td>\n",
       "      <td>1.366509e+06</td>\n",
       "      <td>11.946535</td>\n",
       "      <td>0.021670</td>\n",
       "      <td>0.000093</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.064547</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.274402e+08</td>\n",
       "      <td>1.227227e+08</td>\n",
       "      <td>1.156651e+08</td>\n",
       "      <td>1.231798e+07</td>\n",
       "      <td>7.848929e+06</td>\n",
       "      <td>1.941027e+07</td>\n",
       "      <td>6.014715e+06</td>\n",
       "      <td>1.938021e+07</td>\n",
       "      <td>1.565298e+06</td>\n",
       "      <td>4.843389e+05</td>\n",
       "      <td>1.902355e+07</td>\n",
       "      <td>3.642723e+07</td>\n",
       "      <td>6.331020e+07</td>\n",
       "      <td>3.729649e+07</td>\n",
       "      <td>3.851823e+07</td>\n",
       "      <td>8.222404e+06</td>\n",
       "      <td>1.067265e+07</td>\n",
       "      <td>8.895007e+06</td>\n",
       "      <td>4.603456e+07</td>\n",
       "      <td>5.917809e+07</td>\n",
       "      <td>2.764231e+07</td>\n",
       "      <td>1.359236e+07</td>\n",
       "      <td>3.793929e+07</td>\n",
       "      <td>1.493730e+07</td>\n",
       "      <td>75.555398</td>\n",
       "      <td>46.522534</td>\n",
       "      <td>9.546511</td>\n",
       "      <td>9.546511</td>\n",
       "      <td>9.546511</td>\n",
       "      <td>1.502438e+07</td>\n",
       "      <td>86.341847</td>\n",
       "      <td>0.342715</td>\n",
       "      <td>0.009623</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.245729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-7.019840e+08</td>\n",
       "      <td>-5.072780e+08</td>\n",
       "      <td>-3.542900e+07</td>\n",
       "      <td>-1.921670e+08</td>\n",
       "      <td>-9.210000e+07</td>\n",
       "      <td>-7.581800e+07</td>\n",
       "      <td>-5.946600e+07</td>\n",
       "      <td>-1.681000e+07</td>\n",
       "      <td>-1.034000e+07</td>\n",
       "      <td>-4.778100e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-1.160400e+07</td>\n",
       "      <td>-3.413000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-4.780300e+07</td>\n",
       "      <td>-7.501000e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-3.068000e+06</td>\n",
       "      <td>-2.700000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-2.500000e+05</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>0.000000e+00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.019950e+07</td>\n",
       "      <td>9.130250e+06</td>\n",
       "      <td>8.844000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.112500e+05</td>\n",
       "      <td>-9.400000e+05</td>\n",
       "      <td>1.420000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.010000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.854500e+06</td>\n",
       "      <td>8.100000e+05</td>\n",
       "      <td>4.722500e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>4.100000e+04</td>\n",
       "      <td>9.400000e+04</td>\n",
       "      <td>4.532500e+05</td>\n",
       "      <td>1.634000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>6.722500e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>1.000000e+04</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>3.233500e+07</td>\n",
       "      <td>2.970350e+07</td>\n",
       "      <td>2.840800e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.231500e+06</td>\n",
       "      <td>-1.260000e+05</td>\n",
       "      <td>8.500000e+05</td>\n",
       "      <td>4.100000e+04</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>7.040000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.109450e+07</td>\n",
       "      <td>4.714500e+06</td>\n",
       "      <td>3.028500e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.640000e+05</td>\n",
       "      <td>6.730000e+05</td>\n",
       "      <td>3.528000e+06</td>\n",
       "      <td>7.231000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>3.610000e+05</td>\n",
       "      <td>3.678000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>79.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>1.000000e+04</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>8.887675e+07</td>\n",
       "      <td>8.309500e+07</td>\n",
       "      <td>7.991425e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>4.574750e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>3.543750e+06</td>\n",
       "      <td>2.910000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>3.176000e+06</td>\n",
       "      <td>3.490000e+05</td>\n",
       "      <td>3.423000e+07</td>\n",
       "      <td>1.754600e+07</td>\n",
       "      <td>1.200325e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.188500e+06</td>\n",
       "      <td>3.025750e+06</td>\n",
       "      <td>1.760775e+07</td>\n",
       "      <td>2.541800e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>3.816750e+06</td>\n",
       "      <td>1.424575e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>137.000000</td>\n",
       "      <td>133.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>3.000000e+04</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7.946840e+08</td>\n",
       "      <td>7.949360e+08</td>\n",
       "      <td>7.624930e+08</td>\n",
       "      <td>2.746560e+08</td>\n",
       "      <td>3.023550e+08</td>\n",
       "      <td>1.434511e+09</td>\n",
       "      <td>1.584250e+08</td>\n",
       "      <td>1.481526e+09</td>\n",
       "      <td>5.420200e+07</td>\n",
       "      <td>3.979500e+07</td>\n",
       "      <td>1.485559e+09</td>\n",
       "      <td>1.630344e+09</td>\n",
       "      <td>3.175805e+09</td>\n",
       "      <td>2.890917e+09</td>\n",
       "      <td>1.711586e+09</td>\n",
       "      <td>3.336220e+08</td>\n",
       "      <td>1.271337e+09</td>\n",
       "      <td>2.470140e+08</td>\n",
       "      <td>1.739879e+09</td>\n",
       "      <td>3.349389e+09</td>\n",
       "      <td>2.919676e+09</td>\n",
       "      <td>3.223740e+08</td>\n",
       "      <td>1.820521e+09</td>\n",
       "      <td>1.251578e+09</td>\n",
       "      <td>921.000000</td>\n",
       "      <td>149.000000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>1.584979e+09</td>\n",
       "      <td>998.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         ar_revenue  ar_total_expenses  ar_sale_cost  ar_selling_expenses  \\\n",
       "count  1.725000e+04       1.725000e+04  1.725000e+04         1.725000e+04   \n",
       "mean   8.153776e+07       7.678474e+07  7.320328e+07         2.499198e+06   \n",
       "std    1.274402e+08       1.227227e+08  1.156651e+08         1.231798e+07   \n",
       "min    0.000000e+00      -7.019840e+08 -5.072780e+08        -3.542900e+07   \n",
       "25%    1.019950e+07       9.130250e+06  8.844000e+06         0.000000e+00   \n",
       "50%    3.233500e+07       2.970350e+07  2.840800e+07         0.000000e+00   \n",
       "75%    8.887675e+07       8.309500e+07  7.991425e+07         0.000000e+00   \n",
       "max    7.946840e+08       7.949360e+08  7.624930e+08         2.746560e+08   \n",
       "\n",
       "       ar_management_expenses  ar_sale_profit  ar_balance_of_rvns_and_expns  \\\n",
       "count            1.725000e+04    1.725000e+04                  1.725000e+04   \n",
       "mean             1.082263e+06    4.753016e+06                 -6.424773e+05   \n",
       "std              7.848929e+06    1.941027e+07                  6.014715e+06   \n",
       "min             -1.921670e+08   -9.210000e+07                 -7.581800e+07   \n",
       "25%              0.000000e+00    2.112500e+05                 -9.400000e+05   \n",
       "50%              0.000000e+00    1.231500e+06                 -1.260000e+05   \n",
       "75%              0.000000e+00    4.574750e+06                  0.000000e+00   \n",
       "max              3.023550e+08    1.434511e+09                  1.584250e+08   \n",
       "\n",
       "       ar_profit_before_tax      ar_taxes  ar_other_profit_and_losses  \\\n",
       "count          1.725000e+04  1.725000e+04                1.725000e+04   \n",
       "mean           4.110538e+06  4.188071e+05                3.937304e+04   \n",
       "std            1.938021e+07  1.565298e+06                4.843389e+05   \n",
       "min           -5.946600e+07 -1.681000e+07               -1.034000e+07   \n",
       "25%            1.420000e+05  0.000000e+00                0.000000e+00   \n",
       "50%            8.500000e+05  4.100000e+04                0.000000e+00   \n",
       "75%            3.543750e+06  2.910000e+05                0.000000e+00   \n",
       "max            1.481526e+09  5.420200e+07                3.979500e+07   \n",
       "\n",
       "       ar_net_profit  ab_immobilized_assets  ab_mobile_current_assets  \\\n",
       "count   1.725000e+04           1.725000e+04              1.725000e+04   \n",
       "mean    3.731104e+06           9.007101e+06              3.044379e+07   \n",
       "std     1.902355e+07           3.642723e+07              6.331020e+07   \n",
       "min    -4.778100e+07           0.000000e+00              0.000000e+00   \n",
       "25%     1.010000e+05           0.000000e+00              2.854500e+06   \n",
       "50%     7.040000e+05           0.000000e+00              1.109450e+07   \n",
       "75%     3.176000e+06           3.490000e+05              3.423000e+07   \n",
       "max     1.485559e+09           1.630344e+09              3.175805e+09   \n",
       "\n",
       "       ab_inventory  ab_accounts_receivable  ab_other_current_assets  \\\n",
       "count  1.725000e+04            1.725000e+04             1.725000e+04   \n",
       "mean   1.627389e+07            1.336024e+07             8.096597e+05   \n",
       "std    3.729649e+07            3.851823e+07             8.222404e+06   \n",
       "min    0.000000e+00           -1.160400e+07            -3.413000e+06   \n",
       "25%    8.100000e+05            4.722500e+05             0.000000e+00   \n",
       "50%    4.714500e+06            3.028500e+06             0.000000e+00   \n",
       "75%    1.754600e+07            1.200325e+07             0.000000e+00   \n",
       "max    2.890917e+09            1.711586e+09             3.336220e+08   \n",
       "\n",
       "       ab_cash_and_securities     ab_losses  ab_own_capital  \\\n",
       "count            1.725000e+04  1.725000e+04    1.725000e+04   \n",
       "mean             1.610857e+06  3.307750e+06    1.911815e+07   \n",
       "std              1.067265e+07  8.895007e+06    4.603456e+07   \n",
       "min              0.000000e+00 -4.780300e+07   -7.501000e+07   \n",
       "25%              4.100000e+04  9.400000e+04    4.532500e+05   \n",
