PaymentPrediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Importing modules and dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "ename": "ModuleNotFoundError",
     "evalue": "No module named 'xgboost'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mModuleNotFoundError\u001b[0m                       Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-1-4f0f33728e63>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mseaborn\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0msns\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      6\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 7\u001b[1;33m \u001b[1;32mimport\u001b[0m \u001b[0mxgboost\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0mxgb\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      8\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0msklearn\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmetrics\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0msm\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      9\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[0msklearn\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mensemble\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mRandomForestRegressor\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mModuleNotFoundError\u001b[0m: No module named 'xgboost'"
     ]
    }
   ],
   "source": [
    "# importing all the necessary libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import xgboost as xgb\n",
    "import sklearn.metrics as sm\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.model_selection import RandomizedSearchCV, GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>doc_id</th>\n",
       "      <th>posting_date</th>\n",
       "      <th>document_create_date</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>document type</th>\n",
       "      <th>posting_id</th>\n",
       "      <th>area_business</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>invoice_id</th>\n",
       "      <th>isOpen</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200315290</td>\n",
       "      <td>KWI in</td>\n",
       "      <td>2019-03-07 00:00:00</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>1.928827e+09</td>\n",
       "      <td>2019-02-20</td>\n",
       "      <td>20190220</td>\n",
       "      <td>20190220</td>\n",
       "      <td>20190307.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21227.05</td>\n",
       "      <td>20190220.0</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1.928827e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR foundation</td>\n",
       "      <td>2020-01-16 00:00:00</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930350e+09</td>\n",
       "      <td>2020-01-04</td>\n",
       "      <td>20200103</td>\n",
       "      <td>20200104</td>\n",
       "      <td>20200119.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5669.72</td>\n",
       "      <td>20200104.0</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>1.930350e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200820380</td>\n",
       "      <td>SMITH'S associates</td>\n",
       "      <td>2019-07-08 00:00:00</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>1.929517e+09</td>\n",
       "      <td>2019-06-21</td>\n",
       "      <td>20190619</td>\n",
       "      <td>20190621</td>\n",
       "      <td>20190706.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>67968.48</td>\n",
       "      <td>20190621.0</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1.929517e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR corp</td>\n",
       "      <td>2019-03-22 00:00:00</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>1.928943e+09</td>\n",
       "      <td>2019-03-12</td>\n",
       "      <td>20190311</td>\n",
       "      <td>20190312</td>\n",
       "      <td>20190327.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>61395.73</td>\n",
       "      <td>20190312.0</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>1.928943e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>200230690</td>\n",
       "      <td>DECA in</td>\n",
       "      <td>2019-06-03 00:00:00</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>1.929350e+09</td>\n",
       "      <td>2019-05-21</td>\n",
       "      <td>20190521</td>\n",
       "      <td>20190521</td>\n",
       "      <td>20190608.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1067.72</td>\n",
       "      <td>20190516.0</td>\n",
       "      <td>NAM4</td>\n",
       "      <td>1.929350e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49995</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200726979</td>\n",
       "      <td>BJ'S  us</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930709e+09</td>\n",
       "      <td>2020-03-27</td>\n",
       "      <td>20200326</td>\n",
       "      <td>20200327</td>\n",
       "      <td>20200411.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>261.84</td>\n",
       "      <td>20200327.0</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1.930709e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49996</th>\n",
       "      <td>U001</td>\n",
       "      <td>CCCA02</td>\n",
       "      <td>KRAFT  in</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930694e+09</td>\n",
       "      <td>2020-03-25</td>\n",
       "      <td>20200324</td>\n",
       "      <td>20200325</td>\n",
       "      <td>20200429.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14151.51</td>\n",
       "      <td>20200325.0</td>\n",
       "      <td>NAG2</td>\n",
       "      <td>1.930694e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49997</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200762301</td>\n",
       "      <td>C&amp;S WH corporation</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930606e+09</td>\n",
       "      <td>2020-03-06</td>\n",
       "      <td>20200305</td>\n",
       "      <td>20200306</td>\n",
       "      <td>20200321.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14898.82</td>\n",
       "      <td>20200306.0</td>\n",
       "      <td>NAC6</td>\n",
       "      <td>1.930606e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49998</th>\n",
       "      <td>U001</td>\n",
       "      <td>0100001196</td>\n",
       "      <td>DOLLAR</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930761e+09</td>\n",
       "      <td>2020-04-09</td>\n",
       "      <td>20200408</td>\n",
       "      <td>20200409</td>\n",
       "      <td>20200424.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5127.95</td>\n",
       "      <td>20200409.0</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1.930761e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49999</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200794332</td>\n",
       "      <td>COST systems</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930586e+09</td>\n",
       "      <td>2020-03-03</td>\n",
       "      <td>20200301</td>\n",
       "      <td>20200303</td>\n",
       "      <td>20200318.0</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3382.65</td>\n",
       "      <td>20200303.0</td>\n",
       "      <td>NAAX</td>\n",
       "      <td>1.930586e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>50000 rows × 19 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      business_code cust_number       name_customer           clear_date  \\\n",
       "0              U001  0200315290              KWI in  2019-03-07 00:00:00   \n",
       "1              U001  0200769623  WAL-MAR foundation  2020-01-16 00:00:00   \n",
       "2              U001  0200820380  SMITH'S associates  2019-07-08 00:00:00   \n",
       "3              U001  0200769623        WAL-MAR corp  2019-03-22 00:00:00   \n",
       "4              U001   200230690             DECA in  2019-06-03 00:00:00   \n",
       "...             ...         ...                 ...                  ...   \n",
       "49995          U001  0200726979            BJ'S  us                  NaN   \n",
       "49996          U001      CCCA02           KRAFT  in                  NaN   \n",
       "49997          U001  0200762301  C&S WH corporation                  NaN   \n",
       "49998          U001  0100001196             DOLLAR                   NaN   \n",
       "49999          U001  0200794332        COST systems                  NaN   \n",
       "\n",
       "       buisness_year        doc_id posting_date  document_create_date  \\\n",
       "0             2019.0  1.928827e+09   2019-02-20              20190220   \n",
       "1             2020.0  1.930350e+09   2020-01-04              20200103   \n",
       "2             2019.0  1.929517e+09   2019-06-21              20190619   \n",
       "3             2019.0  1.928943e+09   2019-03-12              20190311   \n",
       "4             2019.0  1.929350e+09   2019-05-21              20190521   \n",
       "...              ...           ...          ...                   ...   \n",
       "49995         2020.0  1.930709e+09   2020-03-27              20200326   \n",
       "49996         2020.0  1.930694e+09   2020-03-25              20200324   \n",
       "49997         2020.0  1.930606e+09   2020-03-06              20200305   \n",
       "49998         2020.0  1.930761e+09   2020-04-09              20200408   \n",
       "49999         2020.0  1.930586e+09   2020-03-03              20200301   \n",
       "\n",
       "       document_create_date.1  due_in_date invoice_currency document type  \\\n",
       "0                    20190220   20190307.0              USD            RV   \n",
       "1                    20200104   20200119.0              USD            RV   \n",
       "2                    20190621   20190706.0              USD            RV   \n",
       "3                    20190312   20190327.0              USD            RV   \n",
       "4                    20190521   20190608.0              USD            RV   \n",
       "...                       ...          ...              ...           ...   \n",
       "49995                20200327   20200411.0              USD            RV   \n",
       "49996                20200325   20200429.0              USD            RV   \n",
       "49997                20200306   20200321.0              USD            RV   \n",
       "49998                20200409   20200424.0              USD            RV   \n",
       "49999                20200303   20200318.0              USD            RV   \n",
       "\n",
       "       posting_id  area_business  total_open_amount  baseline_create_date  \\\n",
       "0             1.0            NaN           21227.05            20190220.0   \n",
       "1             1.0            NaN            5669.72            20200104.0   \n",
       "2             1.0            NaN           67968.48            20190621.0   \n",
       "3             1.0            NaN           61395.73            20190312.0   \n",
       "4             1.0            NaN            1067.72            20190516.0   \n",
       "...           ...            ...                ...                   ...   \n",
       "49995         1.0            NaN             261.84            20200327.0   \n",
       "49996         1.0            NaN           14151.51            20200325.0   \n",
       "49997         1.0            NaN           14898.82            20200306.0   \n",
       "49998         1.0            NaN            5127.95            20200409.0   \n",
       "49999         1.0            NaN            3382.65            20200303.0   \n",
       "\n",
       "      cust_payment_terms    invoice_id  isOpen  \n",
       "0                   NAA8  1.928827e+09       0  \n",
       "1                   NAH4  1.930350e+09       0  \n",
       "2                   NAA8  1.929517e+09       0  \n",
       "3                   NAH4  1.928943e+09       0  \n",
       "4                   NAM4  1.929350e+09       0  \n",
       "...                  ...           ...     ...  \n",
       "49995               NAA8  1.930709e+09       1  \n",
       "49996               NAG2  1.930694e+09       1  \n",
       "49997               NAC6  1.930606e+09       1  \n",
       "49998               NAA8  1.930761e+09       1  \n",
       "49999               NAAX  1.930586e+09       1  \n",
       "\n",
       "[50000 rows x 19 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('Dataset.csv')\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PRE-PROCESSING"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50000 entries, 0 to 49999\n",
      "Data columns (total 19 columns):\n",
      " #   Column                  Non-Null Count  Dtype  \n",
      "---  ------                  --------------  -----  \n",
      " 0   business_code           50000 non-null  object \n",
      " 1   cust_number             50000 non-null  object \n",
      " 2   name_customer           50000 non-null  object \n",
      " 3   clear_date              45181 non-null  object \n",
      " 4   buisness_year           50000 non-null  float64\n",
      " 5   doc_id                  50000 non-null  float64\n",
      " 6   posting_date            50000 non-null  object \n",
      " 7   document_create_date    50000 non-null  int64  \n",
      " 8   document_create_date.1  50000 non-null  int64  \n",
      " 9   due_in_date             50000 non-null  float64\n",
      " 10  invoice_currency        50000 non-null  object \n",
      " 11  document type           50000 non-null  object \n",
      " 12  posting_id              50000 non-null  float64\n",
      " 13  area_business           0 non-null      float64\n",
      " 14  total_open_amount       50000 non-null  float64\n",
      " 15  baseline_create_date    50000 non-null  float64\n",
      " 16  cust_payment_terms      50000 non-null  object \n",
      " 17  invoice_id              49997 non-null  float64\n",
      " 18  isOpen                  50000 non-null  int64  \n",
      "dtypes: float64(8), int64(3), object(8)\n",
      "memory usage: 7.2+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that there are NULL values in clear_date --> which we will use to split in test and train later.     \n",
    "area_business is totally NULL. \n",
    "There are 3 NULL values in invoice_id, we'll look into it later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "business_code                 0\n",
       "cust_number                   0\n",
       "name_customer                 0\n",
       "clear_date                 4819\n",
       "buisness_year                 0\n",
       "doc_id                        0\n",
       "posting_date                  0\n",
       "document_create_date          0\n",
       "document_create_date.1        0\n",
       "due_in_date                   0\n",
       "invoice_currency              0\n",
       "document type                 0\n",
       "posting_id                    0\n",
       "area_business             50000\n",
       "total_open_amount             0\n",
       "baseline_create_date          0\n",
       "cust_payment_terms            0\n",
       "invoice_id                    3\n",
       "isOpen                        0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Date-Time Conversion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0   2019-02-20\n",
       "1   2020-01-04\n",
       "2   2019-06-21\n",
       "3   2019-03-12\n",
       "4   2019-05-21\n",
       "Name: document_create_date.1, dtype: datetime64[ns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['document_create_date.1']=pd.to_datetime(df['document_create_date.1'],format='%Y%m%d')\n",
    "df['document_create_date.1'].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0   2019-03-07\n",
       "1   2020-01-16\n",
       "2   2019-07-08\n",
       "3   2019-03-22\n",
       "4   2019-06-03\n",
       "Name: clear_date, dtype: datetime64[ns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['clear_date']=pd.to_datetime(df['clear_date'],format='%Y-%m-%d %H:%M:%S')\n",
    "df['clear_date'].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0   2019-02-20\n",
       "1   2020-01-04\n",
       "2   2019-06-21\n",
       "3   2019-03-12\n",
       "4   2019-05-21\n",
       "Name: posting_date, dtype: datetime64[ns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['posting_date']=pd.to_datetime(df['posting_date'],format='%Y-%m-%d')\n",
    "df['posting_date'].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0   2019-03-07\n",
       "1   2020-01-19\n",
       "2   2019-07-06\n",
       "3   2019-03-27\n",
       "4   2019-06-08\n",
       "Name: due_in_date, dtype: datetime64[ns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['due_in_date']=pd.to_datetime(df['due_in_date'],format='%Y%m%d')\n",
    "df['due_in_date'].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0   2019-02-20\n",
       "1   2020-01-04\n",
       "2   2019-06-21\n",
       "3   2019-03-12\n",
       "4   2019-05-16\n",
       "Name: baseline_create_date, dtype: datetime64[ns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['baseline_create_date']=pd.to_datetime(df['baseline_create_date'],format='%Y%m%d')\n",
    "df['baseline_create_date'].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50000 entries, 0 to 49999\n",
      "Data columns (total 19 columns):\n",
      " #   Column                  Non-Null Count  Dtype         \n",
      "---  ------                  --------------  -----         \n",
      " 0   business_code           50000 non-null  object        \n",
      " 1   cust_number             50000 non-null  object        \n",
      " 2   name_customer           50000 non-null  object        \n",
      " 3   clear_date              45181 non-null  datetime64[ns]\n",
      " 4   buisness_year           50000 non-null  float64       \n",
      " 5   doc_id                  50000 non-null  float64       \n",
      " 6   posting_date            50000 non-null  datetime64[ns]\n",
      " 7   document_create_date    50000 non-null  int64         \n",
      " 8   document_create_date.1  50000 non-null  datetime64[ns]\n",
