hw2-task.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SkillFactory\n",
    "## Введение в ML, введение в sklearn"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "В этом задании мы с вами рассмотрим данные с конкурса [Задача предсказания отклика клиентов ОТП Банка](http://www.machinelearning.ru/wiki/index.php?title=%D0%97%D0%B0%D0%B4%D0%B0%D1%87%D0%B0_%D0%BF%D1%80%D0%B5%D0%B4%D1%81%D0%BA%D0%B0%D0%B7%D0%B0%D0%BD%D0%B8%D1%8F_%D0%BE%D1%82%D0%BA%D0%BB%D0%B8%D0%BA%D0%B0_%D0%BA%D0%BB%D0%B8%D0%B5%D0%BD%D1%82%D0%BE%D0%B2_%D0%9E%D0%A2%D0%9F_%D0%91%D0%B0%D0%BD%D0%BA%D0%B0_%28%D0%BA%D0%BE%D0%BD%D0%BA%D1%83%D1%80%D1%81%29)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "plt.style.use('ggplot')\n",
    "plt.rcParams['figure.figsize'] = (12,5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Грузим данные"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Считаем описание данных"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "df_descr = pd.read_csv('otp_description.csv', sep='\\t', encoding='utf8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ПОЛЕ</th>\n",
       "      <th>ОПИСАНИЕ</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AGREEMENT_RK</td>\n",
       "      <td>уникальный идентификатор объекта в выборке</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AGE</td>\n",
       "      <td>возраст клиента</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SOCSTATUS_WORK_FL</td>\n",
       "      <td>социальный статус клиента относительно работы ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SOCSTATUS_PENS_FL</td>\n",
       "      <td>социальный статус клиента относительно пенсии ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>GENDER</td>\n",
       "      <td>пол клиента (1- женщины, 0-мужчины)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>CHILD_TOTAL</td>\n",
       "      <td>количество детей клиента</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>DEPENDANTS</td>\n",
       "      <td>количество иждивенцев клиента</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>EDUCATION</td>\n",
       "      <td>образование</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>MARITAL_STATUS</td>\n",
       "      <td>семейное положение</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>GEN_INDUSTRY</td>\n",
       "      <td>отрасль работы клиента</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>GEN_TITLE</td>\n",
       "      <td>должность</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>ORG_TP_STATE</td>\n",
       "      <td>форма собственности компании</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>ORG_TP_FCAPITAL</td>\n",
       "      <td>отношение к иностранному капиталу</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>JOB_DIR</td>\n",
       "      <td>направление деятельности в нутри компании</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>FAMILY_INCOME</td>\n",
       "      <td>семейный доход (несколько категорий)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>PERSONAL_INCOME</td>\n",
       "      <td>личный доход клиента (в рублях)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>REG_ADDRESS_PROVINCE</td>\n",
       "      <td>область регистрации клиента</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>FACT_ADDRESS_PROVINCE</td>\n",
       "      <td>область фактического пребывания клиента</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>POSTAL_ADDRESS_PROVINCE</td>\n",
       "      <td>почтовый адрес область</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>TP_PROVINCE</td>\n",
       "      <td>область торговой точки, где клиент брал послед...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>REGION_NM</td>\n",
       "      <td>регион РФ</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>REG_FACT_FL</td>\n",
       "      <td>адрес регистрации и адрес фактического пребыва...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>FACT_POST_FL</td>\n",
       "      <td>адрес фактического пребывания клиента и его по...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>REG_POST_FL</td>\n",
       "      <td>адрес регистрации клиента и его почтовый адрес...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>REG_FACT_POST_FL</td>\n",
       "      <td>почтовый, фактический и адрес регистрации совп...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>REG_FACT_POST_TP_FL</td>\n",
       "      <td>область регистрации, фактического пребывания, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>FL_PRESENCE_FL</td>\n",
       "      <td>наличие в собственности квартиры (1 - есть, 0 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>OWN_AUTO</td>\n",
       "      <td>кол-во автомобилей в собственности</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>AUTO_RUS_FL</td>\n",
       "      <td>наличие в собственности автомобиля российского...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>HS_PRESENCE_FL</td>\n",
       "      <td>наличие в собственности загородного дома (1 - ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>COT_PRESENCE_FL</td>\n",
       "      <td>наличие в собственности котеджа (1 - есть, 0 -...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>GAR_PRESENCE_FL</td>\n",
       "      <td>наличие в собственности гаража (1 - есть, 0 - ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>LAND_PRESENCE_FL</td>\n",
       "      <td>наличие в собственности земельного участка (1 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>CREDIT</td>\n",
       "      <td>сумма последнего кредита клиента (в рублях)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>TERM</td>\n",
       "      <td>срок кредита</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>FST_PAYMENT</td>\n",
       "      <td>первоначальный взнос (в рублях)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>DL_DOCUMENT_FL</td>\n",
       "      <td>в анкете клиент указал водительское удостовере...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>GPF_DOCUMENT_FL</td>\n",
       "      <td>в анкете клиен указал ГПФ (1 - указал, 0 - не ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>FACT_LIVING_TERM</td>\n",
       "      <td>количество месяцев проживания по месту фактиче...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>WORK_TIME</td>\n",
       "      <td>время работы на текущем месте (в месяцах)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>FACT_PHONE_FL</td>\n",
       "      <td>наличие в заявке телефона по фактическому мест...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>REG_PHONE_FL</td>\n",
       "      <td>наличие в заявке телефона по месту регистрации</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>GEN_PHONE_FL</td>\n",
       "      <td>наличие в заявке рабочего телефона</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>LOAN_NUM_TOTAL</td>\n",
       "      <td>количество ссуд клиента</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>LOAN_NUM_CLOSED</td>\n",
       "      <td>количество погашенных ссуд клиента</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>LOAN_NUM_PAYM</td>\n",
       "      <td>количество платежей, которые сделал клиент</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>LOAN_DLQ_NUM</td>\n",
       "      <td>количество просрочек, допущенных клиентом</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>LOAN_MAX_DLQ</td>\n",
       "      <td>номер максимальной просрочки, допущенной клиентом</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>LOAN_AVG_DLQ_AMT</td>\n",
       "      <td>средняя сумма просрочки (в рублях)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>LOAN_MAX_DLQ_AMT</td>\n",
       "      <td>максимальная сумма просрочки (в рублях)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>PREVIOUS_CARD_NUM_UTILIZED</td>\n",
       "      <td>количество уже утилизированных карт ( если пус...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          ПОЛЕ  \\\n",
       "0                 AGREEMENT_RK   \n",
       "1                          AGE   \n",
       "2            SOCSTATUS_WORK_FL   \n",
       "3            SOCSTATUS_PENS_FL   \n",
       "4                       GENDER   \n",
       "5                  CHILD_TOTAL   \n",
       "6                   DEPENDANTS   \n",
       "7                    EDUCATION   \n",
       "8               MARITAL_STATUS   \n",
       "9                 GEN_INDUSTRY   \n",
       "10                   GEN_TITLE   \n",
       "11                ORG_TP_STATE   \n",
       "12             ORG_TP_FCAPITAL   \n",
       "13                     JOB_DIR   \n",
       "14               FAMILY_INCOME   \n",
       "15             PERSONAL_INCOME   \n",
       "16        REG_ADDRESS_PROVINCE   \n",
       "17       FACT_ADDRESS_PROVINCE   \n",
       "18     POSTAL_ADDRESS_PROVINCE   \n",
       "19                 TP_PROVINCE   \n",
       "20                   REGION_NM   \n",
       "21                 REG_FACT_FL   \n",
       "22                FACT_POST_FL   \n",
       "23                 REG_POST_FL   \n",
       "24            REG_FACT_POST_FL   \n",
       "25         REG_FACT_POST_TP_FL   \n",
       "26              FL_PRESENCE_FL   \n",
       "27                    OWN_AUTO   \n",
       "28                 AUTO_RUS_FL   \n",
       "29              HS_PRESENCE_FL   \n",
       "30             COT_PRESENCE_FL   \n",
       "31             GAR_PRESENCE_FL   \n",
       "32            LAND_PRESENCE_FL   \n",
       "33                      CREDIT   \n",
       "34                        TERM   \n",
       "35                 FST_PAYMENT   \n",
       "36              DL_DOCUMENT_FL   \n",
       "37             GPF_DOCUMENT_FL   \n",
       "38            FACT_LIVING_TERM   \n",
       "39                   WORK_TIME   \n",
       "40               FACT_PHONE_FL   \n",
       "41                REG_PHONE_FL   \n",
       "42                GEN_PHONE_FL   \n",
       "43              LOAN_NUM_TOTAL   \n",
       "44             LOAN_NUM_CLOSED   \n",
       "45               LOAN_NUM_PAYM   \n",
       "46                LOAN_DLQ_NUM   \n",
       "47                LOAN_MAX_DLQ   \n",
       "48            LOAN_AVG_DLQ_AMT   \n",
       "49            LOAN_MAX_DLQ_AMT   \n",
       "50  PREVIOUS_CARD_NUM_UTILIZED   \n",
       "\n",
       "                                             ОПИСАНИЕ  \n",
       "0          уникальный идентификатор объекта в выборке  \n",
       "1                                     возраст клиента  \n",
       "2   социальный статус клиента относительно работы ...  \n",
       "3   социальный статус клиента относительно пенсии ...  \n",
       "4                 пол клиента (1- женщины, 0-мужчины)  \n",
       "5                            количество детей клиента  \n",
       "6                       количество иждивенцев клиента  \n",
       "7                                         образование  \n",
       "8                                  семейное положение  \n",
       "9                              отрасль работы клиента  \n",
       "10                                          должность  \n",
       "11                       форма собственности компании  \n",
       "12                  отношение к иностранному капиталу  \n",
       "13          направление деятельности в нутри компании  \n",
       "14               семейный доход (несколько категорий)  \n",
       "15                    личный доход клиента (в рублях)  \n",
       "16                        область регистрации клиента  \n",
       "17            область фактического пребывания клиента  \n",
       "18                             почтовый адрес область  \n",
       "19  область торговой точки, где клиент брал послед...  \n",
       "20                                          регион РФ  \n",
       "21  адрес регистрации и адрес фактического пребыва...  \n",
       "22  адрес фактического пребывания клиента и его по...  \n",
       "23  адрес регистрации клиента и его почтовый адрес...  \n",
       "24  почтовый, фактический и адрес регистрации совп...  \n",
       "25  область регистрации, фактического пребывания, ...  \n",
       "26  наличие в собственности квартиры (1 - есть, 0 ...  \n",
       "27                кол-во автомобилей в собственности   \n",
       "28  наличие в собственности автомобиля российского...  \n",
       "29  наличие в собственности загородного дома (1 - ...  \n",
       "30  наличие в собственности котеджа (1 - есть, 0 -...  \n",
       "31  наличие в собственности гаража (1 - есть, 0 - ...  \n",
       "32  наличие в собственности земельного участка (1 ...  \n",
       "33        сумма последнего кредита клиента (в рублях)  \n",
       "34                                       срок кредита  \n",
       "35                    первоначальный взнос (в рублях)  \n",
       "36  в анкете клиент указал водительское удостовере...  \n",
       "37  в анкете клиен указал ГПФ (1 - указал, 0 - не ...  \n",
       "38  количество месяцев проживания по месту фактиче...  \n",
       "39          время работы на текущем месте (в месяцах)  \n",
       "40  наличие в заявке телефона по фактическому мест...  \n",
       "41     наличие в заявке телефона по месту регистрации  \n",
       "42                 наличие в заявке рабочего телефона  \n",
       "43                            количество ссуд клиента  \n",
       "44                 количество погашенных ссуд клиента  \n",
       "45         количество платежей, которые сделал клиент  \n",
       "46          количество просрочек, допущенных клиентом  \n",
       "47  номер максимальной просрочки, допущенной клиентом  \n",
       "48                 средняя сумма просрочки (в рублях)  \n",
       "49            максимальная сумма просрочки (в рублях)  \n",
       "50  количество уже утилизированных карт ( если пус...  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_descr"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Считаем обучающую выборки и тестовую (которую мы как бы не видим)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = pd.read_csv('otp_train.csv', sep='\\t', encoding='utf8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15223, 52)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_test = pd.read_csv('otp_test.csv', sep='\\t', encoding='utf8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(14910, 52)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AGREEMENT_RK</th>\n",
       "      <th>TARGET</th>\n",
       "      <th>AGE</th>\n",
       "      <th>SOCSTATUS_WORK_FL</th>\n",
       "      <th>SOCSTATUS_PENS_FL</th>\n",
       "      <th>GENDER</th>\n",
       "      <th>CHILD_TOTAL</th>\n",
       "      <th>DEPENDANTS</th>\n",
       "      <th>EDUCATION</th>\n",
       "      <th>MARITAL_STATUS</th>\n",
       "      <th>...</th>\n",
       "      <th>REG_PHONE_FL</th>\n",
       "      <th>GEN_PHONE_FL</th>\n",
       "      <th>LOAN_NUM_TOTAL</th>\n",
       "      <th>LOAN_NUM_CLOSED</th>\n",
       "      <th>LOAN_NUM_PAYM</th>\n",
       "      <th>LOAN_DLQ_NUM</th>\n",
       "      <th>LOAN_MAX_DLQ</th>\n",
       "      <th>LOAN_AVG_DLQ_AMT</th>\n",
       "      <th>LOAN_MAX_DLQ_AMT</th>\n",
       "      <th>PREVIOUS_CARD_NUM_UTILIZED</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>59910150</td>\n",
       "      <td>0</td>\n",
       "      <td>49</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>Среднее специальное</td>\n",
       "      <td>Состою в браке</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1580</td>\n",
       "      <td>1580</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>59910230</td>\n",
       "      <td>0</td>\n",
       "      <td>32</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>Среднее</td>\n",
       "      <td>Состою в браке</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4020</td>\n",
       "      <td>4020</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>59910525</td>\n",
       "      <td>0</td>\n",
       "      <td>52</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>Неполное среднее</td>\n",
       "      <td>Состою в браке</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>59910803</td>\n",
       "      <td>0</td>\n",
       "      <td>39</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Высшее</td>\n",
       "      <td>Состою в браке</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1589,92333333333</td>\n",
       "      <td>1590</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>59911781</td>\n",
       "      <td>0</td>\n",
       "      <td>30</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Среднее</td>\n",
       "      <td>Состою в браке</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>16</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1152,15</td>\n",
       "      <td>2230</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 52 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   AGREEMENT_RK  TARGET  AGE  SOCSTATUS_WORK_FL  SOCSTATUS_PENS_FL  GENDER  \\\n",
       "0      59910150       0   49                  1                  0       1   \n",
       "1      59910230       0   32                  1                  0       1   \n",
       "2      59910525       0   52                  1                  0       1   \n",
       "3      59910803       0   39                  1                  0       1   \n",
       "4      59911781       0   30                  1                  0       0   \n",
       "\n",
       "   CHILD_TOTAL  DEPENDANTS            EDUCATION  MARITAL_STATUS  \\\n",
       "0            2           1  Среднее специальное  Состою в браке   \n",
       "1            3           3              Среднее  Состою в браке   \n",
       "2            4           0     Неполное среднее  Состою в браке   \n",
       "3            1           1               Высшее  Состою в браке   \n",
       "4            0           0              Среднее  Состою в браке   \n",
       "\n",
       "             ...             REG_PHONE_FL GEN_PHONE_FL LOAN_NUM_TOTAL  \\\n",
       "0            ...                        0            1              1   \n",
       "1            ...                        0            1              1   \n",
       "2            ...                        0            1              2   \n",
       "3            ...                        1            1              1   \n",
       "4            ...                        0            1              2   \n",
       "\n",
       "  LOAN_NUM_CLOSED LOAN_NUM_PAYM LOAN_DLQ_NUM LOAN_MAX_DLQ  LOAN_AVG_DLQ_AMT  \\\n",
       "0               1             6            2            1              1580   \n",
       "1               1             6            1            1              4020   \n",
       "2               1            11            0            0                 0   \n",
       "3               1             6            3            1  1589,92333333333   \n",
       "4               1            16            2            1           1152,15   \n",
       "\n",
       "  LOAN_MAX_DLQ_AMT PREVIOUS_CARD_NUM_UTILIZED  \n",
       "0             1580                        NaN  \n",
       "1             4020                        NaN  \n",
       "2                0                        NaN  \n",
       "3             1590                        NaN  \n",
       "4             2230                        NaN  \n",
       "\n",
       "[5 rows x 52 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Объединим две выборки\n",
    "\n",
    "Так как пока мы пока не умеем работать sklearn  Pipeline, то для того, чтобы после предобработки столбцы в двух выборках находились на своих местах.\n",
    "\n",
    "Для того, чтобы в дальнейшем отделить их введем новый столбец \"sample\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train.loc[:, 'sample'] = 'train'\n",
    "df_test.loc[:, 'sample'] = 'test'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df_test.append(df_train).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(30133, 53)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Чуть-чуть посмотрим на данные"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Посмотрим типы данных и их заполняемость"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 30133 entries, 0 to 30132\n",
      "Data columns (total 53 columns):\n",
      "AGREEMENT_RK                  30133 non-null int64\n",
      "TARGET                        30133 non-null int64\n",
      "AGE                           30133 non-null int64\n",
      "SOCSTATUS_WORK_FL             30133 non-null int64\n",
      "SOCSTATUS_PENS_FL             30133 non-null int64\n",
      "GENDER                        30133 non-null int64\n",
      "CHILD_TOTAL                   30133 non-null int64\n",
      "DEPENDANTS                    30133 non-null int64\n",
      "EDUCATION                     30133 non-null object\n",
      "MARITAL_STATUS                30133 non-null object\n",
      "GEN_INDUSTRY                  27420 non-null object\n",
      "GEN_TITLE                     27420 non-null object\n",
      "ORG_TP_STATE                  27420 non-null object\n",
      "ORG_TP_FCAPITAL               27425 non-null object\n",
      "JOB_DIR                       27420 non-null object\n",
      "FAMILY_INCOME                 30133 non-null object\n",
      "PERSONAL_INCOME               30133 non-null object\n",
      "REG_ADDRESS_PROVINCE          30133 non-null object\n",
      "FACT_ADDRESS_PROVINCE         30133 non-null object\n",
      "POSTAL_ADDRESS_PROVINCE       30133 non-null object\n",
      "TP_PROVINCE                   29543 non-null object\n",
      "REGION_NM                     30131 non-null object\n",
      "REG_FACT_FL                   30133 non-null int64\n",
      "FACT_POST_FL                  30133 non-null int64\n",
      "REG_POST_FL                   30133 non-null int64\n",