       "50%              2.640000e+05  6.730000e+05    3.528000e+06   \n",
       "75%              1.188500e+06  3.025750e+06    1.760775e+07   \n",
       "max              1.271337e+09  2.470140e+08    1.739879e+09   \n",
       "\n",
       "       ab_borrowed_capital  ab_long_term_liabilities  ab_short_term_borrowing  \\\n",
       "count         1.725000e+04              1.725000e+04             1.725000e+04   \n",
       "mean          2.419023e+07              3.149311e+06             4.983973e+06   \n",
       "std           5.917809e+07              2.764231e+07             1.359236e+07   \n",
       "min           0.000000e+00             -3.068000e+06            -2.700000e+06   \n",
       "25%           1.634000e+06              0.000000e+00             0.000000e+00   \n",
       "50%           7.231000e+06              0.000000e+00             3.610000e+05   \n",
       "75%           2.541800e+07              0.000000e+00             3.816750e+06   \n",
       "max           3.349389e+09              2.919676e+09             3.223740e+08   \n",
       "\n",
       "       ab_accounts_payable  ab_other_borrowings       bus_age      ogrn_age  \\\n",
       "count         1.725000e+04         1.725000e+04  32395.000000  32395.000000   \n",
       "mean          1.508316e+07         9.737877e+05     97.121500     78.644822   \n",
       "std           3.793929e+07         1.493730e+07     75.555398     46.522534   \n",
       "min           0.000000e+00        -2.500000e+05      0.000000      0.000000   \n",
       "25%           6.722500e+05         0.000000e+00     35.000000     35.000000   \n",
       "50%           3.678000e+06         0.000000e+00     79.000000     77.000000   \n",
       "75%           1.424575e+07         0.000000e+00    137.000000    133.000000   \n",
       "max           1.820521e+09         1.251578e+09    921.000000    149.000000   \n",
       "\n",
       "       adr_actual_age  head_actual_age  cap_actual_age  ul_capital_sum  \\\n",
       "count    32395.000000     32395.000000    32395.000000    3.239500e+04   \n",
       "mean        15.010341        15.010341       15.010341    1.366509e+06   \n",
       "std          9.546511         9.546511        9.546511    1.502438e+07   \n",
       "min          1.000000         1.000000        1.000000    0.000000e+00   \n",
       "25%          6.000000         6.000000        6.000000    1.000000e+04   \n",
       "50%         14.000000        14.000000       14.000000    1.000000e+04   \n",
       "75%         24.000000        24.000000       24.000000    3.000000e+04   \n",
       "max         33.000000        33.000000       33.000000    1.584979e+09   \n",
       "\n",
       "       ul_founders_cnt  ul_branch_cnt  ul_strategic_flg  ul_systematizing_flg  \\\n",
       "count     32395.000000   32395.000000      32395.000000               32395.0   \n",
       "mean         11.946535       0.021670          0.000093                   0.0   \n",
       "std          86.341847       0.342715          0.009623                   0.0   \n",
       "min           1.000000       0.000000          0.000000                   0.0   \n",
       "25%           1.000000       0.000000          0.000000                   0.0   \n",
       "50%           1.000000       0.000000          0.000000                   0.0   \n",
       "75%           2.000000       0.000000          0.000000                   0.0   \n",
       "max         998.000000      20.000000          1.000000                   0.0   \n",
       "\n",
       "        default_12m  \n",
       "count  32395.000000  \n",
       "mean       0.064547  \n",
       "std        0.245729  \n",
       "min        0.000000  \n",
       "25%        0.000000  \n",
       "50%        0.000000  \n",
       "75%        0.000000  \n",
       "max        1.000000  "
      ]
     },
     "execution_count": 244,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Заметим, что стандратное отклонение у некоторых признаков достаточно большое, пожтому нам нужно будет проводить стандартизацию. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ar_revenue                      12893\n",
       "ar_total_expenses               12779\n",
       "ar_sale_cost                    12664\n",
       "ar_selling_expenses              1994\n",
       "ar_management_expenses           1093\n",
       "ar_sale_profit                   7163\n",
       "ar_balance_of_rvns_and_expns     5063\n",
       "ar_profit_before_tax             6526\n",
       "ar_taxes                         2247\n",
       "ar_other_profit_and_losses        595\n",
       "ar_net_profit                    6231\n",
       "ab_immobilized_assets            3498\n",
       "ab_mobile_current_assets        11222\n",
       "ab_inventory                     9143\n",
       "ab_accounts_receivable           8283\n",
       "ab_other_current_assets          1210\n",
       "ab_cash_and_securities           3677\n",
       "ab_losses                        6175\n",
       "ab_own_capital                   9090\n",
       "ab_borrowed_capital             10261\n",
       "ab_long_term_liabilities         2294\n",
       "ab_short_term_borrowing          4896\n",
       "ab_accounts_payable              8797\n",
       "ab_other_borrowings              1437\n",
       "bus_age                           315\n",
       "ogrn_age                          150\n",
       "adr_actual_age                     33\n",
       "head_actual_age                    33\n",
       "cap_actual_age                     33\n",
       "ul_staff_range                      3\n",
       "ul_capital_sum                   2980\n",
       "ul_founders_cnt                   278\n",
       "ul_branch_cnt                      13\n",
       "ul_strategic_flg                    2\n",
       "ul_systematizing_flg                1\n",
       "default_12m                         2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 245,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.nunique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Итак, мы обнаружиди еще один бесполезный признак, у которого множество значений состоит из одного элемента, это признак 'ul_systematizing_flg', можем его удалить. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [],
   "source": [
    "del train_df['ul_systematizing_flg']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Прежде всего для выявления корреляций и замены пропусков, давайте отбросим аномалии, так как если мы это сделаем после, то мы не сможем посчитать нормальный коэффициент корреляции Пирсона (в его формуле используется среднее значение по признаку, а оно будет не смещенным при наличии выбросов)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Поиск аномалий"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Есть несколько способов для выявления и удаления выбросов, но давайте попробуем сначала самый простой. Из основных способов: \n",
    "1. Z-score, но при своем подсчете использует среднее, так что не найс. \n",
    "1. Функция Антона"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [],
   "source": [
    "Q1 = train_df.quantile(0.01)\n",
    "Q3 = train_df.quantile(0.99)\n",
    "\n",
    "IQR = Q3 - Q1\n",
    "train_df_outliers = train_df[~((train_df < (Q1 - 1.5 * IQR)) |(train_df > (Q3 + 1.5 * IQR))).any(axis=1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Размерность выборки до удаления выбросов: (32395, 35)\n",
      "Размерность выборки после удаления выбросов: (31498, 35)\n"
     ]
    }
   ],
   "source": [
    "print('Размерность выборки до удаления выбросов: {}'.format(train_df.shape))\n",
    "print('Размерность выборки после удаления выбросов: {}'.format(train_df_outliers.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    29422\n",
       "1     2076\n",
       "Name: default_12m, dtype: int64"
      ]
     },
     "execution_count": 249,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df_outliers['default_12m'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Обнаружение корреляций"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Наша задача в этом разделе - обнаружить линейно-зависимые признаки и их удалить. \n",
    "Давайте добавим признаки, которые по нашему мнению, могут обозначать сумму нескольких имеющихся признаков. Например, признак 'ar_total_expenses' скорее всего содержит в себе сумму ar_sale_cost, ar_selling_expenses и ar_management_expenses. Давайте добавим сумму вышеупомянутых признаков в качестве отдельного столбца total_cost, а затем посмотрим на его корреляцию с признаком 'ar_total_expenses'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df_outliers['total_cost'] = train_df_outliers['ar_sale_cost'] + train_df_outliers['ar_selling_expenses'] + train_df_outliers['ar_management_expenses']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Аналогичную операцию проведем и с чистой выручкой. Чистая выручка представляет из себя сумму признаков ar_profit_before_tax и ar_net_profit. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df_outliers['net_profit'] = train_df_outliers['ar_profit_before_tax'] + train_df_outliers['ar_net_profit']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Аналогично..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df_outliers['borrowed_capital'] = train_df_outliers['ab_long_term_liabilities'] + train_df_outliers['ab_short_term_borrowing'] + train_df_outliers['ab_accounts_payable'] + train_df_outliers['ab_other_borrowings']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Аналогично..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df_outliers['revenue'] = train_df_outliers['ar_total_expenses'] + train_df_outliers['ar_sale_profit']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Аналогично...."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df_outliers['mobile_current_assets'] = train_df_outliers['ab_inventory'] + train_df_outliers['ab_accounts_receivable']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {},
   "outputs": [],
   "source": [
    "def find_correlated_features(df_, threshold):\n",
    "    corr_matrix = df_.corr().abs()\n",
    "    corr_matrix = corr_matrix.where(np.triu(np.ones(corr_matrix.shape), k = 1).astype(np.bool))\n",
    "    df_ = corr_matrix.stack().reset_index()\n",
    "    df_.columns = ['feature1', 'feature2', 'corr_coef']\n",
    "    df_ = df_[df_.corr_coef >= threshold].sort_values('corr_coef', ascending=False)\n",
    "    return df_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "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>feature1</th>\n",
       "      <th>feature2</th>\n",
       "      <th>corr_coef</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>total_cost</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>621</th>\n",
       "      <td>head_actual_age</td>\n",