      " 9   due_in_date             50000 non-null  datetime64[ns]\n",
      " 10  invoice_currency        50000 non-null  object        \n",
      " 11  document type           50000 non-null  object        \n",
      " 12  posting_id              50000 non-null  float64       \n",
      " 13  area_business           0 non-null      float64       \n",
      " 14  total_open_amount       50000 non-null  float64       \n",
      " 15  baseline_create_date    50000 non-null  datetime64[ns]\n",
      " 16  cust_payment_terms      50000 non-null  object        \n",
      " 17  invoice_id              49997 non-null  float64       \n",
      " 18  isOpen                  50000 non-null  int64         \n",
      "dtypes: datetime64[ns](5), float64(6), int64(2), object(6)\n",
      "memory usage: 7.2+ MB\n"
     ]
    }
   ],
   "source": [
    "# looking at the converted data-types\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dropping columns with ALL NULL values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(50000, 18)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dropna(axis=1,how='all',inplace=True)\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We dropped area_business"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dropping null values in invoice id\n",
    "invoice id has only 49997 unique values of 50000. So let's analyse it a bit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RV    49997\n",
       "X2        3\n",
       "Name: document type, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"document type\"].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that where the invoice id is missing the document type is X2 and posting date is different from document create date. Since their are only 3 values and invoice_id needs to be unique we can drop these rows"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>doc_id</th>\n",
       "      <th>posting_date</th>\n",
       "      <th>document_create_date</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>document type</th>\n",
       "      <th>posting_id</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>invoice_id</th>\n",
       "      <th>isOpen</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200315290</td>\n",
       "      <td>KWI in</td>\n",
       "      <td>2019-03-07</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>1.928827e+09</td>\n",
       "      <td>2019-02-20</td>\n",
       "      <td>20190220</td>\n",
       "      <td>2019-02-20</td>\n",
       "      <td>2019-03-07</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>21227.05</td>\n",
       "      <td>2019-02-20</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1.928827e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR foundation</td>\n",
       "      <td>2020-01-16</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930350e+09</td>\n",
       "      <td>2020-01-04</td>\n",
       "      <td>20200103</td>\n",
       "      <td>2020-01-04</td>\n",
       "      <td>2020-01-19</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5669.72</td>\n",
       "      <td>2020-01-04</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>1.930350e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200820380</td>\n",
       "      <td>SMITH'S associates</td>\n",
       "      <td>2019-07-08</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>1.929517e+09</td>\n",
       "      <td>2019-06-21</td>\n",
       "      <td>20190619</td>\n",
       "      <td>2019-06-21</td>\n",
       "      <td>2019-07-06</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>67968.48</td>\n",
       "      <td>2019-06-21</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1.929517e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR corp</td>\n",
       "      <td>2019-03-22</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>1.928943e+09</td>\n",
       "      <td>2019-03-12</td>\n",
       "      <td>20190311</td>\n",
       "      <td>2019-03-12</td>\n",
       "      <td>2019-03-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>61395.73</td>\n",
       "      <td>2019-03-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>1.928943e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>200230690</td>\n",
       "      <td>DECA in</td>\n",
       "      <td>2019-06-03</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>1.929350e+09</td>\n",
       "      <td>2019-05-21</td>\n",
       "      <td>20190521</td>\n",
       "      <td>2019-05-21</td>\n",
       "      <td>2019-06-08</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1067.72</td>\n",
       "      <td>2019-05-16</td>\n",
       "      <td>NAM4</td>\n",
       "      <td>1.929350e+09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49992</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200726979</td>\n",
       "      <td>BJ'S  us</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930709e+09</td>\n",
       "      <td>2020-03-27</td>\n",
       "      <td>20200326</td>\n",
       "      <td>2020-03-27</td>\n",
       "      <td>2020-04-11</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>261.84</td>\n",
       "      <td>2020-03-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1.930709e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49993</th>\n",
       "      <td>U001</td>\n",
       "      <td>CCCA02</td>\n",
       "      <td>KRAFT  in</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930694e+09</td>\n",
       "      <td>2020-03-25</td>\n",
       "      <td>20200324</td>\n",
       "      <td>2020-03-25</td>\n",
       "      <td>2020-04-29</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>14151.51</td>\n",
       "      <td>2020-03-25</td>\n",
       "      <td>NAG2</td>\n",
       "      <td>1.930694e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49994</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200762301</td>\n",
       "      <td>C&amp;S WH corporation</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930606e+09</td>\n",
       "      <td>2020-03-06</td>\n",
       "      <td>20200305</td>\n",
       "      <td>2020-03-06</td>\n",
       "      <td>2020-03-21</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>14898.82</td>\n",
       "      <td>2020-03-06</td>\n",
       "      <td>NAC6</td>\n",
       "      <td>1.930606e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49995</th>\n",
       "      <td>U001</td>\n",
       "      <td>0100001196</td>\n",
       "      <td>DOLLAR</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930761e+09</td>\n",
       "      <td>2020-04-09</td>\n",
       "      <td>20200408</td>\n",
       "      <td>2020-04-09</td>\n",
       "      <td>2020-04-24</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5127.95</td>\n",
       "      <td>2020-04-09</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1.930761e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49996</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200794332</td>\n",
       "      <td>COST systems</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>1.930586e+09</td>\n",
       "      <td>2020-03-03</td>\n",
       "      <td>20200301</td>\n",
       "      <td>2020-03-03</td>\n",
       "      <td>2020-03-18</td>\n",
       "      <td>USD</td>\n",
       "      <td>RV</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3382.65</td>\n",
       "      <td>2020-03-03</td>\n",
       "      <td>NAAX</td>\n",
       "      <td>1.930586e+09</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>49997 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      business_code cust_number       name_customer clear_date  buisness_year  \\\n",
       "0              U001  0200315290              KWI in 2019-03-07         2019.0   \n",
       "1              U001  0200769623  WAL-MAR foundation 2020-01-16         2020.0   \n",
       "2              U001  0200820380  SMITH'S associates 2019-07-08         2019.0   \n",
       "3              U001  0200769623        WAL-MAR corp 2019-03-22         2019.0   \n",
       "4              U001   200230690             DECA in 2019-06-03         2019.0   \n",
       "...             ...         ...                 ...        ...            ...   \n",
       "49992          U001  0200726979            BJ'S  us        NaT         2020.0   \n",
       "49993          U001      CCCA02           KRAFT  in        NaT         2020.0   \n",
       "49994          U001  0200762301  C&S WH corporation        NaT         2020.0   \n",
       "49995          U001  0100001196             DOLLAR         NaT         2020.0   \n",
       "49996          U001  0200794332        COST systems        NaT         2020.0   \n",
       "\n",
       "             doc_id posting_date  document_create_date document_create_date.1  \\\n",
       "0      1.928827e+09   2019-02-20              20190220             2019-02-20   \n",
       "1      1.930350e+09   2020-01-04              20200103             2020-01-04   \n",
       "2      1.929517e+09   2019-06-21              20190619             2019-06-21   \n",
       "3      1.928943e+09   2019-03-12              20190311             2019-03-12   \n",
       "4      1.929350e+09   2019-05-21              20190521             2019-05-21   \n",
       "...             ...          ...                   ...                    ...   \n",
       "49992  1.930709e+09   2020-03-27              20200326             2020-03-27   \n",
       "49993  1.930694e+09   2020-03-25              20200324             2020-03-25   \n",
       "49994  1.930606e+09   2020-03-06              20200305             2020-03-06   \n",
       "49995  1.930761e+09   2020-04-09              20200408             2020-04-09   \n",
       "49996  1.930586e+09   2020-03-03              20200301             2020-03-03   \n",
       "\n",
       "      due_in_date invoice_currency document type  posting_id  \\\n",
       "0      2019-03-07              USD            RV         1.0   \n",
       "1      2020-01-19              USD            RV         1.0   \n",
       "2      2019-07-06              USD            RV         1.0   \n",
       "3      2019-03-27              USD            RV         1.0   \n",
       "4      2019-06-08              USD            RV         1.0   \n",
       "...           ...              ...           ...         ...   \n",
       "49992  2020-04-11              USD            RV         1.0   \n",
       "49993  2020-04-29              USD            RV         1.0   \n",
       "49994  2020-03-21              USD            RV         1.0   \n",
       "49995  2020-04-24              USD            RV         1.0   \n",
       "49996  2020-03-18              USD            RV         1.0   \n",
       "\n",
       "       total_open_amount baseline_create_date cust_payment_terms  \\\n",
       "0               21227.05           2019-02-20               NAA8   \n",
       "1                5669.72           2020-01-04               NAH4   \n",
       "2               67968.48           2019-06-21               NAA8   \n",
       "3               61395.73           2019-03-12               NAH4   \n",
       "4                1067.72           2019-05-16               NAM4   \n",
       "...                  ...                  ...                ...   \n",
       "49992             261.84           2020-03-27               NAA8   \n",
       "49993           14151.51           2020-03-25               NAG2   \n",
       "49994           14898.82           2020-03-06               NAC6   \n",
       "49995            5127.95           2020-04-09               NAA8   \n",
       "49996            3382.65           2020-03-03               NAAX   \n",
       "\n",
       "         invoice_id  isOpen  \n",
       "0      1.928827e+09       0  \n",
       "1      1.930350e+09       0  \n",
       "2      1.929517e+09       0  \n",
       "3      1.928943e+09       0  \n",
       "4      1.929350e+09       0  \n",
       "...             ...     ...  \n",
       "49992  1.930709e+09       1  \n",
       "49993  1.930694e+09       1  \n",
       "49994  1.930606e+09       1  \n",
       "49995  1.930761e+09       1  \n",
       "49996  1.930586e+09       1  \n",
       "\n",
       "[49997 rows x 18 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dropna(axis=0,subset=['invoice_id'],inplace=True)\n",
    "df.reset_index(drop=True,inplace=True)\n",
    "df.shape\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Checking for unique values in each column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "business_code - 6\n",
      "cust_number - 1460\n",
      "name_customer - 4291\n",
      "clear_date - 401\n",
      "buisness_year - 2\n",
      "doc_id - 49997\n",
      "posting_date - 502\n",
      "document_create_date - 504\n",
      "document_create_date.1 - 502\n",
      "due_in_date - 543\n",
      "invoice_currency - 2\n",
      "document type - 1\n",
      "posting_id - 1\n",
      "total_open_amount - 45333\n",
      "baseline_create_date - 504\n",
      "cust_payment_terms - 75\n",
      "invoice_id - 49997\n",
      "isOpen - 2\n"
     ]
    }
   ],
   "source": [
    " for cols in df.columns:   \n",
    "    print(f\"{cols} - {df[f'{cols}'].nunique()}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There is only 1 unique value in document type and posting_id. We can drop these."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['document type', 'posting_id']\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Index(['business_code', 'cust_number', 'name_customer', 'clear_date',\n",
       "       'buisness_year', 'doc_id', 'posting_date', 'document_create_date',\n",
       "       'document_create_date.1', 'due_in_date', 'invoice_currency',\n",
       "       'total_open_amount', 'baseline_create_date', 'cust_payment_terms',\n",
       "       'invoice_id', 'isOpen'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unique_cols =  [x for x in df.columns if df[x].nunique()==1] \n",
    "print(unique_cols)\n",
    "df.drop(unique_cols,axis=1,inplace=True)\n",
    "df.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's check for NULL values again."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "business_code             False\n",
       "cust_number               False\n",
       "name_customer             False\n",
       "clear_date                 True\n",
       "buisness_year             False\n",
       "doc_id                    False\n",
       "posting_date              False\n",
       "document_create_date      False\n",
       "document_create_date.1    False\n",
       "due_in_date               False\n",
       "invoice_currency          False\n",
       "total_open_amount         False\n",
       "baseline_create_date      False\n",
       "cust_payment_terms        False\n",
       "invoice_id                False\n",
       "isOpen                    False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().any()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "No null(except clear-date) values so moving on to removing duplicate columns."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Duplicate column removal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# function to find duplicate columns\n",
    "def findDuplicateColumns(df):\n",
    "    duplicatecolumns=set()\n",
    "    for x in range(df.shape[1]):\n",
    "        col1=df.iloc[:,x]\n",
    "        for y in range(x+1,df.shape[1]):\n",
    "            col2=df.iloc[:,y]\n",
    "            if col1.equals(col2):\n",
    "                duplicatecolumns.add(df.columns.values[x])\n",
    "    return list(duplicatecolumns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using user-defined function as the in-built function was computationally expensive."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['posting_date', 'doc_id']"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dr=findDuplicateColumns(df)\n",
    "dr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop(columns=dr,inplace=True) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We dropped posting_date and doc_id.\n",
    "\n",
    "Now, to avoid discrepancy between document_create_date and document_create_date.1 we drop document_create_date and instead use document_create_date.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop(columns=['document_create_date'],inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Also, invoice id is unique for each transaction and don't affect the sales, so we drop it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200315290</td>\n",
       "      <td>KWI in</td>\n",
       "      <td>2019-03-07</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-02-20</td>\n",
       "      <td>2019-03-07</td>\n",
       "      <td>USD</td>\n",
       "      <td>21227.05</td>\n",
       "      <td>2019-02-20</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR foundation</td>\n",
       "      <td>2020-01-16</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-04</td>\n",
       "      <td>2020-01-19</td>\n",
       "      <td>USD</td>\n",
       "      <td>5669.72</td>\n",
       "      <td>2020-01-04</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200820380</td>\n",
       "      <td>SMITH'S associates</td>\n",
       "      <td>2019-07-08</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-06-21</td>\n",
       "      <td>2019-07-06</td>\n",
       "      <td>USD</td>\n",
       "      <td>67968.48</td>\n",
       "      <td>2019-06-21</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR corp</td>\n",
       "      <td>2019-03-22</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-03-12</td>\n",
       "      <td>2019-03-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>61395.73</td>\n",
       "      <td>2019-03-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>200230690</td>\n",
       "      <td>DECA in</td>\n",
       "      <td>2019-06-03</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-05-21</td>\n",
       "      <td>2019-06-08</td>\n",
       "      <td>USD</td>\n",
       "      <td>1067.72</td>\n",
       "      <td>2019-05-16</td>\n",