      "REG_FACT_POST_FL              30133 non-null int64\n",
      "REG_FACT_POST_TP_FL           30133 non-null int64\n",
      "FL_PRESENCE_FL                30133 non-null int64\n",
      "OWN_AUTO                      30133 non-null int64\n",
      "AUTO_RUS_FL                   30133 non-null int64\n",
      "HS_PRESENCE_FL                30133 non-null int64\n",
      "COT_PRESENCE_FL               30133 non-null int64\n",
      "GAR_PRESENCE_FL               30133 non-null int64\n",
      "LAND_PRESENCE_FL              30133 non-null int64\n",
      "CREDIT                        30133 non-null object\n",
      "TERM                          30133 non-null int64\n",
      "FST_PAYMENT                   30133 non-null object\n",
      "DL_DOCUMENT_FL                30133 non-null int64\n",
      "GPF_DOCUMENT_FL               30133 non-null int64\n",
      "FACT_LIVING_TERM              30133 non-null int64\n",
      "WORK_TIME                     27416 non-null float64\n",
      "FACT_PHONE_FL                 30133 non-null int64\n",
      "REG_PHONE_FL                  30133 non-null int64\n",
      "GEN_PHONE_FL                  30133 non-null int64\n",
      "LOAN_NUM_TOTAL                30133 non-null int64\n",
      "LOAN_NUM_CLOSED               30133 non-null int64\n",
      "LOAN_NUM_PAYM                 30133 non-null int64\n",
      "LOAN_DLQ_NUM                  30133 non-null int64\n",
      "LOAN_MAX_DLQ                  30133 non-null int64\n",
      "LOAN_AVG_DLQ_AMT              30133 non-null object\n",
      "LOAN_MAX_DLQ_AMT              30133 non-null object\n",
      "PREVIOUS_CARD_NUM_UTILIZED    600 non-null float64\n",
      "sample                        30133 non-null object\n",
      "dtypes: float64(2), int64(32), object(19)\n",
      "memory usage: 12.2+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Видим, что часть данных - object, скорее всего стоки.\n",
    "\n",
    "\n",
    "Давайте выведем эти значения для каждого столбца"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "==========\n",
      "EDUCATION\n",
      "{'Неполное среднее', 'Высшее', 'Неоконченное высшее', 'Два и более высших образования', 'Среднее специальное', 'Среднее', 'Ученая степень'}\n",
      "\n",
      "\n",
      "==========\n",
      "MARITAL_STATUS\n",
      "{'Гражданский брак', 'Состою в браке', 'Не состоял в браке', 'Разведен(а)', 'Вдовец/Вдова'}\n",
      "\n",
      "\n",
      "==========\n",
      "GEN_INDUSTRY\n",
      "{nan, 'Образование', 'Ресторанный бизнес/Общественное питание', 'Информационные технологии', 'Сельское хозяйство', 'Информационные услуги', 'Строительство', 'Салоны красоты и здоровья', 'Развлечения/Искусство', 'Сборочные производства', 'Управляющая компания', 'Банк/Финансы', 'Коммунальное хоз-во/Дорожные службы', 'Маркетинг', 'Металлургия/Промышленность/Машиностроение', 'Наука', 'Недвижимость', 'Транспорт', 'Энергетика', 'ЧОП/Детективная д-ть', 'Юридические услуги/нотариальные услуги', 'Страхование', 'Химия/Парфюмерия/Фармацевтика', 'Другие сферы', 'Здравоохранение', 'Нефтегазовая промышленность', 'Подбор персонала', 'Торговля', 'Государственная служба', 'СМИ/Реклама/PR-агенства', 'Логистика', 'Туризм'}\n",
      "\n",
      "\n",
      "==========\n",
      "GEN_TITLE\n",
      "{'Служащий', nan, 'Руководитель среднего звена', 'Работник сферы услуг', 'Высококвалифиц. специалист', 'Рабочий', 'Специалист', 'Индивидуальный предприниматель', 'Руководитель низшего звена', 'Другое', 'Партнер', 'Руководитель высшего звена', 'Военнослужащий по контракту'}\n",
      "\n",
      "\n",
      "==========\n",
      "ORG_TP_STATE\n",
      "{nan, 'Частная компания', 'Некоммерческая организация', 'Частная ком. с инос. капиталом', 'Индивидуальный предприниматель', 'Государственная комп./учреж.'}\n",
      "\n",
      "\n",
      "==========\n",
      "ORG_TP_FCAPITAL\n",
      "{nan, 'С участием', 'Без участия'}\n",
      "\n",
      "\n",
      "==========\n",
      "JOB_DIR\n",
      "{nan, 'Юридическая служба', 'Бухгалтерия, финансы, планир.', 'Служба безопасности', 'Реклама и маркетинг', 'Участие в основ. деятельности', 'Снабжение и сбыт', 'Пр-техн. обесп. и телеком.', 'Кадровая служба и секретариат', 'Адм-хоз. и трансп. службы', 'Вспомогательный техперсонал'}\n",
      "\n",
      "\n",
      "==========\n",
      "FAMILY_INCOME\n",
      "{'свыше 50000 руб.', 'от 20000 до 50000 руб.', 'от 5000 до 10000 руб.', 'до 5000 руб.', 'от 10000 до 20000 руб.'}\n",
      "\n",
      "\n",
      "==========\n",
      "PERSONAL_INCOME\n",
      "{'7300', '4515', '13150', '5680', '15380', '13900', '17900', '6400', '44000', '23500', '15200', '4590', '7900', '16800', '45000', '6800', '12800', '35000', '13800', '41000', '43000', '38000', '13600', '7050', '18000', '70000', '5230', '21500', '8726', '14700', '23000', '6670', '6180', '5300', '22955', '16050', '10700', '49000', '9200', '4900', '9100', '23100', '13450', '11400', '40000', '7650', '17800', '13300', '5050', '4600', '12400', '37000', '46000', '7000', '32640', '25000', '42500', '19000', '13000', '77000', '5446', '5790', '11800', '5000', '110000', '4200', '5582', '48000', '10100', '4100', '3400', '12100', '9600', '9330', '2300', '4400', '8000', '19700', '9628', '9700', '20500', '5200', '25800', '9400', '8300', '24', '15000', '14000', '5700', '11600', '19300', '11200', '5360', '20000', '5400', '6500', '47000', '13400', '5800', '11000', '10900', '17700', '17000', '54000', '18200', '3500', '8500', '7800', '51000', '8700', '16100', '1950', '34000', '16200', '10600', '8250', '19500', '12650', '12700', '26500', '12000', '3300', '28000', '16500', '6100', '18500', '9240', '28300', '5350', '11900', '27000', '7101', '12600', '67000', '24500', '5339', '8200', '8350', '41900', '15500', '9900', '36000', '3800', '15400', '3200', '13160', '75000', '14300', '100000', '6608', '7220', '6200', '9500', '7500', '7700', '10200', '67700', '9340', '4000', '3100', '3700', '8600', '14500', '8400', '13500', '12500', '60000', '30000', '7850', '6000', '15300', '16700', '11500', '5460', '170000', '7600', '7640', '29000', '12900', '21800', '4300', '16000', '4800', '6900', '4360', '92000', '17404', '9300', '6700', '6359', '17600', '16900', '7400', '18600', '13700', '18700', '11700', '32000', '3000', '6300', '31000', '4330', '22500', '65000', '3900', '8900', '4700', '7100', '27300', '2800', '39000', '52000', '14358', '150000', '3600', '220000', '80000', '19600', '6600', '5900', '29500', '250000', '11300', '56000', '10300', '10400', '7250', '23800', '5600', '10500', '53000', '10800', '29800', '15560', '10000', '42000', '11640', '20646,16', '17200', '17500', '7830', '26000', '4950', '15600', '8100', '7200', '2000', '8800', '5425', '5500', '50000', '68000', '12300', '19800', '24000', '6299,19', '9800', '5088', '9000', '8066', '13200', '7251', '14800', '7550', '21000', '33000', '2100', '4500', '55000', '160000', '24800', '4340', '5100', '22000'}\n",
      "\n",
      "\n",
      "==========\n",
      "REG_ADDRESS_PROVINCE\n",
      "{'Мордовская республика', 'Санкт-Петербург', 'Ханты-Мансийский АО', 'Татарстан', 'Якутия', 'Волгоградская область', 'Бурятия', 'Тульская область', 'Воронежская область', 'Самарская область', 'Псковская область', 'Карачаево-Черкесия', 'Хабаровский край', 'Белгородская область', 'Тюменская область', 'Удмуртия', 'Амурская область', 'Омская область', 'Иркутская область', 'Пензенская область', 'Горный Алтай', 'Томская область', 'Нижегородская область', 'Мурманская область', 'Москва', 'Башкирия', 'Смоленская область', 'Еврейская АО', 'Челябинская область', 'Свердловская область', 'Кабардино-Балкария', 'Красноярский край', 'Коми', 'Кемеровская область', 'Чувашия', 'Московская область', 'Камчатская область', 'Ульяновская область', 'Сахалинская область', 'Приморский край', 'Хакасия', 'Липецкая область', 'Усть-Ордынский Бурятский АО', 'Курганская область', 'Ивановская область', 'Калужская область', 'Ярославская область', 'Агинский Бурятский АО', 'Саратовская область', 'Брянская область', 'Вологодская область', 'Читинская область', 'Оренбургская область', 'Ямало-Ненецкий АО', 'Владимирская область', 'Ростовская область', 'Новосибирская область', 'Рязанская область', 'Курская область', 'Северная Осетия', 'Алтайский край', 'Тамбовская область', 'Краснодарский край', 'Дагестан', 'Эвенкийский АО', 'Астраханская область', 'Архангельская область', 'Новгородская область', 'Ленинградская область', 'Ставропольский край', 'Кировская область', 'Калмыкия', 'Тверская область', 'Орловская область', 'Калининградская область', 'Карелия', 'Костромская область', 'Марийская республика', 'Магаданская область', 'Адыгея', 'Пермская область'}\n",
      "\n",
      "\n",
      "==========\n",
      "FACT_ADDRESS_PROVINCE\n",
      "{'Мордовская республика', 'Санкт-Петербург', 'Ханты-Мансийский АО', 'Татарстан', 'Якутия', 'Волгоградская область', 'Бурятия', 'Тульская область', 'Воронежская область', 'Самарская область', 'Псковская область', 'Карачаево-Черкесия', 'Хабаровский край', 'Белгородская область', 'Тюменская область', 'Удмуртия', 'Амурская область', 'Омская область', 'Иркутская область', 'Пензенская область', 'Горный Алтай', 'Томская область', 'Нижегородская область', 'Мурманская область', 'Москва', 'Башкирия', 'Смоленская область', 'Еврейская АО', 'Челябинская область', 'Свердловская область', 'Кабардино-Балкария', 'Красноярский край', 'Коми', 'Кемеровская область', 'Чувашия', 'Московская область', 'Камчатская область', 'Ульяновская область', 'Сахалинская область', 'Приморский край', 'Хакасия', 'Липецкая область', 'Усть-Ордынский Бурятский АО', 'Курганская область', 'Ивановская область', 'Калужская область', 'Ярославская область', 'Агинский Бурятский АО', 'Саратовская область', 'Брянская область', 'Вологодская область', 'Читинская область', 'Оренбургская область', 'Ямало-Ненецкий АО', 'Владимирская область', 'Ростовская область', 'Новосибирская область', 'Рязанская область', 'Курская область', 'Северная Осетия', 'Алтайский край', 'Тамбовская область', 'Краснодарский край', 'Дагестан', 'Эвенкийский АО', 'Астраханская область', 'Архангельская область', 'Новгородская область', 'Ленинградская область', 'Ставропольский край', 'Кировская область', 'Калмыкия', 'Тверская область', 'Орловская область', 'Калининградская область', 'Карелия', 'Костромская область', 'Марийская республика', 'Магаданская область', 'Адыгея', 'Пермская область'}\n",
      "\n",
      "\n",
      "==========\n",
      "POSTAL_ADDRESS_PROVINCE\n",
      "{'Мордовская республика', 'Санкт-Петербург', 'Ханты-Мансийский АО', 'Татарстан', 'Якутия', 'Волгоградская область', 'Бурятия', 'Тульская область', 'Воронежская область', 'Самарская область', 'Псковская область', 'Карачаево-Черкесия', 'Хабаровский край', 'Белгородская область', 'Тюменская область', 'Удмуртия', 'Амурская область', 'Омская область', 'Иркутская область', 'Пензенская область', 'Горный Алтай', 'Томская область', 'Нижегородская область', 'Мурманская область', 'Москва', 'Башкирия', 'Смоленская область', 'Еврейская АО', 'Челябинская область', 'Свердловская область', 'Кабардино-Балкария', 'Красноярский край', 'Коми', 'Кемеровская область', 'Чувашия', 'Московская область', 'Камчатская область', 'Ульяновская область', 'Сахалинская область', 'Приморский край', 'Хакасия', 'Липецкая область', 'Усть-Ордынский Бурятский АО', 'Курганская область', 'Ивановская область', 'Калужская область', 'Ярославская область', 'Агинский Бурятский АО', 'Саратовская область', 'Брянская область', 'Вологодская область', 'Читинская область', 'Оренбургская область', 'Ямало-Ненецкий АО', 'Владимирская область', 'Ростовская область', 'Новосибирская область', 'Рязанская область', 'Курская область', 'Северная Осетия', 'Алтайский край', 'Тамбовская область', 'Краснодарский край', 'Эвенкийский АО', 'Астраханская область', 'Архангельская область', 'Новгородская область', 'Ленинградская область', 'Ставропольский край', 'Кировская область', 'Калмыкия', 'Тверская область', 'Орловская область', 'Калининградская область', 'Карелия', 'Костромская область', 'Марийская республика', 'Магаданская область', 'Адыгея', 'Пермская область'}\n",
      "\n",
      "\n",
      "==========\n",
      "TP_PROVINCE\n",
      "{nan, 'Тамбовская область', 'Краснодарский край', 'Ульяновская область', 'Сахалинская область', 'Мордовская республика', 'Горный Алтай', 'Приморский край', 'Астраханская область', 'Липецкая область', 'Томская область', 'Санкт-Петербург', 'Архангельская область', 'Новгородская область', 'Ханты-Мансийский АО', 'Нижегородская область', 'Ставропольский край', 'Мурманская область', 'Татарстан', 'Ивановская область', 'Калужская область', 'Якутия', 'Башкирия', 'Кировская область', 'Курганская область', 'Ярославская область', 'Москва', 'Саратовская область', 'Смоленская область', 'Брянская область', 'Волгоградская область', 'Еврейская АО', 'Вологодская область', 'Тверская область', 'Орловская область', 'Бурятия', 'Челябинская область', 'Читинская область', 'Оренбургская область', 'Тульская область', 'Калининградская область', 'Воронежская область', 'Карелия', 'Костромская область', 'Марийская республика', 'Свердловская область', 'Ростовская область', 'Владимирская область', 'Кабардино-Балкария', 'Красноярский край', 'Самарская область', 'Магаданская область', 'Псковская область', 'Коми', 'Кемеровская область', 'Чувашия', 'Хабаровский край', 'Белгородская область', 'Адыгея', 'Новосибирская область', 'Тюменская область', 'Удмуртия', 'Курская область', 'Рязанская область', 'Амурская область', 'Камчатская область', 'Омская область', 'Пермская область', 'Иркутская область', 'Алтайский край', 'Пензенская область'}\n",
      "\n",
      "\n",
      "==========\n",
      "REGION_NM\n",
      "{nan, 'ЦЕНТРАЛЬНЫЙ 1', 'ЗАПАДНО-СИБИРСКИЙ', 'ПОВОЛЖСКИЙ', 'ЮЖНЫЙ', 'ВОСТОЧНО-СИБИРСКИЙ', 'УРАЛЬСКИЙ', 'ДАЛЬНЕВОСТОЧНЫЙ', 'СЕВЕРО-ЗАПАДНЫЙ', 'ПРИВОЛЖСКИЙ', 'ЦЕНТРАЛЬНЫЙ 2', 'ЦЕНТРАЛЬНЫЙ ОФИС'}\n",
      "\n",
      "\n",
      "==========\n",
      "CREDIT\n",
      "{'29430', '6510', '4727', '16280', '69000', '30596', '14107', '23280', '19627', '18027', '42689', '10917', '5745', '21520', '16099', '15720,91', '8950', '30603', '10700', '8080', '4750', '61832', '2719,2', '10040', '3659,5', '31263', '6609', '5790', '3860', '3411', '21943', '6454,8', '11830', '14560', '10021', '7609,5', '4489', '9994', '6230', '3690', '19457', '9349', '7296', '8278', '70047', '21328', '12080', '5120', '22469', '22977', '11230', '17220', '8910', '7066', '12910', '8549', '20155', '5665', '29461,34', '37560', '7220', '4435', '5408', '35960', '3123', '11853', '6345', '9430,38', '8373', '20475', '47900', '28971', '19205', '16197', '17605', '22482', '8041', '20310', '26489', '8088', '41500', '11039', '25260', '10773', '19970', '6793', '27980', '2090', '5184', '28890', '3014', '14460', '6787,8', '12384', '31990,6', '23800', '18740,5', '7090', '19316', '42740', '8880', '22398', '8253', '6860', '22139', '3215', '7285', '98042', '40574', '13650', '9350', '9720', '8360', '18643', '7876', '22147', '5938', '15036', '19491', '15319', '17595', '8140', '2936', '24179', '13150', '19900', '4756,5', '13180', '7890', '6455', '6491', '26695,8', '29388', '18450', '24502', '19046', '16547,3', '3646', '23358', '15164', '12710', '27598', '56673', '4372,2', '12246', '5285', '14940', '3743', '25000', '26710', '16278', '26900', '13040', '20755', '25271', '13321', '3598', '34641', '16395', '11315', '14260', '33320', '9009', '9268', '15150', '12650', '12097', '21620', '35950', '15820', '37016', '20820,44', '7084,28', '5487', '22140', '13190', '28090', '4130', '8850', '12345', '6678', '13179', '31294', '6820', '25631', '6639', '26174', '14047', '20370', '13847', '2250', '86900', '24490', '22962', '28098', '10585', '5259', '13260', '8775', '25240', '3745', '13591', '6647', '20400', '11871', '19390', '5931,8', '79770', '6146', '21204', '14646', '5355', '7978', '37640,77', '13623', '2119', '4454', '22356', '53169', '86041', '5150', '5570', '19440', '23351', '82700,9', '27060', '83106', '29382', '26000', '10374', '19930', '8770', '16870', '9456', '14060', '29460', '13710', '4405', '6561,9', '47223', '22575', '11637', '81423', '24858', '39980', '23577', '10424,6', '9536', '21146,4', '6255,15', '18450,8', '11394', '11519,1', '28160', '17193', '30422', '36177', '23485', '11322', '19080', '62000', '3501', '9804', '10061', '35450', '13830', '23166', '31560', '5688', '9985', '21890', '23197', '10855', '32310', '7999', '16988', '27990', '15390', '85270', '4089', '6653', '10830', '9438', '21340', '15099', '6643', '15653', '25960', '4060', '9679', '20225', '74700', '8311,29', '16766', '23852', '13320', '8741', '2380', '2844', '12700', '36490', '21405', '14930', '26510', '9718', '23391', '2979', '23793', '22509', '14765', '100000', '19590', '9557,1', '5482,3', '3497', '10990', '23147', '5632', '4495', '13289', '12799', '11136', '4973', '21760', '4300', '28162', '13835', '7478', '12315', '9760', '6899', '21624', '7399', '13620', '7100', '15017', '28684', '6360', '32576,5', '11595', '11202,3', '4199', '24250', '6513', '48800', '14718', '7954', '33680', '98030', '49227', '4966', '17263', '7713,42', '11298', '21170', '13926', '24945', '22050', '29653', '38810', '16070', '32800', '6898', '20317', '10778', '21460', '18566', '15115', '20862', '10233', '9504', '3270', '34150', '19740', '17091,9', '24000', '25980', '24260', '9927', '6340', '16420', '26543', '4098', '4499', '12636', '5805', '17040', '21382', '2700', '6145', '14840', '12591', '4527', '18150', '5929,95', '13259', '9588', '43353', '2299', '13050', '24599', '40200', '10735', '51900', '4375', '8150', '23516,84', '7310', '5085', '31415', '13430', '5019', '33598', '25587', '10019', '25570', '6097,5', '23300', '59224', '5960', '14288', '19300', '4440', '24897', '22999', '15512', '27200', '26260', '18702', '16107', '32100', '18445', '7189,6', '14690', '2224,8', '3790', '3520,8', '17940', '5984', '9950', '4212,9', '14044', '19945,35', '63306', '7780', '16848', '21740', '23126', '3035', '8230', '9728', '2570', '2333', '22168', '14185', '2360', '10238,82', '4257', '20129', '17080', '88312', '21300', '20294', '10737', '6480', '2440', '20179,3', '4452,91', '3825', '22428', '17076,2', '5329', '11576', '19324', '3205', '25317', '10300', '19550', '18855', '32828', '15470', '5520', '6279', '16799,4', '22075', '5616,15', '38108', '11113', '24740', '49750', '8386', '22011', '7872', '11383', '19993,4', '9590', '7105', '13520', '22394', '13480', '44415', '33400', '4228', '25900', '17567', '5297', '16999', '15147', '9347', '14243', '16960', '18296', '10540', '24929', '2268', '11765', '6018,39', '14629', '19286', '72885,57', '29950', '36400', '15975', '8912', '25454', '97026', '9689', '10950', '6096,13', '8648,5', '5300', '3339', '8610', '17496', '14730', '12998', '23945', '31810', '20768', '11368', '16833', '17593', '15045', '4996,8', '8300', '15121', '28310', '20381', '66578', '27228', '5482', '16830', '26397', '9680', '5579', '6299,72', '19950', '29691', '11943', '10349', '8597', '6396', '14796', '41021', '5504', '17370', '9979', '6867', '19720', '3490,91', '4535', '4320', '23406', '13324', '20290', '2860', '8730', '28790', '23148', '13112', '10836', '76550', '10381,5', '2957', '37508', '11304', '21009', '21860', '22898', '39028', '2872', '17550', '16512', '15293', '14521', '22616', '11700', '3000', '40050', '8929', '16189', '38520', '2125', '66504', '5548', '19844', '8297', '5650', '20679', '20079', '17350', '11635', '17560', '8470', '9990', '10800', '16114', '10052', '6675', '16519,02', '8797', '24145', '7068', '22105', '11395', '2428,2', '3034', '24085', '32397', '36194', '11042', '7752', '32292', '7605', '5366', '10631', '20490', '14750', '2447', '13098', '46385', '4781', '14050', '15360', '35000', '6024', '4291', '12490,9', '19250', '6430', '31596,92', '14580', '2190', '17316', '8210', '9199', '11295', '24210', '3301', '9130', '8127', '11135', '18810', '22864', '36730', '5660', '15625', '39780', '15700', '4242', '11954', '5370', '14960', '15000', '20630', '5360', '16489,9', '20477', '19150', '20620', '15255', '17272', '3377', '4939', '6619', '5871', '2781', '2799', '34391', '10073', '31509', '3808', '5748,3', '12606', '17361,76', '23226', '5738', '23783', '43732,8', '18582', '13934', '3731', '14360', '24752', '38058', '27298', '14170', '21704', '11723,2', '11740', '25732,2', '8750', '3575', '14400', '11398', '13110', '7560', '22561', '4570', '40084', '7875', '13665', '35035,2', '29250', '26440', '5451', '4041,9', '20979', '18340', '25830', '24120', '6161', '18784', '15041', '8398', '7381', '6027', '28450', '16979', '14420', '30471', '9862', '12514', '5103', '11850', '9532', '3289', '14110', '3725', '16005', '22198,1', '10530', '7888', '5517', '8410', '10021,5', '12170', '2165,16', '8298', '5625', '10990,2', '6925', '3376,8', '67950', '18913', '9040', '12060', '31680', '3849', '41470', '19245', '6984', '39900', '41170', '20698,1', '27012', '4462', '6982', '14917', '10734', '30940', '24850', '16965', '21128', '8857', '44000', '21895', '2088', '4072', '6990,92', '4590', '3730', '6445', '12593,7', '17230', '34162', '45750', '9963', '15440', '4428,1', '35091', '23625', '24103', '17650', '14383', '2928', '23735', '22760', '11790', '19000', '37400', '12262', '9533', '8830', '10755', '20431', '82800', '9629', '2310', '9278', '23146', '19700', '7798', '11692', '6190,92', '11200', '12304', '2759', '7390', '9095', '22600', '5684', '52469,87', '6964', '28224', '7040', '28970', '14321', '6373', '9240', '5276', '14596', '26410', '12597,3', '56964,6', '18048', '15500', '8620', '10550', '25440', '9500', '19960', '22324', '18940', '26199', '22499', '4860', '23251', '86071', '16002', '8590', '13287', '22700', '2305,8', '17400', '8241', '44879', '33470', '10188', '9296', '19719', '5300,9', '3230', '29640', '10656', '20770', '6099', '48572', '7878', '9814', '15602', '34555', '4169', '89200', '30696', '62700', '6380,91', '2792', '9430', '4293', '5865', '9840,6', '3898', '4170', '21570', '27166', '27589', '2600', '5970', '10630', '17200', '26950', '30800', '31580