       "      <td>cap_actual_age</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>612</th>\n",
       "      <td>adr_actual_age</td>\n",
       "      <td>cap_actual_age</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>611</th>\n",
       "      <td>adr_actual_age</td>\n",
       "      <td>head_actual_age</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>revenue</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>527</th>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>389</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.998305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>256</th>\n",
       "      <td>ar_profit_before_tax</td>\n",
       "      <td>net_profit</td>\n",
       "      <td>0.997970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>337</th>\n",
       "      <td>ar_net_profit</td>\n",
       "      <td>net_profit</td>\n",
       "      <td>0.997567</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>658</th>\n",
       "      <td>total_cost</td>\n",
       "      <td>revenue</td>\n",
       "      <td>0.997491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>revenue</td>\n",
       "      <td>0.997491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>0.997491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>total_cost</td>\n",
       "      <td>0.997491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>0.995991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>total_cost</td>\n",
       "      <td>0.995991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>0.993717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>revenue</td>\n",
       "      <td>0.993717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>233</th>\n",
       "      <td>ar_profit_before_tax</td>\n",
       "      <td>ar_net_profit</td>\n",
       "      <td>0.991102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>321</th>\n",
       "      <td>ar_net_profit</td>\n",
       "      <td>ab_losses</td>\n",
       "      <td>0.979982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>491</th>\n",
       "      <td>ab_losses</td>\n",
       "      <td>net_profit</td>\n",
       "      <td>0.979814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>240</th>\n",
       "      <td>ar_profit_before_tax</td>\n",
       "      <td>ab_losses</td>\n",
       "      <td>0.975494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>515</th>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>0.907017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>572</th>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>0.907017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>171</th>\n",
       "      <td>ar_sale_profit</td>\n",
       "      <td>ar_profit_before_tax</td>\n",
       "      <td>0.863693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588</th>\n",
       "      <td>bus_age</td>\n",
       "      <td>ogrn_age</td>\n",
       "      <td>0.862825</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>ar_sale_profit</td>\n",
       "      <td>net_profit</td>\n",
       "      <td>0.854270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>387</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>0.850601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>372</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>0.850601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>529</th>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.847267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>664</th>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.847267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>174</th>\n",
       "      <td>ar_sale_profit</td>\n",
       "      <td>ar_net_profit</td>\n",
       "      <td>0.839975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>412</th>\n",
       "      <td>ab_inventory</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.835632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>366</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>ab_inventory</td>\n",
       "      <td>0.832604</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>181</th>\n",
       "      <td>ar_sale_profit</td>\n",
       "      <td>ab_losses</td>\n",
       "      <td>0.824288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>367</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>0.806294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>434</th>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.805924</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>0.768721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>574</th>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.766307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>347</th>\n",
       "      <td>ab_immobilized_assets</td>\n",
       "      <td>ab_own_capital</td>\n",
       "      <td>0.732683</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>0.722301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>417</th>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>0.722301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>0.719284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>388</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>revenue</td>\n",
       "      <td>0.719284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.716834</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>665</th>\n",
       "      <td>revenue</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.716834</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>420</th>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>0.714944</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>0.701495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>385</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>total_cost</td>\n",
       "      <td>0.701495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.698951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>659</th>\n",
       "      <td>total_cost</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.698951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>0.698438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.696831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>476</th>\n",
       "      <td>ab_losses</td>\n",
       "      <td>ab_own_capital</td>\n",
       "      <td>0.689706</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>322</th>\n",
       "      <td>ar_net_profit</td>\n",
       "      <td>ab_own_capital</td>\n",
       "      <td>0.682026</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>0.679551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>0.679551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>663</th>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>revenue</td>\n",
       "      <td>0.679551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>528</th>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>revenue</td>\n",
       "      <td>0.679551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>509</th>\n",
       "      <td>ab_own_capital</td>\n",
       "      <td>net_profit</td>\n",
       "      <td>0.677697</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>525</th>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>total_cost</td>\n",
       "      <td>0.675423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>0.675423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>0.675423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>657</th>\n",
       "      <td>total_cost</td>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>0.675423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>ab_inventory</td>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>0.671383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>410</th>\n",
       "      <td>ab_inventory</td>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>0.671383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>ar_profit_before_tax</td>\n",
       "      <td>ab_own_capital</td>\n",
       "      <td>0.670862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>ab_borrowed_capital</td>\n",
       "      <td>0.666729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>borrowed_capital</td>\n",
       "      <td>0.666729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>394</th>\n",
       "      <td>ab_inventory</td>\n",
       "      <td>ab_own_capital</td>\n",
       "      <td>0.649080</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>172</th>\n",
       "      <td>ar_sale_profit</td>\n",
       "      <td>ar_taxes</td>\n",
       "      <td>0.645529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>433</th>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>revenue</td>\n",
       "      <td>0.639473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>0.639473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>231</th>\n",
       "      <td>ar_profit_before_tax</td>\n",
       "      <td>ar_taxes</td>\n",
       "      <td>0.636972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>573</th>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>revenue</td>\n",
       "      <td>0.630962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>ar_revenue</td>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>0.630962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>570</th>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>total_cost</td>\n",
       "      <td>0.630023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>0.630023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>0.625987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>ar_sale_cost</td>\n",
       "      <td>ab_accounts_payable</td>\n",
       "      <td>0.623357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>ar_total_expenses</td>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>0.623134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>430</th>\n",
       "      <td>ab_accounts_receivable</td>\n",
       "      <td>total_cost</td>\n",
       "      <td>0.623134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>349</th>\n",
       "      <td>ab_immobilized_assets</td>\n",
       "      <td>ab_long_term_liabilities</td>\n",
       "      <td>0.622376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>371</th>\n",
       "      <td>ab_mobile_current_assets</td>\n",
       "      <td>ab_own_capital</td>\n",
       "      <td>0.617288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <td>ab_own_capital</td>\n",
       "      <td>mobile_current_assets</td>\n",
       "      <td>0.616702</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     feature1                  feature2  corr_coef\n",
       "66          ar_total_expenses                total_cost   1.000000\n",
       "621           head_actual_age            cap_actual_age   1.000000\n",