       "      <td>NAM4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49992</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200726979</td>\n",
       "      <td>BJ'S  us</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-03-27</td>\n",
       "      <td>2020-04-11</td>\n",
       "      <td>USD</td>\n",
       "      <td>261.84</td>\n",
       "      <td>2020-03-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49993</th>\n",
       "      <td>U001</td>\n",
       "      <td>CCCA02</td>\n",
       "      <td>KRAFT  in</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-03-25</td>\n",
       "      <td>2020-04-29</td>\n",
       "      <td>USD</td>\n",
       "      <td>14151.51</td>\n",
       "      <td>2020-03-25</td>\n",
       "      <td>NAG2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49994</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200762301</td>\n",
       "      <td>C&amp;S WH corporation</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-03-06</td>\n",
       "      <td>2020-03-21</td>\n",
       "      <td>USD</td>\n",
       "      <td>14898.82</td>\n",
       "      <td>2020-03-06</td>\n",
       "      <td>NAC6</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49995</th>\n",
       "      <td>U001</td>\n",
       "      <td>0100001196</td>\n",
       "      <td>DOLLAR</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-04-09</td>\n",
       "      <td>2020-04-24</td>\n",
       "      <td>USD</td>\n",
       "      <td>5127.95</td>\n",
       "      <td>2020-04-09</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49996</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200794332</td>\n",
       "      <td>COST systems</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-03-03</td>\n",
       "      <td>2020-03-18</td>\n",
       "      <td>USD</td>\n",
       "      <td>3382.65</td>\n",
       "      <td>2020-03-03</td>\n",
       "      <td>NAAX</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>49997 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      business_code cust_number       name_customer clear_date  buisness_year  \\\n",
       "0              U001  0200315290              KWI in 2019-03-07         2019.0   \n",
       "1              U001  0200769623  WAL-MAR foundation 2020-01-16         2020.0   \n",
       "2              U001  0200820380  SMITH'S associates 2019-07-08         2019.0   \n",
       "3              U001  0200769623        WAL-MAR corp 2019-03-22         2019.0   \n",
       "4              U001   200230690             DECA in 2019-06-03         2019.0   \n",
       "...             ...         ...                 ...        ...            ...   \n",
       "49992          U001  0200726979            BJ'S  us        NaT         2020.0   \n",
       "49993          U001      CCCA02           KRAFT  in        NaT         2020.0   \n",
       "49994          U001  0200762301  C&S WH corporation        NaT         2020.0   \n",
       "49995          U001  0100001196             DOLLAR         NaT         2020.0   \n",
       "49996          U001  0200794332        COST systems        NaT         2020.0   \n",
       "\n",
       "      document_create_date.1 due_in_date invoice_currency  total_open_amount  \\\n",
       "0                 2019-02-20  2019-03-07              USD           21227.05   \n",
       "1                 2020-01-04  2020-01-19              USD            5669.72   \n",
       "2                 2019-06-21  2019-07-06              USD           67968.48   \n",
       "3                 2019-03-12  2019-03-27              USD           61395.73   \n",
       "4                 2019-05-21  2019-06-08              USD            1067.72   \n",
       "...                      ...         ...              ...                ...   \n",
       "49992             2020-03-27  2020-04-11              USD             261.84   \n",
       "49993             2020-03-25  2020-04-29              USD           14151.51   \n",
       "49994             2020-03-06  2020-03-21              USD           14898.82   \n",
       "49995             2020-04-09  2020-04-24              USD            5127.95   \n",
       "49996             2020-03-03  2020-03-18              USD            3382.65   \n",
       "\n",
       "      baseline_create_date cust_payment_terms  isOpen  \n",
       "0               2019-02-20               NAA8       0  \n",
       "1               2020-01-04               NAH4       0  \n",
       "2               2019-06-21               NAA8       0  \n",
       "3               2019-03-12               NAH4       0  \n",
       "4               2019-05-16               NAM4       0  \n",
       "...                    ...                ...     ...  \n",
       "49992           2020-03-27               NAA8       1  \n",
       "49993           2020-03-25               NAG2       1  \n",
       "49994           2020-03-06               NAC6       1  \n",
       "49995           2020-04-09               NAA8       1  \n",
       "49996           2020-03-03               NAAX       1  \n",
       "\n",
       "[49997 rows x 12 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.drop(columns=['invoice_id'],inplace=True) \n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TARGET VARIABLE AND SORT\n",
    "Our model cant take in dates for prediction so our target column will not be clear_date. Instead it will be the delay column as delay will be in int format which our model can predict.\n",
    "\n",
    "clear_date - due_date will tell us how much delay was there in paying. \n",
    " \n",
    " \"-\" indicates it has been paid off before due date. \n",
    " \"+\" indicates it has been paid off after the due date"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "unsupported operand type(s) for -: 'str' and 'float'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\core\\ops\\array_ops.py\u001b[0m in \u001b[0;36mna_arithmetic_op\u001b[1;34m(left, right, op, is_cmp)\u001b[0m\n\u001b[0;32m    142\u001b[0m     \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 143\u001b[1;33m         \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mexpressions\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mevaluate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mop\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mleft\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mright\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    144\u001b[0m     \u001b[1;32mexcept\u001b[0m \u001b[0mTypeError\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\core\\computation\\expressions.py\u001b[0m in \u001b[0;36mevaluate\u001b[1;34m(op, a, b, use_numexpr)\u001b[0m\n\u001b[0;32m    232\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0muse_numexpr\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 233\u001b[1;33m             \u001b[1;32mreturn\u001b[0m \u001b[0m_evaluate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mop\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mop_str\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mb\u001b[0m\u001b[1;33m)\u001b[0m  \u001b[1;31m# type: ignore\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    234\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0m_evaluate_standard\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mop\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mop_str\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mb\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\core\\computation\\expressions.py\u001b[0m in \u001b[0;36m_evaluate_numexpr\u001b[1;34m(op, op_str, a, b)\u001b[0m\n\u001b[0;32m    118\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mresult\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 119\u001b[1;33m         \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_evaluate_standard\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mop\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mop_str\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mb\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    120\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\core\\computation\\expressions.py\u001b[0m in \u001b[0;36m_evaluate_standard\u001b[1;34m(op, op_str, a, b)\u001b[0m\n\u001b[0;32m     67\u001b[0m     \u001b[1;32mwith\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0merrstate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mall\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m\"ignore\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 68\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mop\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mb\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     69\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mTypeError\u001b[0m: unsupported operand type(s) for -: 'str' and 'float'",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[1;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-4-ff19d77e2f63>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mdf\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'delay'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdf\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'clear_date'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m-\u001b[0m\u001b[0mdf\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'due_in_date'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdays\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\core\\ops\\common.py\u001b[0m in \u001b[0;36mnew_method\u001b[1;34m(self, other)\u001b[0m\n\u001b[0;32m     63\u001b[0m         \u001b[0mother\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mitem_from_zerodim\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mother\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     64\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 65\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mmethod\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mother\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     66\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     67\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mnew_method\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\core\\ops\\__init__.py\u001b[0m in \u001b[0;36mwrapper\u001b[1;34m(left, right)\u001b[0m\n\u001b[0;32m    341\u001b[0m         \u001b[0mlvalues\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mextract_array\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mleft\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mextract_numpy\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mTrue\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    342\u001b[0m         \u001b[0mrvalues\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mextract_array\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mright\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mextract_numpy\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mTrue\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 343\u001b[1;33m         \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0marithmetic_op\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlvalues\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrvalues\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mop\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    344\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    345\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0mleft\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_construct_result\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mresult\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mname\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mres_name\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\core\\ops\\array_ops.py\u001b[0m in \u001b[0;36marithmetic_op\u001b[1;34m(left, right, op)\u001b[0m\n\u001b[0;32m    188\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    189\u001b[0m         \u001b[1;32mwith\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0merrstate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mall\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m\"ignore\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 190\u001b[1;33m             \u001b[0mres_values\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mna_arithmetic_op\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlvalues\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrvalues\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mop\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    191\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    192\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mres_values\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\core\\ops\\array_ops.py\u001b[0m in \u001b[0;36mna_arithmetic_op\u001b[1;34m(left, right, op, is_cmp)\u001b[0m\n\u001b[0;32m    148\u001b[0m             \u001b[1;31m#  will handle complex numbers incorrectly, see GH#32047\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    149\u001b[0m             \u001b[1;32mraise\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 150\u001b[1;33m         \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmasked_arith_op\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mleft\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mright\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mop\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    151\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    152\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mis_cmp\u001b[0m \u001b[1;32mand\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mis_scalar\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mresult\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mor\u001b[0m \u001b[0mresult\u001b[0m \u001b[1;32mis\u001b[0m \u001b[0mNotImplemented\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\pandas\\core\\ops\\array_ops.py\u001b[0m in \u001b[0;36mmasked_arith_op\u001b[1;34m(x, y, op)\u001b[0m\n\u001b[0;32m     90\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mmask\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0many\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     91\u001b[0m             \u001b[1;32mwith\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0merrstate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mall\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m\"ignore\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 92\u001b[1;33m                 \u001b[0mresult\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mmask\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mop\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mxrav\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mmask\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0myrav\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mmask\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     93\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     94\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mTypeError\u001b[0m: unsupported operand type(s) for -: 'str' and 'float'"
     ]
    }
   ],
   "source": [
    "df['delay']=(df['clear_date']-df['due_in_date']).dt.days"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sorting in ascending order by document_create_date.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>26302</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR foundation</td>\n",
       "      <td>2019-01-09</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>25749.68</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36165</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200762301</td>\n",
       "      <td>C&amp;S WH co</td>\n",
       "      <td>2019-01-15</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>18116.87</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAC6</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38783</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200955827</td>\n",
       "      <td>KE</td>\n",
       "      <td>2019-01-16</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>32856.32</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4134</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744019</td>\n",
       "      <td>TARG trust</td>\n",
       "      <td>2019-01-18</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>11982.59</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8597</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200794332</td>\n",
       "      <td>COST corp</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>32715.47</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAAX</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46773</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200592182</td>\n",
       "      <td>DECA us</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>2020-05-24</td>\n",
       "      <td>USD</td>\n",
       "      <td>1226.00</td>\n",
       "      <td>2020-05-01</td>\n",
       "      <td>NAM4</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45434</th>\n",
       "      <td>U001</td>\n",
       "      <td>200769623</td>\n",
       "      <td>WAL-MAR co</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>2020-05-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>36513.74</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45919</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200915438</td>\n",
       "      <td>GROC trust</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-13</td>\n",
       "      <td>2020-05-28</td>\n",
       "      <td>USD</td>\n",
       "      <td>15220.00</td>\n",
       "      <td>2020-05-13</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48707</th>\n",
       "      <td>U001</td>\n",
       "      <td>100030964</td>\n",
       "      <td>NATURA corporation</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-16</td>\n",
       "      <td>2020-05-31</td>\n",
       "      <td>USD</td>\n",
       "      <td>17893.25</td>\n",
       "      <td>2020-05-16</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45610</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200915438</td>\n",