', '21901', '3545', '6703', '6081', '21780', '18825', '14678,63', '21597', '16530', '7815', '2028,56', '7598', '3199,2', '25122', '3919', '27816', '4961', '10506,76', '30925', '19050', '27795', '40080', '83871', '11090', '16190', '4493', '7650', '39627', '9610', '30350', '5364', '28400', '5707,8', '4410', '26560', '13678', '9826', '20826', '17424', '15347', '21685', '20575', '11778', '11070', '5640', '2831,97', '8892', '20590', '20140', '6182', '4592', '19840', '5949', '6726,1', '4499,2', '10210', '3413', '6035', '11601', '26276', '16883', '14206', '25829,1', '12870', '33097', '7848', '5839', '17865', '18999,2', '10447', '4191', '14717', '9167', '4000', '12365,2', '5409', '10478', '54700', '33678,2', '16948', '8199', '3150', '6420', '27381', '17170', '28208', '27599', '14349', '15680', '4837', '32617', '11219', '18700', '24330', '8630', '11926', '9840', '15999', '12059,1', '3231', '7563', '22930', '19464', '12945', '11154', '8453', '14611', '35369,4', '67500', '4630', '23395', '36290', '10268', '10528,3', '34120', '10245', '7407', '15883', '21238', '8225', '10646', '4070', '10610', '13456', '9338', '18528', '18468', '19343', '30825', '20063', '8050', '17956', '17098,2', '10024', '4099,9', '11727', '3010', '4986', '17451', '34891', '8697', '2947', '9986', '23500', '9791', '8833', '10134', '29052', '64670', '21500', '5724,2', '9597', '16544', '17448', '8033', '9996', '3850,85', '5050', '4658', '16989,92', '4534', '9270', '17111', '10198', '13140', '12493', '16467', '6748', '16310', '3450', '39500', '3916', '20252', '21920', '34295', '14669', '4889', '11955', '20263', '7863', '4209,2', '5768', '6088', '10010', '45297', '79910', '4801', '8381', '25301', '19622', '13599', '5384', '22252', '5499', '23680', '4880', '13139', '7700', '21050', '21927', '6142', '13825', '14766', '2490,92', '19929', '4302', '31587', '19520', '15784', '11789', '11968', '4840', '10757', '8057', '3448,8', '29680', '8393', '13640', '25998', '4150,95', '8132', '38070', '5419', '6993', '15192', '4967', '11996', '12140', '51750', '23104', '9972', '10609,2', '74690', '4505', '10442', '11977', '7891', '22904', '30681', '8543', '14168', '5031', '15610', '28810', '9799', '5169', '26085', '28220', '3616', '6758', '18868', '16450', '11842', '3750,91', '7829', '6390', '17840', '5820', '24475', '28421', '29782', '25683', '3892', '8487', '32914,31', '5391', '14030', '22761', '4026,55', '62550', '10759', '13141', '7473', '17025', '3050', '3737,6', '20297', '12711', '10463', '7690', '2065,5', '23598', '11916', '3547,42', '12580', '7088', '29780', '16995', '28955', '17688', '17991', '23185', '12450', '4619', '84570', '5895', '10080', '10398', '12434', '19460', '29040', '3335', '20989', '7682', '15889', '7787', '8991', '17020', '35550', '20464', '26547', '18711,05', '24506', '23688', '52930', '24480', '10775', '18692', '19375', '2190,2', '34479', '17802', '55846', '41600', '4729,72', '37500', '5890', '21082', '14160', '24864', '23259', '6297', '11590,92', '9263', '22102', '3886', '78120', '12720', '27900', '17572,68', '24015', '9652', '11303', '15948', '3744', '8380', '10640', '9248', '2691,9', '7498,9', '35529', '84630', '19170', '27877', '5351', '19751', '27986', '3894', '17280', '7709', '9301', '18361', '3942', '24690', '8905', '5888', '12908', '30344', '12554', '9250', '3537', '17254', '11910', '5088', '75581', '8771', '18240,91', '3396', '18390', '7130', '6747', '59629', '9360', '6461', '12290', '22080', '116360', '7229', '15290,7', '6616', '19047', '6967', '4599', '8268', '3899', '10331', '9960', '20350', '27356', '28750', '4720', '17946,2', '19419', '45280', '15079,91', '10085,9', '12331', '18654', '5661', '3882', '15270', '8037', '35700', '7649', '14780', '4990', '5175', '7676', '17372', '3140', '33292', '4978', '7327,8', '7472', '21845', '5426', '31460', '21762', '27617,1', '33111', '3500', '9930', '17509', '4362', '16052', '24328', '12830', '6181', '10295', '9774', '19184', '42928', '20880', '3420', '4701,6', '6650', '15300', '20903,5', '4041', '20277', '13887,68', '3138', '10874', '5320', '19404', '8690', '21889,92', '11698', '21630', '17888', '13700', '14470', '20980', '3276', '8951', '16390', '18045', '12318', '6790', '17592', '18535', '17276', '17270', '3600', '37017', '8424', '25180', '43573', '29500', '11260', '22180', '13960', '5773', '14660', '6069', '4190', '9526', '17981,9', '2095', '17339', '5032', '2699', '11940', '3959', '3645', '3070', '6889', '3848', '12599', '41929', '20305', '40521', '6136', '5478', '20856', '4982,4', '50678', '6870', '24630', '12782', '4125', '26730', '11290', '9349,92', '9820', '11409', '8441', '23437', '9540', '9042', '7892', '34990', '7541', '19797,1', '7632', '7336,3', '10730', '19782', '5328', '7488', '23347', '11138', '6269', '8010', '10512', '15990', '21380', '3305', '6652', '16879', '12499', '18881', '25641', '20298', '6517', '9889', '10446', '8515', '36805', '9468', '19798', '11000', '29898', '11175', '13669', '7478,2', '14199', '13479', '12935', '6906', '5930', '11204', '23840', '19580', '3383', '15170', '74466', '31520', '23109', '15840', '26970,6', '10738,5', '4938', '13438,3', '12810,94', '21800', '7838', '30590,92', '13566', '3890', '20229', '8047', '16785', '23550', '22400', '20863', '17610', '5545', '14295', '4088', '16191', '15696', '16814', '4369', '10000', '12906', '37069', '71243', '20132', '21600', '22450', '9733', '15600', '28070', '19522', '2680', '8405', '26140', '10440', '23854', '12038', '6799', '11240', '27560', '11034', '18105', '22000', '17330', '22711,21', '7290', '16950', '96000', '16704', '4680', '11410', '7632,92', '12760', '25315', '7523', '6970', '8604', '18850', '19169', '49500', '18175', '19189', '18840', '14290', '15647', '26928', '17140', '24130', '3904,3', '8700', '3752', '5999', '2421', '17380', '5531,8', '10890', '13199', '12860', '22040', '41125,88', '9554', '4508', '17765', '20046', '23382', '8400', '14658', '12789', '39240', '23954', '28879', '18050', '20800', '21899,94', '26595', '5755', '16865', '10189', '12417', '13399', '5272', '30510', '16367', '24263', '9997', '5337', '28715', '21400', '6260', '5435', '12310', '4090', '16653', '25313', '94330', '3048', '4689', '5037', '25300', '6181,09', '9380', '18490,91', '12160', '18777', '17997', '30200', '29384', '22085', '9470', '18004', '14348', '10846', '23696', '5500', '2599,2', '15650', '11405', '5922', '12628,91', '5369', '5942', '10417', '83407', '4812', '14570', '21120', '7900', '13383', '4107', '18006', '18960', '18669,2', '10071', '7170', '9899', '15640', '8049', '6995', '4290', '11286', '11750,31', '8651', '23784', '17134', '18197', '27810', '14207,1', '81743,87', '7107', '8412', '5465,26', '8421', '12848', '7410', '4140', '7180,9', '23530', '20086', '21590', '20550', '16398', '36350', '28059,81', '5749', '6250', '15806,5', '13815', '32400', '5225', '15904', '20007', '3430', '11890', '5083', '26200', '6916', '18934', '11199', '13921,36', '18789', '7539', '23389', '4314', '14916', '16990', '20390', '13409', '8200', '8490', '18780', '3553', '3385', '16949', '12818', '10804', '5941', '3160', '15034', '8491,82', '13922', '3990', '19938', '9298', '4436', '22536', '30325', '6692', '27369', '3735', '9971,3', '25783', '4365', '30735', '25508', '9072', '58640', '4800', '13100', '21940', '11489,91', '13897', '23230', '8280', '21802', '4548', '32700', '5687', '5152', '30430', '37150', '11551', '5612', '31885', '10352', '9669', '8970', '18693', '24923', '10853', '6597', '16701', '15250', '3540,2', '8846', '12463', '25920', '15096', '7200', '17599', '29699', '2050', '24272', '65400', '10539', '8886', '21736,2', '9417', '47789', '4947', '13734', '9902', '11345', '18949', '26872', '14709', '4326', '33330', '17415', '3894,2', '37881', '18999', '25625', '35767,4', '2150', '12854', '3187', '13170', '5333', '5195', '17090', '15280', '14949', '23713', '17173', '6935', '7509', '3117', '12549,07', '10998', '14590', '11450', '11796', '23790', '29478', '20706', '19301', '10163', '13793', '21980', '14795', '18100', '87880', '13695', '54000', '2070', '4380', '41837', '2489', '20536', '15070', '27769', '6210', '13916', '17320', '13880', '17250', '8999', '10430', '32790,9', '13981,8', '82162', '23315', '14280', '30324', '9891', '6947', '12560,01', '11288', '5149', '22570', '14860', '8641', '6900', '82910', '5207', '11391', '44500', '24400', '10647', '18414', '17906', '11180', '8998', '35500', '6450', '11498', '51300', '30960', '11958', '4013', '15216,06', '5488', '15320', '5280', '10149', '26800', '4908', '14872', '10668', '11370', '24470', '13318', '5678', '60888', '15309', '12702', '2910', '25765', '15690', '17558', '26596', '21484', '18398', '24800', '15850', '9448', '4660', '11749', '4366,8', '20910', '6575', '15750,91', '6941', '4490', '34618', '19040', '33358,84', '4745', '40700', '21349', '16250', '3308', '15325', '16098', '7369', '15098', '21850,94', '9538', '7181', '4220', '12557', '29006', '4645', '7599', '12200', '27498', '21670', '7857', '22210', '7588', '3560', '14320', '22214,69', '25800', '9400', '27971', '14497', '7869,61', '5030', '4420', '25780', '26450', '42228', '21665', '7991', '70410', '35142', '22498,2', '12989', '5310', '13921,52', '24962', '27010', '25989', '21827', '3033', '10616,4', '3722', '8438', '23320', '18631', '4386', '5827', '8580', '23044', '8568', '92000', '5288', '27570', '23170', '25983', '5292', '43376', '3801', '16834', '4275', '14398', '18724', '13221', '10987', '18162', '21730', '6489', '25225', '16890', '29334', '4930', '21499', '6108', '4546', '12940', '10286', '26390', '6938', '18510', '28900', '12283', '30930', '17146', '16320', '49838', '25680', '4280', '22970', '24220', '21646', '4807', '12397', '14094', '13205', '20691', '5353,5', '3173', '6780,95', '6831', '13600', '22799', '76360', '14200', '9573,1', '15658', '8198', '11040', '22263', '13440', '9768', '6247', '7667', '19897,1', '33500', '6141,75', '17702', '4595', '10483', '2183', '15914', '19769', '19710', '5602', '3639', '10555', '8389', '6966', '24721', '37800', '9351', '10617', '31060', '7439', '6865', '13728', '18370', '8195', '3296', '7495', '13326', '9837', '13288', '21510', '15621', '26641', '7174,6', '11651', '16765', '24270', '5390', '22413', '26080', '25897', '39590', '29000', '16290', '8416', '6651', '4118', '42320', '7417,5', '14880', '3827', '6848', '6939', '6567', '14725', '5848', '17117', '26329', '12730', '4150', '3330', '10470', '2341', '6246', '31790', '40824', '19415', '8055', '3699', '19809', '17100', '4430', '4950', '8330', '24550', '6590', '11270', '2689', '15780', '14891', '3760', '5870', '4563', '6975', '12251', '10389', '26485', '27680', '6930', '17425', '7063', '49490', '28933', '3448', '7729', '4638', '6856', '6319', '23390', '61600', '19489,91', '22747,4', '4885', '14209', '22872', '21390', '10898', '6599', '34821', '31040', '8299', '23792', '5040', '16140', '24100,83', '26020', '6036', '15214', '10841', '13860', '13400', '7348,52', '26380', '17247', '16090', '14035', '16996', '6364', '13671', '16200', '26930', '27450', '12110', '14144', '39090', '16381', '12396', '9445', '7098', '6273', '3015,2', '23154', '5706', '13225', '4064', '21440', '12797', '4935', '8715', '29295', '31404', '4458', '24741', '11655', '27493', '20296', '13948,2', '3930', '15580', '4620', '13820', '28674', '9525', '20698,2', '10536', '5431,2', '23745', '10690', '13166', '28215', '19746', '15899', '20430', '2745,9', '12090', '3876', '13380,56', '35326', '6103', '3960', '13507', '28410', '3141', '4579', '12465', '16324', '22900', '35530', '7337', '5077', '6470', '4428', '15315', '23930', '21481', '30627', '28932', '11317', '27886', '8550', '7794', '5550', '13488', '4455', '24399', '8883', '17379,99', '6873', '12284', '12309', '24030', '9375', '8990,9', '20160', '24465', '6910', '19210', '4994', '3998', '19895', '19008', '24600', '3442', '35563', '26034', '28042', '3520', '6660', '19529', '24889', '3237', '2998', '7120', '13441', '21838', '4350', '6166', '27846', '17466', '10566', '18970', '5260', '19470', '27448', '19857', '19750', '17130', '28110', '22621', '19475', '21750', '17045', '3979', '5899', '10600', '32907,5', '18990,5', '7920', '10712', '41700', '13747', '29612', '17150', '29300', '7845', '3877', '13680', '5669', '3200', '9189', '7112', '22640', '8829', '19020', '21312', '12317', '2679', '4370', '22302', '22490,91', '28041', '20091', '24980', '11620', '16491', '20987', '12019', '3815', '18358', '34199', '34154', '21840', '7687', '4476', '8355', '7497', '5765', '26720', '29376', '11377', '10675', '17366', '18328,6', '8006,1', '5508', '5162', '29526,12', '13890', '15639', '24190', '25999', '41994', '4775', '6007', '33183,4', '77210', '22781', '31484,1', '12991,82', '24771', '11480', '5555', '8694', '7830', '12070', '3370', '17430', '21814', '9759', '4281', '11690', '20943', '24047', '13193', '25177', '10548', '5806', '11748', '62800', '7140', '10981,82', '7394', '3681', '12835', '28781', '5525', '20338', '5129', '16980', '7032', '17530', '21498', '14498', '119700', '27970', '17347', '14013', '9988', '3067', '3910', '26315', '2151', '60080', '12868', '31300', '5605', '3499', '11571', '19400', '3093', '3865', '28469', '3961', '14385', '15912', '21086', '13302,83', '8735', '5489', '3768', '23950', '33850', '21097', '13312', '17920', '4117,5', '17988', '32810', '17950', '10570', '6739', '61000', '34000', '3924', '5516', '8598', '3134', '8350', '23425', '6201', '21482', '4351', '15759', '12780', '17273', '24283', '9212', '13689,86', '36980', '6298,6', '19799', '11192', '19342', '3923', '12737', '44760', '8489', '22290', '37466', '27180', '34999', '6071', '37235', '13235', '11289', '14890', '32000', '20326', '28288', '6504', '14600', '13494', '3114', '4827', '52000', '46729', '14487', '32200', '12220', '17756', '12655', '6193', '18329', '9059,92', '15985', '23040', '5610', '9606', '10090', '7144', '5695', '3285', '3780', '24990', '20644', '5450', '32370', '11496', '5799,2', '24223', '10999', '33100', '3670', '41171', '6770', '17640', '32940', '11063', '12640', '16110', '7510', '22756', '9150', '96490', '29220', '13421', '21877', '34804', '27780', '13598', '19674', '28990,91', '7224', '17790', '21464', '18330', '7080', '28420', '13653', '19240', '21640', '26340', '16551', '33821', '4250', '8322,77', '30826', '17896', '44495', '2045', '16284', '5240', '2633', '11243', '2961', '6991', '8570', '13841', '22480', '2500', '21588', '19045', '6940', '4168,8', '10795', '14350', '12197', '13580', '21960', '12889', '30820', '4888', '21580', '6746', '7898', '19680', '18730', '15398', '44700', '15468', '23820', '8091', '11600,6', '23890', '3784', '20424', '25713', '5670', '3310', '11385', '18290', '26790', '9574', '8190', '10591', '4095', '21123', '11786,85', '20699,1', '28273', '11194,06', '10423', '6928', '16571', '5345', '25720', '9949', '9813', '29198', '15110', '16189,4', '7340', '25158', '16620', '2498', '27690', '15279', '8780', '21270', '16220', '9809', '32844', '8633,7', '23998', '20488', '21770', '34499,2', '3498', '17549', '11990,95', '27820', '8390', '6658', '21771', '23290', '16849', '11529', '13078', '37480', '3701', '23304,6', '3390', '28530', '5850', '3854', '26580', '24667', '17091', '30799', '31850', '15186', '3149', '43010', '4248', '17162', '23980', '33120', '58900', '9405', '7030', '22300', '14284', '20999', '3590', '4690', '6288', '17631', '13433', '12570', '6089', '26817', '12896', '23830', '3547', '7183', '20977', '7614', '15190,5', '20499', '13136', '5144', '25109', '16780', '4704', '12057', '11558', '53499', '17541', '5078', '14850', '21209', '19260', '3786', '20659', '16824', '5197', '18240', '33725', '35670', '28590', '12950', '13920', '6522', '9422', '8644', '24500', '11764', '3410', '16263', '11099', '22430', '8725', '9560', '25785', '6309', '25050', '15442', '16743', '11808', '9128', '61915', '12249', '25199', '30418', '32670,79', '5070', '11528', '12290,9', '4491', '13905', '13745', '17966', '9406', '33340', '16635', '6367', '40180', '7530', '14365', '3490,92', '10216', '20495', '18599', '21550', '36308', '8900', '3769', '39000', '4850', '94720', '27541', '9551', '19340', '10990,94', '29742', '13748', '53364', '7645', '3393', '4470', '53400', '16789,98', '16690', '6883', '3143', '2000', '20190', '3001', '3069,9', '19035', '8901', '18610', '21866', '39830', '7547,2', '4208', '22987', '18130', '18696', '21480', '3415', '22792', '8835', '28565', '15190', '10089', '5680', '19060', '4550', '6494', '7511', '24059,83', '9477', '22233', '8221', '13006', '23992', '28241', '8840', '29234', '15090', '13990,92', '7980', '17649', '21430', '18270', '27939', '19021', '31450', '5681', '11077', '11363,99', '7998', '11310', '3941', '19699', '25662', '9490', '41210', '30370', '14637', '18198', '9009,91', '18749', '3651', '24174', '24479', '6733', '10315', '13679', '11280', '2219', '2160', '19079', '35600', '3840', '29661', '24810', '15165', '16370', '17815', '17529,91', '27196,1', '10322', '5456', '8754', '24459', '30131', '15830', '8220', '19439', '15400', '10359', '22221', '8917', '13443', '12687', '5380', '3515', '51370', '73996', '20272', '16399', '7404', '26910', '16265', '32444', '12575', '9991', '23990,83', '35787', '18750', '20580', '7395', '7228,2', '6435', '50020', '11308', '40900', '5780', '5097', '28340', '14639', '12980', '23663', '29788', '5174', '11529,91', '6948', '13965', '7909', '11359', '5793,3', '18180', '12472', '29445', '89196', '8056', '4018', '5736,6', '30980', '4882', '26819', '77860', '12213', '26555', '11390', '6490,5', '18580', '3987', '8800', '38832', '11990', '24682', '37531', '39600', '12454', '9397', '24772', '6698', '9030', '16648', '22672', '21083', '23920', '12820', '62100', '29970', '31610', '75535,08', '4878', '17476', '14493', '26462', '12566', '6470,9', '20670', '13950', '12389', '32975,3', '23310', '4103', '73300', '16245', '15556', '22167,6', '14469', '17869', '11710', '6783', '19977', '9748', '11130', '3542', '17556', '25280', '8244', '15789', '12228', '53162', '14461', '13990', '4758', '6204', '20877', '15880', '2559', '13122', '6393', '21974', '7902', '20237', '5005,8', '9198', '9251', '23128', '3989', '16416', '87673', '8918', '6190', '19630', '9823', '17043', '5614', '15030', '7981', '11422', '3424', '10438', '9592', '24900', '33476', '4830', '9341,5', '11539', '23667', '29905', '3375', '4360', '6990', '30879,55', '13829', '4652', '5389', '19490', '7715', '15838', '7947', '3168', '13392', '17789,2', '2660', '8399', '14659,2', '4743,68', '10925', '30590', '23711', '4240,91', '5399', '10496', '14065', '37360', '16600', '23268', '38406', '21166', '17099,1', '24003', '17050', '26461,91', '7336', '20520', '4005', '4336,2', '11477', '6749', '25881', '6039', '7581', '14724', '19781', '3195,9', '5238', '17540', '19200', '23861', '20767', '29120', '6127', '7173', '11426', '20598', '14610', '14046', '33450', '20090', '14619', '6875', '12639', '6546', '21778', '33030', '8038', '37780', '16650', '5270', '14994,6', '3464', '70300', '15084', '6545', '4200', '6518', '6721', '55500', '51570', '17075', '3970', '30455', '6379', '11126', '28899', '7940', '10240', '5997', '15127,4', '4670', '22674', '22390', '18096', '29750', '15844', '5800', '14100', '14495,35', '3290', '19925', '26837', '20443', '11100', '12000', '8858,8', '13770', '10980', '35522', '26988', '7281,94', '30050', '31933', '23411', '8498', '8978', '17998', '3879,17', '27620', '25740', '10720', '16820', '8652', '6983', '14965', '71740', '12348', '24112', '26590', '4120', '84500', '11302', '21999,2', '21490', '11753', '10789', '6192', '18980', '23200', '87570', '22121,1', '2555', '12546', '8745', '10968', '6680', '16070,5', '6850', '2490', '17089,9', '16868', '26896', '24650', '16781', '10048', '12597', '11450,9', '77762', '24070', '96200', '7299', '18001,28', '6247,8', '11446', '29697', '9412', '3280', '15984', '16045', '5995,46', '24290', '19572,1', '5155', '6498', '15598', '12990', '15727', '24367', '27600', '87786', '23202', '4596', '8440', '3663', '18578', '25632', '3930,81', '8763', '6665', '18590', '10854', '8090', '5025', '25400', '44690', '19790', '23975', '14456', '9246', '25385', '6969', '26961', '8099', '7514', '5023', '22530', '2031', '7968', '14130', '14513', '44225', '83000', '21654', '22320', '11489', '22338,94', '27260', '8878', '6116', '14847', '7513', '47600', '14410', '30196', '19843', '19730', '7499', '9340', '28431', '16498', '28557', '10250', '20636,91', '3700', '5830', '21536', '65250', '7198', '23377', '21105', '35596', '5533', '26120', '4810', '18230', '25538', '11949', '4186', '20658,2', '9907,2', '8639', '2850', '4035,2', '15100,82', '11456', '22500', '12325', '7885', '31200', '23393', '19600', '24301', '9845', '11466', '5940', '9440', '18707,1', '71400', '13959', '2120', '23400', '15127,38', '6350', '11189', '4004', '20786', '29707', '9793', '20648', '27020', '9702', '5179', '12340', '3798,9', '9573', '4040', '25840', '12904', '8156', '22240', '14636', '17397', '16647', '33190', '16499', '6049', '17346', '13439', '2421,9', '12843', '21150', '10780', '18187', '5760', '5995', '17704', '3299', '5282', '23882', '4653', '17240', '4177', '16740', '15755', '14169,1', '23453', '15875', '11718,94', '3619', '12750', '3933', '16715', '5589', '18010', '12260', '4240', '6699', '15481', '21470', '5128', '11445', '8595', '28728', '11877', '15555', '6070', '8640', '6696', '5987', '8308,35', '29858', '7839', '6756', '11223', '8698,1', '27446', '7750', '4773', '27860', '41451', '3949,92', '38718', '5445', '25937', '3493', '6584', '22317', '5287', '22460', '7711', '14397', '25146', '7590,31', '7586', '17297,6', '6168', '14648', '17190', '30974', '14760', '5910', '22361', '7977', '6030', '7640', '78000', '31833', '18538', '5090', '39160', '9160', '24408', '44440', '10990,92', '2165,4', '9570', '24760', '31474', '9999', '21592', '13060', '5109', '23750', '80000', '4770', '16362', '19382', '2039', '3875', '12127', '3984', '14361', '50652,94', '26642', '26886', '4799', '4949', '8064', '16649', '14356', '32819', '4915', '21015', '16392', '19890', '15740', '5330', '4752', '23650', '52561', '8184', '15602,5', '4244', '17561,8', '29386', '17006', '30686', '33255