       "612            adr_actual_age            cap_actual_age   1.000000\n",
       "611            adr_actual_age           head_actual_age   1.000000\n",
       "34                 ar_revenue                   revenue   1.000000\n",
       "527       ab_borrowed_capital          borrowed_capital   1.000000\n",
       "389  ab_mobile_current_assets     mobile_current_assets   0.998305\n",
       "256      ar_profit_before_tax                net_profit   0.997970\n",
       "337             ar_net_profit                net_profit   0.997567\n",
       "658                total_cost                   revenue   0.997491\n",
       "69          ar_total_expenses                   revenue   0.997491\n",
       "0                  ar_revenue         ar_total_expenses   0.997491\n",
       "31                 ar_revenue                total_cost   0.997491\n",
       "36          ar_total_expenses              ar_sale_cost   0.995991\n",
       "100              ar_sale_cost                total_cost   0.995991\n",
       "1                  ar_revenue              ar_sale_cost   0.993717\n",
       "103              ar_sale_cost                   revenue   0.993717\n",
       "233      ar_profit_before_tax             ar_net_profit   0.991102\n",
       "321             ar_net_profit                 ab_losses   0.979982\n",
       "491                 ab_losses                net_profit   0.979814\n",
       "240      ar_profit_before_tax                 ab_losses   0.975494\n",
       "515       ab_borrowed_capital       ab_accounts_payable   0.907017\n",
       "572       ab_accounts_payable          borrowed_capital   0.907017\n",
       "171            ar_sale_profit      ar_profit_before_tax   0.863693\n",
       "588                   bus_age                  ogrn_age   0.862825\n",
       "197            ar_sale_profit                net_profit   0.854270\n",
       "387  ab_mobile_current_assets          borrowed_capital   0.850601\n",
       "372  ab_mobile_current_assets       ab_borrowed_capital   0.850601\n",
       "529       ab_borrowed_capital     mobile_current_assets   0.847267\n",
       "664          borrowed_capital     mobile_current_assets   0.847267\n",
       "174            ar_sale_profit             ar_net_profit   0.839975\n",
       "412              ab_inventory     mobile_current_assets   0.835632\n",
       "366  ab_mobile_current_assets              ab_inventory   0.832604\n",
       "181            ar_sale_profit                 ab_losses   0.824288\n",
       "367  ab_mobile_current_assets    ab_accounts_receivable   0.806294\n",
       "434    ab_accounts_receivable     mobile_current_assets   0.805924\n",
       "375  ab_mobile_current_assets       ab_accounts_payable   0.768721\n",
       "574       ab_accounts_payable     mobile_current_assets   0.766307\n",
       "347     ab_immobilized_assets            ab_own_capital   0.732683\n",
       "432    ab_accounts_receivable          borrowed_capital   0.722301\n",
       "417    ab_accounts_receivable       ab_borrowed_capital   0.722301\n",
       "11                 ar_revenue  ab_mobile_current_assets   0.719284\n",
       "388  ab_mobile_current_assets                   revenue   0.719284\n",
       "35                 ar_revenue     mobile_current_assets   0.716834\n",
       "665                   revenue     mobile_current_assets   0.716834\n",
       "420    ab_accounts_receivable       ab_accounts_payable   0.714944\n",
       "46          ar_total_expenses  ab_mobile_current_assets   0.701495\n",
       "385  ab_mobile_current_assets                total_cost   0.701495\n",
       "70          ar_total_expenses     mobile_current_assets   0.698951\n",
       "659                total_cost     mobile_current_assets   0.698951\n",
       "80               ar_sale_cost  ab_mobile_current_assets   0.698438\n",
       "104              ar_sale_cost     mobile_current_assets   0.696831\n",
       "476                 ab_losses            ab_own_capital   0.689706\n",
       "322             ar_net_profit            ab_own_capital   0.682026\n",
       "33                 ar_revenue          borrowed_capital   0.679551\n",
       "18                 ar_revenue       ab_borrowed_capital   0.679551\n",
       "663          borrowed_capital                   revenue   0.679551\n",
       "528       ab_borrowed_capital                   revenue   0.679551\n",
       "509            ab_own_capital                net_profit   0.677697\n",
       "525       ab_borrowed_capital                total_cost   0.675423\n",
       "53          ar_total_expenses       ab_borrowed_capital   0.675423\n",
       "68          ar_total_expenses          borrowed_capital   0.675423\n",
       "657                total_cost          borrowed_capital   0.675423\n",
       "395              ab_inventory       ab_borrowed_capital   0.671383\n",
       "410              ab_inventory          borrowed_capital   0.671383\n",
       "241      ar_profit_before_tax            ab_own_capital   0.670862\n",
       "87               ar_sale_cost       ab_borrowed_capital   0.666729\n",
       "102              ar_sale_cost          borrowed_capital   0.666729\n",
       "394              ab_inventory            ab_own_capital   0.649080\n",
       "172            ar_sale_profit                  ar_taxes   0.645529\n",
       "433    ab_accounts_receivable                   revenue   0.639473\n",
       "13                 ar_revenue    ab_accounts_receivable   0.639473\n",
       "231      ar_profit_before_tax                  ar_taxes   0.636972\n",
       "573       ab_accounts_payable                   revenue   0.630962\n",
       "21                 ar_revenue       ab_accounts_payable   0.630962\n",
       "570       ab_accounts_payable                total_cost   0.630023\n",
       "56          ar_total_expenses       ab_accounts_payable   0.630023\n",
       "82               ar_sale_cost    ab_accounts_receivable   0.625987\n",
       "90               ar_sale_cost       ab_accounts_payable   0.623357\n",
       "48          ar_total_expenses    ab_accounts_receivable   0.623134\n",
       "430    ab_accounts_receivable                total_cost   0.623134\n",
       "349     ab_immobilized_assets  ab_long_term_liabilities   0.622376\n",
       "371  ab_mobile_current_assets            ab_own_capital   0.617288\n",
       "512            ab_own_capital     mobile_current_assets   0.616702"
      ]
     },
     "execution_count": 256,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "find_correlated_features(train_df_outliers, 0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {},
   "outputs": [],
   "source": [
    "del train_df_outliers['head_actual_age']\n",
    "del train_df_outliers['cap_actual_age']\n",
    "del train_df_outliers['bus_age']\n",
    "del train_df_outliers['ab_borrowed_capital']\n",
    "del train_df_outliers['ar_revenue']\n",
    "del train_df_outliers['ar_total_expenses']\n",
    "del train_df_outliers['ar_taxes']\n",
    "del train_df_outliers['ab_mobile_current_assets']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {},
   "outputs": [],
   "source": [
    "# удаляем признаки, которые мы создали сами для проверки линейной зависимости\n",
    "del train_df_outliers['revenue'] \n",
    "del train_df_outliers['net_profit']\n",
    "del train_df_outliers['total_cost']\n",
    "del train_df_outliers['mobile_current_assets']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {},
   "outputs": [],
   "source": [
    "del train_df_outliers['ab_losses']\n",
    "del train_df_outliers['ar_profit_before_tax']\n",
    "del train_df_outliers['ab_accounts_payable']\n",
    "del train_df_outliers['ar_net_profit']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(31498, 24)"
      ]
     },
     "execution_count": 260,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df_outliers.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Типы признаков"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 31498 entries, 0 to 32394\n",
      "Data columns (total 24 columns):\n",
      " #   Column                        Non-Null Count  Dtype  \n",
      "---  ------                        --------------  -----  \n",
      " 0   ar_sale_cost                  16473 non-null  float64\n",
      " 1   ar_selling_expenses           16473 non-null  float64\n",
      " 2   ar_management_expenses        16473 non-null  float64\n",
      " 3   ar_sale_profit                16473 non-null  float64\n",
      " 4   ar_balance_of_rvns_and_expns  16473 non-null  float64\n",
      " 5   ar_other_profit_and_losses    16473 non-null  float64\n",
      " 6   ab_immobilized_assets         16473 non-null  float64\n",
      " 7   ab_inventory                  16473 non-null  float64\n",
      " 8   ab_accounts_receivable        16473 non-null  float64\n",
      " 9   ab_other_current_assets       16473 non-null  float64\n",
      " 10  ab_cash_and_securities        16473 non-null  float64\n",
      " 11  ab_own_capital                16473 non-null  float64\n",
      " 12  ab_long_term_liabilities      16473 non-null  float64\n",
      " 13  ab_short_term_borrowing       16473 non-null  float64\n",
      " 14  ab_other_borrowings           16473 non-null  float64\n",
      " 15  ogrn_age                      31498 non-null  int64  \n",
      " 16  adr_actual_age                31498 non-null  int64  \n",
      " 17  ul_staff_range                31498 non-null  object \n",
      " 18  ul_capital_sum                31498 non-null  float64\n",
      " 19  ul_founders_cnt               31498 non-null  int64  \n",
      " 20  ul_branch_cnt                 31498 non-null  int64  \n",
      " 21  ul_strategic_flg              31498 non-null  int64  \n",
      " 22  default_12m                   31498 non-null  int64  \n",
      " 23  borrowed_capital              16473 non-null  float64\n",
      "dtypes: float64(17), int64(6), object(1)\n",
      "memory usage: 6.0+ MB\n"
     ]
    }
   ],
   "source": [
    "train_df_outliers.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Все признаки числовые, кроме одного, ul_staff_range."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1-100]      28657\n",
       "(100-500]     2529\n",
       "> 500          312\n",
       "Name: ul_staff_range, dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df_outliers['ul_staff_range'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Этот признак можно закодировать через LabelEncoder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "le = preprocessing.LabelEncoder()\n",
    "train_df_outliers['ul_staff_range'] = le.fit_transform(train_df_outliers['ul_staff_range'] )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Стандартизация"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>ar_sale_cost</th>\n",