       "      <td>GROC</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-19</td>\n",
       "      <td>2020-06-03</td>\n",
       "      <td>USD</td>\n",
       "      <td>8326.87</td>\n",
       "      <td>2020-05-19</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>49997 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      business_code cust_number       name_customer clear_date  buisness_year  \\\n",
       "26302          U001  0200769623  WAL-MAR foundation 2019-01-09         2019.0   \n",
       "36165          U001  0200762301           C&S WH co 2019-01-15         2019.0   \n",
       "38783          U001  0200955827                 KE  2019-01-16         2019.0   \n",
       "4134           U001  0200744019          TARG trust 2019-01-18         2019.0   \n",
       "8597           U001  0200794332           COST corp 2019-01-14         2019.0   \n",
       "...             ...         ...                 ...        ...            ...   \n",
       "46773          U001  0200592182             DECA us        NaT         2020.0   \n",
       "45434          U001   200769623          WAL-MAR co        NaT         2020.0   \n",
       "45919          U001  0200915438          GROC trust        NaT         2020.0   \n",
       "48707          U001   100030964  NATURA corporation        NaT         2020.0   \n",
       "45610          U001  0200915438               GROC         NaT         2020.0   \n",
       "\n",
       "      document_create_date.1 due_in_date invoice_currency  total_open_amount  \\\n",
       "26302             2018-12-30  2019-01-14              USD           25749.68   \n",
       "36165             2018-12-30  2019-01-14              USD           18116.87   \n",
       "38783             2018-12-30  2019-01-14              USD           32856.32   \n",
       "4134              2018-12-30  2019-01-14              USD           11982.59   \n",
       "8597              2018-12-30  2019-01-14              USD           32715.47   \n",
       "...                      ...         ...              ...                ...   \n",
       "46773             2020-05-12  2020-05-24              USD            1226.00   \n",
       "45434             2020-05-12  2020-05-27              USD           36513.74   \n",
       "45919             2020-05-13  2020-05-28              USD           15220.00   \n",
       "48707             2020-05-16  2020-05-31              USD           17893.25   \n",
       "45610             2020-05-19  2020-06-03              USD            8326.87   \n",
       "\n",
       "      baseline_create_date cust_payment_terms  isOpen  delay  \n",
       "26302           2018-12-30               NAH4       0   -5.0  \n",
       "36165           2018-12-30               NAC6       0    1.0  \n",
       "38783           2018-12-30               NAA8       0    2.0  \n",
       "4134            2018-12-30               NAA8       0    4.0  \n",
       "8597            2018-12-30               NAAX       0    0.0  \n",
       "...                    ...                ...     ...    ...  \n",
       "46773           2020-05-01               NAM4       1    NaN  \n",
       "45434           2020-05-12               NAH4       1    NaN  \n",
       "45919           2020-05-13               NAA8       1    NaN  \n",
       "48707           2020-05-16               NAA8       1    NaN  \n",
       "45610           2020-05-19               NAA8       1    NaN  \n",
       "\n",
       "[49997 rows x 13 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sort_values(by='document_create_date.1',inplace=True) # sorting in ascending order by document_create_date.1\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SPLITTING\n",
    "Creating test set on null clear_date"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200772595</td>\n",
       "      <td>SAFEW associates</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>16157.16</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140105847</td>\n",
       "      <td>GORDON F co</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-10</td>\n",
       "      <td>CAD</td>\n",
       "      <td>722.50</td>\n",
       "      <td>2020-02-29</td>\n",
       "      <td>CA10</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744019</td>\n",
       "      <td>TARG co</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>4206.28</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200943275</td>\n",
       "      <td>US  in</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>51877.23</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200726979</td>\n",
       "      <td>BJ'S  llc</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>365.88</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4814</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200592182</td>\n",
       "      <td>DECA us</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>2020-05-24</td>\n",
       "      <td>USD</td>\n",
       "      <td>1226.00</td>\n",
       "      <td>2020-05-01</td>\n",
       "      <td>NAM4</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4815</th>\n",
       "      <td>U001</td>\n",
       "      <td>200769623</td>\n",
       "      <td>WAL-MAR co</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>2020-05-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>36513.74</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4816</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200915438</td>\n",
       "      <td>GROC trust</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-13</td>\n",
       "      <td>2020-05-28</td>\n",
       "      <td>USD</td>\n",
       "      <td>15220.00</td>\n",
       "      <td>2020-05-13</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4817</th>\n",
       "      <td>U001</td>\n",
       "      <td>100030964</td>\n",
       "      <td>NATURA corporation</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-16</td>\n",
       "      <td>2020-05-31</td>\n",
       "      <td>USD</td>\n",
       "      <td>17893.25</td>\n",
       "      <td>2020-05-16</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4818</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200915438</td>\n",
       "      <td>GROC</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-19</td>\n",
       "      <td>2020-06-03</td>\n",
       "      <td>USD</td>\n",
       "      <td>8326.87</td>\n",
       "      <td>2020-05-19</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4819 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     business_code cust_number       name_customer clear_date  buisness_year  \\\n",
       "0             U001  0200772595    SAFEW associates        NaT         2020.0   \n",
       "1             CA02  0140105847         GORDON F co        NaT         2020.0   \n",
       "2             U001  0200744019             TARG co        NaT         2020.0   \n",
       "3             U001  0200943275              US  in        NaT         2020.0   \n",
       "4             U001  0200726979           BJ'S  llc        NaT         2020.0   \n",
       "...            ...         ...                 ...        ...            ...   \n",
       "4814          U001  0200592182             DECA us        NaT         2020.0   \n",
       "4815          U001   200769623          WAL-MAR co        NaT         2020.0   \n",
       "4816          U001  0200915438          GROC trust        NaT         2020.0   \n",
       "4817          U001   100030964  NATURA corporation        NaT         2020.0   \n",
       "4818          U001  0200915438               GROC         NaT         2020.0   \n",
       "\n",
       "     document_create_date.1 due_in_date invoice_currency  total_open_amount  \\\n",
       "0                2020-02-27  2020-03-13              USD           16157.16   \n",
       "1                2020-02-27  2020-03-10              CAD             722.50   \n",
       "2                2020-02-27  2020-03-13              USD            4206.28   \n",
       "3                2020-02-27  2020-03-13              USD           51877.23   \n",
       "4                2020-02-27  2020-03-13              USD             365.88   \n",
       "...                     ...         ...              ...                ...   \n",
       "4814             2020-05-12  2020-05-24              USD            1226.00   \n",
       "4815             2020-05-12  2020-05-27              USD           36513.74   \n",
       "4816             2020-05-13  2020-05-28              USD           15220.00   \n",
       "4817             2020-05-16  2020-05-31              USD           17893.25   \n",
       "4818             2020-05-19  2020-06-03              USD            8326.87   \n",
       "\n",
       "     baseline_create_date cust_payment_terms  isOpen  delay  \n",
       "0              2020-02-27               NAA8       1    NaN  \n",
       "1              2020-02-29               CA10       1    NaN  \n",
       "2              2020-02-27               NAA8       1    NaN  \n",
       "3              2020-02-27               NAA8       1    NaN  \n",
       "4              2020-02-27               NAA8       1    NaN  \n",
       "...                   ...                ...     ...    ...  \n",
       "4814           2020-05-01               NAM4       1    NaN  \n",
       "4815           2020-05-12               NAH4       1    NaN  \n",
       "4816           2020-05-13               NAA8       1    NaN  \n",
       "4817           2020-05-16               NAA8       1    NaN  \n",
       "4818           2020-05-19               NAA8       1    NaN  \n",
       "\n",
       "[4819 rows x 13 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data = df[df.clear_date.isnull()].reset_index()\n",
    "test_data.drop(columns=['index'],inplace=True)\n",
    "test_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "ts = test_data.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Splitting test_set into x_test and y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      NaN\n",
       "1      NaN\n",
       "2      NaN\n",
       "3      NaN\n",
       "4      NaN\n",
       "        ..\n",
       "4814   NaN\n",
       "4815   NaN\n",
       "4816   NaN\n",
       "4817   NaN\n",
       "4818   NaN\n",
       "Name: delay, Length: 4819, dtype: float64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test = test_data[\"delay\"]\n",
    "y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200772595</td>\n",
       "      <td>SAFEW associates</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>16157.16</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140105847</td>\n",
       "      <td>GORDON F co</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-10</td>\n",
       "      <td>CAD</td>\n",
       "      <td>722.50</td>\n",
       "      <td>2020-02-29</td>\n",
       "      <td>CA10</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744019</td>\n",
       "      <td>TARG co</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>4206.28</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200943275</td>\n",
       "      <td>US  in</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>51877.23</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200726979</td>\n",
       "      <td>BJ'S  llc</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>365.88</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4814</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200592182</td>\n",
       "      <td>DECA us</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>2020-05-24</td>\n",
       "      <td>USD</td>\n",
       "      <td>1226.00</td>\n",
       "      <td>2020-05-01</td>\n",
       "      <td>NAM4</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4815</th>\n",
       "      <td>U001</td>\n",
       "      <td>200769623</td>\n",
       "      <td>WAL-MAR co</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>2020-05-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>36513.74</td>\n",
       "      <td>2020-05-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4816</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200915438</td>\n",
       "      <td>GROC trust</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-13</td>\n",
       "      <td>2020-05-28</td>\n",
       "      <td>USD</td>\n",
       "      <td>15220.00</td>\n",
       "      <td>2020-05-13</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4817</th>\n",
       "      <td>U001</td>\n",
       "      <td>100030964</td>\n",
       "      <td>NATURA corporation</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-16</td>\n",
       "      <td>2020-05-31</td>\n",
       "      <td>USD</td>\n",
       "      <td>17893.25</td>\n",
       "      <td>2020-05-16</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4818</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200915438</td>\n",
       "      <td>GROC</td>\n",
       "      <td>NaT</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-05-19</td>\n",
       "      <td>2020-06-03</td>\n",
       "      <td>USD</td>\n",
       "      <td>8326.87</td>\n",
       "      <td>2020-05-19</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4819 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     business_code cust_number       name_customer clear_date  buisness_year  \\\n",
       "0             U001  0200772595    SAFEW associates        NaT         2020.0   \n",
       "1             CA02  0140105847         GORDON F co        NaT         2020.0   \n",
       "2             U001  0200744019             TARG co        NaT         2020.0   \n",
       "3             U001  0200943275              US  in        NaT         2020.0   \n",
       "4             U001  0200726979           BJ'S  llc        NaT         2020.0   \n",
       "...            ...         ...                 ...        ...            ...   \n",
       "4814          U001  0200592182             DECA us        NaT         2020.0   \n",
       "4815          U001   200769623          WAL-MAR co        NaT         2020.0   \n",
       "4816          U001  0200915438          GROC trust        NaT         2020.0   \n",
       "4817          U001   100030964  NATURA corporation        NaT         2020.0   \n",
       "4818          U001  0200915438               GROC         NaT         2020.0   \n",
       "\n",
       "     document_create_date.1 due_in_date invoice_currency  total_open_amount  \\\n",
       "0                2020-02-27  2020-03-13              USD           16157.16   \n",
       "1                2020-02-27  2020-03-10              CAD             722.50   \n",
       "2                2020-02-27  2020-03-13              USD            4206.28   \n",
       "3                2020-02-27  2020-03-13              USD           51877.23   \n",
       "4                2020-02-27  2020-03-13              USD             365.88   \n",
       "...                     ...         ...              ...                ...   \n",
       "4814             2020-05-12  2020-05-24              USD            1226.00   \n",
       "4815             2020-05-12  2020-05-27              USD           36513.74   \n",
       "4816             2020-05-13  2020-05-28              USD           15220.00   \n",
       "4817             2020-05-16  2020-05-31              USD           17893.25   \n",
       "4818             2020-05-19  2020-06-03              USD            8326.87   \n",
       "\n",
       "     baseline_create_date cust_payment_terms  isOpen  \n",
       "0              2020-02-27               NAA8       1  \n",
       "1              2020-02-29               CA10       1  \n",
       "2              2020-02-27               NAA8       1  \n",
       "3              2020-02-27               NAA8       1  \n",
       "4              2020-02-27               NAA8       1  \n",
       "...                   ...                ...     ...  \n",
       "4814           2020-05-01               NAM4       1  \n",
       "4815           2020-05-12               NAH4       1  \n",
       "4816           2020-05-13               NAA8       1  \n",
       "4817           2020-05-16               NAA8       1  \n",
       "4818           2020-05-19               NAA8       1  \n",
       "\n",
       "[4819 rows x 12 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test = test_data.iloc[:,:-1].copy()\n",
    "x_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Creating a seprate data frame out of which we will extract val1, val2 and final train set.\n",
    "WHOLE DATA ---> [TRAIN_DATA + VAL1 (1.5 MONTHS) + VAL2 (1.5 MONTHS)] + TEST_DATA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR foundation</td>\n",
       "      <td>2019-01-09</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>25749.68</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200762301</td>\n",
       "      <td>C&amp;S WH co</td>\n",
       "      <td>2019-01-15</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>18116.87</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAC6</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200955827</td>\n",
       "      <td>KE</td>\n",
       "      <td>2019-01-16</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>32856.32</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744019</td>\n",
       "      <td>TARG trust</td>\n",
       "      <td>2019-01-18</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>11982.59</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200794332</td>\n",
       "      <td>COST corp</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>32715.47</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAAX</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45173</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140106408</td>\n",
       "      <td>WAL-M us</td>\n",
       "      <td>2020-03-20</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-09</td>\n",
       "      <td>CAD</td>\n",
       "      <td>2878.29</td>\n",