', '2200', '5994', '11133', '4744', '4853', '15003', '7300', '14150', '17531', '6202', '6264', '9998', '6797', '4838', '39110', '23399,2', '5505', '22945', '17231', '22427', '18119', '28059', '13020', '14096', '13184', '2800,85', '19983', '4600', '16418', '34650', '21586', '16501', '3032', '3955', '32846', '5912', '3680', '15756', '7150', '7454', '22617', '23152', '11966', '11025', '83194', '5210', '19980', '3874', '5751', '8480', '24340', '20480', '15085', '14942', '49798,1', '7695', '35825', '2048', '28842', '29280', '10939', '16160', '10820', '14387', '30197', '4039', '17199', '14845', '16350', '3816', '18689', '22951', '6720', '13448', '16991', '18630', '20928', '31806', '10760', '56310', '6197', '5963', '5750', '2595', '14680', '5410', '5730', '5058', '4910', '4664', '4276', '18030', '5095', '16436', '6796', '8617', '18470', '13690', '3762', '21861', '8160', '6380', '27456', '6876', '6053', '20660', '27490', '2489,4', '9080', '9790', '16479', '29206', '4390,91', '6499', '10590', '99313', '13489,91', '5432', '7452', '9835', '6782', '12995', '34493', '7985', '10704', '23900', '11768', '2223', '5268,47', '24881', '28700', '8648', '13850', '31588', '15097', '15314', '4746', '13200', '8097', '18206', '19430', '9140', '24233', '12845', '32114', '25730', '29100', '25749', '3654', '2704', '31490', '10508', '29663', '23110', '12120', '9599', '38700', '6560', '19610', '9341', '2089', '12699', '49743', '3682', '43880', '15367', '8940', '78390', '19350', '20510', '14609', '40756', '18891', '16773', '11362', '19417', '10940', '6291', '37495,2', '6839', '5309', '11279', '28417', '12672', '3316', '11428', '3407', '16380', '7151', '16783', '7022,7', '18304', '47000', '9230,9', '14994', '18885', '26871', '9586', '7995', '5398', '18880', '15313', '5980', '4679', '76300', '33280', '10860', '14279', '22323', '22455', '37456', '29070', '23459', '9980', '14996', '16598', '15226', '35320', '11250', '49860', '46881', '31466', '12136', '8666', '15763', '17778', '4488', '25855', '6189,92', '45438', '25167', '21347', '16170', '28280', '13776', '25459,3', '28247', '2499', '29997,24', '13550', '6561', '2439', '10114', '6119', '5746', '70245', '35770', '5600', '22230', '10904', '21190', '10390', '4010', '26941', '22495', '30220', '8148', '3492', '13420', '28666', '22944', '12550', '5015', '5496', '13216', '12451', '26337,6', '6217', '6313', '6021', '5346', '18495', '3895', '4510', '10991', '15363', '18319', '22057', '5759', '27673', '4110', '12972', '35201', '8445', '37350', '20433', '4735', '12428,7', '9719', '10525', '34466,4', '12455', '27854', '11350', '22194', '16796', '22798', '8170', '34695', '12959', '5494', '7317', '21726', '41100', '11795', '11799', '7796', '3898,1', '15950', '8906', '9120', '11225', '20779', '9137', '19924', '15299', '3294', '6893,7', '36570', '11804', '13469', '19899', '7315', '13577', '6779', '10880', '13350', '6100', '6766,4', '7357', '29380', '23396', '35610', '16550', '70936,32', '22497', '12425', '33459', '13263', '13254', '4485', '32076', '39720', '7747', '4389', '2650', '18900', '30756', '6495', '12645', '10310', '6959', '63000', '7084', '62468', '20320', '4339', '13325', '16485', '8879', '19597', '4674', '7984', '12822', '10699', '14900', '20753', '9499', '4498', '4048', '5931', '7209', '25520', '23797', '7235', '32090', '16257', '30946,29', '4703', '25447', '25490', '17573', '58947', '26960', '6105', '79380', '3190', '7115', '5936,81', '4079', '7362', '17670', '2450', '7550', '17839,92', '9370', '2871', '16961,1', '19187', '9530', '3255', '3868', '7516', '10340', '18882', '11980', '17804', '27208', '17760', '11876', '19889,92', '5639', '2619', '25020', '12898', '10130', '28778,3', '7693', '6460', '15570', '8902', '8435', '7116', '10665', '30390', '11860', '11050', '20205', '17450', '38408', '6527', '2000,51', '11697,1', '50568', '17290', '4642', '8250', '15832', '10818', '13688', '7874', '17970', '63776', '4640', '20169', '12902', '14219', '3222,9', '58490', '5636', '14446,3', '8540', '13240', '7480', '30031', '16060', '7655', '13563', '6835', '4399', '10750', '29948', '4826', '6920', '14999', '10319', '35943', '7698', '26150', '12189', '22090', '20009', '3834', '4964', '13030', '12029', '27962', '27250', '10520', '4207', '3540', '28116', '34700', '5900', '11580', '13290', '2620', '63700', '10400', '12803,4', '41620', '29390', '9010', '5221', '9879', '14925', '15199', '16371', '10505,5', '7988', '13388', '16169', '50000', '8536', '46649', '31895,48', '14530', '16720', '21330', '21042', '34100', '32900', '31020', '22145,1', '8292', '20562,2', '9315', '11159', '10499', '13178', '14230', '38800', '9594', '23009', '4077', '11400', '26100', '5745,12', '13130', '22969', '11760', '12210', '94358', '8025', '10298', '3245', '44660', '20180', '18550', '25870', '7551', '2797', '6479', '26740', '6580', '11600', '7194,72', '7629', '6381', '18113', '33660', '11190', '27796', '33570', '10767', '6205', '21847', '35520', '12508,2', '14978', '15960', '19570', '23760', '11488', '2366', '12132', '15160', '9675', '8556', '24314', '4841', '15790', '22896', '6163', '19125', '81612', '20220', '14758', '5010', '19810', '30000', '18380', '3110', '27736', '11687', '17112', '4608', '27449', '10448', '11670', '11349', '19130', '11897', '7651', '25639', '21160', '3650', '19358', '53091', '14097', '7493', '19880', '5965', '11323', '16105', '11460', '4909', '9241', '16379', '17432', '30428', '25390,94', '3883', '12077', '17952', '22800', '3195', '11093', '17300', '24389', '7191', '5877', '9117', '14406', '4477', '8286', '25695', '9800', '7132', '16447', '2940', '17630', '29900', '14245', '28221', '10190', '25299', '14990', '72890', '6544', '5020', '16039', '19540', '29880', '15393', '14328', '13730', '12646', '10714', '23370', '3764', '23490', '23608', '12443', '13080', '8262', '4575', '23865', '17671,2', '3869', '28991', '13360', '6792', '9769', '2060', '9577,8', '5807', '23103', '9931', '23971', '9830', '18318', '4021,92', '8549,9', '22345', '26330', '8777,1', '33978,63', '13025', '35775', '7490', '9444', '17609', '7302', '4899', '14620', '22410', '16662', '23944', '5879', '3100', '46156', '5261', '17880', '6594', '22959', '20519', '35955', '10526', '23480', '13733', '25406', '5740', '29493', '9759,4', '18710', '9457', '23030', '22940', '10354', '14798', '15496', '6374,78', '5721', '11648', '29700', '30837', '12280', '29677', '29600', '17990', '11674', '9730', '6741', '4870', '28210', '42000', '5220', '7810', '13541', '60500', '3614', '4189', '16616', '13298', '85100', '14950', '7949', '3284', '26999', '30920', '8564', '16141', '21487', '27348', '42408', '13558', '13999', '17955', '22880', '9395', '4780', '11779', '25727', '76000', '18820', '19051', '3599', '5599', '5825', '4968', '11672', '15890,9', '63657', '15340', '8442', '46000', '12092', '4299', '15907', '24711', '19819', '6033', '11236', '34699', '14175', '73230', '31800', '48000', '13332', '16860', '8360,92', '20882', '23579', '4804', '17483', '27062', '20646', '20699', '5770', '24840', '10456', '28922', '13157', '13910', '10323', '9235', '18500', '27000', '22440', '7166', '3561', '28079,84', '75000', '14390', '2389', '9158', '25105,1', '4672', '8984', '2730', '16908', '27670', '16490', '2890', '4097', '15890', '21388', '18698', '17251', '17398,2', '2529,61', '23690,91', '2870', '9648', '2790', '19999', '14449', '9429', '27347', '34740', '10366', '21615', '97000', '33475', '5990', '6559', '9515', '17528', '14005', '10852', '2970', '45863', '3321', '10770', '6863', '2889', '15245', '9034', '15201', '8675', '33830', '18425', '9295', '22310', '3729', '13063', '7201', '6570', '23478', '16145', '17775', '13435', '8980', '7433', '11227', '13985', '6219', '31991', '26099', '31730,91', '32348', '17780', '11619,84', '24605', '15380', '12435', '2259,9', '26658,9', '8977', '79200', '25574', '41400', '13120', '7749', '8108', '26219', '17789', '8149', '29336,4', '3107,7', '27065', '12498', '4310', '16940', '6266', '10225', '28517', '28050', '41490', '8955', '29450', '16244,9', '23020', '14599', '21648', '23240', '14472', '11196', '17842', '4941', '17590', '3300', '23990', '5448', '10339', '5339', '3717', '3152', '25498', '19521', '4738', '2180', '36093', '50400', '57581', '6999,3', '2629,8', '12310,91', '30190', '5228', '22660', '5232', '24768', '4952', '12512', '28039', '7187', '3824', '10971', '23660', '7240', '14340', '44697', '8303', '5634', '10708', '45034', '59009', '20053', '20690', '12159', '6467', '4730', '4114', '50098', '22550', '23124', '4725', '17082', '17895', '2990', '14698', '5005', '7880', '12300', '9899,92', '15788', '4743', '8034,2', '6120', '3516', '8825', '16992', '18964', '29205', '6411,72', '54105', '11526', '15354', '3320', '4661', '15887', '71500', '4911', '10373', '22692', '5645', '41397', '6781', '34905', '6958,92', '67795', '7050', '8680', '46800', '5357', '11399', '35580,91', '31500', '25278,14', '29790', '6556', '12191', '3861', '12240', '19883', '73306', '60300', '20113', '16272', '6999,9', '14755', '12044', '12173', '17490', '18660', '32955', '4580', '17624', '15480', '17820', '4268', '7289', '20000', '30600', '43631', '35250', '13462', '26690', '16591', '59590', '5988', '40490,91', '10361', '7020,9', '18140', '7570', '8544', '28502', '13567', '12150', '6625', '39800', '20750', '5798', '14696', '14500', '7996', '28440', '17453', '10972', '19320', '11769', '23070,6', '42335', '11978', '28661', '22072', '10884', '3660', '52509', '15128', '27234', '12360', '25904', '8718,1', '5119', '4274', '35940', '2099', '24750', '3275', '6300', '17420', '12219', '42300', '18740', '18549', '12608', '23410', '2930', '4377', '24844', '22850', '17398,1', '7710', '17460', '9219', '34200', '20409', '4390', '39020', '17901', '9390', '8720', '6209', '16218', '17873,75', '12195', '26889', '9371', '23999', '3640', '10084', '9004', '14651', '6809', '22100', '7319', '22830', '95441', '10995', '4777', '11455', '26547,3', '19736', '56130', '30230', '9443', '5607', '9621', '5011', '9955', '9053', '10140', '3892,2', '17324', '13699', '26022', '19145', '21298', '93890', '21506', '26890', '4879', '4589', '16050', '4641', '7441', '8369', '20200', '5712', '18269', '17800', '24552', '16020', '17439', '22590', '4241', '9107', '47195,92', '12511', '18067', '13590', '12695', '6881', '22848', '24132', '16621', '58541', '14201', '5099', '6785', '2007', '37450', '2879', '20450', '8117', '5542', '6278', '31278', '9074', '18575,2', '5388', '38625', '9237', '15745', '11247', '19772', '12206', '6200', '25500', '6477', '14292', '20653', '16150', '4185', '12816', '16698', '3119', '4182', '22097', '5294', '29094', '46500', '18819', '52200', '14052', '23137', '3314', '6840', '6578', '26339', '17763', '4379', '11297', '27428', '12762', '5170', '7970', '20925', '7092', '16786', '23749,76', '11586', '11300', '9866', '27990,91', '6885', '18074', '4729', '19360', '2399', '21237', '21999', '9614', '25080', '8381,5', '14490', '29799', '5312', '5789', '35400', '48700', '50940', '15857', '12478', '6415', '4134,4', '4057', '13995', '3090', '32540', '20874', '5742', '39388', '14148', '4816', '18725', '8755', '3470', '12696', '6745', '19734', '7258', '9217', '7470', '52295', '7020', '18217', '9970', '6186,7', '3398', '24795', '6697', '32650', '71600', '9670', '13490', '18472', '17510', '6507', '14938', '12840', '13402', '8844', '15460', '32980', '4845', '9789', '8460', '15897', '4328,1', '24983', '23470', '14396', '14193', '3337', '22330', '9282', '6377', '2590', '5577', '5377', '14891,92', '5757,1', '21060', '17962', '7990', '26650', '4809', '5906', '6290', '8500', '28242', '22810', '18471', '25990', '6075', '8941', '23659', '11296', '5840,5', '10231', '18861', '17633', '7360', '6386', '5837', '19998', '5620', '8320', '5051', '74319', '5089', '6819', '18251', '26379,91', '10740,64', '6348', '25515', '14355', '4504', '28172', '22286', '16998', '6256,8', '3541', '7160', '30597', '10785', '14993', '24450', '5143', '4047', '44030', '22280', '26321', '15638', '23029', '26457', '35992', '7720', '7763', '16985', '37517', '21850', '10500', '9846', '4887', '3999', '15033', '18719', '7469', '17388', '12480', '10490', '3076', '4463', '9176', '7137', '21496', '20249', '5580', '16627', '21619', '16897', '27480', '15863', '4879,8', '28460', '10113', '12520', '30951', '13301', '9127', '12742', '4139', '91168', '12800', '8228', '15870', '8186', '7204', '6469', '2243,7', '13843', '19860', '18099', '5231', '15730', '5017,9', '26770', '2390', '12919', '9287', '14644', '2975', '20088', '17750', '9224', '8456', '13476', '22635', '11646', '12385', '5932', '7450', '12236', '8270', '58500', '5239', '14885', '22124', '2855', '27922', '34195', '18800', '7870', '4607', '7420', '44560', '7975', '7500', '20710', '22598,1', '9520', '5824', '3870', '13498', '6080', '34148', '5676', '17520', '15336', '9754', '30580', '8459', '13740', '14510', '10075', '13465', '9940', '17660', '6361', '10191', '32770', '13527', '10467,1', '18440', '5860', '21218', '11685', '16687', '7425', '10205,5', '8698', '18497', '5540', '7368', '20583', '16613', '27470', '2189', '21919', '2400', '57209', '12734', '2412', '16092', '18360', '10740', '18895', '2680,4', '22950', '15704', '2100', '9389', '24344', '3535', '13366', '4522', '22905', '4460', '11495', '3128', '23575', '20268', '24466', '7240,5', '35165', '7580', '3358', '21717', '4765', '8869', '6280', '12690', '11730', '3389', '7735', '38300', '21597,8', '4995', '13195', '13495', '35640', '22250', '7011', '15498', '5190', '9031', '22098', '28350', '20790', '9600', '6282', '17811', '18138', '4210', '8717', '21660', '20069,33', '4221', '19830', '23488', '13980', '5989', '24940', '34449', '17298', '7908', '28722', '28745', '4675', '5840', '4767', '20720', '21180', '3750', '22091', '10910', '18410', '5189', '8068', '13500', '27781', '11321', '18049', '51700,5', '28950', '12572', '10790', '25640', '7376', '32938', '25420', '17795', '33881', '25360', '29830', '18060', '13300,91', '5950', '4212', '5436', '13165', '11565', '65000', '5301', '13016', '8788', '12598,2', '74482', '3109', '30340', '40480', '22960', '23715', '4387', '51140', '5449', '14927,83', '22522', '28850', '12460', '24933', '23552', '24833', '4345', '17459', '7574', '48600', '7610', '17329,92', '11899', '15599', '6375', '9681,5', '13005', '11819', '2238', '3220', '5847', '32391', '11725', '5480', '3850', '42669', '22488', '7982', '27850', '6090', '89344', '9910', '6537', '20645', '13275', '15748', '6502', '13450', '15800', '22618', '7459,6', '13250', '10419', '40037', '2075', '28800', '22293', '3538', '37437,66', '31538', '17261', '11800', '5000', '17851', '3072', '9365', '4392', '14370', '17875', '24035,5', '4400', '8419', '4726', '13768', '13848', '9467', '6500', '12590', '7440', '30011', '7047', '27248', '26500', '4928,96', '14316', '9738', '11989', '4962', '9461,5', '23499', '16519', '5265', '12948', '6327', '31726', '10420', '13780,82', '10431', '11259', '35669', '2241', '30170', '19560', '29690', '32525', '6490', '35100', '24647', '3095', '11217', '4629', '29200', '10805', '11212,06', '7604', '13148', '23050', '7997', '40500', '15897,92', '6679', '8530', '5921', '10706', '26051', '9595', '15327', '11798', '8508', '8762', '10418', '22881', '2995', '23451', '4008', '12716', '22162', '6672', '4934', '5855', '20680', '2511,82', '15010', '19608', '11416', '11430', '14006', '36617', '20505', '5335', '33862', '11855', '14522', '22589', '7180', '8260', '7621', '17930', '4500,95', '21473', '19099', '4319', '5898', '10058,5', '20324,92', '23475', '11049', '22082', '45110', '16507', '13670', '29650', '7620', '9839', '23057', '17850', '26479', '25227', '9956', '16040', '4497', '4132,5', '42436', '6776', '7060', '14040', '9700', '14854', '13940', '7637,63', '2430', '14849', '15843', '5198', '5295', '15859', '16081', '6240', '7207', '7101', '5535', '4265', '7416', '4401', '14598', '9290', '15450', '3800', '7760', '2350', '12447', '17310', '3610', '4398', '21214', '13355', '19615', '17546', '6828', '12631', '3928', '15299,9', '34900', '4576,5', '19805', '6040', '30100', '31091', '2372', '18392,91', '19224', '31875,3', '23772,27', '24423', '9425', '6605', '26999,2', '13001', '17949', '10115', '4921', '7430', '10791', '26370', '14399', '31900', '4700', '13888', '8806', '36701', '15520', '15247', '5720', '16576,2', '10338', '4202', '13356', '20330', '13930', '25263', '32335', '16880', '9497', '17500', '35991', '6189,3', '35145', '38778,01', '5968', '11986', '29471', '17770', '14571', '5775', '25950', '4315', '3145', '3548', '30150', '37857,2', '5867', '13686', '19230', '4115', '5131', '4644', '19341,2', '16470', '16400', '29177', '25896', '23989', '8943', '16748,1', '6191', '21100', '4349', '6999', '16438', '14781', '24110', '23100', '4413', '5735,7', '19966', '18389', '3710', '17716', '26055', '18573', '12944', '77000', '3353', '20300', '16260', '22870', '15998', '4617', '3565', '6520', '17752', '26820', '23940', '13889', '5472', '35744', '9447', '12884', '17720', '13138', '21880', '18170', '3199', '63271', '10497', '21673', '18517,1', '4284', '17697', '16872,3', '24867', '14848', '16449', '37656', '10439', '4858', '7766', '8935', '16733', '31491', '15230', '3061,45', '6918,34', '12495', '11518', '16198', '11021', '10410', '9259', '19791', '4212,8', '7342', '2290,7', '15739', '6289', '12651', '9755', '3360', '18734', '8744', '10687', '26865', '3311', '38570', '29980', '8078', '3434', '23204', '4979', '22299', '26461', '50950', '6060', '14910', '13294', '6998', '3354', '12790', '28490', '18481', '29592', '28341', '25812,5', '10964', '4867', '14550', '4160', '18515,03', '5748', '12342', '25298', '3405', '48580', '14856', '24119', '43950', '19989', '25211', '6960', '5757', '7128', '27645', '7549,89', '37305', '20295', '29949,7', '27580', '8285', '14632', '3990,92', '12239', '4021', '9983', '35900', '22528', '7182', '9810', '7327', '47529,91', '18085', '12330', '6109', '3418,5', '11666', '32990', '14092', '8000', '28060', '6139', '18426', '2934', '12509,9', '4195', '8792', '41907', '9919', '15940', '6843', '10935', '16745', '26774', '89018', '11560', '10116', '8128', '11598', '39276', '6730', '11950', '16214', '15427', '11898', '14520', '30599', '5966,7', '12427,1', '12784', '25350', '17308', '17911', '3030', '98996', '2635', '17422', '5415', '70360', '11293', '13230', '15624', '30313', '6530', '21595', '31333', '13010', '27340', '20565', '11266', '14661', '13570', '24405,4', '8997', '30738', '30720', '13624', '10864', '21858', '34990,8', '7507', '16285', '7349', '6093', '10088', '24950', '3391', '5199', '8979', '18499', '5130', '7110', '2745', '4984', '10605', '10230', '6050', '8100', '21295', '19728', '32481,83', '5016', '5290', '37100', '10274', '19797', '8689', '3792,92', '25991', '14338', '13992', '8188,04', '5913', '76730', '4851,81', '11724', '25760', '4040,1', '9013', '11167', '8658', '11650', '6800', '37818', '13785', '38000', '15670', '10457', '46600', '84800', '29001', '10930,52', '21080', '15865,65', '21930', '16389', '10635', '21936', '2900', '26818', '24195', '23108', '6056', '37000', '28930', '10479', '34280', '3250', '10394', '3392', '24362', '6550', '9276', '5200', '4983', '16590', '26138', '22384', '16730', '21450', '28755', '17367', '15078', '10693', '24519', '39638', '16875', '50970', '12490', '18216', '8899', '4904', '6220', '9173,88', '19905', '10672', '2245', '23025', '15732', '26379', '87600', '3180', '79959', '41132', '7540', '13689', '88000', '18906', '7850', '7249', '23600', '11434', '5018', '17580', '5405', '10870', '12872', '32706', '93947', '3940', '19567', '61290', '7147', '8625', '18012', '20587', '19278', '10279', '7769', '48920', '4540', '7515', '14083', '22264', '9538,9', '14747,1', '21865', '21989', '16520', '16838', '15565,5', '18990', '18090', '5232,2', '27150', '47691', '8874', '4140,2', '26084', '20280', '7905', '7713', '4754', '19620', '13918', '30110', '9924', '6299', '5832', '10697', '10799', '6170', '9069', '6330', '48900', '10912', '45952', '3528', '14820', '4919', '32449', '21870', '11589', '21408', '7272', '9529', '9169', '89640', '6985', '3950', '15372', '9534', '10680', '4891', '18176', '19626', '8904', '42940', '5761,9', '34140', '4740', '23822', '61540', '22890', '26660,2', '4252', '20059', '21069,8', '7611', '25168', '18399', '5700', '12393', '6138', '41730', '9825', '16100', '45665', '10147', '7019', '30300', '17568,3', '18602', '3356', '27472', '7625', '7313', '9061', '6496,1', '14300', '5568', '13365', '7455', '16669', '25068,94', '14031', '3717,9', '11720', '18972', '30520', '3798', '4768', '9928', '17039,91', '18225', '17060', '20592', '14990,95', '14645', '5544', '26090', '18797', '2999', '11389,2', '12947', '3550', '29345', '12559', '4533', '4366', '16900', '13727', '10542', '20143', '50330', '36260', '17097', '7089', '21350', '3789', '18376', '9432', '19886', '6838', '8589', '15619', '3130', '4479', '7350', '9007', '8849', '2909', '34916', '75500', '16760', '20530', '7380', '61250', '36800', '5718', '22927', '13455', '34630', '16460', '10471', '19840,5', '7908,1', '20985,62', '24420', '6640', '13999,2', '9278,1', '17325', '54738', '7590', '17594', '5796', '24198,2', '17428', '18857', '17360', '36527', '3580', '11148', '6710', '16063,29', '16605', '19765', '23000', '6058,1', '3059', '3385,92', '25250', '43373', '30198', '20209', '42930', '6199', '27501', '15403,02', '10752', '4930,95', '10670', '79679', '24520', '25899', '88300', '9545', '20850', '2072,7', '21220', '22558', '13949', '6397', '28861', '13641,3', '99308', '18460', '15795', '29037,12', '9550', '27088', '27040', '10105', '7589', '22803', '16407', '14453', '6270', '5562', '9917', '8960', '23780', '29789,91', '10618', '11389', '11333', '13227,7', '24570', '24295', '27518', '8237', '26700', '15927', '19143', '19660', '7860', '20498,1', '18720', '8948', '32280', '19513', '21197', '24835', '21187', '3720', '21721,1', '40410', '6624', '3668', '4403', '8095', '3271,5', '4492', '30729', '17503', '13630', '9710', '35332', '17998,2', '7740', '19095', '20356', '14331', '29590', '14310', '10290', '11305', '21000', '6101', '4898', '19990', '7498', '19976', '18592', '6245', '5510', '3269', '3040', '12890,92', '22275', '29443', '6712', '46921', '5817', '3033,9', '10151', '20379,5', '22236', '19482', '96860', '25820', '5880', '11639', '8118', '8430', '12506', '6342', '7679', '31646', '24778', '10978', '4262', '7189', '21724', '7017', '31096', '11660,93', '29400', '2300', '6118', '3509', '6492', '7230', '19503', '21994', '21250', '28983', '31050', '19218', '7910', '2819', '22490', '11396', '2878', '7899', '5185', '10022', '7703,1', '19446', '11487', '7965', '86618', '34806', '19789', '19839', '7578', '19082,3', '3674', '9622,77', '20715', '21809', '13989,4', '12749', '12900', '26529', '24680', '15359', '22419', '16634,5', '9647', '3879', '20820,2', '4673', '10348', '15980', '7558', '3555', '3006', '11150', '7593', '4647', '25966', '13386', '12366', '6585', '39450', '6887', '24458', '56000', '3385,83', '4648', '8070', '45672', '12741', '8723', '9398', '13720', '12745', '13709', '18068', '3185', '19665', '24239', '35598', '27714', '4920', '15856', '12474', '9000', '25182,4', '13993', '31792', '18080', '70500', '16064', '17545', '27740', '20343', '9995', '11960', '4572', '11269', '13900', '10450', '20931', '18185', '78150', '9200', '4900', '9899,1', '5689', '11376', '10575', '23912', '12100', '23706', '17248', '14875', '5105,7', '28026', '22670', '3137', '2255,4', '13899', '44100', '24275', '7123', '21950', '16750', '17000', '11441', '7197', '20250', '15027', '6370', '12600', '26300', '7799', '25540', '21038', '29199', '15930', '6125', '12541', '26400', '19220', '14991', '4761', '31990', '4135', '4969', '37415', '31910', '17470', '12411', '21246', '3530', '11665', '3929', '9971,89', '13173', '7730', '38880', '10650', '5455,8', '23737', '13649', '12130', '6457', '26750', '24777', '18760,75', '31694', '32499', '6744', '9450', '6281', '25140', '7486,1', '8490,92', '3980', '13302', '24587', '18790', '2720', '2907', '3028', '5977', '4788', '3820', '13456,82', '8532', '3715', '8076', '21808', '3387', '3779', '7443', '11510', '15065', '23216', '6980', '18667', '6598', '2610', '27110', '28641', '37200', '14481', '38725', '6858', '2569', '20700', '81000', '5100', '4290,95', '14595', '22110', '10260', '25290', '12903', '14540,04', '3399', '6400', '10029,92', '57593', '10158', '19991', '4239', '31250', '16317', '7582', '22955', '19005', '49000', '18220', '16450,71', '32970', '6525', '14177', '21938', '60720', '8120', '10194', '15067', '6295', '31817,2', '24380', '15760', '9060', '18624', '46960', '21243', '8581', '28580', '39422', '3489', '8361', '30918', '21154', '27096', '12250', '22990', '27270', '7739', '22047', '21984,55', '20640', '19410', '6514', '99540', '6945', '18020', '10936', '10646,2', '6299,2', '5060', '34396', '9290,91', '17340', '11997,2', '13410', '11502', '7919', '98644', '2724', '34557,1', '17575', '14140', '4080', '59760', '19280', '9019', '11098,8', '28258', '8609,92', '11470', '17073', '21200', '17826', '23060', '7400', '21420', '3039', '10538', '6462', '4963', '4329', '10413', '3748', '19611', '7163', '16823', '21475', '26920', '21728', '6600', '3570', '36136', '8185', '13310', '12914', '6740', '6997', '2671,2', '23160', '4196', '17399,2', '4951', '17910', '10124', '22525', '11630', '7002', '6336', '4050', '10495', '26780', '14197', '7169', '21398', '2798', '13491', '7422', '26490', '15629', '18171', '13952,73', '4500', '27880', '13351', '13090', '27190', '2158', '15435', '4475', '26600', '10584', '18054', '29294', '16798,2', '16977', '8560', '11817', '28170', '25588', '13644', '13800', '22627', '26524', '15615', '12518,5', '9215', '7559', '3340', '11936', '14678', '7788', '6691', '5928,1', '12050', '99000', '21910', '9690', '10001', '31702', '3240', '18648', '3380', '8275', '33510', '24300', '19106', '26708', '6150', '16859', '33700', '3080', '11690,91', '10910,6', '21112', '12708', '5178', '4892', '19502', '20561', '11999', '4760', '7353', '7556', '34500', '7776', '20540', '13463', '12833', '11751', '42746', '13975', '7755', '40970', '5710', '24281', '7391', '5424', '34766', '5530', '3636,9', '11085', '18413', '30990', '7135', '16322', '10782', '10970', '46540', '42950', '26395', '6301', '25506,19', '9801', '19110', '15895', '49100', '21479', '25340', '90590', '4633', '26355', '40499', '25700', '53000', '8699', '11514', '13536', '13115', '7824', '6236', '2263', '26538', '6707', '20900', '23498', '49653', '28338', '70705,96', '24717', '13336', '15920', '13750', '12209,5', '6719', '30630', '18239', '13978', '16222', '4188', '30361', '19820', '14800', '18984', '4590,2', '27510', '17379', '21746', '13021', '10480', '10616', '4042', '3752,9', '8870', '5859', '6311', '15810', '8856', '20990', '29875', '4059', '22225', '3920', '7648', '2599', '12849', '16428', '9510', '12400', '3045', '12199', '3419', '20939', '6048', '8240', '25150', '11429', '5490', '11499', '14081', '6347', '31090', '5694', '19650', '28509', '15220', '31498', '11642,3', '9180', '16893', '26875', '7040,92', '55174', '6129', '17215', '19377', '4408', '4843', '11935', '17648', '3063', '4785', '8420', '2330', '35060', '9620', '12095', '9695', '7677', '25145,79', '3159', '29975', '2990,92', '7369,7', '66680', '8492', '23774', '5140', '18547', '3167', '4890,2', '5404', '28594', '16540', '37710', '10695,13', '15221', '17980', '9153', '7812', '7279', '22307', '2800', '10560', '12347', '12061', '3628', '27790', '4217', '15523', '9655,2', '6237,03', '7950', '26365,6', '9496', '31882', '17863', '9308', '22599', '6354', '5043', '91536', '25343', '50177', '17049', '20751', '9085', '3103,2', '6475', '31716', '8071', '3239', '30615', '9280', '39025', '94300', '15575', '20100', '9750', '5715', '10466', '19089', '5219', '5799', '11540', '9230', '17900', '17110', '15503', '20242', '10169', '11880', '5298,1', '26346', '4872', '26054', '3357', '3429', '21216', '5849', '32500', '16128', '8619', '13389', '10248,2', '3909', '16455', '29760', '20995', '18997', '14947', '9367', '21494', '15036,21', '19254', '3400', '7790', '11340', '30697', '13989', '13882,92', '7039', '9880', '7199', '13387', '10185', '8499', '14000', '11719', '19547', '14998', '11677', '20536,1', '17700', '10280', '13719', '17854', '10347', '7858', '12542', '22070', '21881', '6950,7', '28000', '18144', '5267', '11755', '11754', '2572', '12829', '19083', '14593', '32593', '6893', '17995', '5076', '12148', '8098', '20789', '6682', '7077', '6000', '13326,5', '11081,52', '15725', '7937', '9798', '27995', '11927', '14866', '20196', '4520', '11212', '27313', '51200', '6338', '26049', '6073', '6880', '88440', '23037', '7639', '6700', '22395', '17860', '5064', '23390,91', '12680', '5243', '9260', '5590', '3885', '13134', '21700', '4525', '28308', '16185', '3051', '33344', '15540', '22192', '7373', '12890', '3935', '21124', '7250', '16615', '8204', '7157', '2573', '29012', '53673', '13592', '18463', '17538', '10938', '23539', '15616', '46695', '3328', '30060', '11833', '15806', '21990', '9420', '21650', '10850', '19120', '9342', '3359', '17279,2', '26170', '78400', '7820', '45000', '11195', '4487,4', '5242,3', '14787', '3128,2', '6890', '15272', '4842', '7307', '5936', '19128,41', '5135', '13300', '8377', '11520', '2955', '21589,9', '25491', '6407,5', '27360', '2282', '14625', '8529', '4628', '6115', '11170', '11160', '3718', '6698,3', '4175', '47309', '18649', '6662', '31502', '5250', '3703', '3675', '6004', '5588', '19854', '7505', '37790', '21030', '28238', '15822', '33760', '22789', '56518', '11048', '7448', '16063', '11500', '28500', '21710', '9054', '3170', '16880,91', '19497', '3348', '8603', '9546', '15552', '11689', '15685', '4769', '20440', '11750', '5110', '20393', '11930', '12598,1', '33996', '31000', '3023,58', '32765', '13284', '7619', '11020', '4858,1', '16330', '16364', '3690,82', '25200', '8448', '10599', '15240', '18210', '12265', '11524,79', '9393', '16480', '13366,2', '40313', '80440', '6130,8', '5160', '17810,82', '9749', '7795', '6978', '17999', '38032', '8785', '9870', '8811', '30813', '5699', '16010', '10840', '28016', '5869', '18690', '5560', '7083,05', '23220', '14770', '31190', '20040', '8517', '9913', '11264,25', '12328', '16975', '26760', '3322', '7630', '14618', '13539,2', '4650', '16194', '4918', '18849', '19920', '39327,2', '7642', '10150', '6539', '16358', '20650', '3089', '6732', '10844', '19133', '10551', '13232,88', '8606', '18022', '90747,9', '13941', '18670', '14089', '14165', '29622', '7660,42', '11680', '22458', '11211', '5222', '7385', '24027', '9555', '10174', '20748', '10580', '13846', '18040', '30420', '18300', '30360', '25386', '8990', '13344', '26164', '27925', '14430', '11536', '3880', '24061', '17099', '23517', '16692', '10490,91', '6529', '6562', '2240', '6346', '9300', '23352', '9990,91', '13492', '3046', '18782,2', '16440', '3900', '19589', '5683', '8832', '27540', '9858', '21696', '19037,2', '18530', '61487', '10170', '4980', '7334', '3617', '11911', '14440', '12289', '10953', '7847', '40534', '8169,55', '2435', '9090', '10256,1', '12290,94', '23450', '6622', '8102', '20601', '16629', '6822', '43996', '9923', '8047,1', '28402', '13489', '9416', '14670', '6283', '14700', '2540', '10803,84', '98136', '30299', '27372', '8235', '27637', '14190', '5772', '9460', '30999', '29475', '8180', '13635', '13010,94', '2578', '18143', '20500', '15035', '10648', '30656', '8957', '5055', '10262,2', '12850', '2691', '8920', '2457', '10248', '37880', '11188,14', '63320', '14688', '2104', '4074', '23397', '28294', '4741', '6552', '20375', '4710', '35793', '20060', '6919', '12470', '37440', '3060', '6690', '12899', '7460', '29075', '17838', '21381,12', '18890', '93000', '9876', '17505', '16680', '6944', '11116', '36267', '11810', '9484', '27890', '4835', '10671', '4082', '12424', '21688', '10428', '26430', '28230', '4030', '32508,2', '36086', '23511', '4180', '18899', '6904', '22160', '16112', '13615', '9021', '28640', '8669', '18286', '19800', '18553', '14921', '25702', '8790', '12166', '30224', '35360', '3222', '5955', '3771', '25480', '7745', '5304', '14189', '10112,8', '25600', '12488,25', '25584', '28600', '12530', '5797', '11920', '25460', '21280', '45500', '4260', '4789,92', '8290', '19085', '7544', '16899', '3329', '8295', '14270', '8860', '13875', '8753', '7770', '12410', '56899', '10529', '10330', '6549,83', '21286', '8650', '2248', '10097', '17769', '16840', '5438', '3196', '14443', '26990', '19295', '30673', '39741', '14004', '12198', '10525,01', '27400', '15291', '26082', '14744', '60829', '16045,2', '7597', '16448', '5835', '33192', '7236', '27149', '15246', '8638', '9097', '7895', '10060', '11151', '2974,15', '20883', '4624', '2520', '9892', '27945', '10200', '23758', '26241,6', '80973', '9116', '4970', '19959', '4311', '6503', '4820', '29630', '5723', '24830', '3951', '5818,5', '13265', '9729', '78800', '17714', '17499', '26060', '14337', '5247', '10659', '15210', '15292', '32822', '9170', '17885', '16657', '9045', '16175', '5193', '29266', '14205', '16850', '12755', '46671', '6452', '7070', '9660', '6903', '14257,91', '5897', '2030', '43682', '22498', '58670', '21784', '19591,36', '11330', '20313', '61920', '74650', '7999,1', '6419,78', '4611', '9325', '13390', '16425', '11995', '24199', '31672', '4999', '7638', '14090', '10769', '22269', '29694', '6789', '9920', '4580,92', '7348', '93400', '35683', '9025', '15200', '4031', '65629,5', '6136,58', '17106', '20095', '38850', '6481,8', '22770', '58567', '25100', '29990', '18632', '10920', '9908', '18224', '14489,2', '18490', '27759', '4905', '11623', '11938', '12999', '18686', '5440', '12599,9', '31015', '14980', '13460', '14390,92', '28820', '12610', '22200', '15466', '18190', '16312', '15020', '38120', '13490,2', '18962', '33800', '21021,9', '11080', '6715,2', '47247', '30796', '32175', '30500', '10445', '10287', '13447', '14020', '9900', '41990', '13229', '29580', '22692,96', '29877,3', '7247', '7210', '9144', '18250', '19100', '38553', '10989', '22860', '21265', '75424', '24280', '15430', '4243', '6165', '32387', '16882', '13382', '24357', '4318,2', '16199,2', '14133', '8432', '27800', '20992,78', '4998', '12211', '10837', '4297', '22329,67', '12996', '14494', '22369', '39865', '8102,08', '18650', '15560', '32750', '34720', '47704', '27700', '10655', '10523', '14570,01', '11325', '5937', '3307', '9989', '6323,33', '22560', '19129', '16630', '40110', '13009', '23095', '19850', '35750', '29470', '26232', '15350', '5611', '78550', '25081,98', '16800', '12024', '5230', '15524', '15910', '8406', '3099', '6950', '15771', '12713', '5430', '41936', '11669', '20609', '8020', '3460', '9264,69', '20570', '4560', '24580', '4598,6', '5920', '17390', '10779', '6263', '16993', '14948', '9850', '9310', '5080', '22750', '31224', '5575', '19542', '17198', '17712', '18491', '9842', '6429', '23186', '9650', '23720', '7800', '23985', '5180', '3996,9', '21578', '8450', '18598', '3604', '18374', '3490', '13891,5', '25330', '10020', '25179', '62240', '13278', '13039', '90000', '14250', '18599,4', '19745', '84000', '10990,91', '18668', '16776', '23850', '19290', '25940', '16660', '23355', '20792', '18560', '32415', '17600', '5279', '9977', '6620', '22349', '3179', '4312', '12398', '14380', '8188', '10193', '27070', '4869', '8396', '13780', '27934', '5340', '32580', '11748,2', '12581,89', '3031', '3630', '18246', '13895', '31317', '19877', '17960', '5296', '9664,5', '22588', '24017', '22947', '2470', '24370', '10432', '10320', '13499', '10220', '5218', '15695', '2880', '21075', '10628', '34020', '5176,8', '6272', '5045', '16300', '11550', '21915', '23584', '7859', '12412', '25887', '26183,5', '21062', '2950', '25492', '16065', '18000', '6449', '56700', '51685', '4940', '20133', '5630', '12099,1', '2676', '16030', '17480', '20355', '38200', '9651', '14391', '7000', '22920', '4922', '8398,1', '21900', '27080', '11615,4', '8890', '97220', '10492', '8547', '64500', '33990', '18540', '9464', '9973', '3484', '24675,85', '8429', '12953', '8340', '25898', '19235', '4290,9', '6979', '14527', '19500', '27950', '17622', '26115', '7471', '5347', '10057', '6519', '7431', '21690', '11055', '6892', '23440', '6630', '8600', '12500', '18280', '23365', '3602', '8067', '10801', '16011', '14240', '4936', '29944', '27706', '4986,5', '3983', '11691', '5306', '18600', '3740', '8889', '56340', '24346', '2638', '16596', '19177', '7618', '12320', '34422', '7849', '3770', '12380', '7386,75', '5778', '15490', '9550,82', '15389', '15120', '7680', '19269', '20600', '8233,95', '5908', '10992', '21320', '3474', '18898', '22015', '7035', '11776', '11892', '43200', '11440', '9396', '17690', '20871', '22515', '65747', '7370', '31070', '5268', '42135', '22212', '39254', '18799', '2751', '8110', '21410', '6440', '29599', '12960', '5690', '12725', '26280', '27005', '3312', '41000', '26547,1', '5889', '6180', '6130', '16691', '13872', '17613', '19940', '19593', '28999', '12012', '20833', '24700', '23297', '10350', '3889', '6596', '12598', '2748', '50929,82', '11590', '15102', '15269', '24236', '26848', '18615', '2580', '10900', '6365', '2420', '4790', '6023,7', '34190', '6610', '7797,6', '8341', '5573', '11900', '15770', '9630', '15290', '4890', '10576', '5299', '2690', '23130', '10360', '12670', '6399', '20698', '5208', '28055', '22980', '9890', '10238', '34390', '9098', '16579', '23399', '6757', '12810', '21207', '5427', '16000', '23689', '19689', '5594', '25137', '72000', '6780', '14989', '56277', '9596', '4330', '14204', '29615', '23331', '10969', '4610', '27550', '29285', '10034', '29065', '16080', '21717,92', '16298', '4270', '41498', '6228', '4099', '8370', '14098,1', '13208', '3647', '13530', '13998', '22244', '32041', '4290,91', '21805', '15370', '18338', '31696', '8988', '52250', '16999,82', '2899', '20610', '24131', '7280', '14438', '33000', '12048', '55000', '30391', '16199', '13698', '8793', '11699', '28428', '29933,4', '4542', '7452,4', '18458', '23825', '19748', '4100', '88555', '2495', '10879', '41211', '14773', '10429', '13198', '4602', '13790', '32935', '21355', '16561', '29942', '16227', '23595', '7917', '21985', '45900', '23684', '26550', '21505', '10342', '16669,84', '9452,97', '12053', '15237,1', '11392', '7459', '5708', '5211', '5400', '21240', '13280', '17890', '10422', '13315', '4779', '18200', '23795', '4896', '16910', '4109', '14480', '11970', '19770', '3690,92', '8289', '23094', '36000', '9326', '36910', '21810', '36900', '8011', '37752', '5829', '18498', '16700', '14215', '84558', '12138', '3943', '3510', '9345', '19420', '13197', '39690', '28775', '44290', '13018,1', '27898', '21279', '14588', '6329', '22650', '15582', '3897', '9567', '4122', '21195', '19315', '14857', '4945', '24059', '14528', '7492', '15040', '31499', '14388,3', '22188', '19993', '3591', '10025', '4292', '4165', '10265', '3350', '11583', '8767', '15016', '43515', '12390', '38830', '59398,2', '6854', '12537', '25627', '13248', '27370', '92951', '10660', '64660', '7262', '27342,5', '12339', '9580', '4469', '21290', '40581', '18260', '69282', '24871', '7335', '18334', '20802', '21907', '13792', '25936', '40000', '4923', '25897,4', '7741', '90200', '27130', '31014', '13000', '38244', '9427', '32490', '9655', '7867', '24200', '12510', '4450', '17184', '13267', '50879,83', '13087', '24125', '23121', '21948', '30545', '3549', '2873', '8389,91', '15482', '2009', '15824', '51000', '5988,46', '26989', '10960', '5754', '13769,2', '18910', '19498', '33088,5', '16500', '7208', '27744', '22150', '35097,66', '27750', '27283', '2270', '9898', '2881', '18204', '27686,2', '10611', '33126', '26499', '60000', '28186', '5794', '7595', '4148', '9480', '9855', '6020', '14450', '4581,87', '19296', '14441', '24970', '16732', '33300', '19011', '13155', '16180', '28845', '13257', '22195', '30372', '23677', '15750', '26997', '5359', '3749', '22520', '6424', '6543', '8171', '3188', '22662', '6834', '3799', '23661', '45820', '27905', '21598', '4869,66', '11335', '11030', '45160', '17203', '7190', '4056', '8994', '16920', '9498', '24949,09', '24640', '10046', '3711', '10128,8', '25599', '19670', '31255', '4480', '20852', '25320', '29490', '8129', '18896', '32850', '21658', '27170', '14080', '28945', '18350', '17399,1', '8995', '70000', '14196', '12405', '18400', '36500', '6750', '9227', '27895', '9100', '5253', '8346', '6540', '14574', '19795', '4555', '16240', '11060', '18603', '15215', '42180', '33920', '10100', '27500', '33900', '8116', '4230', '35990', '69188', '19116', '38340', '6086', '5091', '20050', '5397', '8685', '4020', '14119', '8515,7', '19760', '50700', '17615', '7366', '16736', '19044', '23700', '5350', '14976', '20397', '2514', '19140', '22656', '28472', '13191', '4852', '44900', '3238', '27398', '11115', '7259', '5592', '11787', '15547', '19289', '12350', '15900', '13137', '25220', '12232', '38250', '7600', '15705', '5857', '12678', '20179', '2322', '24495', '6830', '27124', '12421', '23140', '22645', '28551', '10435', '18115', '7660', '16048', '7291,85', '25992', '14330', '14229', '2152', '3906', '15518', '31400', '32450', '3480', '22213', '20960', '5606', '2135,14', '4683', '6538', '69500', '17501', '23647', '4131', '12616', '17010', '7152', '24160', '31221', '5852', '9050', '11490', '18121', '3440', '16670', '6323', '19602', '27350', '76738', '34335', '15100', '9654', '32293', '26091', '18209', '9576', '25623', '4671', '13496'}\n",
      "\n",
      "\n",
      "==========\n",
      "FST_PAYMENT\n",