       "      <th>ar_selling_expenses</th>\n",
       "      <th>ar_management_expenses</th>\n",
       "      <th>ar_sale_profit</th>\n",
       "      <th>ar_balance_of_rvns_and_expns</th>\n",
       "      <th>ar_other_profit_and_losses</th>\n",
       "      <th>ab_immobilized_assets</th>\n",
       "      <th>ab_inventory</th>\n",
       "      <th>ab_accounts_receivable</th>\n",
       "      <th>ab_other_current_assets</th>\n",
       "      <th>ab_cash_and_securities</th>\n",
       "      <th>ab_own_capital</th>\n",
       "      <th>ab_long_term_liabilities</th>\n",
       "      <th>ab_short_term_borrowing</th>\n",
       "      <th>ab_other_borrowings</th>\n",
       "      <th>ogrn_age</th>\n",
       "      <th>adr_actual_age</th>\n",
       "      <th>ul_staff_range</th>\n",
       "      <th>ul_capital_sum</th>\n",
       "      <th>ul_founders_cnt</th>\n",
       "      <th>ul_branch_cnt</th>\n",
       "      <th>ul_strategic_flg</th>\n",
       "      <th>default_12m</th>\n",
       "      <th>borrowed_capital</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>1.647300e+04</td>\n",
       "      <td>31498.000000</td>\n",
       "      <td>31498.000000</td>\n",
       "      <td>31498.000000</td>\n",
       "      <td>3.149800e+04</td>\n",
       "      <td>31498.000000</td>\n",
       "      <td>31498.0</td>\n",
       "      <td>31498.0</td>\n",
       "      <td>31498.000000</td>\n",
       "      <td>1.647300e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>6.735824e+07</td>\n",
       "      <td>1.961989e+06</td>\n",
       "      <td>7.038106e+05</td>\n",
       "      <td>4.058385e+06</td>\n",
       "      <td>-5.882518e+05</td>\n",
       "      <td>2.364803e+04</td>\n",
       "      <td>6.516087e+06</td>\n",
       "      <td>1.424668e+07</td>\n",
       "      <td>1.099652e+07</td>\n",
       "      <td>3.256347e+05</td>\n",
       "      <td>1.301645e+06</td>\n",
       "      <td>1.601239e+07</td>\n",
       "      <td>2.170261e+06</td>\n",
       "      <td>4.322812e+06</td>\n",
       "      <td>3.419119e+05</td>\n",
       "      <td>77.620516</td>\n",
       "      <td>14.971998</td>\n",
       "      <td>1.829513</td>\n",
       "      <td>6.561572e+05</td>\n",
       "      <td>5.676360</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.065909</td>\n",
       "      <td>1.965916e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.072794e+08</td>\n",
       "      <td>9.671108e+06</td>\n",
       "      <td>4.366645e+06</td>\n",
       "      <td>9.032176e+06</td>\n",
       "      <td>4.556331e+06</td>\n",
       "      <td>1.497990e+05</td>\n",
       "      <td>2.202381e+07</td>\n",
       "      <td>2.493200e+07</td>\n",
       "      <td>2.313252e+07</td>\n",
       "      <td>2.364454e+06</td>\n",
       "      <td>3.191715e+06</td>\n",
       "      <td>3.379051e+07</td>\n",
       "      <td>8.400831e+06</td>\n",
       "      <td>1.058665e+07</td>\n",
       "      <td>1.791719e+06</td>\n",
       "      <td>46.401355</td>\n",
       "      <td>9.529689</td>\n",
       "      <td>0.549557</td>\n",
       "      <td>3.846601e+06</td>\n",
       "      <td>37.640188</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.248127</td>\n",
       "      <td>3.389751e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-5.309500e+07</td>\n",
       "      <td>-1.406100e+07</td>\n",
       "      <td>-8.335000e+06</td>\n",
       "      <td>-3.957000e+07</td>\n",
       "      <td>-6.972900e+07</td>\n",
       "      <td>-7.690000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-1.160400e+07</td>\n",
       "      <td>-3.413000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-5.851700e+07</td>\n",
       "      <td>-3.068000e+06</td>\n",
       "      <td>-2.700000e+06</td>\n",
       "      <td>-2.500000e+05</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>8.301000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.080000e+05</td>\n",
       "      <td>-8.800000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>7.620000e+05</td>\n",
       "      <td>4.300000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>3.600000e+04</td>\n",
       "      <td>4.100000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>34.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000e+04</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.503000e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.650600e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.168000e+06</td>\n",
       "      <td>-1.220000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>4.367000e+06</td>\n",
       "      <td>2.794000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.440000e+05</td>\n",
       "      <td>3.102000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>3.140000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000e+04</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6.626000e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.334100e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>4.242000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.530000e+05</td>\n",
       "      <td>1.611100e+07</td>\n",
       "      <td>1.081700e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.061000e+06</td>\n",
       "      <td>1.505400e+07</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>3.493000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>131.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.500000e+04</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.231900e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7.604440e+08</td>\n",
       "      <td>1.494320e+08</td>\n",
       "      <td>7.787000e+07</td>\n",
       "      <td>1.127170e+08</td>\n",
       "      <td>5.991900e+07</td>\n",
       "      <td>2.511000e+06</td>\n",
       "      <td>3.074030e+08</td>\n",
       "      <td>3.006660e+08</td>\n",
       "      <td>3.364400e+08</td>\n",
       "      <td>4.617300e+07</td>\n",
       "      <td>4.442300e+07</td>\n",
       "      <td>4.620080e+08</td>\n",
       "      <td>1.318680e+08</td>\n",
       "      <td>1.468000e+08</td>\n",
       "      <td>3.080800e+07</td>\n",
       "      <td>149.000000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>6.500792e+07</td>\n",
       "      <td>816.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.505650e+08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       ar_sale_cost  ar_selling_expenses  ar_management_expenses  \\\n",
       "count  1.647300e+04         1.647300e+04            1.647300e+04   \n",
       "mean   6.735824e+07         1.961989e+06            7.038106e+05   \n",
       "std    1.072794e+08         9.671108e+06            4.366645e+06   \n",
       "min   -5.309500e+07        -1.406100e+07           -8.335000e+06   \n",
       "25%    8.301000e+06         0.000000e+00            0.000000e+00   \n",
       "50%    2.650600e+07         0.000000e+00            0.000000e+00   \n",
       "75%    7.334100e+07         0.000000e+00            0.000000e+00   \n",
       "max    7.604440e+08         1.494320e+08            7.787000e+07   \n",
       "\n",
       "       ar_sale_profit  ar_balance_of_rvns_and_expns  \\\n",
       "count    1.647300e+04                  1.647300e+04   \n",
       "mean     4.058385e+06                 -5.882518e+05   \n",
       "std      9.032176e+06                  4.556331e+06   \n",
       "min     -3.957000e+07                 -6.972900e+07   \n",
       "25%      2.080000e+05                 -8.800000e+05   \n",
       "50%      1.168000e+06                 -1.220000e+05   \n",
       "75%      4.242000e+06                  0.000000e+00   \n",
       "max      1.127170e+08                  5.991900e+07   \n",
       "\n",
       "       ar_other_profit_and_losses  ab_immobilized_assets  ab_inventory  \\\n",
       "count                1.647300e+04           1.647300e+04  1.647300e+04   \n",
       "mean                 2.364803e+04           6.516087e+06  1.424668e+07   \n",
       "std                  1.497990e+05           2.202381e+07  2.493200e+07   \n",
       "min                 -7.690000e+05           0.000000e+00  0.000000e+00   \n",
       "25%                  0.000000e+00           0.000000e+00  7.620000e+05   \n",
       "50%                  0.000000e+00           0.000000e+00  4.367000e+06   \n",
       "75%                  0.000000e+00           1.530000e+05  1.611100e+07   \n",
       "max                  2.511000e+06           3.074030e+08  3.006660e+08   \n",
       "\n",
       "       ab_accounts_receivable  ab_other_current_assets  \\\n",
       "count            1.647300e+04             1.647300e+04   \n",
       "mean             1.099652e+07             3.256347e+05   \n",
       "std              2.313252e+07             2.364454e+06   \n",
       "min             -1.160400e+07            -3.413000e+06   \n",
       "25%              4.300000e+05             0.000000e+00   \n",
       "50%              2.794000e+06             0.000000e+00   \n",
       "75%              1.081700e+07             0.000000e+00   \n",
       "max              3.364400e+08             4.617300e+07   \n",
       "\n",
       "       ab_cash_and_securities  ab_own_capital  ab_long_term_liabilities  \\\n",
       "count            1.647300e+04    1.647300e+04              1.647300e+04   \n",
       "mean             1.301645e+06    1.601239e+07              2.170261e+06   \n",
       "std              3.191715e+06    3.379051e+07              8.400831e+06   \n",
       "min              0.000000e+00   -5.851700e+07             -3.068000e+06   \n",
       "25%              3.600000e+04    4.100000e+05              0.000000e+00   \n",
       "50%              2.440000e+05    3.102000e+06              0.000000e+00   \n",
       "75%              1.061000e+06    1.505400e+07              0.000000e+00   \n",
       "max              4.442300e+07    4.620080e+08              1.318680e+08   \n",
       "\n",
       "       ab_short_term_borrowing  ab_other_borrowings      ogrn_age  \\\n",
       "count             1.647300e+04         1.647300e+04  31498.000000   \n",
       "mean              4.322812e+06         3.419119e+05     77.620516   \n",
       "std               1.058665e+07         1.791719e+06     46.401355   \n",
       "min              -2.700000e+06        -2.500000e+05      0.000000   \n",
       "25%               0.000000e+00         0.000000e+00     34.000000   \n",
       "50%               3.140000e+05         0.000000e+00     75.000000   \n",
       "75%               3.493000e+06         0.000000e+00    131.000000   \n",
       "max               1.468000e+08         3.080800e+07    149.000000   \n",
       "\n",