       "      <td>2020-02-28</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>11.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45174</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200721222</td>\n",
       "      <td>GO associates</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-05-02</td>\n",
       "      <td>USD</td>\n",
       "      <td>10770.76</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAGD</td>\n",
       "      <td>0</td>\n",
       "      <td>-50.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45175</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744707</td>\n",
       "      <td>PERFOR</td>\n",
       "      <td>2020-03-16</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>39408.22</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45176</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200759878</td>\n",
       "      <td>SA in</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>70504.55</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45177</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200700919</td>\n",
       "      <td>US  co</td>\n",
       "      <td>2020-03-20</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-18</td>\n",
       "      <td>USD</td>\n",
       "      <td>90195.02</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAD1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>45178 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      business_code cust_number       name_customer clear_date  buisness_year  \\\n",
       "0              U001  0200769623  WAL-MAR foundation 2019-01-09         2019.0   \n",
       "1              U001  0200762301           C&S WH co 2019-01-15         2019.0   \n",
       "2              U001  0200955827                 KE  2019-01-16         2019.0   \n",
       "3              U001  0200744019          TARG trust 2019-01-18         2019.0   \n",
       "4              U001  0200794332           COST corp 2019-01-14         2019.0   \n",
       "...             ...         ...                 ...        ...            ...   \n",
       "45173          CA02  0140106408            WAL-M us 2020-03-20         2020.0   \n",
       "45174          U001  0200721222       GO associates 2020-03-13         2020.0   \n",
       "45175          U001  0200744707             PERFOR  2020-03-16         2020.0   \n",
       "45176          U001  0200759878               SA in 2020-03-13         2020.0   \n",
       "45177          U001  0200700919              US  co 2020-03-20         2020.0   \n",
       "\n",
       "      document_create_date.1 due_in_date invoice_currency  total_open_amount  \\\n",
       "0                 2018-12-30  2019-01-14              USD           25749.68   \n",
       "1                 2018-12-30  2019-01-14              USD           18116.87   \n",
       "2                 2018-12-30  2019-01-14              USD           32856.32   \n",
       "3                 2018-12-30  2019-01-14              USD           11982.59   \n",
       "4                 2018-12-30  2019-01-14              USD           32715.47   \n",
       "...                      ...         ...              ...                ...   \n",
       "45173             2020-02-27  2020-03-09              CAD            2878.29   \n",
       "45174             2020-02-27  2020-05-02              USD           10770.76   \n",
       "45175             2020-02-27  2020-03-13              USD           39408.22   \n",
       "45176             2020-02-27  2020-03-13              USD           70504.55   \n",
       "45177             2020-02-27  2020-03-18              USD           90195.02   \n",
       "\n",
       "      baseline_create_date cust_payment_terms  isOpen  delay  \n",
       "0               2018-12-30               NAH4       0   -5.0  \n",
       "1               2018-12-30               NAC6       0    1.0  \n",
       "2               2018-12-30               NAA8       0    2.0  \n",
       "3               2018-12-30               NAA8       0    4.0  \n",
       "4               2018-12-30               NAAX       0    0.0  \n",
       "...                    ...                ...     ...    ...  \n",
       "45173           2020-02-28               CA10       0   11.0  \n",
       "45174           2020-02-27               NAGD       0  -50.0  \n",
       "45175           2020-02-27               NAA8       0    3.0  \n",
       "45176           2020-02-27               NAH4       0    0.0  \n",
       "45177           2020-02-27               NAD1       0    2.0  \n",
       "\n",
       "[45178 rows x 13 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data2 = df[df.clear_date.notnull()].reset_index() #train set on non-null clear date\n",
    "train_data2.drop(columns=['index'],inplace=True)\n",
    "train_data2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2020-01-12 00:00:00\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "40518"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Making use of offset to find date 1.5 months prior to 2020-02-27\n",
    "x = train_data2['document_create_date.1'].iloc[-1] - pd.DateOffset(months=1, days=15)\n",
    "y = train_data2['document_create_date.1'].loc[train_data2['document_create_date.1'] <= x].iloc[-1]\n",
    "print(y) \n",
    "train_data2['document_create_date.1'].loc[train_data2['document_create_date.1'] == y].last_valid_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This gives the date which is 1.5 months prior to 2020-02-27.\n",
    "\n",
    "val2 will start from 2020-01-13 to 2020-02-27"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200789077</td>\n",
       "      <td>US  corp</td>\n",
       "      <td>2020-01-24</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>2020-01-28</td>\n",
       "      <td>USD</td>\n",
       "      <td>32547.72</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR in</td>\n",
       "      <td>2020-01-23</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>2020-01-28</td>\n",
       "      <td>USD</td>\n",
       "      <td>4388.75</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR co</td>\n",
       "      <td>2020-01-23</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>2020-01-28</td>\n",
       "      <td>USD</td>\n",
       "      <td>6040.68</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR in</td>\n",
       "      <td>2020-01-23</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>2020-01-28</td>\n",
       "      <td>USD</td>\n",
       "      <td>71525.23</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200117492</td>\n",
       "      <td>PE in</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>2020-01-28</td>\n",
       "      <td>USD</td>\n",
       "      <td>231.24</td>\n",
       "      <td>2020-01-13</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>-15.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4654</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140106408</td>\n",
       "      <td>WAL-M us</td>\n",
       "      <td>2020-03-20</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-09</td>\n",
       "      <td>CAD</td>\n",
       "      <td>2878.29</td>\n",
       "      <td>2020-02-28</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>11.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4655</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200721222</td>\n",
       "      <td>GO associates</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-05-02</td>\n",
       "      <td>USD</td>\n",
       "      <td>10770.76</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAGD</td>\n",
       "      <td>0</td>\n",
       "      <td>-50.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4656</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744707</td>\n",
       "      <td>PERFOR</td>\n",
       "      <td>2020-03-16</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>39408.22</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4657</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200759878</td>\n",
       "      <td>SA in</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>70504.55</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4658</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200700919</td>\n",
       "      <td>US  co</td>\n",
       "      <td>2020-03-20</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>2020-03-18</td>\n",
       "      <td>USD</td>\n",
       "      <td>90195.02</td>\n",
       "      <td>2020-02-27</td>\n",
       "      <td>NAD1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4659 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     business_code cust_number  name_customer clear_date  buisness_year  \\\n",
       "0             U001  0200789077       US  corp 2020-01-24         2020.0   \n",
       "1             U001  0200769623     WAL-MAR in 2020-01-23         2020.0   \n",
       "2             U001  0200769623     WAL-MAR co 2020-01-23         2020.0   \n",
       "3             U001  0200769623     WAL-MAR in 2020-01-23         2020.0   \n",
       "4             U001  0200117492          PE in 2020-01-13         2020.0   \n",
       "...            ...         ...            ...        ...            ...   \n",
       "4654          CA02  0140106408       WAL-M us 2020-03-20         2020.0   \n",
       "4655          U001  0200721222  GO associates 2020-03-13         2020.0   \n",
       "4656          U001  0200744707        PERFOR  2020-03-16         2020.0   \n",
       "4657          U001  0200759878          SA in 2020-03-13         2020.0   \n",
       "4658          U001  0200700919         US  co 2020-03-20         2020.0   \n",
       "\n",
       "     document_create_date.1 due_in_date invoice_currency  total_open_amount  \\\n",
       "0                2020-01-13  2020-01-28              USD           32547.72   \n",
       "1                2020-01-13  2020-01-28              USD            4388.75   \n",
       "2                2020-01-13  2020-01-28              USD            6040.68   \n",
       "3                2020-01-13  2020-01-28              USD           71525.23   \n",
       "4                2020-01-13  2020-01-28              USD             231.24   \n",
       "...                     ...         ...              ...                ...   \n",
       "4654             2020-02-27  2020-03-09              CAD            2878.29   \n",
       "4655             2020-02-27  2020-05-02              USD           10770.76   \n",
       "4656             2020-02-27  2020-03-13              USD           39408.22   \n",
       "4657             2020-02-27  2020-03-13              USD           70504.55   \n",
       "4658             2020-02-27  2020-03-18              USD           90195.02   \n",
       "\n",
       "     baseline_create_date cust_payment_terms  isOpen  delay  \n",
       "0              2020-01-13               NAA8       0   -4.0  \n",
       "1              2020-01-13               NAH4       0   -5.0  \n",
       "2              2020-01-13               NAH4       0   -5.0  \n",
       "3              2020-01-13               NAH4       0   -5.0  \n",
       "4              2020-01-13               NAA8       0  -15.0  \n",
       "...                   ...                ...     ...    ...  \n",
       "4654           2020-02-28               CA10       0   11.0  \n",
       "4655           2020-02-27               NAGD       0  -50.0  \n",
       "4656           2020-02-27               NAA8       0    3.0  \n",
       "4657           2020-02-27               NAH4       0    0.0  \n",
       "4658           2020-02-27               NAD1       0    2.0  \n",
       "\n",
       "[4659 rows x 13 columns]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# dataframe for val2\n",
    "val2 = train_data2.iloc[40519: , :].copy()\n",
    "val2.reset_index(drop=True,inplace=True)\n",
    "val2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Removing val2 from train_data2 and storing into new dataframe to extract val1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR foundation</td>\n",
       "      <td>2019-01-09</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>25749.68</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200762301</td>\n",
       "      <td>C&amp;S WH co</td>\n",
       "      <td>2019-01-15</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>18116.87</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAC6</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200955827</td>\n",
       "      <td>KE</td>\n",
       "      <td>2019-01-16</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>32856.32</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744019</td>\n",
       "      <td>TARG trust</td>\n",
       "      <td>2019-01-18</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>11982.59</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200794332</td>\n",
       "      <td>COST corp</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>32715.47</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAAX</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40514</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200726979</td>\n",
       "      <td>BJ'S  corporation</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>2020.42</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40515</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR foundation</td>\n",
       "      <td>2020-01-24</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>256.91</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40516</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR systems</td>\n",
       "      <td>2020-01-24</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>256.91</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40517</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR corp</td>\n",
       "      <td>2020-01-22</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>1118.71</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40518</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200707822</td>\n",
       "      <td>PUBLI</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>135465.20</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>40497 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      business_code cust_number       name_customer clear_date  buisness_year  \\\n",
       "0              U001  0200769623  WAL-MAR foundation 2019-01-09         2019.0   \n",
       "1              U001  0200762301           C&S WH co 2019-01-15         2019.0   \n",
       "2              U001  0200955827                 KE  2019-01-16         2019.0   \n",
       "3              U001  0200744019          TARG trust 2019-01-18         2019.0   \n",
       "4              U001  0200794332           COST corp 2019-01-14         2019.0   \n",
       "...             ...         ...                 ...        ...            ...   \n",
       "40514          U001  0200726979   BJ'S  corporation 2020-01-27         2020.0   \n",
       "40515          U001  0200769623  WAL-MAR foundation 2020-01-24         2020.0   \n",
       "40516          U001  0200769623     WAL-MAR systems 2020-01-24         2020.0   \n",
       "40517          U001  0200769623        WAL-MAR corp 2020-01-22         2020.0   \n",
       "40518          U001  0200707822              PUBLI  2020-01-27         2020.0   \n",
       "\n",
       "      document_create_date.1 due_in_date invoice_currency  total_open_amount  \\\n",
       "0                 2018-12-30  2019-01-14              USD           25749.68   \n",
       "1                 2018-12-30  2019-01-14              USD           18116.87   \n",
       "2                 2018-12-30  2019-01-14              USD           32856.32   \n",
       "3                 2018-12-30  2019-01-14              USD           11982.59   \n",
       "4                 2018-12-30  2019-01-14              USD           32715.47   \n",
       "...                      ...         ...              ...                ...   \n",
       "40514             2020-01-12  2020-01-27              USD            2020.42   \n",
       "40515             2020-01-12  2020-01-27              USD             256.91   \n",
       "40516             2020-01-12  2020-01-27              USD             256.91   \n",
       "40517             2020-01-12  2020-01-27              USD            1118.71   \n",
       "40518             2020-01-12  2020-01-27              USD          135465.20   \n",
       "\n",
       "      baseline_create_date cust_payment_terms  isOpen  delay  \n",
       "0               2018-12-30               NAH4       0   -5.0  \n",
       "1               2018-12-30               NAC6       0    1.0  \n",
       "2               2018-12-30               NAA8       0    2.0  \n",
       "3               2018-12-30               NAA8       0    4.0  \n",
       "4               2018-12-30               NAAX       0    0.0  \n",
       "...                    ...                ...     ...    ...  \n",
       "40514           2020-01-12               NAA8       0    0.0  \n",
       "40515           2020-01-12               NAH4       0   -3.0  \n",
       "40516           2020-01-12               NAH4       0   -3.0  \n",
       "40517           2020-01-12               NAH4       0   -5.0  \n",
       "40518           2020-01-12               NAA8       0    0.0  \n",
       "\n",
       "[40497 rows x 13 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data1 = pd.concat([train_data2, val2]).drop_duplicates(keep=False)\n",
    "train_data1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2019-11-27 00:00:00\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "37251"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x2 = train_data1['document_create_date.1'].iloc[-1] - pd.DateOffset(months=1, days=15)\n",
    "y2 = train_data1['document_create_date.1'].loc[train_data1['document_create_date.1'] <= x2].iloc[-1]\n",
    "print(y2)\n",
    "train_data1['document_create_date.1'].loc[train_data1['document_create_date.1'] == y2].last_valid_index() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This gives the date which is 1.5 months prior to 2020-01-13 (start of val2).\n",