      "{'948', '4846,96', '2647', '1444', '460', '1763', '999,2', '1996', '168', '1875', '14200', '648', '1895', '120', '2200,91', '10700', '9190', '4750', '9200', '4900', '6882', '687', '710', '2784', '5790', '13360', '3411', '12100', '2604', '2060', '2001', '800,82', '2605', '3212', '1608', '7794,92', '680', '1160', '2482', '8389', '316', '16750', '17000', '8', '8688', '46832', '1503', '880,2', '1310,62', '588', '1915', '5120', '10489,92', '11230', '1392', '2013', '26802', '6370', '3426', '7495', '2140', '825', '12910', '9444', '12600', '3822', '5075,84', '599,95', '2355', '5234', '7000,5', '2320', '515', '3495', '18588', '5786', '3100', '561', '2315', '2458', '1216', '4916', '13818', '2528', '4335', '1500,92', '10177', '4194,3', '9020', '652', '481,8', '550', '439,95', '10650', '20890', '409', '2499,8', '2173', '5352', '3678', '9457', '341,1', '5784', '3371', '4567', '160', '1098', '959,3', '757', '723', '1530', '832', '443', '1005', '725', '3330', '2090', '3014', '2500,2', '3980', '23800', '1888', '6000,1', '210', '9730', '4551', '4870', '5767', '2720', '3028', '553', '2985', '3215', '6837', '4788', '1290', '346', '362', '1340,92', '4430', '234', '4950', '3779', '13650', '3284', '2394', '6598', '511', '2610', '276,6', '2689', '3760', '426', '551', '2569', '1806', '1219', '6975', '12251', '625', '5100', '586', '1865', '10260', '2843', '3599', '1892', '1866', '6491', '6400', '1591', '306', '3448', '4239', '299,1', '2019', '1300,92', '6398', '2776', '9096', '1845', '855', '4925', '797', '1082', '692', '779', '14940', '1255', '1500,2', '25000', '8120', '230,3', '610', '1434', '4027', '4593', '1369', '468,1', '1819', '3172', '1890', '358,1', '1765', '1196', '765', '1103', '1831', '463,2', '1453', '799,5', '1935', '1386', '1239', '13400', '1139', '1046', '7623,3', '1950', '3489', '1597', '2053', '580', '1047,82', '598', '1112', '2345', '1990', '6777', '27000', '2454', '6945', '213', '875', '1524', '687,7', '10270', '10015', '626', '2233', '1182', '774', '5060', '7165', '2383', '4823', '4000,92', '5812', '1952', '2730', '4935', '6534', '2890', '1666', '747', '1175', '1737', '1999,8', '2870', '2790', '6820', '3198', '4080', '834', '3930', '1823', '880', '1087', '9288', '2250', '5990', '760', '518', '7400', '2970', '3039', '1717', '1050,92', '15939', '3745', '1040', '895', '1629', '950,92', '2651', '1785', '9108', '555', '3960', '6600', '3570', '4579', '1609', '3829,6', '9685', '2222', '2857', '1610,4', '1534', '1093', '5150', '42906', '868', '5722,21', '512', '3108', '53370', '5375', '2500,4', '26000', '2614', '1000,92', '5248,91', '4050', '5550', '4091', '1824', '250,6', '2798', '6326', '4455', '1039', '1815', '3228', '4500', '788', '1099,9', '1015', '8263', '11320', '850', '3454', '9375', '5038,1', '9639', '2754', '2845', '6910', '1814', '1075', '2198', '894', '685', '4851', '29080,92', '758', '1002', '3442', '1291', '811', '50500', '818', '1417', '3520', '3340', '732', '360', '2130', '10227,2', '3003,3', '2998', '2926,4', '41580', '11076', '4310', '1528', '711', '1252', '4350', '1410', '3240', '8650,48', '28257', '380', '3380', '7464', '5979', '5260', '911', '3363', '3080', '11337', '2403', '22399', '7964', '620', '11660', '21648', '18375', '10557', '4060', '1640', '13613', '637,4', '21750', '962', '2380', '651', '5899', '12700', '988', '3482', '7776', '3300', '769', '9922', '1655', '217,5', '669,3', '772,3', '10000,1', '3717', '7633', '5305', '2287', '7845', '40664', '1025', '19521', '3200', '698', '1351', '2180', '1399,3', '4155', '9391', '4370', '5710', '1547', '532', '470', '7125', '6820,45', '1969', '4014', '3135', '3815', '5022', '4300', '1190', '41774', '3000,4', '1484', '7240', '816', '4476', '332', '475', '13526', '1600', '982', '421', '5246,5', '9411', '3596', '917', '2500,8', '1512', '2442', '7148,1', '1159', '2340', '5726,3', '7100', '0', '1289', '933', '1605', '4199', '1110', '1780', '6513', '3043', '48800', '1821', '3469', '2688', '4850,8', '4775', '1625,2', '6007', '14220', '1708', '1897', '7477', '13536', '1018', '2553', '4364', '2263', '16070', '1680', '660', '1125', '936', '1522', '2990', '3370', '177', '1008', '606', '3183', '13750', '1148', '7880', '12300', '653', '49038', '3270', '1784,81', '24000', '718', '1195', '11425,8', '4743', '444,1', '6340', '1540,84', '1517', '36330', '3516', '4098', '140000', '4499', '3236', '14935', '2700,75', '3320', '2700', '429', '1700,92', '980', '1840', '8680', '2550', '1249,95', '2205', '2599', '925', '2077', '29032', '1129', '2698,89', '635', '1116', '2584', '3795', '1482', '845', '1900', '2911', '5844', '3910', '10000,92', '50', '2151', '3045', '612', '1290,6', '675', '364', '3499', '12044', '4815', '2040', '905', '1281', '4580', '2299,8', '372', '14385', '305,1', '1556,8', '5437', '58800', '1339,9', '1338', '20000', '11673', '2170', '5398,1', '677,3', '2098', '451', '34000', '2688,69', '1297', '941', '31350', '3790', '1176', '3134', '540', '830', '828', '392', '2330', '10527', '3333', '3000,91', '694', '366', '1172', '1026', '14500', '9695', '1326', '691', '2016', '2570', '3159', '2159', '2333', '2370', '7185', '43359', '1077', '1389,8', '2816', '798', '2360', '1356', '1233,2', '2864', '2259', '3660', '877', '1319', '4948', '5000,91', '5140', '4274', '2440', '6480', '498,6', '2099', '3275', '6300', '344,8', '11221,5', '628', '1200,8', '1884', '2800', '1343', '860', '2930', '1180', '51127', '3205', '3114', '431', '5735', '10560', '772', '2156', '4565', '1924', '2136', '3585', '2571', '4053', '5811', '4390', '7950', '5610', '8710', '10090', '19958', '415', '3995', '739', '745', '1812', '1100', '3780', '20644', '371', '1207', '5450', '3159,86', '1153,5', '383', '1288', '7447', '964', '1460', '627', '2703', '524,85', '2335', '3670', '2613', '469', '2420,1', '667', '910', '1161', '1509', '1500,95', '549', '3739', '5814', '1912', '7547,8', '9750', '1682', '1796,94', '240,6', '1648', '3800,2', '9955', '10140', '1328', '1340', '17110', '1229', '30660', '1549', '1910', '1004', '10950', '3260', '3357', '5300', '564', '2115', '1565', '1390', '3339', '1808', '14448', '1820', '7080', '554', '3524', '2210', '529', '4523', '24500,03', '4250', '7082', '3400', '696', '2655', '3605', '2045', '1869', '13519', '5240', '1400', '8300', '7528', '9770', '8499', '1519', '14000', '2294', '6785', '2007', '387', '1760', '2500', '2988', '3000,2', '19950', '1260', '615', '1860,9', '2879', '3000,95', '10795', '1200,2', '1348', '6396', '2000,79', '8117', '2596', '13580', '28000', '10152', '1703,2', '868,42', '5504', '1041', '446,05', '1231', '2763', '9979', '7707', '11754', '6646', '2572', '984', '1000', '16000,1', '1462', '955', '1383', '330', '893', '4320', '583', '7506', '748', '6200', '820', '3086', '2860', '3318,4', '1318', '2916', '9568', '4185', '5670', '6000', '39487', '510,9', '3310', '3119', '9574', '1939', '8190', '363', '4095', '642', '9418', '3416', '2248,1', '46500', '5985', '1450', '2289', '605', '9047', '3073,8', '3000,92', '2000,6', '62250', '1899,9', '7174,87', '6700', '449', '545', '1841,91', '5445,92', '1639', '542,2', '5108', '8945', '870', '3000', '9272', '2796', '859', '3885', '3270,82', '2125', '10465', '5170', '2299,9', '3257,04', '2643', '3935', '6151', '2348,8', '565', '5650', '2498', '21094', '7250', '2323', '10041', '3414', '1309,65', '1107', '10800', '1158', '2399', '3669,6', '3088', '1244,7', '1255,41', '3498', '14786', '1520', '5036', '650,95', '3328', '11292', '8390', '629', '39910', '42518', '7142', '3090', '1399,8', '2084', '4587', '1261', '11382', '2199,95', '0,95', '3470', '49200', '3390', '1278', '5850', '1546', '2324', '7020', '5366', '1739,69', '6357', '1365', '1880,2', '2999,9', '3359', '3398', '1090', '1394', '639', '644', '755', '1470', '3149', '7820', '2132', '1064,2', '1121', '2409', '45000', '7670', '35000', '559', '3590', '4690', '299,92', '6890', '2190', '2892', '8342,38', '1876,25', '1059', '9838', '13300', '929', '2586', '1395', '2955', '1500', '1871,75', '2414', '1475', '10007', '5744,5', '6641', '1786,9', '3337', '1111', '2698,93', '5660', '2590', '4086', '1789', '2835', '6115', '2017', '5245', '6257', '15000', '2157', '1045', '690,95', '1961', '2532', '7990', '1722', '3502', '2482,84', '4809', '1084', '1322,9', '8500', '6290', '1866,6', '2373', '410', '7049', '1123', '3703', '4318', '229,5', '5250', '1257', '54910', '6075', '2201', '3685', '786', '2278', '24500', '260', '3410', '3278,9', '9445,95', '2419', '567', '8750', '9788', '507', '1000,95', '2142', '3575', '767', '355', '450', '4363', '1671', '11500', '5460', '1275', '8320', '3997', '2004', '238', '2175', '3170', '2055', '1268', '837,91', '1370', '55982', '11528', '19556', '11689', '10000,2', '14806', '8008', '495', '4504', '1570', '448,9', '1327', '5143', '15091', '31000', '6399,9', '845,5', '973', '4712', '11020', '8900', '2578,8', '1094', '53530', '2702', '2129', '3403', '2131', '11044', '25131,59', '546', '2900,5', '399', '56638', '2121', '496', '2593', '1422,6', '858', '2014', '4184', '10500', '9602', '5517', '1264', '643,7', '1999', '3999', '18719', '7795', '1956', '2989', '2000', '5580', '9040', '8522,2', '12144,42', '3265', '4521', '3650,1', '6532', '1179', '1588', '488', '1828', '1455', '12520', '622', '1566', '927', '2888', '3415', '5560', '479', '3136', '2901', '1072', '4550', '1210', '1622', '2088', '4590', '3730', '3830', '1854', '1080', '10515', '835', '10051', '1273', '10', '6670', '3534', '2000,92', '920', '7630', '489', '4650', '1198', '557', '7764', '19000', '2898', '2387', '2390', '1989', '2310', '9490', '50120', '2975', '39400', '5148', '11200', '2092,94', '1493', '4000,5', '1266', '3297', '2160', '1331', '1818', '3840', '323', '1293', '2855', '2830', '1921', '5192', '1000,67', '2674', '26410', '3137,9', '1197', '18800', '1703', '754,2', '500', '1103,85', '15500', '10550', '75570', '7420', '2885', '2395', '9500', '669', '744', '1049', '2770', '7500', '1106', '4139,36', '990', '556,2', '4225,44', '815', '9520', '377', '1500,5', '3870', '9991', '1700', '4395', '1171', '792', '681,2', '8816', '3880', '1720', '6320', '7395', '808', '2041', '9754', '3846', '2247', '2076', '1883', '2230', '814,2', '1560', '6010,3', '1557', '2240', '3107', '1978', '485', '1901', '9300', '2685', '5097', '468', '12335', '956', '419', '3230', '729', '4937', '8040', '1433', '1279', '6948', '7953', '3900', '4517', '885', '15886', '6509', '7270', '3209', '3202', '21696', '272', '3763', '10467', '4018', '10205,5', '1573', '3169', '7816', '480', '5540', '2600', '16515', '9267', '1619', '14440', '6933', '2400', '1940', '8800', '649', '1353', '2522', '2803', '4953', '777', '1830', '1922', '5038', '11990', '3639,5', '960', '39600', '3545', '7945', '1948,92', '508,79', '176', '2100', '2290', '8000,91', '3186', '1721', '1355', '1930', '1876', '9254', '457,5', '589', '1567', '1800', '1420', '2710', '455', '2540', '11090', '3288', '16481', '5761', '2034', '2000,85', '4995', '1781,83', '3568', '4742', '2078', '14190', '4989,92', '729,1', '8180', '650', '2000,4', '7124', '5074', '9600', '764', '1842', '2347', '1936', '3104', '4210', '20595', '1265', '2608', '3374', '1776', '1287', '742', '2559', '2027,7', '5989', '14753', '999', '3413', '4275,2', '54929', '700,83', '950', '4675', '1580', '510', '3060', '699', '1030', '3750', '9037', '2265', '1408', '2000,9', '1572', '299', '13807', '1135', '4000', '13500', '3369', '8274', '2840', '4045', '4830', '337', '6625,16', '2331', '878', '33590', '1960', '1140', '2775', '413', '6174', '3150', '2630', '2805', '874', '11116', '6990', '15680', '4360', '34698,2', '1800,1', '0,92', '674', '1201', '5950', '15140', '2632', '433', '65000', '4082', '2660', '1035', '1680,82', '3899,95', '4030', '1894', '2149', '1474,3', '2047', '422,1', '2000,7', '440', '4387', '1050', '5000,87', '3787', '869', '1071', '340', '16484', '7659,01', '953', '405', '52182', '459', '60432', '1294', '381', '7034', '2410', '2466', '2280', '1185', '528,5', '2765', '4005', '2238', '800', '2517', '60245', '8790', '3220', '5847', '787', '3010', '388', '520', '1081', '3850', '780', '1350', '1179,6', '7261', '251,1', '337,1', '384', '6127', '2947', '386', '6995,92', '1999,2', '4011', '3092', '579', '2690,92', '485,2', '4260', '6875', '21500', '9910', '630', '20645', '1399,7', '14640', '977', '1170', '9996', '2469', '5050', '2413', '824', '1491', '2716', '28800', '1533', '3538', '2810', '11800', '5000', '9270', '10330', '1800,2', '8650', '4200', '2248', '942', '1001,7', '10097', '724', '13140', '3970', '1099', '4400', '6748', '10240', '1310', '4346', '8082', '1016', '3450', '48052,1', '1850', '1458', '842', '1488', '3319,7', '1120', '6500', '6160', '10933', '3765', '706', '5800', '1877', '14100', '1717,92', '1058', '3290', '1638', '1768', '2665', '2114', '8125', '12000', '1302', '539,09', '1499', '1060', '1710', '3976,32', '1452', '10980', '3100,3', '2849', '5499', '2520', '46300', '643', '7700', '10200', '10431', '618', '947', '1366,05', '571', '8820', '2241', '5145', '2861', '2325', '6832', '22097,99', '934', '3023', '4820', '633', '1089', '1269', '2112', '974', '965,92', '54470', '1393', '3126', '2615', '3252', '3095', '11027', '4206,8', '2358', '4629', '2006', '413,1', '2460', '3500,6', '6541', '4138', '2966,2', '3249', '1992', '2046', '1625', '1384', '2749', '776', '1065', '2368', '10729', '16657', '1751', '249,3', '2555', '10706', '900,2', '861', '4000,9', '2547', '1543,5', '3666,2', '2491', '4944', '465', '296,8', '2832', '2490', '4352,37', '54230', '2030', '938', '2981', '8923', '5902', '636', '230', '5521', '2480', '749', '849', '1504', '1515,7', '9325', '888', '2616', '1705,5', '457', '1124', '5820', '3280', '25089', '13772,82', '6789', '7567', '1693', '1155', '3597', '279', '5155', '677', '7631', '3658', '17930', '780,1', '10759', '456', '1279,9', '280', '619', '3050', '53541', '3569', '2863', '4319', '2283', '5033', '522,4', '1400,95', '737', '268', '2490,95', '1235', '1361', '681', '1700,91', '407', '2923', '2467', '13670', '1215', '1648,5', '799', '3449', '2997', '25400', '1188', '5928', '1013', '12999', '3364', '10725', '509', '1130', '5440', '336', '1165', '3901', '435', '9700', '10981', '14990,94', '4619', '10502', '8099', '2430', '5023', '1630', '2031', '1486', '9784', '3335', '1860', '1786', '11080', '26565', '5198', '847', '1885', '6240', '44963', '1259', '7207', '799,92', '2740', '18152', '2020', '915', '63219', '9900', '3800', '717,2', '2350', '1820,94', '1691,6', '3000,68', '3610', '736', '2686', '4398', '7531', '3700', '432', '7520', '1469', '1653', '9068', '862', '2372', '1295', '5890', '1958', '987', '3595', '1132', '1555', '9425', '1168', '734', '3235', '2999,6', '1213', '3444', '2231', '6297', '7882', '8164', '1767', '560', '783', '1463', '20235', '985', '1701', '1050,33', '3433', '2850', '12211', '19151', '1550', '22500', '2384', '4700', '638,7', '2560', '1247', '9248', '9845', '1584', '5940', '3916', '1230', '200', '3078', '2186', '1000,2', '2649', '756', '11233', '9964', '1055', '9793', '2780', '1498', '32750', '1685', '2696', '750', '17500', '839', '464', '2472,5', '2192', '2962,1', '1949', '4289', '4040', '2349,6', '5838', '508,5', '9250', '3537', '1872', '1689', '4315', '1245', '1805', '590', '2352', '1270', '2508', '4340', '3087', '1979', '7805', '2758', '16400', '1690', '2025', '5760', '1108', '1735', '1535', '3299', '891', '5611', '4653', '4177', '6999', '4498,91', '2598', '3619', '567,3', '1143', '3899', '971', '728', '3099', '6950', '9960', '9663', '3210', '10695', '1141', '29860', '4720', '3710', '3049', '1477', '4000,91', '3084', '6699', '3148', '2286', '3810', '8640', '404,1', '2220', '838', '4990', '1070', '197', '3140', '4145', '5472', '4978', '2587', '2327', '2338', '138', '4081', '848', '887', '705', '1957', '1849', '7750', '3500', '7800', '479,8', '2954', '746', '368', '8450', '239', '5667', '1730', '1632', '391,2', '292,2', '3490', '10497', '1980', '2260', '1138', '9824', '2029', '775', '1770', '1284', '7586', '6650', '14760', '1898', '2865', '1670', '11024', '3138', '1596', '515,85', '5320', '1574,1', '2024', '2433', '608', '981', '10723', '1740', '854', '5279', '13700', '1487', '1973', '1150', '3360', '11924', '576', '4686', '9570', '269', '1899,8', '3600', '3311', '524', '738', '2712', '29500', '5340', '4770', '223', '2039', '1300', '795', '4597', '281', '3435', '3984', '249', '5177', '4190', '3907', '7118', '2807', '1202', '2095', '8618', '4983,2', '4949', '2356', '8064', '1256', '5000,1', '513', '2166', '1847', '2470', '1715', '4752', '309', '319', '2880', '1899', '3439', '2365', '1479', '1545', '5244', '1677', '4160', '2753', '6929', '2200', '1250', '2228', '5045', '970', '2179', '1987', '655', '18735', '1747', '3124', '6844', '1364', '1695', '7300', '2375', '10425', '2640', '369', '3902', '1497', '7128', '293', '2235', '3749,74', '2950', '686', '7541', '18000', '863', '2670', '1999,94', '5002', '646,5', '1914', '727', '2708', '2506', '5226,9', '6923', '4600', '1092', '2729', '7000', '2667', '10512', '1759', '1750', '1457', '516', '965', '4287', '2052', '855,9', '1329', '21900', '4708', '3680', '8000', '1400,1', '6196', '225', '1994', '5210', '640', '2934', '3874', '3484', '8480', '4000,22', '11000', '1034', '553,6', '11560', '3165', '9664', '352', '1226', '6730', '2477', '694,2', '1725', '525', '6519', '1731', '697', '682', '3698', '16350', '531', '6481,31', '2920', '3030', '12500', '1745', '2285', '1708,5', '3602', '2635', '740', '2096', '1199', '1834', '5532,66', '900', '715', '9871', '283', '5750', '1481', '1802', '998', '2595', '40100', '3451', '5730', '4739', '2750', '4986,5', '4910', '10095', '2588', '2015', '5095', '2949,7', '3740', '375', '668,71', '22400', '3621,7', '1635', '1095', '1809', '558', '17610', '16086', '9039', '7507', '2244', '7822', '1686', '3391', '9790', '8734', '10696', '508', '247', '6554', '1271', '2011', '5130', '1119', '7110', '1248', '10000', '6782', '1051', '1373', '538', '54080', '4817', '2745', '735', '1864', '10230', '23900', '2223', '11440', '6050', '15600', '8100', '1381', '1238', '1146', '5081', '4393', '10008', '2680', '1512,1', '35800', '1933', '4746', '1702', '13200', '19430', '2751', '1649', '5631', '50442', '634,2', '1093,4', '22000', '3105', '5690', '3425', '256,2', '445', '912', '60488', '709', '4680', '41000', '6800', '38000', '1246', '15670', '1341', '1495', '2197', '2668', '1371', '582', '2455', '1771', '10098', '889', '18850', '2900', '1516', '75606', '596', '1507', '13203', '8940', '1515', '12012', '350', '19355', '1514', '10350', '1841', '544', '25405', '390', '1029,41', '320', '2444', '3250', '846', '1559', '1752', '3392', '6550', '6603', '420', '5200', '439', '2510', '416', '752', '301', '574', '1598', '2580', '10900', '6365', '2420', '8700', '1962', '3752', '5999', '6610', '349', '7995', '11900', '5980', '6317', '13539', '3609', '9016', '2245', '11762,27', '10434', '1360', '3180', '11548', '2690', '499', '365', '3277', '353', '400', '6957', '8400', '425', '594', '2493', '1440', '1312', '1651', '7836', '1799', '6410', '4198', '5609', '395', '5461,11', '658', '2000,2', '16000', '2357', '1508', '1525', '5594', '3688', '943', '8751', '5383,85', '1480', '1271,1', '2022', '1225', '100', '315', '250', '4330', '857', '13776', '2499', '1581', '7243', '2535', '10267', '2869,9', '2062', '3097,8', '1385', '2199', '1305', '4828', '48210', '4697', '4270', '1204', '2178', '517', '10637,7', '5746', '6228', '5600', '2825,5', '290', '2958', '781', '25300', '3298', '1670,69', '2437', '1243', '7238,86', '2206', '18589', '1726', '7883', '7669', '359', '1000,89', '3492', '3774', '4980,91', '5500', '2000,22', '326', '1012', '1411', '2369', '542', '1899,1', '2899', '6217', '323,25', '790', '868,83', '2279', '979', '258', '2846', '4235', '55000', '34492', '375,2', '684', '60360', '600', '4510', '2625', '541', '2567', '2564', '4919', '1991', '4110', '940', '49356', '1063', '7900', '695', '4100', '3298,92', '2495', '7170', '130', '1152', '1660', '4546,8', '930', '159', '2213', '3950', '1036', '376,21', '10394,5', '10525', '13790', '393', '1330', '8879,48', '9528', '1147', '810', '1009', '602', '1704', '7913', '1418', '5998', '958', '2549', '20550', '1151', '3157', '26550', '1873', '1056', '15000,4', '5044', '50642', '39804', '1634', '903', '1253,7', '2772,13', '5700', '2080', '5708', '1571', '5400', '6138', '1042', '19924', '1251', '3294', '1955', '3187,74', '11890', '5373', '2271', '449,1', '1719', '1240', '2398', '16100', '3975', '438', '10147', '1561', '1779', '7019', '780,3', '437', '1880', '3266', '6100', '34923', '7811', '1928', '8200', '1431', '3385', '1380', '813', '3677', '770', '3160', '7747', '2650', '606,2', '2429', '3990', '4768', '2858', '3304', '18900', '6495', '21839', '1303', '3794', '3293', '3943', '3735', '70940', '26090', '3510', '9661,66', '631', '4197', '2568', '3550', '4800', '2328', '8349', '13100', '2459', '645', '1523', '345,3', '2415', '539', '616', '514', '1020', '2653', '7298', '1374,88', '1700,95', '1222', '1102', '833', '4268,72', '2457,9', '1595', '6445,86', '363,5', '4498', '4048', '1299', '2980', '3130', '1398', '7350', '1835', '3679', '13395', '5752', '2288', '778', '75500', '1392,4', '1544', '2218', '2062,5', '2538', '3463', '1575', '7200', '592', '5718', '3194', '2760', '1178', '2050', '1430', '3350', '690', '10136', '2450', '1181', '16293', '473', '890', '5001,9', '4343', '7550', '695,2', '10460', '527', '700', '7590', '9530', '6000,95', '342', '873', '949', '1679', '373', '918', '2150', '2044', '573', '15701', '2155', '1368', '8328', '4992', '23000', '2293', '840', '1272', '2405,69', '478', '1937', '40000', '519', '2237', '1069', '2852', '311', '13000', '2253', '276', '659', '3264', '599', '2952', '930,91', '2067', '613', '3015', '49582', '3872', '4267', '4450', '2518', '1540', '7116', '1788', '1810', '446,88', '2500,95', '33694', '1620', '1887', '530', '11860', '3245,95', '1258', '2487', '11597', '1590', '1568', '2070', '712', '1254', '663', '1048', '1778', '8250', '3158', '6325', '9472', '989', '2135', '1501', '16500', '1829,43', '1813', '1292', '379', '10162', '623', '6387', '9550', '1975', '2270', '804,5', '9898', '285', '4353', '1337', '2079,02', '2316', '3346', '2811', '1313', '1010', '4640', '1003', '4434', '5562', '1004,2', '1674', '9480', '1126', '4632', '8540', '555,2', '1424', '7480', '2937', '4444', '1790', '3538,92', '1920', '6835', '6900', '4399', '10750', '17669', '4806', '1870', '44500', '2848', '400,2', '1599', '334', '7553', '8998', '7519', '54870', '1982', '676', '3020', '1855', '1011', '952', '5361', '5900', '1306', '2515', '3120', '6543', '2620', '575,2', '726', '10400', '3720', '638', '4666', '3216', '2530', '2346', '4403', '2826', '995', '383,2', '1333', '2910', '3214', '503', '3657', '50000', '1438', '7740', '9498', '1360,92', '3711', '827', '29590', '25599', '6599,8', '4480', '1563', '899', '3037', '21000', '677,5', '3620', '24800', '4660', '4898', '19990', '2574', '6575', '458', '4490', '3687,5', '3408,8', '9699,69', '3318', '4348', '38800', '6712', '3064', '16250', '10619', '5817', '3055', '18400', '1556', '6750', '720', '2110', '436', '535', '1267', '4358', '2504', '367', '2028', '570', '3162,71', '656', '3223', '3245', '2204', '1656', '10100', '2485', '2697', '1661', '424', '730', '48156,13', '4230', '688', '3969', '2300', '1399', '1889', '9400', '1115', '2797', '1490', '793', '1405', '1510', '1000,1', '11600', '389', '1442', '969,8', '4020', '7230', '3100,1', '1970', '7095', '205', '1829', '10026', '5703', '552', '921', '2878', '23700', '5350', '1874,7', '12358', '2514', '1280', '2375,68', '418', '1220', '14221', '5387', '8688,09', '490', '853', '672', '1363', '1053', '1650', '300', '601', '7259', '5010', '15900', '30000', '3110', '13126', '2935', '1615', '7600', '2877', '2634', '40758', '500,39', '637,3', '2778', '1445', '2010', '2437,5', '2597', '562', '5705', '7660', '339', '1320', '3555', '2887', '1619,95', '150', '1837', '1489', '3718,2', '3650', '2152', '3006', '4715', '867', '4647', '1200', '5999,2', '1610', '4245', '3480', '16668,82', '2364', '1376', '754', '713', '4536', '1997', '6000,81', '720,2', '1374', '485,8', '370', '2200,05', '1694', '3195', '2167', '10516', '4263', '11490', '5000,8', '3440', '4280', '670', '759', '6323', '1091', '2915', '269,1', '3334', '27350', '2698', '4798', '304', '9800', '3511', '771', '480,9', '247,2', '1379', '9000', '430', '2940', '1995', '704', '10397', '4186,59', '51929', '18080', '1959', '3173', '2239', '30303'}\n",
      "\n",
      "\n",
      "==========\n",
      "LOAN_AVG_DLQ_AMT\n",