       "       adr_actual_age  ul_staff_range  ul_capital_sum  ul_founders_cnt  \\\n",
       "count    31498.000000    31498.000000    3.149800e+04     31498.000000   \n",
       "mean        14.971998        1.829513    6.561572e+05         5.676360   \n",
       "std          9.529689        0.549557    3.846601e+06        37.640188   \n",
       "min          1.000000        0.000000    0.000000e+00         1.000000   \n",
       "25%          6.000000        2.000000    1.000000e+04         1.000000   \n",
       "50%         14.000000        2.000000    1.000000e+04         1.000000   \n",
       "75%         24.000000        2.000000    2.500000e+04         2.000000   \n",
       "max         33.000000        2.000000    6.500792e+07       816.000000   \n",
       "\n",
       "       ul_branch_cnt  ul_strategic_flg   default_12m  borrowed_capital  \n",
       "count        31498.0           31498.0  31498.000000      1.647300e+04  \n",
       "mean             0.0               0.0      0.065909      1.965916e+07  \n",
       "std              0.0               0.0      0.248127      3.389751e+07  \n",
       "min              0.0               0.0      0.000000      0.000000e+00  \n",
       "25%              0.0               0.0      0.000000      1.503000e+06  \n",
       "50%              0.0               0.0      0.000000      6.626000e+06  \n",
       "75%              0.0               0.0      0.000000      2.231900e+07  \n",
       "max              0.0               0.0      1.000000      4.505650e+08  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df_outliers.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_x = train_df_outliers.drop('default_12m',axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "scaler.fit(train_x)\n",
    "train_x = pd.DataFrame(scaler.transform(train_x), index=train_x.index, columns=train_x.columns)\n",
    "train_x['default_12m'] = train_df_outliers['default_12m']\n",
    "train_df_std = train_x.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Обработка пропусков"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Количество пропущенных значений:\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>ar_sale_cost</th>\n",
       "      <th>ab_other_current_assets</th>\n",
       "      <th>borrowed_capital</th>\n",
       "      <th>ab_other_borrowings</th>\n",
       "      <th>ab_short_term_borrowing</th>\n",
       "      <th>ar_selling_expenses</th>\n",
       "      <th>ab_own_capital</th>\n",
       "      <th>ab_cash_and_securities</th>\n",
       "      <th>ab_long_term_liabilities</th>\n",
       "      <th>ab_accounts_receivable</th>\n",
       "      <th>ab_inventory</th>\n",
       "      <th>ab_immobilized_assets</th>\n",
       "      <th>ar_other_profit_and_losses</th>\n",
       "      <th>ar_balance_of_rvns_and_expns</th>\n",
       "      <th>ar_sale_profit</th>\n",
       "      <th>ar_management_expenses</th>\n",
       "      <th>ogrn_age</th>\n",
       "      <th>adr_actual_age</th>\n",
       "      <th>ul_staff_range</th>\n",
       "      <th>ul_capital_sum</th>\n",
       "      <th>ul_founders_cnt</th>\n",
       "      <th>ul_branch_cnt</th>\n",
       "      <th>ul_strategic_flg</th>\n",
       "      <th>default_12m</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>15025</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   ar_sale_cost  ab_other_current_assets  borrowed_capital  \\\n",
       "0         15025                    15025             15025   \n",
       "\n",
       "   ab_other_borrowings  ab_short_term_borrowing  ar_selling_expenses  \\\n",
       "0                15025                    15025                15025   \n",
       "\n",
       "   ab_own_capital  ab_cash_and_securities  ab_long_term_liabilities  \\\n",
       "0           15025                   15025                     15025   \n",
       "\n",
       "   ab_accounts_receivable  ab_inventory  ab_immobilized_assets  \\\n",
       "0                   15025         15025                  15025   \n",
       "\n",
       "   ar_other_profit_and_losses  ar_balance_of_rvns_and_expns  ar_sale_profit  \\\n",
       "0                       15025                         15025           15025   \n",
       "\n",
       "   ar_management_expenses  ogrn_age  adr_actual_age  ul_staff_range  \\\n",
       "0                   15025         0               0               0   \n",
       "\n",
       "   ul_capital_sum  ul_founders_cnt  ul_branch_cnt  ul_strategic_flg  \\\n",
       "0               0                0              0                 0   \n",
       "\n",
       "   default_12m  \n",
       "0            0  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"\\nКоличество пропущенных значений:\")\n",
    "train_df_std.isna().sum().to_frame().sort_values(0, ascending = False).T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Пропусков в данных сликшом много.  \n",
    "1 вариант: удалить столбцы с пропусками  \n",
    "2 вариант:  заменить пропуски на среднее  \n",
    "3 вариант:  удалить строки с NaN\n",
    "4 вариант: заполнить пропуски через KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1 вариант \n",
    "train_df_std_1 = train_df_std.dropna(axis = 1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 2 вариант \n",
    "train_df_std_2 = train_df_std.fillna(train_df_std.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 3 вариант \n",
    "train_df_std_3 = train_df_std.dropna() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 4 вариант \n",
    "columns_with_nan = train_df_std.columns[train_df_std.isna().any()].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-40-6e0ad69687a6>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0msklearn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mimpute\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mKNNImputer\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mimputer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mKNNImputer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mn_neighbors\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m train_df_std_4_nan = pd.DataFrame(imputer.fit_transform(train_df_std), \n\u001b[0m\u001b[1;32m      4\u001b[0m                                   index=train_df_std.index, columns=train_df_std.columns)\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.8/site-packages/sklearn/base.py\u001b[0m in \u001b[0;36mfit_transform\u001b[0;34m(self, X, y, **fit_params)\u001b[0m\n\u001b[1;32m    697\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0my\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    698\u001b[0m             \u001b[0;31m# fit method of arity 1 (unsupervised transformation)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 699\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtransform\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    700\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    701\u001b[0m             \u001b[0;31m# fit method of arity 2 (supervised transformation)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.8/site-packages/sklearn/impute/_knn.py\u001b[0m in \u001b[0;36mtransform\u001b[0;34m(self, X)\u001b[0m\n\u001b[1;32m    295\u001b[0m             \u001b[0mforce_all_finite\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mforce_all_finite\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    296\u001b[0m             reduce_func=process_chunk)\n\u001b[0;32m--> 297\u001b[0;31m         \u001b[0;32mfor\u001b[0m \u001b[0mchunk\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mgen\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    298\u001b[0m             \u001b[0;31m# process_chunk modifies X in place. No return value.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    299\u001b[0m             \u001b[0;32mpass\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.8/site-packages/sklearn/metrics/pairwise.py\u001b[0m in \u001b[0;36mpairwise_distances_chunked\u001b[0;34m(X, Y, reduce_func, metric, n_jobs, working_memory, **kwds)\u001b[0m\n\u001b[1;32m   1621\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1622\u001b[0m             \u001b[0mX_chunk\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0msl\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1623\u001b[0;31m         D_chunk = pairwise_distances(X_chunk, Y, metric=metric,\n\u001b[0m\u001b[1;32m   1624\u001b[0m                                      n_jobs=n_jobs, **kwds)\n\u001b[1;32m   1625\u001b[0m         if ((X is Y or Y is None)\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.8/site-packages/sklearn/utils/validation.py\u001b[0m in \u001b[0;36minner_f\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     61\u001b[0m             \u001b[0mextra_args\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m-\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mall_args\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mextra_args\u001b[0m \u001b[0;34m<=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 63\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     64\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     65\u001b[0m             \u001b[0;31m# extra_args > 0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.8/site-packages/sklearn/metrics/pairwise.py\u001b[0m in \u001b[0;36mpairwise_distances\u001b[0;34m(X, Y, metric, n_jobs, force_all_finite, **kwds)\u001b[0m\n\u001b[1;32m   1788\u001b[0m         \u001b[0mfunc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpartial\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdistance\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcdist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmetric\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mmetric\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1789\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1790\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0m_parallel_pairwise\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mY\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mn_jobs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1791\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1792\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.8/site-packages/sklearn/metrics/pairwise.py\u001b[0m in \u001b[0;36m_parallel_pairwise\u001b[0;34m(X, Y, func, n_jobs, **kwds)\u001b[0m\n\u001b[1;32m   1357\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1358\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0meffective_n_jobs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mn_jobs\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1359\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mY\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1360\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1361\u001b[0m     \u001b[0;31m# enforce a threading backend to prevent data communication overhead\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.8/site-packages/sklearn/utils/validation.py\u001b[0m in \u001b[0;36minner_f\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     61\u001b[0m             \u001b[0mextra_args\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m-\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mall_args\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mextra_args\u001b[0m \u001b[0;34m<=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 63\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     64\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     65\u001b[0m             \u001b[0;31m# extra_args > 0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.8/site-packages/sklearn/metrics/pairwise.py\u001b[0m in \u001b[0;36mnan_euclidean_distances\u001b[0;34m(X, Y, squared, missing_values, copy)\u001b[0m\n\u001b[1;32m    416\u001b[0m     \u001b[0mYY\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mY\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0mY\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    417\u001b[0m     \u001b[0mdistances\u001b[0m \u001b[0;34m-=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mXX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmissing_Y\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mT\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 418\u001b[0;31m     \u001b[0mdistances\u001b[0m \u001b[0;34m-=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmissing_X\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mYY\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mT\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    419\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    420\u001b[0m     \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdistances\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdistances\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "from sklearn.impute import KNNImputer\n",
    "imputer = KNNImputer(n_neighbors=2)\n",
    "train_df_std_4_nan = pd.DataFrame(imputer.fit_transform(train_df_std), \n",
    "                                  index=train_df_std.index, columns=train_df_std.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns_without_nan = train_df_std.columns[train_df_std.notna().any()].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df_std_without_nan = train_df_std_1[columns_without_nan]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df_std_4 = pd.concat([train_df_std_4_nan, train_df_std_without_nan], sort = False, axis = 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "При 1 варианте accuracy равна: 0.9301587301587302  \n",
    "При 2 варианте accuracy равна: 0.930031746031746  \n",
    "При 3 варианте accuracy равна: 0.9463462005341102"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Бейзлайн с этими признаками"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    29422\n",
       "1     2076\n",
       "Name: default_12m, dtype: int64"
      ]
     },
     "execution_count": 318,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df_std['default_12m'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9301587301587302"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "X_train, X_test, y_train, y_test = train_test_split(.drop('default_12m',axis=1),\n",
    "                                                  train_df_std_1['default_12m'],\n",
    "                                                    random_state=42, shuffle=True)\n",
    "\n",
    "logit = LogisticRegression(max_iter=1000).fit(X_train, y_train)\n",
    "y_pred = logit.predict(X_test)\n",
    "accuracy_score(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Скоринговая карта"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Подсчет WoE и IV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calc_woe_iv(df, feature, target):\n",
    "    \n",
    "    vals = list(df[feature].unique())\n",
    "    count_all = []\n",
    "    default = []\n",
    "    non_default = []\n",
    "    \n",
    "    for i in vals:\n",
    "        count_all.append(len(df[df[feature] == i]))\n",
    "        default.append(len(df[(df[target] == 1) & (df[feature] == i)]))\n",
    "        non_default.append(len(df[(df[target] == 0) & (df[feature] == i)]))\n",
    "    \n",
    "    data = {'value': vals, 'count_all': count_all, 'default': default, 'non_default': non_default}\n",
    "    df_woe_iv = pd.DataFrame(data)\n",
    "        \n",
    "    df_woe_iv['Distr_non_default'] = df_woe_iv['non_default'] / df_woe_iv['non_default'].sum()\n",
    "    df_woe_iv['Distr_default'] = df_woe_iv['default'] / df_woe_iv['default'].sum()\n",
    "    df_woe_iv['WoE'] = np.log(df_woe_iv['Distr_non_default'] / df_woe_iv['Distr_default'])\n",
    "    df_woe_iv = df_woe_iv.replace({'WoE': {np.inf: 0, -np.inf: 0}})\n",
    "    df_woe_iv['IV'] = (df_woe_iv['Distr_non_default'] - df_woe_iv['Distr_default']) * df_woe_iv['WoE']\n",
    "    iv = df_woe_iv['IV'].sum()\n",
    "    \n",
    "    return df_woe_iv, iv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(    value  count_all  default  non_default  Distr_non_default  Distr_default  \\\n",
       " 0       0      32071     2085        29986           0.989506       0.997131   \n",
       " 1       1        216        6          210           0.006930       0.002869   \n",
       " 2       2         42        0           42           0.001386       0.000000   \n",
       " 3       5         11        0           11           0.000363       0.000000   \n",
       " 4       4         11        0           11           0.000363       0.000000   \n",
       " 5       3         19        0           19           0.000627       0.000000   \n",
       " 6      10          6        0            6           0.000198       0.000000   \n",
       " 7       8          5        0            5           0.000165       0.000000   \n",
       " 8      12          3        0            3           0.000099       0.000000   \n",
       " 9       6          7        0            7           0.000231       0.000000   \n",
       " 10     20          2        0            2           0.000066       0.000000   \n",
       " 11     17          1        0            1           0.000033       0.000000   \n",
       " 12     11          1        0            1           0.000033       0.000000   \n",
       " \n",
       "          WoE        IV  \n",
       " 0  -0.007676  0.000059  \n",
       " 1   0.881711  0.003580  \n",
       " 2   0.000000  0.000000  \n",
       " 3   0.000000  0.000000  \n",
       " 4   0.000000  0.000000  \n",
       " 5   0.000000  0.000000  \n",
       " 6   0.000000  0.000000  \n",
       " 7   0.000000  0.000000  \n",
       " 8   0.000000  0.000000  \n",
       " 9   0.000000  0.000000  \n",
       " 10  0.000000  0.000000  \n",
       " 11  0.000000  0.000000  \n",
       " 12  0.000000  0.000000  ,\n",
       " 0.0036385636149679447)"
      ]
     },
     "execution_count": 262,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "calc_woe_iv(train_df, 'ul_branch_cnt', 'default_12m')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calc_iv_features(df, target):\n",
    "    \n",
    "    ivs = []\n",
    "    feature_names = df.columns\n",
    "    for i in feature_names:\n",
    "        df_woe_iv = calc_woe_iv(df, i, target)[0]\n",
    "        ivs.append(df_woe_iv['IV'].sum())\n",
    "    \n",
    "    data = {'feature_name': feature_names, 'iv': ivs}\n",
    "    df_iv = pd.DataFrame(data)\n",
    "    return df_iv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {},
   "outputs": [],
   "source": [
    "del train_df_outliers['ul_staff_range']\n",
    "#del train_df_outliers['default_12m']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {},
   "outputs": [],
   "source": [
    "number_bins = range(1, 20)\n",
    "ivvv = []\n",
    "\n",
    "for feature in train_df_outliers.columns:\n",
    "    if feature == 'default_12m':\n",
    "        continue\n",
    "    name = 'cut_' + str(feature)\n",
    "    for i in number_bins:\n",
    "        train_df_outliers[name] = pd.cut(train_df_outliers[feature], i)\n",
    "        ivvv.append(calc_woe_iv(train_df_outliers, name, 'default_12m')[1])\n",
    "\n",
    "    bin_iv = {}\n",
    "    for A, B in zip(number_bins, ivvv):\n",
    "        bin_iv[A] = B\n",
    "    cut_bin = max(bin_iv, key=bin_iv.get)\n",
    "\n",
    "    train_df_outliers[name] = pd.cut(train_df_outliers[feature], cut_bin)\n",
    "    del train_df_outliers[feature]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/user/opt/anaconda3/lib/python3.8/site-packages/pandas/core/arraylike.py:358: RuntimeWarning: divide by zero encountered in log\n",
      "  result = getattr(ufunc, method)(*inputs, **kwargs)\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>feature_name</th>\n",
       "      <th>iv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>default_12m</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>cut_ar_sale_cost</td>\n",
       "      <td>0.114166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>cut_ar_selling_expenses</td>\n",
       "      <td>0.002176</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>cut_ar_management_expenses</td>\n",
       "      <td>0.003116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>cut_ar_sale_profit</td>\n",
       "      <td>0.045288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>cut_ar_balance_of_rvns_and_expns</td>\n",
       "      <td>0.072686</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>cut_ar_other_profit_and_losses</td>\n",
       "      <td>0.010265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>cut_ab_immobilized_assets</td>\n",
       "      <td>0.041659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>cut_ab_inventory</td>\n",
       "      <td>0.092467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>cut_ab_accounts_receivable</td>\n",
       "      <td>0.007831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>cut_ab_other_current_assets</td>\n",
       "      <td>0.024790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>cut_ab_cash_and_securities</td>\n",
       "      <td>0.036218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>cut_ab_own_capital</td>\n",
       "      <td>0.072998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>cut_ab_long_term_liabilities</td>\n",
       "      <td>0.008637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>cut_ab_short_term_borrowing</td>\n",
       "      <td>0.010717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>cut_ab_other_borrowings</td>\n",
       "      <td>0.018103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>cut_ogrn_age</td>\n",