    "\n",
    "val1 will start from 2019-11-28 to 2020-01-12."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200763152</td>\n",
       "      <td>PERFOR co</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>21748.48</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR systems</td>\n",
       "      <td>2019-12-10</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>70698.34</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR co</td>\n",
       "      <td>2019-12-18</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>897.46</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140106408</td>\n",
       "      <td>WAL-M systems</td>\n",
       "      <td>2019-12-19</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-16</td>\n",
       "      <td>CAD</td>\n",
       "      <td>2367.26</td>\n",
       "      <td>2019-12-06</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744019</td>\n",
       "      <td>TARG trust</td>\n",
       "      <td>2019-12-16</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>34027.91</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3241</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200726979</td>\n",
       "      <td>BJ'S  corporation</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>2020.42</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3242</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR foundation</td>\n",
       "      <td>2020-01-24</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>256.91</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3243</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR systems</td>\n",
       "      <td>2020-01-24</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>256.91</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3244</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR corp</td>\n",
       "      <td>2020-01-22</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>1118.71</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3245</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200707822</td>\n",
       "      <td>PUBLI</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>2020.0</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>2020-01-27</td>\n",
       "      <td>USD</td>\n",
       "      <td>135465.20</td>\n",
       "      <td>2020-01-12</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3246 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     business_code cust_number       name_customer clear_date  buisness_year  \\\n",
       "0             U001  0200763152           PERFOR co 2019-12-13         2019.0   \n",
       "1             U001  0200769623     WAL-MAR systems 2019-12-10         2019.0   \n",
       "2             U001  0200769623          WAL-MAR co 2019-12-18         2019.0   \n",
       "3             CA02  0140106408       WAL-M systems 2019-12-19         2019.0   \n",
       "4             U001  0200744019          TARG trust 2019-12-16         2019.0   \n",
       "...            ...         ...                 ...        ...            ...   \n",
       "3241          U001  0200726979   BJ'S  corporation 2020-01-27         2020.0   \n",
       "3242          U001  0200769623  WAL-MAR foundation 2020-01-24         2020.0   \n",
       "3243          U001  0200769623     WAL-MAR systems 2020-01-24         2020.0   \n",
       "3244          U001  0200769623        WAL-MAR corp 2020-01-22         2020.0   \n",
       "3245          U001  0200707822              PUBLI  2020-01-27         2020.0   \n",
       "\n",
       "     document_create_date.1 due_in_date invoice_currency  total_open_amount  \\\n",
       "0                2019-11-28  2019-12-13              USD           21748.48   \n",
       "1                2019-11-28  2019-12-13              USD           70698.34   \n",
       "2                2019-11-28  2019-12-13              USD             897.46   \n",
       "3                2019-11-28  2019-12-16              CAD            2367.26   \n",
       "4                2019-11-28  2019-12-13              USD           34027.91   \n",
       "...                     ...         ...              ...                ...   \n",
       "3241             2020-01-12  2020-01-27              USD            2020.42   \n",
       "3242             2020-01-12  2020-01-27              USD             256.91   \n",
       "3243             2020-01-12  2020-01-27              USD             256.91   \n",
       "3244             2020-01-12  2020-01-27              USD            1118.71   \n",
       "3245             2020-01-12  2020-01-27              USD          135465.20   \n",
       "\n",
       "     baseline_create_date cust_payment_terms  isOpen  delay  \n",
       "0              2019-11-28               NAA8       0    0.0  \n",
       "1              2019-11-28               NAH4       0   -3.0  \n",
       "2              2019-11-28               NAH4       0    5.0  \n",
       "3              2019-12-06               CA10       0    3.0  \n",
       "4              2019-11-28               NAA8       0    3.0  \n",
       "...                   ...                ...     ...    ...  \n",
       "3241           2020-01-12               NAA8       0    0.0  \n",
       "3242           2020-01-12               NAH4       0   -3.0  \n",
       "3243           2020-01-12               NAH4       0   -3.0  \n",
       "3244           2020-01-12               NAH4       0   -5.0  \n",
       "3245           2020-01-12               NAA8       0    0.0  \n",
       "\n",
       "[3246 rows x 13 columns]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val1 = train_data1.iloc[37251: , :].copy()\n",
    "val1.reset_index(drop=True,inplace=True)\n",
    "val1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Removing val1 from the dataframe and making the final train set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR foundation</td>\n",
       "      <td>2019-01-09</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>25749.68</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200762301</td>\n",
       "      <td>C&amp;S WH co</td>\n",
       "      <td>2019-01-15</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>18116.87</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAC6</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200955827</td>\n",
       "      <td>KE</td>\n",
       "      <td>2019-01-16</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>32856.32</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744019</td>\n",
       "      <td>TARG trust</td>\n",
       "      <td>2019-01-18</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>11982.59</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200794332</td>\n",
       "      <td>COST corp</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>USD</td>\n",
       "      <td>32715.47</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>NAAX</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37266</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR trust</td>\n",
       "      <td>2019-12-09</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>31752.94</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37267</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200744019</td>\n",
       "      <td>TARG systems</td>\n",
       "      <td>2019-12-12</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>45591.41</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37268</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200769623</td>\n",
       "      <td>WAL-MAR corporation</td>\n",
       "      <td>2019-12-16</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>61336.27</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>NAH4</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37269</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140105847</td>\n",
       "      <td>GORDON F</td>\n",
       "      <td>2019-12-17</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-08</td>\n",
       "      <td>CAD</td>\n",
       "      <td>13937.60</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37270</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200726979</td>\n",
       "      <td>BJ'S  co</td>\n",
       "      <td>2019-12-16</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>USD</td>\n",
       "      <td>445.79</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>NAA8</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>37251 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      business_code cust_number        name_customer clear_date  \\\n",
       "0              U001  0200769623   WAL-MAR foundation 2019-01-09   \n",
       "1              U001  0200762301            C&S WH co 2019-01-15   \n",
       "2              U001  0200955827                  KE  2019-01-16   \n",
       "3              U001  0200744019           TARG trust 2019-01-18   \n",
       "4              U001  0200794332            COST corp 2019-01-14   \n",
       "...             ...         ...                  ...        ...   \n",
       "37266          U001  0200769623        WAL-MAR trust 2019-12-09   \n",
       "37267          U001  0200744019         TARG systems 2019-12-12   \n",
       "37268          U001  0200769623  WAL-MAR corporation 2019-12-16   \n",
       "37269          CA02  0140105847            GORDON F  2019-12-17   \n",
       "37270          U001  0200726979             BJ'S  co 2019-12-16   \n",
       "\n",
       "       buisness_year document_create_date.1 due_in_date invoice_currency  \\\n",
       "0             2019.0             2018-12-30  2019-01-14              USD   \n",
       "1             2019.0             2018-12-30  2019-01-14              USD   \n",
       "2             2019.0             2018-12-30  2019-01-14              USD   \n",
       "3             2019.0             2018-12-30  2019-01-14              USD   \n",
       "4             2019.0             2018-12-30  2019-01-14              USD   \n",
       "...              ...                    ...         ...              ...   \n",
       "37266         2019.0             2019-11-28  2019-12-13              USD   \n",
       "37267         2019.0             2019-11-28  2019-12-13              USD   \n",
       "37268         2019.0             2019-11-28  2019-12-13              USD   \n",
       "37269         2019.0             2019-11-28  2019-12-08              CAD   \n",
       "37270         2019.0             2019-11-28  2019-12-13              USD   \n",
       "\n",
       "       total_open_amount baseline_create_date cust_payment_terms  isOpen  \\\n",
       "0               25749.68           2018-12-30               NAH4       0   \n",
       "1               18116.87           2018-12-30               NAC6       0   \n",
       "2               32856.32           2018-12-30               NAA8       0   \n",
       "3               11982.59           2018-12-30               NAA8       0   \n",
       "4               32715.47           2018-12-30               NAAX       0   \n",
       "...                  ...                  ...                ...     ...   \n",
       "37266           31752.94           2019-11-28               NAH4       0   \n",
       "37267           45591.41           2019-11-28               NAA8       0   \n",
       "37268           61336.27           2019-11-28               NAH4       0   \n",
       "37269           13937.60           2019-11-28               CA10       0   \n",
       "37270             445.79           2019-11-28               NAA8       0   \n",
       "\n",
       "       delay  \n",
       "0       -5.0  \n",
       "1        1.0  \n",
       "2        2.0  \n",
       "3        4.0  \n",
       "4        0.0  \n",
       "...      ...  \n",
       "37266   -4.0  \n",
       "37267   -1.0  \n",
       "37268    3.0  \n",
       "37269    9.0  \n",
       "37270    3.0  \n",
       "\n",
       "[37251 rows x 13 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data = pd.concat([train_data1, val1]).drop_duplicates(keep=False)\n",
    "train_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We successfully splitted the data into train, val1, val2, and test set."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# EDA on train set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## UNIVARIATE ANALYSIS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 37251 entries, 0 to 37270\n",
      "Data columns (total 13 columns):\n",
      " #   Column                  Non-Null Count  Dtype         \n",
      "---  ------                  --------------  -----         \n",
      " 0   business_code           37251 non-null  object        \n",
      " 1   cust_number             37251 non-null  object        \n",
      " 2   name_customer           37251 non-null  object        \n",
      " 3   clear_date              37251 non-null  datetime64[ns]\n",
      " 4   buisness_year           37251 non-null  float64       \n",
      " 5   document_create_date.1  37251 non-null  datetime64[ns]\n",
      " 6   due_in_date             37251 non-null  datetime64[ns]\n",
      " 7   invoice_currency        37251 non-null  object        \n",
      " 8   total_open_amount       37251 non-null  float64       \n",
      " 9   baseline_create_date    37251 non-null  datetime64[ns]\n",
      " 10  cust_payment_terms      37251 non-null  object        \n",
      " 11  isOpen                  37251 non-null  int64         \n",
      " 12  delay                   37251 non-null  float64       \n",
      "dtypes: datetime64[ns](4), float64(3), int64(1), object(5)\n",
      "memory usage: 4.0+ MB\n"
     ]
    }
   ],
   "source": [
    "train_data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see there are no null values in our dataset. There are only 3 continuous variable column i.e. buisness_year, total_open_amount and delay,"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### business_code\n",
    "company code of the account"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "U001    33713\n",
       "CA02     2884\n",
       "U013      520\n",
       "U002      124\n",
       "U005        8\n",
       "U007        2\n",
       "Name: business_code, dtype: int64"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.business_code.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='business_code', ylabel='count'>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(x = train_data['business_code'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As there are only 6 values we can apply encoding on it later."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### cust_number\n",
    "customer number given to all the customers of the Account."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0200769623    8140\n",
       "0200726979    1415\n",
       "0200762301    1023\n",
       "0200759878     939\n",
       "0200794332     788\n",
       "              ... \n",
       "0140105033       1\n",
       "0200420725       1\n",
       "200783189        1\n",
       "100031920        1\n",
       "0100006823       1\n",
       "Name: cust_number, Length: 1359, dtype: int64"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.cust_number.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1359"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(dict(train_data.cust_number.value_counts()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are 1359 unique customers or comapny that we have transactions with.\n",
    "\n",
    "cust_number represent account no of different companies. For eg WAL-MAR has many varities that is WAL-MAR corporation, WAL-MAR systems etc but they have the same cust_number."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### name_customer\n",
    "name of the customer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "WAL-MAR corporation    843\n",
       "WAL-MAR systems        831\n",
       "WAL-MAR foundation     826\n",
       "WAL-MAR llc            819\n",
       "WAL-MAR co             809\n",
       "                      ... \n",
       "CHENS E systems          1\n",
       "FEDER corp               1\n",
       "PERISH corp              1\n",
       "GEOR trust               1\n",
       "WR us                    1\n",
       "Name: name_customer, Length: 3865, dtype: int64"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.name_customer.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Companies Like WAL-MAR, WAL-MAR systems, etc belong to company WAL-MAR and have a unique customer_num to their name hence this column will be dropped."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### clear_date\n",
    "The date on which the customer clears an invoice, or in simple terms, they make the full payment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2019-01-04 00:00:00')"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.clear_date.min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2019-12-31 00:00:00')"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.clear_date.max()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our clear_date ranges for 362 days."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5     3786\n",
       "4     3585\n",
       "11    3524\n",
       "8     3503\n",
       "10    3454\n",
       "9     3405\n",
       "7     3402\n",
       "3     3307\n",
       "6     3294\n",
       "2     2770\n",
       "12    1689\n",
       "1     1532\n",
       "Name: clear_date, dtype: int64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clear_month = train_data.clear_date.dt.month\n",
    "clear_month.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='clear_date', ylabel='count'>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(x=clear_month, palette=\"hls\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "May is the month where the companies completed their maximum transations."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### buisness_year\n",