      "{'1344,33', '1171,686', '1441,57', '1068,47', '2989,98', '437,26', '605,496', '1340', '1241,32428571429', '4680', '1830,2', '1910', '4260', '1010,43', '2820', '3509,22', '1552,325', '2185', '3260', '630', '1538,69', '753', '2701,26', '1390', '856,64', '2509,38', '2186,585', '2262,12', '1313,33333333333', '756,1', '1742,41', '6130', '1164,99666666667', '1170', '567,51', '2900', '1820', '339,83', '710', '1196,43', '2210', '1292,59', '1405,53', '619,73', '2810', '2372,77', '2549,52', '2158,72', '827,57', '2293,65', '2598,9', '4250', '4200', '1088,48', '3250', '802,36', '3400', '3970', '2368,84', '2060', '518,9', '812,11', '5048,19', '911,793333333333', '1639,99', '6689,82', '1310', '2127,1', '4400', '821,93', '2510', '1400', '2112,17', '596,29', '680', '136,76', '1160', '989,39', '1850', '892,18', '2961', '2959', '1120', '6500', '2580', '1760', '2500', '1260', '1411,31', '709,96', '880,2', '673,02', '1007,36', '2656,68', '1789,04', '699,78', '1649,37', '1107,5', '2276,54666666667', '849,91', '367,35', '1141,06', '824,581428571429', '2002,45', '2140', '991,19', '1386,96', '727,63', '3691,2675', '1060', '3209,5', '1710', '1608,255', '5603,82', '1335,47', '2794,22333333333', '1000', '3127,19', '769,43', '1920,9575', '4827,666', '1459,99', '568,7', '2441,4', '1360', '330', '3180', '4320', '1007,8', '2320', '2269,735', '6200', '2520', '820', '1831,51333333333', '722,59', '2860', '1768,07', '951', '400', '3100', '1304,0625', '3310', '2326,81', '1440', '38,32', '6490', '687,11', '776,59', '1258,69', '782,21', '2240,34', '908,78', '1614,36', '1426,72', '224,59', '1325,63', '1179,394', '1450', '1152,15', '3530', '2732,2', '1193,9675', '3303,2', '1697,22', '1589,92333333333', '761,866666666667', '550', '3782,39', '2524,28', '1480', '4120', '578,06', '870', '100', '3202,2', '3000', '3940', '2379,46', '1889,43', '804,32', '433,306666666667', '5170', '1530', '7340', '1666,28', '5650', '2178', '2090', '3330', '2365,47', '2232,28333333333', '6260', '1776,96666666667', '1259,43', '1391,47', '3589,09', '988,85', '2258,3', '917,57', '2720', '1299,55', '2490', '553', '860,505', '2030', '2349,51', '4445,6', '62,46', '1179,86', '1745,7', '1283,89', '1959,99', '984,79', '1290', '3820', '2755,21', '3086,66666666667', '1922,69', '980,42', '2679,82', '1520', '2359,898', '559,94', '2480', '869,26', '1088,05', '3768,7', '5500', '2094,3', '2610', '1113,335', '2782,89', '1029,75', '6390', '3470', '3760', '478,13', '1468,38', '5870', '790', '651,625', '2409,01', '3390', '3509,805', '2590,15', '510,136666666667', '5850', '1276,66666666667', '750,175', '598,5', '620,75', '1582,435', '1312,44', '516,18', '4472,91333333333', '1724,15333333333', '543,17', '600', '1413,765', '543,33', '976,68', '1090', '2699,88', '1716,52', '1470', '1093,75', '875,09', '825,10875', '1489,13', '940', '270', '1529,06', '1439,9', '7670', '2263,2', '444', '5526,72666666667', '3428,56', '989,15', '3590', '930', '4290', '772,73', '2585', '4690', '1300,386', '1200,99', '857,956666666667', '3914,24', '3050', '855', '2518,18', '1140,2', '2190', '2044,155', '1330', '810', '1109,855', '3072,475', '1215', '3964,85', '1875,70333333333', '1500', '1679,2', '1848,05', '610', '2287,02', '2340,06777777778', '4140', '2247,51', '2066,905', '2960', '1317,06', '1324,41285714286', '1740,19', '2573,8', '1003,255', '1208,09', '1348,68', '954,65', '1953,6', '2273,72', '882,01', '59,68', '2590', '5040', '682,56', '9060', '1130', '777,84', '957,59', '1890', '862,43', '1314,11', '2742,25', '1373,99333333333', '2477,64', '1176,82', '4729,07', '5545,0225', '87,39', '3539,82', '5370', '970,155', '2430', '2080', '1059,47', '929,51', '1630', '2751,25', '1860', '1240', '758,82', '1950', '957,4', '4530', '63,31', '1539,2', '13920', '580', '2101,256', '1052,63', '1990', '1880', '610,986666666667', '2259,35', '1549,18', '855,675', '1002,60333333333', '2740', '1288,7', '779,415', '2020', '599,19', '755,063333333333', '3410', '2031,34333333333', '1359,6', '1380', '3433,42', '1677,48', '3501,45', '669,12', '2350', '438,19', '674,263333333333', '1197,38333333333', '355,04', '2910,25', '3610', '770', '3160', '990,85', '110,91', '1555,00090909091', '1593,76', '872,065', '2650', '2890', '1259,39', '5830', '450', '3990', '1155,02142857143', '2870', '1077,605', '683,9975', '2790', '3170', '2547,85', '1259,71333333333', '4001,34', '806,19', '863,68', '1370', '3510', '5269,78', '684,89', '880', '2139,63375', '4810', '1160,96', '4620', '451,495', '631', '1222,06', '457,71', '1275,785', '5110', '2250', '760', '1570', '2151,50333333333', '1636,04', '2739,76', '560', '2235,225', '2464,55', '1683,49', '2957,71', '2970', '642,31', '502,1', '532,24', '749,82', '2239,37', '2419,9', '1459,54666666667', '2462,1', '240', '2850', '1020', '1371,86', '892,95', '1040', '3024', '1010,55', '758,95', '1550', '766,26', '1591,94', '1932,5', '1761,03', '4700', '1094', '1139,49', '2560', '879,74', '3960', '668,53', '1539,13', '728,49', '918,255', '3570', '1374,28', '837,48', '1230', '1205,4725', '2120', '704,93', '1817,76', '1892,238', '1398,58', '2980', '2647,9', '769,915', '3716,35', '1329,54', '724,38', '859,84', '2361,22', '2816,25', '3018,275', '2690,33333333333', '1894,04', '647,42', '5160', '714', '2492,54', '5075,99', '1108,96333333333', '750', '4884,95', '763,185', '1010,27', '115,65', '1516,872', '2538', '4164,34', '1581,82', '1601,63', '2000', '3172,83', '4050', '2760', '134,78', '3190', '2050', '2858,6', '1430', '1070,05', '1033', '690', '762,93', '5550', '2105', '2450', '4158,85', '1095,61', '890', '5872,64', '4055,03', '789,23', '2202,69', '1478,49', '472,11', '5560', '1270', '700', '1966,99', '310,126', '1399,25', '3971,6', '446,31', '1015', '1258,585', '586,646666666667', '1690', '5760', '850', '4550', '1210', '2315,64857142857', '2150', '1429,99', '818,74', '1314,58666666667', '946,49', '2005,15', '6041,86', '3626,91', '324,16', '1844,885', '658,42', '1080', '5397,15', '3506,07', '1269,82', '7010', '4508,2', '1120,93', '5394,29', '3089,155', '920', '918,55', '840', '791,98', '3520', '6660', '360', '2130', '656,705', '1433,4', '840,7', '4650', '3777,93333333333', '1175,57', '5430', '4240', '4310', '4376,4', '1199,27', '1410', '4350', '3240', '3106,29', '2390', '481,506', '673,89', '1706,3', '380', '1144,03', '8020', '3460', '1713,74', '3810', '6070', '2184,66', '552,57', '2220', '1286,73', '1116,49', '1070', '3283,1', '1085,5', '2310', '11310', '4450', '1138,2', '3140', '1540', '1810,155', '560,36', '2201,11', '1810', '3080', '1230,63', '1620', '530', '620', '1033,905', '4115,33333333333', '879', '1841,9', '1939,61', '1149,495', '1590', '1867,66166666667', '1640', '1597,92', '2059,46', '1717,47333333333', '2070', '3500', '2160', '2009', '425,97', '3769,3625', '2380', '1093,22', '3840', '433,3425', '4760', '1793,66', '407,8', '1730', '2830', '3300', '1409,965', '3490', '752,923076923077', '1804,9', '2070,52', '782,705', '816,83', '1980', '2386,505', '823,54', '2260', '500', '3037,58', '2270', '177,34', '1004,225', '3200', '216,59', '1145,85', '2180', '948,58', '1066,005', '1770', '1010', '2770', '3420', '4640', '4370', '1439,94', '2237,245', '678,2', '6650', '990', '1065,62', '1095,89', '2916,18', '979,63', '1931,44', '6080', '94,89', '1670', '6133,41', '670,085', '1700', '3102,49333333333', '1720', '5428,62', '3880', '1833,77', '862,5', '5320', '4300', '2230', '1190', '1920', '1560', '1790', '648,68', '1594,914', '2406,8', '1870', '1402,43', '1740', '1878,86', '2240', '5780', '830,7', '1260,55', '2090,53', '1600', '901,64', '3230', '1150', '3360', '1512', '2370,01', '1883,04', '2340', '846,55', '1427,295', '386,666666666667', '1006,59', '3900', '7100', '0', '3315', '3540', '3020', '1110', '1780', '3600', '1626', '18307,16', '1098,8', '702,565', '569,121428571429', '2979,31', '4770', '3120', '526,563333333333', '846,74', '796,59', '1692,49', '630,49', '3637,43', '1300', '778,32', '1639,24571428571', '1258,49', '763,3', '3630', '2084,32', '798,53', '554,36', '4190', '2530', '791,56', '480', '2600', '1680', '660', '3959', '600,93', '3070', '768,97', '913,4975', '1411,835', '2990', '310', '2701,5', '2470', '2910', '1949,8', '1102,7', '2119,77', '2400', '1940', '2097,97333333333', '604,43', '865,05', '1830', '3659,0625', '3201,95', '673,6825', '6310', '2796,66', '1829,73', '2880', '2568,45', '1288,898', '960', '1508,79', '4160', '966,953333333333', '2018,255', '2200', '2503,19', '3322,7775', '1250', '366,58', '767,566', '1879,28', '2100', '981,71', '2290', '970', '1097,86', '499,69', '2702,4', '3620', '1496,3275', '836,6', '4660', '1930', '289,02', '588,9', '2088,86', '2700', '1675,75', '2640', '1985,3', '1792,71', '1778,635', '980', '1800', '1402,92', '1420', '1264,23', '1215,45777777778', '5510', '1119,27', '776,3', '859,92', '1272,72', '3040', '533', '1840', '965,1', '1154,34', '2710', '2550', '1671,7', '1957,21', '2950', '1113,74', '2540', '665,57', '1387,15', '1509,44', '327,98', '2670', '1309,89', '720', '2429,09', '1900', '1090,8', '2083', '2110', '4940', '3910', '821,45', '3702,8', '2629,87', '4600', '1686,92', '820,124', '1321,68833333333', '1670,18', '558,06625', '1069,6', '367', '836,79', '198,88', '1247,92', '4220', '570', '1750', '1266,34', '1149,395', '2261,19', '650', '2554,005', '1127,76', '2040', '2272,5', '2051,5', '1259,1', '1418,86', '2128,74', '730', '3279,705', '2467,33', '904,08', '1134,58', '2300', '1477,37', '1296,74', '1066,605', '497,62', '4210', '2425', '1369,755', '1155,125', '640', '1490', '2466,8625', '5698,46', '887,79', '812,693333333333', '1510', '2343,41', '269,51', '578,3', '4020', '5030', '1388,83769230769', '1072,34', '1970', '2476,77', '959,955', '2170', '2142,2', '1570,46', '2013,19', '549,88', '605,32', '1258,95', '740,575', '1070,52', '1599,9', '517,51', '2925', '1182,23', '845,55', '677,474', '950', '1751,37', '1731,06', '4710', '2707,25', '1280', '1108,574', '1056,95', '1580', '1220', '1424,9175', '510', '574,21', '3060', '894,29', '540', '830', '1030', '3750', '490', '2265', '975', '2082,47857142857', '2330', '682', '1626,75', '1053', '5076,65', '1749,69', '957,46', '1650', '966,07', '3119,99', '300', '7780', '1135', '2569,05', '726,57', '4000', '2920', '3030', '1077,03', '2840', '1439,815', '1040,78666666667', '3110', '2570', '740', '1319,56', '2269,565', '1899,7', '1733,88', '1615', '805,86', '2510,2', '2370', '900', '3860,18', '1140', '2010', '1634,81', '1960', '1147,5', '3150', '1845,5', '2360', '2630', '2441,465', '875,2', '3890', '5410', '849,19', '3660', '5140', '2750', '1605,95', '767,7', '3635,77', '2497,09', '1563,45', '1237,24', '849,085', '2440', '1320', '1492,404', '613,21', '49,35', '702,23', '1498,26', '1986,88', '1412,55', '1851,55', '3237,94', '663,05', '1221,92', '604,353333333333', '1663,27', '2157,32', '535,53', '1909,16', '1200', '1880,59', '0,04', '2154,02', '2800', '860', '2930', '1610', '1180', '2660', '1684,04', '896,1', '0,71', '3480', '3770', '2779,4', '4030', '4565', '370,77', '4630', '455,32', '1296,4', '1535,89', '440', '1091,2', '4390', '809,42', '1002,22', '2163,94', '1050', '830,53', '603,55', '3870,25', '986,606666666667', '560,86', '429,11', '509,52', '735', '582,386666666667', '315,796666666667', '909,424285714286', '370', '1100', '3780', '1930,595', '1305,97', '51,8', '619,4', '1288,63', '4070', '3440', '957,63', '588,13', '670', '1798,49', '1942,72', '651,16', '1460', '2410', '9280', '2280', '3670', '1410,53333333333', '1781,225', '1144,98', '613,12', '1386,115', '2680', '4920', '567,16', '800', '896,31', '3651,445', '910', '430', '2940', '1352,85714285714', '1283,93', '14870', '3010', '28012,2', '3850', '780', '3762,5', '1350', '14990', '2676,2125'}\n",
      "\n",
      "\n",
      "==========\n",
      "LOAN_MAX_DLQ_AMT\n",
      "{'1344,33', '1324,5', '1068,47', '2989,98', '437,26', '1340', '4680', '1910', '4260', '2820', '3509,22', '1538,69', '3260', '630', '2701,26', '1390', '856,64', '2509,38', '2262,12', '756,1', '1742,41', '6130', '1170', '567,51', '2900', '1820', '2102,75', '331', '339,83', '710', '1196,43', '2210', '1292,59', '1405,53', '2969,75', '619,73', '2810', '2372,77', '2549,52', '2158,72', '827,57', '2589,2', '2293,65', '2598,9', '4250', '4200', '3250', '802,36', '660,86', '3400', '605,59', '3970', '2368,84', '2060', '518,9', '812,11', '5048,19', '4400', '1639,99', '6689,82', '1310', '2127,1', '2902,89', '821,93', '2510', '1400', '2112,17', '596,29', '3450', '680', '136,76', '1160', '2642,83', '1850', '989,39', '892,18', '2959', '1120', '6500', '2580', '1760', '2500', '1260', '1411,31', '880,2', '673,02', '1450,22', '2656,68', '1789,04', '1110,39', '2618,78', '699,78', '1649,37', '1395,7', '849,91', '367,35', '1141,06', '2002,45', '2140', '991,19', '1060', '1710', '5603,82', '18310', '1489,45', '1335,47', '1452', '1000', '3127,19', '769,43', '1459,99', '2441,4', '568,7', '1360', '330', '3180', '4320', '1007,8', '2320', '6200', '2520', '820', '4880', '722,59', '2860', '1768,07', '400', '3100', '3310', '1440', '38,32', '954,34', '6490', '840,4', '687,11', '776,59', '908,78', '1614,36', '1426,72', '224,59', '1325,63', '1450', '3530', '2732,2', '3303,2', '1697,22', '550', '3782,39', '2524,28', '1480', '620,18', '4120', '578,06', '870', '100', '3202,2', '1901,3', '3000', '3940', '2379,46', '1889,43', '804,32', '724,93', '5170', '5470', '1530', '1729,66', '7340', '1666,28', '5650', '2178', '2090', '3330', '2365,47', '883,33', '6260', '1259,43', '1391,47', '3589,09', '988,85', '2258,3', '917,57', '2720', '1299,55', '2490', '553', '2030', '2349,51', '62,46', '1179,86', '1745,7', '1959,99', '3820', '1290', '2755,21', '1870,69', '559,94', '980,42', '2679,82', '1520', '2480', '869,26', '1088,05', '3768,7', '5500', '2094,3', '2610', '1029,75', '6390', '3470', '3760', '478,13', '1468,38', '5870', '3390', '790', '2409,01', '776,6', '2590,15', '5850', '598,5', '620,75', '1312,44', '516,18', '543,17', '1365', '543,33', '600', '976,68', '1090', '2699,88', '1716,52', '1470', '1093,75', '875,09', '7890', '1489,13', '940', '961,94', '270', '1529,06', '3357,52', '1439,9', '7670', '2263,2', '444', '989,15', '3428,56', '1819,26', '919,88', '3590', '930', '4290', '772,73', '1860,72', '4690', '1427,4', '1200,99', '490,88', '3914,24', '3050', '2518,18', '1140,2', '2190', '1650,53', '1330', '810', '3964,85', '1500', '1679,2', '1848,05', '610', '2287,02', '4140', '2247,51', '1740,19', '2960', '1317,06', '828,51', '2573,8', '1208,09', '954,65', '1348,68', '1953,6', '59,68', '2590', '5040', '9060', '1130', '777,84', '957,59', '1046,01', '1890', '2858,62', '862,43', '2742,25', '2477,64', '1176,82', '4729,07', '87,39', '826,77', '3539,82', '5370', '2430', '2080', '1479,51', '1059,47', '1630', '1154,9', '2751,25', '1860', '1240', '1950', '957,4', '1185,5', '4530', '63,31', '1539,2', '13920', '580', '1052,63', '1990', '1880', '1549,18', '2740', '849,07', '1288,7', '2020', '599,19', '2289,89', '3410', '1359,6', '1380', '1828,89', '3433,42', '1677,48', '3501,45', '669,12', '2350', '1386,14', '438,19', '1066', '355,04', '3610', '770', '3160', '990,85', '110,91', '1593,76', '3111,86', '2650', '586,5', '3700', '1259,39', '2890', '450', '3990', '5830', '2870', '2064,08', '2790', '2547,85', '3170', '4001,34', '806,19', '863,68', '1370', '3510', '5269,78', '684,89', '880', '631', '1160,96', '4620', '1222,06', '457,71', '5110', '2250', '760', '1570', '1840,14', '1636,04', '2739,76', '560', '783', '937,79', '1683,49', '2957,71', '2970', '642,31', '502,1', '749,82', '2239,37', '4960', '2419,9', '2850', '2462,1', '240', '1020', '1010,55', '1371,86', '892,95', '1040', '1056,64', '1086,57', '758,95', '1550', '766,26', '1591,94', '1139,49', '1761,03', '4700', '1094', '2560', '879,74', '3960', '1249,61', '668,53', '1539,13', '728,49', '3570', '1374,28', '918,28', '837,48', '1230', '1672,88', '2120', '704,93', '1817,76', '1398,58', '2980', '2647,9', '3716,35', '724,38', '1351,06', '2361,22', '2816,25', '647,42', '1894,04', '937,6', '5160', '714', '2492,54', '5075,99', '4884,95', '750', '1010,27', '115,65', '1783,55', '4164,34', '1581,82', '1601,63', '2000', '3172,83', '4050', '2760', '2050', '134,78', '3190', '2858,6', '1430', '5580', '1070,05', '1033', '690', '762,93', '5550', '2450', '4158,85', '1095,61', '890', '590', '5872,64', '4055,03', '789,23', '2202,69', '1478,49', '472,11', '5560', '1270', '700', '1966,99', '3971,6', '446,31', '1015', '1690', '5760', '850', '4550', '1210', '1105,07', '2150', '1429,99', '1370,25', '818,74', '946,49', '2005,15', '6041,86', '3626,91', '604,76', '324,16', '658,42', '1080', '5397,15', '3506,07', '7010', '4508,2', '1120,93', '481,51', '5394,29', '920', '918,55', '840', '791,98', '3520', '6660', '360', '2130', '1433,4', '4650', '3285,74', '1175,57', '5430', '4240', '4310', '1199,27', '1410', '4350', '3240', '3106,29', '2390', '1706,3', '673,89', '1930,47', '1144,03', '8020', '3460', '3810', '6070', '2184,66', '552,57', '2220', '1116,49', '856,3', '1070', '3283,1', '1085,5', '2310', '11310', '4450', '1138,2', '3140', '1540', '560,36', '2201,11', '790,77', '1810', '3080', '1230,63', '1620', '530', '620', '879', '1841,9', '1939,61', '1590', '1597,92', '1640', '2059,46', '2070', '3500', '2160', '2009', '2380', '1093,22', '3840', '4760', '1793,66', '407,8', '1730', '2830', '3300', '1804,9', '3490', '816,83', '1980', '797,16', '823,54', '2260', '500', '2270', '177,34', '3200', '589,45', '216,59', '1145,85', '2180', '2497', '948,58', '1770', '4446,3', '1010', '2770', '3420', '4640', '4370', '1439,94', '678,2', '6650', '990', '3780,1', '1065,62', '1095,89', '2916,18', '470', '979,63', '1931,44', '6080', '94,89', '1670', '6133,41', '1700', '3880', '1720', '5428,62', '1833,77', '862,5', '4300', '5320', '2230', '1190', '1920', '1560', '1790', '648,68', '2406,8', '1870', '2240', '1740', '1402,43', '830,7', '5780', '1260,55', '2090,53', '1600', '901,64', '3230', '1150', '473,16', '3360', '1883,04', '2340', '846,55', '1006,59', '3900', '7100', '0', '2281,1', '3540', '3020', '1110', '1780', '3600', '1098,8', '5340', '2220,72', '4770', '2620', '3120', '796,59', '846,74', '1692,49', '630,49', '1302,38', '3637,43', '1300', '778,32', '1258,49', '3630', '2084,32', '798,53', '554,36', '4190', '2530', '791,56', '480', '2600', '1680', '660', '3959', '609,52', '943,56', '600,93', '3070', '768,97', '2990', '310', '2470', '2910', '2398,8', '1102,7', '2119,77', '2400', '1949,8', '1940', '581,67', '604,43', '3018,62', '1830', '2568,45', '3201,95', '6310', '2796,66', '2880', '960', '5980,18', '1508,79', '4160', '2200', '2503,19', '366,58', '1250', '2100', '981,71', '2290', '970', '1097,86', '499,69', '2702,4', '3620', '836,6', '4660', '1930', '2979,35', '289,02', '2088,86', '2700', '755,76', '2640', '2058,71', '1985,3', '1792,71', '980', '1800', '1420', '4490', '1119,27', '859,92', '776,3', '3040', '1272,72', '533', '1840', '965,1', '1154,34', '2710', '2550', '1671,7', '3920', '1957,21', '2375,59', '1113,74', '489,13', '2950', '1519,64', '665,57', '2540', '1387,15', '1509,44', '327,98', '2670', '1309,89', '720', '2429,09', '1900', '1090,8', '2083', '2110', '2782,93', '4940', '3910', '3702,8', '2629,87', '4600', '1686,92', '1670,18', '367', '1069,6', '3089,5', '918,83', '836,79', '1247,92', '4220', '570', '1750', '1266,34', '1387,47', '2261,19', '650', '1127,76', '2040', '2051,5', '1259,1', '1418,86', '730', '2467,33', '904,08', '1134,58', '2300', '3560', '1477,37', '1296,74', '497,62', '4210', '2425', '910,63', '640', '1490', '1012,9', '5698,46', '887,79', '1510', '2343,41', '269,51', '578,3', '4020', '5030', '1970', '2170', '2142,2', '2013,19', '549,88', '605,32', '1070,52', '1599,9', '517,51', '1300,16', '2707,25', '950', '1751,37', '1731,06', '4710', '1056,95', '1280', '1580', '1220', '510', '3060', '574,21', '894,29', '540', '830', '1030', '997,66', '3750', '490', '2265', '975', '2330', '1387,3', '1626,75', '682', '5076,65', '1749,69', '1053', '1650', '966,07', '300', '2569,05', '736,3', '726,57', '4000', '2920', '3030', '1077,03', '2840', '3110', '2570', '740', '1319,56', '1733,88', '1615', '1090,01', '4830', '2510,2', '2370', '2064,03', '900', '3860,18', '1140', '2010', '1960', '3150', '1845,5', '2360', '2630', '875,2', '3890', '5410', '849,19', '3660', '5140', '2750', '1563,45', '767,7', '3635,77', '2497,09', '2440', '1320', '613,21', '49,35', '702,23', '1498,26', '1408,32', '1986,88', '1851,55', '3237,94', '663,05', '1221,92', '572,22', '1663,27', '2157,32', '535,53', '1909,16', '1200', '1880,59', '0,04', '2154,02', '2800', '860', '2930', '1610', '1180', '2660', '896,1', '0,71', '3480', '3770', '4030', '2779,4', '4565', '370,77', '1296,4', '440', '4390', '809,42', '1002,22', '2163,94', '1050', '830,53', '603,55', '3870,25', '429,11', '370', '1100', '3780', '1837,31', '2044,32', '51,8', '1288,63', '4070', '3440', '957,63', '588,13', '670', '1798,49', '1942,72', '651,16', '1460', '2410', '1819,09', '9280', '2280', '3670', '1144,98', '2680', '567,16', '4920', '800', '896,31', '910', '3220', '430', '2940', '1283,93', '14870', '3010', '28012,2', '3850', '780', '3762,5', '1350', '14990'}\n",
      "\n",
      "\n",
      "==========\n",
      "sample\n",
      "{'train'}\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for i in df_train.columns: # перебираем все столбцы\n",
    "    if str(df_train[i].dtype) == 'object': # если тип столбца - object\n",
    "        print('='*10)\n",
    "        print(i) # выводим название столбца\n",
    "        print(set(df_train[i])) # выводим все его значения (но делаем set - чтоб значения не повторялись)\n",
    "        print('\\n') # выводим пустую строку"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Mожно заметить что некоторые переменные, которые обозначены как строки (например PERSONAL_INCOME) на самом деле числа, но по какой-то причине были распознаны как строки\n",
    "\n",
    "Причина же что использовалась запятая для разделения не целой части числа.."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Перекодировать их можно например так:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0         7000\n",
       "1         4100\n",
       "2        40000\n",
       "3        20000\n",
       "4        17000\n",
       "5         7000\n",
       "6        10000\n",
       "7         7500\n",
       "8         9000\n",
       "9         4000\n",
       "10       15000\n",
       "11       12000\n",
       "12        8000\n",
       "13       20000\n",
       "14       18500\n",
       "15       14000\n",
       "16        6000\n",
       "17        9000\n",
       "18        5000\n",