       "      <td>0.249547</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>cut_adr_actual_age</td>\n",
       "      <td>0.055967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>cut_ul_capital_sum</td>\n",
       "      <td>0.006924</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>cut_ul_founders_cnt</td>\n",
       "      <td>0.004577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>cut_ul_branch_cnt</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>cut_ul_strategic_flg</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>cut_borrowed_capital</td>\n",
       "      <td>0.014603</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        feature_name        iv\n",
       "0                        default_12m  0.000000\n",
       "1                   cut_ar_sale_cost  0.114166\n",
       "2            cut_ar_selling_expenses  0.002176\n",
       "3         cut_ar_management_expenses  0.003116\n",
       "4                 cut_ar_sale_profit  0.045288\n",
       "5   cut_ar_balance_of_rvns_and_expns  0.072686\n",
       "6     cut_ar_other_profit_and_losses  0.010265\n",
       "7          cut_ab_immobilized_assets  0.041659\n",
       "8                   cut_ab_inventory  0.092467\n",
       "9         cut_ab_accounts_receivable  0.007831\n",
       "10       cut_ab_other_current_assets  0.024790\n",
       "11        cut_ab_cash_and_securities  0.036218\n",
       "12                cut_ab_own_capital  0.072998\n",
       "13      cut_ab_long_term_liabilities  0.008637\n",
       "14       cut_ab_short_term_borrowing  0.010717\n",
       "15           cut_ab_other_borrowings  0.018103\n",
       "16                      cut_ogrn_age  0.249547\n",
       "17                cut_adr_actual_age  0.055967\n",
       "18                cut_ul_capital_sum  0.006924\n",
       "19               cut_ul_founders_cnt  0.004577\n",
       "20                 cut_ul_branch_cnt  0.000000\n",
       "21              cut_ul_strategic_flg  0.000000\n",
       "22              cut_borrowed_capital  0.014603"
      ]
     },
     "execution_count": 268,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "calc_iv_features(train_df_outliers, 'default_12m')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "     class 0       0.95      1.00      0.97      3899\n",
      "     class 1       0.00      0.00      0.00       220\n",
      "\n",
      "    accuracy                           0.95      4119\n",
      "   macro avg       0.47      0.50      0.49      4119\n",
      "weighted avg       0.90      0.95      0.92      4119\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "target_names = ['class 0', 'class 1']\n",
    "print(classification_report(y_test, y_pred, target_names=target_names))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Скоринговая карта "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''\n",
    "coefficients = pd.DataFrame({\"Feature\":X_train.columns,\"Coefficients\":np.transpose(clf_lr.coef_[0] )})\n",
    "print(coefficients.head(10))\n",
    "print(clf_lr.intercept_)\n",
    "beta = coefficients[coefficients['Feature']=='A3'].Coefficients.iloc[0]\n",
    "alpha = 0\n",
    "n = X_train.shape[1]\n",
    "import math\n",
    "factor = 40/math.log(2)\n",
    "offset = 600 -57.7 * math.log(50)\n",
    "o = offset/n\n",
    "beta = 0.2\n",
    "\n",
    "for kk in final_iv[final_iv['VAR_NAME']=='A3'].WOE:\n",
    "    ll = round(beta*kk+alpha/n,3)\n",
    "    tt=ll*factor\n",
    "    score = tt + o\n",
    "    print(round(score))\n",
    "print(final_iv[final_iv['VAR_NAME']=='A3'])\n",
    "\n",
    "'''"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Начало их кода "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tIkwdsWrN92b"
   },
   "source": [
    "### Строим бэйзлайн модель c 3-ми фичами"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "A_XNOaYeN92b"
   },
   "outputs": [],
   "source": [
    "features = ['bus_age','ul_capital_sum', 'ul_founders_cnt']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "GzVaIYtKN92e"
   },
   "outputs": [],
   "source": [
    "features_df = train_df.loc[:,features]\n",
    "features_df['default_12m'] = train_df['default_12m']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "id": "iGuLdQOKN92h",
    "outputId": "c5179ed0-cea4-4309-ed35-9cf804d37728"
   },
   "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>bus_age</th>\n",
       "      <th>ul_capital_sum</th>\n",
       "      <th>ul_founders_cnt</th>\n",
       "      <th>default_12m</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>198</td>\n",
       "      <td>50000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>256</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>33</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>185</td>\n",
       "      <td>300000.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>220</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   bus_age  ul_capital_sum  ul_founders_cnt  default_12m\n",
       "0      198         50000.0                2            0\n",
       "1      256         10000.0                5            0\n",
       "2       33         10000.0                2            0\n",
       "3      185        300000.0                1            0\n",
       "4      220         10000.0                2            0"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "zsaqlg0rN92k"
   },
   "source": [
    "### Построим карту коллеряций"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 173
    },
    "id": "PuTGDJNlN92l",
    "outputId": "b01d4f62-00b9-4024-ad02-8755cc37a3a1"
   },
   "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>bus_age</th>\n",
       "      <th>ul_capital_sum</th>\n",
       "      <th>ul_founders_cnt</th>\n",
       "      <th>default_12m</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bus_age</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.066929</td>\n",
       "      <td>0.209451</td>\n",
       "      <td>-0.109389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ul_capital_sum</th>\n",
       "      <td>0.066929</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.029165</td>\n",
       "      <td>-0.019452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ul_founders_cnt</th>\n",
       "      <td>0.209451</td>\n",
       "      <td>0.029165</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.023244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>default_12m</th>\n",
       "      <td>-0.109389</td>\n",
       "      <td>-0.019452</td>\n",
       "      <td>-0.023244</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  bus_age  ul_capital_sum  ul_founders_cnt  default_12m\n",
       "bus_age          1.000000        0.066929         0.209451    -0.109389\n",
       "ul_capital_sum   0.066929        1.000000         0.029165    -0.019452\n",
       "ul_founders_cnt  0.209451        0.029165         1.000000    -0.023244\n",
       "default_12m     -0.109389       -0.019452        -0.023244     1.000000"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features_df.corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "DjygjQIM8u4Y"
   },
   "source": [
    "### Валидация результата"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "NaH5PC4wN92n",
    "outputId": "f14f8971-208c-4db8-d1d7-d5d73361ab2c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6483643068016243"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(features_df.drop('default_12m',axis=1),\n",
    "                                                    features_df['default_12m'],\n",
    "                                                    random_state=42, shuffle=True)\n",
    "\n",
    "logit = LogisticRegression(C=1,solver='lbfgs',max_iter=1000)\n",
    "cv_baseline = cross_val_score(logit, X_train, y_train, cv=5, scoring='roc_auc').mean()\n",
    "cv_baseline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true,
    "id": "13BN0BG1N92q"
   },
   "source": [
    "ROC/AUC бэйзлайн модели - 64,83%"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JKsc2E3d8_ZN"
   },
   "source": [
    "### Построение модели на всех данных"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "8k4uVr9Q8sYs",
    "outputId": "a043e8ac-4279-4e7b-bb51-9b618cc74006"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=1, class_weight=None, dual=False, fit_intercept=True,\n",
       "                   intercept_scaling=1, l1_ratio=None, max_iter=1000,\n",
       "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
       "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
       "                   warm_start=False)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "logit.fit(features_df.drop('default_12m',axis=1), features_df['default_12m'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "f5Y9G9HCGo_c"
   },
   "outputs": [],
   "source": [
    "logit.predict(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OQ1Y1r6E9bvN"
   },
   "source": [
    "### Формирование файла с ответом"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "_9YdWZm8-7k0"
   },
   "outputs": [],
   "source": [
    "answ_df = test_df[['record_id']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 115
    },
    "id": "RizmO5g4_wPO",
    "outputId": "1cf87965-5f56-4a89-f90e-572bcf5d3736"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "answ_df['id'] = answ_df['record_id']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 98
    },
    "id": "r7fiDf_l9igh",
    "outputId": "f29b6e54-12ee-4b9d-88f5-3b58036adc18"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.7/site-packages/pandas/core/frame.py:3997: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  errors=errors,\n"
     ]
    }
   ],
   "source": [
    "\n",
    "answ_df.drop('record_id',axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ye8YVTYG90_6"
   },
   "outputs": [],
   "source": [
    "cutoff = y_train.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 115
    },
    "id": "GtV2vm1o9tMF",
    "outputId": "271a9a2e-3fe6-4182-a75a-9cdb60029d6b"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "answ_df['predict'] = (logit.predict_proba(test_df.loc[:,features])[:,1]>cutoff).astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "BAet0WO6-A_5"
   },
   "outputs": [],
   "source": [
    "answ_df.to_csv('PD-submit.csv',index=False, sep=';')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "G7OdXngvN92r"
   },
   "source": [
    "Для загрузки своего рузультата на https://dsbattle.com/hackathons/juniords-new/ вам нужно создать csv файл с двумя колонками (id, predict), разделитель - точка с запятой (;)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "PD_jun_ds",
   "provenance": [],
   "toc_visible": true
  },
  "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}