    "indicates the year of clear date"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2019.0    37251\n",
       "Name: buisness_year, dtype: int64"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.buisness_year.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As there is only 1 buisness_year we will drop this column."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### document_create_date.1\n",
    "The date on which the invoice document was created"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2018-12-30 00:00:00')"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data[\"document_create_date.1\"].min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2019-11-28 00:00:00')"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data[\"document_create_date.1\"].max()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It ranges for 334 days, almost less than a year"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### invoice_currency\n",
    "The currency of the invoice amount in the document for the invoice"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "USD    34365\n",
       "CAD     2886\n",
       "Name: invoice_currency, dtype: int64"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data[\"invoice_currency\"].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As there are two currencies involved, we can convert either one them. However, as there is not much difference between USD and CAD, there is not much use of converting them."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### total_open_amount\n",
    "The amount that is yet to be paid for that invoice"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count      37251.000000\n",
       "mean       32245.973581\n",
       "std        40201.159144\n",
       "min            2.400000\n",
       "25%         4604.675000\n",
       "50%        17260.020000\n",
       "75%        46953.185000\n",
       "max      1011888.450000\n",
       "Name: total_open_amount, dtype: object"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data['total_open_amount'].describe().apply(lambda x: format(x,'f'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Automatically binning the amount using pd.qcut."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2.399, 801.76]           3726\n",
       "(82056.05, 1011888.45]    3725\n",
       "(55340.22, 82056.05]      3725\n",
       "(39301.59, 55340.22]      3725\n",
       "(25241.13, 39301.59]      3725\n",
       "(17260.02, 25241.13]      3725\n",
       "(12714.26, 17260.02]      3725\n",
       "(6921.84, 12714.26]       3725\n",
       "(2882.6, 6921.84]         3725\n",
       "(801.76, 2882.6]          3725\n",
       "Name: total_open_amount, dtype: int64"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "open_amount_bins = pd.qcut(train_data['total_open_amount'],q=10)\n",
    "open_amount_bins.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "amount_bins = [0,5000,10000,50000,100000,1100000]\n",
    "new_open_amount_bin = pd.cut(train_data['total_open_amount'], bins = amount_bins)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='total_open_amount', ylabel='count'>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xticks(fontsize=10, rotation=90)\n",
    "sns.countplot(x=open_amount_bins, palette=\"hls\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The number of companies in each distribution is same."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### baseline_create_date\n",
    "The date on which the Invoice was created."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140105847</td>\n",
       "      <td>GORDON F in</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-12</td>\n",
       "      <td>CAD</td>\n",
       "      <td>18061.92</td>\n",
       "      <td>2019-01-02</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140105686</td>\n",
       "      <td>SYSC co</td>\n",
       "      <td>2019-01-17</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-31</td>\n",
       "      <td>2019-01-13</td>\n",
       "      <td>CAD</td>\n",
       "      <td>24557.30</td>\n",
       "      <td>2019-01-03</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140104429</td>\n",
       "      <td>COSTCO llc</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2018-12-31</td>\n",
       "      <td>2019-01-13</td>\n",
       "      <td>CAD</td>\n",
       "      <td>4375.60</td>\n",
       "      <td>2019-01-03</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>183</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140104429</td>\n",
       "      <td>COSTCO corporation</td>\n",
       "      <td>2019-01-18</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-01-02</td>\n",
       "      <td>2019-01-19</td>\n",
       "      <td>CAD</td>\n",
       "      <td>19318.06</td>\n",
       "      <td>2019-01-09</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140104340</td>\n",
       "      <td>COLABOR  corporation</td>\n",
       "      <td>2019-01-17</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-01-02</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>CAD</td>\n",
       "      <td>18294.15</td>\n",
       "      <td>2019-01-04</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37185</th>\n",
       "      <td>U001</td>\n",
       "      <td>0200592182</td>\n",
       "      <td>DECA in</td>\n",
       "      <td>2019-12-02</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-27</td>\n",
       "      <td>2019-12-09</td>\n",
       "      <td>USD</td>\n",
       "      <td>70.80</td>\n",
       "      <td>2019-11-16</td>\n",
       "      <td>NAM4</td>\n",
       "      <td>0</td>\n",
       "      <td>-7.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37202</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0100034820</td>\n",
       "      <td>MTY TIKI llc</td>\n",
       "      <td>2019-12-06</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-27</td>\n",
       "      <td>2019-12-09</td>\n",
       "      <td>CAD</td>\n",
       "      <td>13630.00</td>\n",
       "      <td>2019-11-29</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37203</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140104429</td>\n",
       "      <td>COSTCO corp</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-27</td>\n",
       "      <td>2019-12-09</td>\n",
       "      <td>CAD</td>\n",
       "      <td>19751.31</td>\n",
       "      <td>2019-11-29</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37254</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140104440</td>\n",
       "      <td>SO corp</td>\n",
       "      <td>2019-12-19</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-16</td>\n",
       "      <td>CAD</td>\n",
       "      <td>272728.34</td>\n",
       "      <td>2019-12-06</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37262</th>\n",
       "      <td>CA02</td>\n",
       "      <td>0140106408</td>\n",
       "      <td>WAL-M foundation</td>\n",
       "      <td>2019-12-19</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-10</td>\n",
       "      <td>CAD</td>\n",
       "      <td>25489.96</td>\n",
       "      <td>2019-11-30</td>\n",
       "      <td>CA10</td>\n",
       "      <td>0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3983 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      business_code cust_number         name_customer clear_date  \\\n",
       "23             CA02  0140105847           GORDON F in 2019-01-14   \n",
       "94             CA02  0140105686               SYSC co 2019-01-17   \n",
       "133            CA02  0140104429            COSTCO llc 2019-01-14   \n",
       "183            CA02  0140104429    COSTCO corporation 2019-01-18   \n",
       "185            CA02  0140104340  COLABOR  corporation 2019-01-17   \n",
       "...             ...         ...                   ...        ...   \n",
       "37185          U001  0200592182               DECA in 2019-12-02   \n",
       "37202          CA02  0100034820          MTY TIKI llc 2019-12-06   \n",
       "37203          CA02  0140104429           COSTCO corp 2019-12-13   \n",
       "37254          CA02  0140104440               SO corp 2019-12-19   \n",
       "37262          CA02  0140106408      WAL-M foundation 2019-12-19   \n",
       "\n",
       "       buisness_year document_create_date.1 due_in_date invoice_currency  \\\n",
       "23            2019.0             2018-12-30  2019-01-12              CAD   \n",
       "94            2019.0             2018-12-31  2019-01-13              CAD   \n",
       "133           2019.0             2018-12-31  2019-01-13              CAD   \n",
       "183           2019.0             2019-01-02  2019-01-19              CAD   \n",
       "185           2019.0             2019-01-02  2019-01-14              CAD   \n",
       "...              ...                    ...         ...              ...   \n",
       "37185         2019.0             2019-11-27  2019-12-09              USD   \n",
       "37202         2019.0             2019-11-27  2019-12-09              CAD   \n",
       "37203         2019.0             2019-11-27  2019-12-09              CAD   \n",
       "37254         2019.0             2019-11-28  2019-12-16              CAD   \n",
       "37262         2019.0             2019-11-28  2019-12-10              CAD   \n",
       "\n",
       "       total_open_amount baseline_create_date cust_payment_terms  isOpen  \\\n",
       "23              18061.92           2019-01-02               CA10       0   \n",
       "94              24557.30           2019-01-03               CA10       0   \n",
       "133              4375.60           2019-01-03               CA10       0   \n",
       "183             19318.06           2019-01-09               CA10       0   \n",
       "185             18294.15           2019-01-04               CA10       0   \n",
       "...                  ...                  ...                ...     ...   \n",
       "37185              70.80           2019-11-16               NAM4       0   \n",
       "37202           13630.00           2019-11-29               CA10       0   \n",
       "37203           19751.31           2019-11-29               CA10       0   \n",
       "37254          272728.34           2019-12-06               CA10       0   \n",
       "37262           25489.96           2019-11-30               CA10       0   \n",
       "\n",
       "       delay  \n",
       "23       2.0  \n",
       "94       4.0  \n",
       "133      1.0  \n",
       "183     -1.0  \n",
       "185      3.0  \n",
       "...      ...  \n",
       "37185   -7.0  \n",
       "37202   -3.0  \n",
       "37203    4.0  \n",
       "37254    3.0  \n",
       "37262    9.0  \n",
       "\n",
       "[3983 rows x 13 columns]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.loc[train_data[\"document_create_date.1\"]!=train_data[\"baseline_create_date\"]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Baseline create date and document create date are same except if the company is canadian there is a delay in filing the document and creating an invoice. \n",
    "\n",
    "It's not that important to our model so we'll drop it."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### cust_payment_terms\n",
    "Business terms and agreements between customers and accounts on discounts and days of payment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NAA8    15127\n",
       "NAH4     9756\n",
       "CA10     2815\n",
       "NAC6     1174\n",
       "NAM4     1012\n",
       "        ...  \n",
       "NAUW        1\n",
       "B030        1\n",
       "NATL        1\n",
       "NAVC        1\n",
       "NAD8        1\n",
       "Name: cust_payment_terms, Length: 68, dtype: int64"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.cust_payment_terms.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1359"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(dict(train_data.cust_number.value_counts()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It is same as cust_number."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### isOpen\n",
    "- Tells whether a transaction is open or closed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    37251\n",
       "Name: isOpen, dtype: int64"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.isOpen.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "isOpen is 0 for all the train set as all the invoices has been closed. So we can drop this column."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Delay\n",
    "The negative delay specifies that the amount was cleared before the due date and thus there was no delay. Therefore we have capped the negative values to 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13430"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.delay[train_data.delay<0].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data['delay']=train_data.delay.apply(lambda x: 0 if x<0 else x)\n",
    "train_data.delay[train_data.delay<0].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    37251.000000\n",
       "mean         2.822126\n",
       "std          9.368047\n",
       "min          0.000000\n",
       "25%          0.000000\n",
       "50%          0.000000\n",
       "75%          2.000000\n",
       "max        277.000000\n",
       "Name: delay, dtype: float64"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.delay.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MULTIVARIATE ANALYSIS"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Relation b/w business_code with invoice_currency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "USD    34365\n",
       "CAD     2886\n",
       "Name: invoice_currency, dtype: int64"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data['invoice_currency'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "business_code  invoice_currency\n",
       "CA02           CAD                  2884\n",
       "U001           USD                 33713\n",
       "U002           USD                   124\n",
       "U005           USD                     8\n",
       "U007           USD                     2\n",
       "U013           USD                   518\n",
       "               CAD                     2\n",
       "Name: invoice_currency, dtype: int64"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.groupby(\"business_code\").invoice_currency.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that business code first letter describes in which currency the transaction took place except for 2 entries. Let's explore them further"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "invoice_currency  business_code\n",
       "CAD               CA02              2884\n",
       "                  U013                 2\n",
       "USD               U001             33713\n",
       "                  U013               518\n",
       "                  U002               124\n",
       "                  U005                 8\n",
       "                  U007                 2\n",
       "Name: business_code, dtype: int64"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.groupby(\"invoice_currency\").business_code.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "business_code U013 is an US based company but paying in CAD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "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>business_code</th>\n",
       "      <th>cust_number</th>\n",
       "      <th>name_customer</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>buisness_year</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>invoice_currency</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>baseline_create_date</th>\n",
       "      <th>cust_payment_terms</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6441</th>\n",
       "      <td>U013</td>\n",
       "      <td>0100052318</td>\n",
       "      <td>PRATT' trust</td>\n",
       "      <td>2019-03-26</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-03-01</td>\n",
       "      <td>2019-03-31</td>\n",
       "      <td>CAD</td>\n",
       "      <td>1815.04</td>\n",
       "      <td>2019-03-01</td>\n",
       "      <td>NAVE</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37009</th>\n",
       "      <td>U013</td>\n",
       "      <td>0100052318</td>\n",
       "      <td>PRATT' co</td>\n",
       "      <td>2019-12-24</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>2019-11-26</td>\n",
       "      <td>2019-12-26</td>\n",
       "      <td>CAD</td>\n",
       "      <td>3373.05</td>\n",
       "      <td>2019-11-26</td>\n",
       "      <td>NAVE</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      business_code cust_number name_customer clear_date  buisness_year  \\\n",
       "6441           U013  0100052318  PRATT' trust 2019-03-26         2019.0   \n",
       "37009          U013  0100052318     PRATT' co 2019-12-24         2019.0   \n",
       "\n",
       "      document_create_date.1 due_in_date invoice_currency  total_open_amount  \\\n",
       "6441              2019-03-01  2019-03-31              CAD            1815.04   \n",
       "37009             2019-11-26  2019-12-26              CAD            3373.05   \n",
       "\n",
       "      baseline_create_date cust_payment_terms  isOpen  delay  \n",
       "6441            2019-03-01               NAVE       0    0.0  \n",
       "37009           2019-11-26               NAVE       0    0.0  "
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp = train_data.loc[train_data['business_code'] == 'U013']\n",
    "temp.loc[temp['invoice_currency']=='CAD']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that only PRATT company is the only US based company paying in CAD.\n",
    "\n",
    "Morever they are only 2 transactions so it's not as significant. They don't even have any delay so this won't affect our model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Relation between total_open_amount and delay"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>delay</th>\n",
       "      <th>(0, 15]</th>\n",
       "      <th>(15, 30]</th>\n",