       "19       15000\n",
       "20        6100\n",
       "21       22180\n",
       "22        9000\n",
       "23        8700\n",
       "24        7000\n",
       "25        6000\n",
       "26       15000\n",
       "27       12000\n",
       "28       11000\n",
       "29        8000\n",
       "         ...  \n",
       "30103    12000\n",
       "30104    10000\n",
       "30105    10000\n",
       "30106     8000\n",
       "30107    11000\n",
       "30108     7000\n",
       "30109    10000\n",
       "30110     8000\n",
       "30111    25000\n",
       "30112    10000\n",
       "30113    12000\n",
       "30114     7000\n",
       "30115    12000\n",
       "30116    14500\n",
       "30117    12000\n",
       "30118     8000\n",
       "30119     7000\n",
       "30120    10000\n",
       "30121    10000\n",
       "30122    10000\n",
       "30123    15000\n",
       "30124     8100\n",
       "30125     9500\n",
       "30126    13000\n",
       "30127    10000\n",
       "30128    12000\n",
       "30129    10000\n",
       "30130     6000\n",
       "30131     4500\n",
       "30132    20000\n",
       "Name: PERSONAL_INCOME, Length: 30133, dtype: object"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#df['PERSONAL_INCOME'].map(lambda x: x.replace(',', '.')).astype('float')\n",
    "df['PERSONAL_INCOME']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Такой эффект наблюдается в столбцах `PERSONAL_INCOME`, `CREDIT`, `FST_PAYMENT`, `LOAN_AVG_DLQ_AMT`, `LOAN_MAX_DLQ_AMT`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Теперь ваше небольшое исследование"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "#### Задание 1. Есть ли пропуски в данных? Что с ними сделать?\n",
    "\n",
    "(единственного верного ответа нет - аргументируйте)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 30133 entries, 0 to 30132\n",
      "Data columns (total 53 columns):\n",
      "AGREEMENT_RK                  30133 non-null int64\n",
      "TARGET                        30133 non-null int64\n",
      "AGE                           30133 non-null int64\n",
      "SOCSTATUS_WORK_FL             30133 non-null int64\n",
      "SOCSTATUS_PENS_FL             30133 non-null int64\n",
      "GENDER                        30133 non-null int64\n",
      "CHILD_TOTAL                   30133 non-null int64\n",
      "DEPENDANTS                    30133 non-null int64\n",
      "EDUCATION                     30133 non-null object\n",
      "MARITAL_STATUS                30133 non-null object\n",
      "GEN_INDUSTRY                  27420 non-null object\n",
      "GEN_TITLE                     27420 non-null object\n",
      "ORG_TP_STATE                  27420 non-null object\n",
      "ORG_TP_FCAPITAL               27425 non-null object\n",
      "JOB_DIR                       27420 non-null object\n",
      "FAMILY_INCOME                 30133 non-null object\n",
      "PERSONAL_INCOME               30133 non-null object\n",
      "REG_ADDRESS_PROVINCE          30133 non-null object\n",
      "FACT_ADDRESS_PROVINCE         30133 non-null object\n",
      "POSTAL_ADDRESS_PROVINCE       30133 non-null object\n",
      "TP_PROVINCE                   29543 non-null object\n",
      "REGION_NM                     30131 non-null object\n",
      "REG_FACT_FL                   30133 non-null int64\n",
      "FACT_POST_FL                  30133 non-null int64\n",
      "REG_POST_FL                   30133 non-null int64\n",
      "REG_FACT_POST_FL              30133 non-null int64\n",
      "REG_FACT_POST_TP_FL           30133 non-null int64\n",
      "FL_PRESENCE_FL                30133 non-null int64\n",
      "OWN_AUTO                      30133 non-null int64\n",
      "AUTO_RUS_FL                   30133 non-null int64\n",
      "HS_PRESENCE_FL                30133 non-null int64\n",
      "COT_PRESENCE_FL               30133 non-null int64\n",
      "GAR_PRESENCE_FL               30133 non-null int64\n",
      "LAND_PRESENCE_FL              30133 non-null int64\n",
      "CREDIT                        30133 non-null object\n",
      "TERM                          30133 non-null int64\n",
      "FST_PAYMENT                   30133 non-null object\n",
      "DL_DOCUMENT_FL                30133 non-null int64\n",
      "GPF_DOCUMENT_FL               30133 non-null int64\n",
      "FACT_LIVING_TERM              30133 non-null int64\n",
      "WORK_TIME                     27416 non-null float64\n",
      "FACT_PHONE_FL                 30133 non-null int64\n",
      "REG_PHONE_FL                  30133 non-null int64\n",
      "GEN_PHONE_FL                  30133 non-null int64\n",
      "LOAN_NUM_TOTAL                30133 non-null int64\n",
      "LOAN_NUM_CLOSED               30133 non-null int64\n",
      "LOAN_NUM_PAYM                 30133 non-null int64\n",
      "LOAN_DLQ_NUM                  30133 non-null int64\n",
      "LOAN_MAX_DLQ                  30133 non-null int64\n",
      "LOAN_AVG_DLQ_AMT              30133 non-null object\n",
      "LOAN_MAX_DLQ_AMT              30133 non-null object\n",
      "PREVIOUS_CARD_NUM_UTILIZED    600 non-null float64\n",
      "sample                        30133 non-null object\n",
      "dtypes: float64(2), int64(32), object(19)\n",
      "memory usage: 12.2+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()\n",
    "# Видим что есть пропуски в данных, не все данные значимы, например в столбцах GEN_INDUSTRY GEN_TITLE ORG_TP_STATE  и так далее\n",
    "\n",
    "# Исходя из описания названий столбцов, можно сделать вывод о том, что большинство пропущенных данных имеют текстовый формат,\n",
    "# а часть имеет числовой\n",
    "# Следовательно для текстовых строк можно закодировать категориальные признаки (например с помощью OneHotEncoding),\n",
    "# а затем заполнить недостающие данные средними значениями для этой категории, а можно просто удалить эти строки\n",
    "# Для числовых значений сразу вычислить средние значения либо медиану"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "#### Задание 2. Есть ли категориальные признаки? Что с ними делать?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Категориальные признаки присутствуют, например: GEN_TITLE должность, REG_ADDRESS_PROVINCE область регистрации клиента и т.д.\n",
    "# Их можно закодировать с помощью функции OneHotEncoding\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Задание 3. Фунция предобработки"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Напишите функцию, которая бы\n",
    "\n",
    "* Удаляло идентификатор `AGREEMENT_RK`\n",
    "* Избавлялась от проблем с '.' и ',' в стобцах PERSONAL_INCOME, CREDIT, FST_PAYMENT, LOAN_AVG_DLQ_AMT, LOAN_MAX_DLQ_AMT\n",
    "* Что-то делала с пропусками\n",
    "* Кодировала категориальные признаки\n",
    "\n",
    "В результате, ваш датафрейм должен содержать только числа и не содержать пропусков!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Задание 3\n",
    "\n",
    "def preproc_data(df_input):\n",
    "    df_output = df_input.copy()\n",
    "    \n",
    "    #Удаляем идентификатор AGREEMENT_RK\n",
    "    df_output = df_output.drop(['AGREEMENT_RK'], axis=1)\n",
    "    \n",
    "    # Избавляемся от проблем с \",\" и \".\"\n",
    "    df_output['PERSONAL_INCOME']=df_output['PERSONAL_INCOME'].map(lambda x: x.replace(',', '.')).astype('float')\n",
    "    df_output['CREDIT']=df_output['CREDIT'].map(lambda x: x.replace(',', '.')).astype('float')\n",
    "    df_output['FST_PAYMENT']=df_output['FST_PAYMENT'].map(lambda x: x.replace(',', '.')).astype('float')\n",
    "    df_output['LOAN_AVG_DLQ_AMT']=df_output['LOAN_AVG_DLQ_AMT'].map(lambda x: x.replace(',', '.')).astype('float')\n",
    "    df_output['LOAN_MAX_DLQ_AMT']=df_output['LOAN_MAX_DLQ_AMT'].map(lambda x: x.replace(',', '.')).astype('float')\n",
    "    \n",
    "    # Заменяем пропуски в столбце WORK_TIME средними значениями по столбцу\n",
    "    df_output.loc[df_output['WORK_TIME'].isnull(), 'WORK_TIME'] = df_output['WORK_TIME'].mean()\n",
    "    \n",
    "    \n",
    "    \n",
    "    \n",
    "    #Так как у оставшихся столбцов с пропусками тип данных object - удаляем все оставшиеся строки с пропусками\n",
    "    df_output = df_output.drop(['PREVIOUS_CARD_NUM_UTILIZED','GEN_INDUSTRY', 'GEN_TITLE', 'ORG_TP_STATE', 'ORG_TP_FCAPITAL', 'JOB_DIR', 'TP_PROVINCE', 'REGION_NM'], axis=1)\n",
    "    \n",
    "    df_output.info()\n",
    "    # Ищем столбцы с категориальными признаками и кодируем их с помощью OneHotEncoding кроме столбца simple с маркерами test/train\n",
    "    df_output = pd.get_dummies(df_output, columns=['EDUCATION', 'MARITAL_STATUS', 'FAMILY_INCOME', 'REG_ADDRESS_PROVINCE', 'FACT_ADDRESS_PROVINCE', 'POSTAL_ADDRESS_PROVINCE'])\n",
    "   \n",
    "    return df_output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 30133 entries, 0 to 30132\n",
      "Data columns (total 44 columns):\n",
      "TARGET                     30133 non-null int64\n",
      "AGE                        30133 non-null int64\n",
      "SOCSTATUS_WORK_FL          30133 non-null int64\n",
      "SOCSTATUS_PENS_FL          30133 non-null int64\n",
      "GENDER                     30133 non-null int64\n",
      "CHILD_TOTAL                30133 non-null int64\n",
      "DEPENDANTS                 30133 non-null int64\n",
      "EDUCATION                  30133 non-null object\n",
      "MARITAL_STATUS             30133 non-null object\n",
      "FAMILY_INCOME              30133 non-null object\n",
      "PERSONAL_INCOME            30133 non-null float64\n",
      "REG_ADDRESS_PROVINCE       30133 non-null object\n",
      "FACT_ADDRESS_PROVINCE      30133 non-null object\n",
      "POSTAL_ADDRESS_PROVINCE    30133 non-null object\n",
      "REG_FACT_FL                30133 non-null int64\n",
      "FACT_POST_FL               30133 non-null int64\n",
      "REG_POST_FL                30133 non-null int64\n",
      "REG_FACT_POST_FL           30133 non-null int64\n",
      "REG_FACT_POST_TP_FL        30133 non-null int64\n",
      "FL_PRESENCE_FL             30133 non-null int64\n",
      "OWN_AUTO                   30133 non-null int64\n",
      "AUTO_RUS_FL                30133 non-null int64\n",
      "HS_PRESENCE_FL             30133 non-null int64\n",
      "COT_PRESENCE_FL            30133 non-null int64\n",
      "GAR_PRESENCE_FL            30133 non-null int64\n",
      "LAND_PRESENCE_FL           30133 non-null int64\n",
      "CREDIT                     30133 non-null float64\n",
      "TERM                       30133 non-null int64\n",
      "FST_PAYMENT                30133 non-null float64\n",
      "DL_DOCUMENT_FL             30133 non-null int64\n",
      "GPF_DOCUMENT_FL            30133 non-null int64\n",
      "FACT_LIVING_TERM           30133 non-null int64\n",
      "WORK_TIME                  30133 non-null float64\n",
      "FACT_PHONE_FL              30133 non-null int64\n",
      "REG_PHONE_FL               30133 non-null int64\n",
      "GEN_PHONE_FL               30133 non-null int64\n",
      "LOAN_NUM_TOTAL             30133 non-null int64\n",
      "LOAN_NUM_CLOSED            30133 non-null int64\n",
      "LOAN_NUM_PAYM              30133 non-null int64\n",
      "LOAN_DLQ_NUM               30133 non-null int64\n",
      "LOAN_MAX_DLQ               30133 non-null int64\n",
      "LOAN_AVG_DLQ_AMT           30133 non-null float64\n",
      "LOAN_MAX_DLQ_AMT           30133 non-null float64\n",
      "sample                     30133 non-null object\n",
      "dtypes: float64(6), int64(31), object(7)\n",
      "memory usage: 10.1+ MB\n"
     ]
    }
   ],
   "source": [
    "\n",
    "df_preproc = df.pipe(preproc_data)\n",
    "df_train_preproc = df_preproc.query('sample == \"train\"').drop(['sample'], axis=1)\n",
    "df_test_preproc = df_preproc.query('sample == \"test\"').drop(['sample'], axis=1)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Задание 4. Отделите целевую переменную и остальные признаки\n",
    "\n",
    "Должно получится:\n",
    "* 2 матрицы: X и X_test\n",
    "* 2 вектора: y и y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = df_test_preproc['TARGET']\n",
    "from sklearn.model_selection import train_test_split\n",
    "X, X_test, y, y_test = train_test_split(df_test_preproc, y, test_size=0.33, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Задание 5. Обучение и оценка качества разных моделей"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Ildar\\Anaconda3\\lib\\site-packages\\sklearn\\cross_validation.py:41: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. Also note that the interface of the new CV iterators are different from that of this module. This module will be removed in 0.20.\n",
      "  \"This module will be removed in 0.20.\", DeprecationWarning)\n"
     ]
    }
   ],
   "source": [
    "y = df_test_preproc['TARGET']\n",
    "\n",
    "from sklearn.cross_validation import train_test_split\n",
    "from sklearn.metrics import precision_score, recall_score, accuracy_score, classification_report\n",
    "\n",
    "X, X_test, y, y_test = train_test_split(df_test_preproc, y, test_size=0.33, random_state=42)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# test_size=0.3, random_state=42\n",
    "\n",
    "## Your Code Here\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "precision 1.0\n",
      "recall 1.0\n",
      "accuracy 1.0\n",
      "f1 1.0\n"
     ]
    }
   ],
   "source": [
    "# Попробовать следующие \"черные ящики\": интерфейс одинаковый \n",
    "# Постепенно мы узнаем, что они делают а сейчас учимся понимать какой работает качественнее\n",
    "#     fit, \n",
    "#     predict, \n",
    "#     predict_proba\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.metrics import f1_score\n",
    "\n",
    "model = DecisionTreeClassifier()                    # инициализируем модель\n",
    "model.fit(X, y)                                     # обучаем модель\n",
    "predict = model.predict(X_test)                     # делаем предсказание\n",
    "predict_proba1 = model.predict_proba(X_test)\n",
    "print('precision', precision_score(y_test, predict))\n",
    "print('recall', recall_score(y_test, predict))\n",
    "print('accuracy', accuracy_score(y_test, predict))\n",
    "print('f1', f1_score(y_test, predict))\n",
    "## Your Code Herepredict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Ildar\\Anaconda3\\lib\\site-packages\\sklearn\\ensemble\\weight_boosting.py:29: DeprecationWarning: numpy.core.umath_tests is an internal NumPy module and should not be imported. It will be removed in a future NumPy release.\n",
      "  from numpy.core.umath_tests import inner1d\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "precision 1.0\n",
      "recall 0.8885245901639345\n",
      "accuracy 0.9861816703921967\n",
      "f1 0.9409722222222222\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "model = RandomForestClassifier()                    # инициализируем модель\n",
    "model.fit(X, y)                                     # обучаем модель\n",
    "predict = model.predict(X_test)                     # делаем предсказание\n",
    "predict_proba2 = model.predict_proba(X_test)\n",
    "print('precision', precision_score(y_test, predict))\n",
    "print('recall', recall_score(y_test, predict))\n",
    "print('accuracy', accuracy_score(y_test, predict))\n",
    "print('f1', f1_score(y_test, predict))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "precision 0.9958677685950413\n",
      "recall 0.7901639344262295\n",
      "accuracy 0.9735826051615525\n",
      "f1 0.8811700182815355\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "model = LogisticRegression()                    # инициализируем модель\n",
    "model.fit(X, y)                                     # обучаем модель\n",
    "predict = model.predict(X_test)                     # делаем предсказание\n",
    "predict_proba3 = model.predict_proba(X_test)\n",
    "print('precision', precision_score(y_test, predict))\n",
    "print('recall', recall_score(y_test, predict))\n",
    "print('accuracy', accuracy_score(y_test, predict))\n",
    "print('f1', f1_score(y_test, predict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Посчитать метрики стандартные\n",
    "# accuracy, precision, recall\n",
    "## Посчитано выше :)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Визуалищировать эти метрики всех моделей на одном графике (чтоб визуально сравнить)\n",
    "# Возможно вас удивит качество! Но задача подобрана специально ;) Такое качество тоже бывает\n",
    "## Your Code Here\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "roc_auc для DecisionTreeClassifier 1.0\n",
      "roc_auc для RandomForestClassifier 0.9999132984245411\n",
      "roc_auc для LogisticRegression 0.9979440698784278\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x1800 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Потроить roc-кривые всех можелей на одном графике\n",
    "# Вывести roc_auc каждой моделе\n",
    "# Подпишите оси и линии\n",
    "%matplotlib inline\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn.metrics import roc_auc_score, roc_curve\n",
    "fpr1, tpr1, thresholds1 = roc_curve(y_test, predict_proba1[:,1])\n",
    "fpr2, tpr2, thresholds2 = roc_curve(y_test, predict_proba2[:,1])\n",
    "fpr3, tpr3, thresholds3 = roc_curve(y_test, predict_proba3[:,1])\n",
    "plt.figure(figsize=(25, 25))\n",
    "plt.plot(fpr1, tpr1,'r',label='DecisionTreeClassifier')\n",
    "plt.plot(fpr2, tpr2,'b',label='RandomForestClassifier')\n",
    "plt.plot(fpr3, tpr3,'g',label='LogisticRegression')\n",
    "plt.plot([0, 1], [0, 1])\n",
    "plt.ylabel('tpr')\n",
    "plt.xlabel('fpr')\n",
    "plt.grid(True)\n",
    "plt.title('ROC curve')\n",
    "plt.xlim((-0.01, 1.01))\n",
    "plt.ylim((-0.01, 1.01))\n",
    "plt.legend()\n",
    "\n",
    "print ('roc_auc для DecisionTreeClassifier', roc_auc_score(y_test, predict_proba1[:,1]))\n",
    "print ('roc_auc для RandomForestClassifier', roc_auc_score(y_test, predict_proba2[:,1]))\n",
    "print ('roc_auc для LogisticRegression', roc_auc_score(y_test, predict_proba3[:,1]))\n",
    "## Your Code Here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Кросс валидация для DecisionTreeClassifier [1. 1. 1. 1. 1. 1. 1. 1. 1. 1.]\n",
      "Кросс валидация для RandomForestClassifier [0.99999009 1.         0.99998513 1.         1.         1.\n",
      " 0.9999554  1.         0.99960848 0.99963285]\n",
      "Кросс валидация для LogisticRegression [0.79349787 0.70708693 0.70467836 0.88101893 1.         0.87541877\n",
      " 0.99670929 0.9528992  0.70047577 0.98018377]\n",
      "Среднее значение rog_auc DecisionTreeClassifier= 1.0\n",
      "Среднее значение rog_auc RandomForestClassifier= 0.9999171951083323\n",
      "Среднее значение rog_auc LogisticRegression= 0.8591968886413186\n"
     ]
    }
   ],
   "source": [
    "from sklearn.cross_validation import cross_val_score\n",
    "# Сделать k-fold (10 фолдов) кросс-валидацию каждой модели\n",
    "# И посчитать средний roc_auc\n",
    "import numpy\n",
    "rfc = DecisionTreeClassifier()\n",
    "a=cross_val_score(\n",
    "    rfc,\n",
    "    X,\n",
    "    y,\n",
    "    scoring='roc_auc',\n",
    "    cv=10\n",
    ")\n",
    "rfc = RandomForestClassifier()\n",
    "b=cross_val_score(\n",
    "    rfc,\n",
    "    X,\n",
    "    y,\n",
    "    scoring='roc_auc',\n",
    "    cv=10\n",
    ")\n",
    "rfc = LogisticRegression()\n",
    "c=cross_val_score(\n",
    "    rfc,\n",
    "    X,\n",
    "    y,\n",
    "    scoring='roc_auc',\n",
    "    cv=10\n",
    ")\n",
    "print ('Кросс валидация для DecisionTreeClassifier',a)\n",
    "print ('Кросс валидация для RandomForestClassifier',b)\n",
    "print ('Кросс валидация для LogisticRegression',c)\n",
    "print ('Среднее значение rog_auc DecisionTreeClassifier=',numpy.mean(a))\n",
    "print ('Среднее значение rog_auc RandomForestClassifier=',numpy.mean(b))\n",
    "print ('Среднее значение rog_auc LogisticRegression=',numpy.mean(c))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 0, 0, 0], dtype=int64)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Взять лучшую модель и сделать предсказания (с вероятностями (!!!)) для test выборки\n",
    "model = DecisionTreeClassifier()\n",
    "model.fit(X, y)\n",
    "predict = model.predict(X_test)\n",
    "\n",
    "predict\n",
    "## Your Code Here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "roc_auc для DecisionTreeClassifier 1.0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x20de1f3ae48>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x1800 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Померить roc_auc на тесте\n",
    "print ('roc_auc для DecisionTreeClassifier', roc_auc_score(y_test, predict))\n",
    "# Вывести текстом и на графике =)\n",
    "fpr, tpr, thresholds = roc_curve(y_test, predict)\n",
    "plt.figure(figsize=(25, 25))\n",
    "plt.plot(fpr, tpr,'r',label='DecisionTreeClassifier')\n",
    "plt.plot([0, 1], [0, 1])\n",
    "plt.ylabel('tpr')\n",
    "plt.xlabel('fpr')\n",
    "plt.grid(True)\n",
    "plt.title('ROC curve')\n",
    "plt.xlim((-0.01, 1.01))\n",
    "plt.ylim((-0.01, 1.01))\n",
    "plt.legend()\n",
    "## Your Code Here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Больше обязательных заданий нет, но ниже вы можете провести исследование, поиграться с данными\n",
    "### Это возможность написать код, который я посмотрю и в случае чего откомметирую. ;)\n",
    "### Это не оценивается и остается на ваше усмотрение. Просто дополнительная возможность"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "### И далее ;)"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {
    "height": "31px",
    "width": "253px"
   },
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": "block",
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}