       "      <th>(30, 45]</th>\n",
       "      <th>(45, 60]</th>\n",
       "      <th>(60, 300]</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_open_amount</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>(0, 5000]</th>\n",
       "      <td>3226</td>\n",
       "      <td>130</td>\n",
       "      <td>109</td>\n",
       "      <td>102</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>(5000, 10000]</th>\n",
       "      <td>1467</td>\n",
       "      <td>76</td>\n",
       "      <td>61</td>\n",
       "      <td>36</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>(10000, 50000]</th>\n",
       "      <td>6223</td>\n",
       "      <td>293</td>\n",
       "      <td>183</td>\n",
       "      <td>113</td>\n",
       "      <td>66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>(50000, 100000]</th>\n",
       "      <td>2507</td>\n",
       "      <td>68</td>\n",
       "      <td>19</td>\n",
       "      <td>12</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>(100000, 1100000]</th>\n",
       "      <td>1152</td>\n",
       "      <td>20</td>\n",
       "      <td>12</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "delay              (0, 15]  (15, 30]  (30, 45]  (45, 60]  (60, 300]\n",
       "total_open_amount                                                  \n",
       "(0, 5000]             3226       130       109       102         98\n",
       "(5000, 10000]         1467        76        61        36         20\n",
       "(10000, 50000]        6223       293       183       113         66\n",
       "(50000, 100000]       2507        68        19        12         15\n",
       "(100000, 1100000]     1152        20        12         8          3"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "delay_bins = [0,15,30,45,60,300]\n",
    "delay_bucket = pd.cut(train_data['delay'], bins = delay_bins)\n",
    "pd.crosstab(index = new_open_amount_bin, columns =delay_bucket)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='total_open_amount', ylabel='count'>"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,5))\n",
    "plt.xlabel(\"Amount bin\")\n",
    "sns.countplot(x = new_open_amount_bin, hue=delay_bucket)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see from the graph for each bucket the amount was returned in 0-15 delay days."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "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>buisness_year</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>isOpen</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>buisness_year</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_open_amount</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.047619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>isOpen</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>delay</th>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.047619</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   buisness_year  total_open_amount  isOpen     delay\n",
       "buisness_year                NaN                NaN     NaN       NaN\n",
       "total_open_amount            NaN           1.000000     NaN -0.047619\n",
       "isOpen                       NaN                NaN     NaN       NaN\n",
       "delay                        NaN          -0.047619     NaN  1.000000"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x1eb6f757220>"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x1152 with 20 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.pairplot(train_data, height=4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can infer: The total open amount decreses as the delay increases. There is negative co-relation."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# FEATURE ENGINEERING"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "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>cust_number</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0200769623</td>\n",
       "      <td>2019-01-09</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>25749.68</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0200762301</td>\n",
       "      <td>2019-01-15</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>18116.87</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0200955827</td>\n",
       "      <td>2019-01-16</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>32856.32</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0200744019</td>\n",
       "      <td>2019-01-18</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>11982.59</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0200794332</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>32715.47</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37266</th>\n",
       "      <td>0200769623</td>\n",
       "      <td>2019-12-09</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>31752.94</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37267</th>\n",
       "      <td>0200744019</td>\n",
       "      <td>2019-12-12</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>45591.41</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37268</th>\n",
       "      <td>0200769623</td>\n",
       "      <td>2019-12-16</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>61336.27</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37269</th>\n",
       "      <td>0140105847</td>\n",
       "      <td>2019-12-17</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-08</td>\n",
       "      <td>13937.60</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37270</th>\n",
       "      <td>0200726979</td>\n",
       "      <td>2019-12-16</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>445.79</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>37251 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      cust_number clear_date document_create_date.1 due_in_date  \\\n",
       "0      0200769623 2019-01-09             2018-12-30  2019-01-14   \n",
       "1      0200762301 2019-01-15             2018-12-30  2019-01-14   \n",
       "2      0200955827 2019-01-16             2018-12-30  2019-01-14   \n",
       "3      0200744019 2019-01-18             2018-12-30  2019-01-14   \n",
       "4      0200794332 2019-01-14             2018-12-30  2019-01-14   \n",
       "...           ...        ...                    ...         ...   \n",
       "37266  0200769623 2019-12-09             2019-11-28  2019-12-13   \n",
       "37267  0200744019 2019-12-12             2019-11-28  2019-12-13   \n",
       "37268  0200769623 2019-12-16             2019-11-28  2019-12-13   \n",
       "37269  0140105847 2019-12-17             2019-11-28  2019-12-08   \n",
       "37270  0200726979 2019-12-16             2019-11-28  2019-12-13   \n",
       "\n",
       "       total_open_amount  delay  \n",
       "0               25749.68    0.0  \n",
       "1               18116.87    1.0  \n",
       "2               32856.32    2.0  \n",
       "3               11982.59    4.0  \n",
       "4               32715.47    0.0  \n",
       "...                  ...    ...  \n",
       "37266           31752.94    0.0  \n",
       "37267           45591.41    0.0  \n",
       "37268           61336.27    3.0  \n",
       "37269           13937.60    9.0  \n",
       "37270             445.79    3.0  \n",
       "\n",
       "[37251 rows x 6 columns]"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.drop(columns=['business_code', 'name_customer','buisness_year','invoice_currency','baseline_create_date', 'cust_payment_terms', 'isOpen'], inplace=True)\n",
    "train_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Extracting day and month from document_create_date.1 and due_in_date."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data[\"doc_create_day\"] = train_data[\"document_create_date.1\"].dt.day\n",
    "train_data[\"doc_create_month\"] = train_data[\"document_create_date.1\"].dt.month\n",
    "\n",
    "train_data[\"due_day\"] = train_data[\"due_in_date\"].dt.day\n",
    "train_data[\"due_month\"] = train_data[\"due_in_date\"].dt.month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "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>cust_number</th>\n",
       "      <th>clear_date</th>\n",
       "      <th>document_create_date.1</th>\n",
       "      <th>due_in_date</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>delay</th>\n",
       "      <th>doc_create_day</th>\n",
       "      <th>doc_create_month</th>\n",
       "      <th>due_day</th>\n",
       "      <th>due_month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0200769623</td>\n",
       "      <td>2019-01-09</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>25749.68</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0200762301</td>\n",
       "      <td>2019-01-15</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>18116.87</td>\n",
       "      <td>1.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0200955827</td>\n",
       "      <td>2019-01-16</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>32856.32</td>\n",
       "      <td>2.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0200744019</td>\n",
       "      <td>2019-01-18</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>11982.59</td>\n",
       "      <td>4.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0200794332</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>2018-12-30</td>\n",
       "      <td>2019-01-14</td>\n",
       "      <td>32715.47</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37266</th>\n",
       "      <td>0200769623</td>\n",
       "      <td>2019-12-09</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>31752.94</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37267</th>\n",
       "      <td>0200744019</td>\n",
       "      <td>2019-12-12</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>45591.41</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37268</th>\n",
       "      <td>0200769623</td>\n",
       "      <td>2019-12-16</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>61336.27</td>\n",
       "      <td>3.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37269</th>\n",
       "      <td>0140105847</td>\n",
       "      <td>2019-12-17</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-08</td>\n",
       "      <td>13937.60</td>\n",
       "      <td>9.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>8</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37270</th>\n",
       "      <td>0200726979</td>\n",
       "      <td>2019-12-16</td>\n",
       "      <td>2019-11-28</td>\n",
       "      <td>2019-12-13</td>\n",
       "      <td>445.79</td>\n",
       "      <td>3.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>37251 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      cust_number clear_date document_create_date.1 due_in_date  \\\n",
       "0      0200769623 2019-01-09             2018-12-30  2019-01-14   \n",
       "1      0200762301 2019-01-15             2018-12-30  2019-01-14   \n",
       "2      0200955827 2019-01-16             2018-12-30  2019-01-14   \n",
       "3      0200744019 2019-01-18             2018-12-30  2019-01-14   \n",
       "4      0200794332 2019-01-14             2018-12-30  2019-01-14   \n",
       "...           ...        ...                    ...         ...   \n",
       "37266  0200769623 2019-12-09             2019-11-28  2019-12-13   \n",
       "37267  0200744019 2019-12-12             2019-11-28  2019-12-13   \n",
       "37268  0200769623 2019-12-16             2019-11-28  2019-12-13   \n",
       "37269  0140105847 2019-12-17             2019-11-28  2019-12-08   \n",
       "37270  0200726979 2019-12-16             2019-11-28  2019-12-13   \n",
       "\n",
       "       total_open_amount  delay  doc_create_day  doc_create_month  due_day  \\\n",
       "0               25749.68    0.0              30                12       14   \n",
       "1               18116.87    1.0              30                12       14   \n",
       "2               32856.32    2.0              30                12       14   \n",
       "3               11982.59    4.0              30                12       14   \n",
       "4               32715.47    0.0              30                12       14   \n",
       "...                  ...    ...             ...               ...      ...   \n",
       "37266           31752.94    0.0              28                11       13   \n",
       "37267           45591.41    0.0              28                11       13   \n",
       "37268           61336.27    3.0              28                11       13   \n",
       "37269           13937.60    9.0              28                11        8   \n",
       "37270             445.79    3.0              28                11       13   \n",
       "\n",
       "       due_month  \n",
       "0              1  \n",
       "1              1  \n",
       "2              1  \n",
       "3              1  \n",
       "4              1  \n",
       "...          ...  \n",
       "37266         12  \n",
       "37267         12  \n",
       "37268         12  \n",
       "37269         12  \n",
       "37270         12  \n",
       "\n",
       "[37251 rows x 10 columns]"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "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>cust_number</th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>delay</th>\n",
       "      <th>doc_create_day</th>\n",
       "      <th>doc_create_month</th>\n",
       "      <th>due_day</th>\n",
       "      <th>due_month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0200769623</td>\n",
       "      <td>25749.68</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0200762301</td>\n",
       "      <td>18116.87</td>\n",
       "      <td>1.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0200955827</td>\n",
       "      <td>32856.32</td>\n",
       "      <td>2.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0200744019</td>\n",
       "      <td>11982.59</td>\n",
       "      <td>4.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0200794332</td>\n",
       "      <td>32715.47</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37266</th>\n",
       "      <td>0200769623</td>\n",
       "      <td>31752.94</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37267</th>\n",
       "      <td>0200744019</td>\n",
       "      <td>45591.41</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37268</th>\n",
       "      <td>0200769623</td>\n",
       "      <td>61336.27</td>\n",
       "      <td>3.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37269</th>\n",
       "      <td>0140105847</td>\n",
       "      <td>13937.60</td>\n",
       "      <td>9.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>8</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37270</th>\n",
       "      <td>0200726979</td>\n",
       "      <td>445.79</td>\n",
       "      <td>3.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>37251 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      cust_number  total_open_amount  delay  doc_create_day  doc_create_month  \\\n",
       "0      0200769623           25749.68    0.0              30                12   \n",
       "1      0200762301           18116.87    1.0              30                12   \n",
       "2      0200955827           32856.32    2.0              30                12   \n",
       "3      0200744019           11982.59    4.0              30                12   \n",
       "4      0200794332           32715.47    0.0              30                12   \n",
       "...           ...                ...    ...             ...               ...   \n",
       "37266  0200769623           31752.94    0.0              28                11   \n",
       "37267  0200744019           45591.41    0.0              28                11   \n",
       "37268  0200769623           61336.27    3.0              28                11   \n",
       "37269  0140105847           13937.60    9.0              28                11   \n",
       "37270  0200726979             445.79    3.0              28                11   \n",
       "\n",
       "       due_day  due_month  \n",
       "0           14          1  \n",
       "1           14          1  \n",
       "2           14          1  \n",
       "3           14          1  \n",
       "4           14          1  \n",
       "...        ...        ...  \n",
       "37266       13         12  \n",
       "37267       13         12  \n",
       "37268       13         12  \n",
       "37269        8         12  \n",
       "37270       13         12  \n",
       "\n",
       "[37251 rows x 7 columns]"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.drop(columns=['clear_date', 'document_create_date.1', 'due_in_date'], inplace=True)\n",
    "train_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Extracting the integer part from cust_number"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>delay</th>\n",
       "      <th>doc_create_day</th>\n",
       "      <th>doc_create_month</th>\n",
       "      <th>due_day</th>\n",
       "      <th>due_month</th>\n",
       "      <th>cn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>25749.68</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>18116.87</td>\n",
       "      <td>1.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>0200762301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>32856.32</td>\n",
       "      <td>2.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>0200955827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11982.59</td>\n",
       "      <td>4.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>0200744019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>32715.47</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>0200794332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37266</th>\n",
       "      <td>31752.94</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37267</th>\n",
       "      <td>45591.41</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>0200744019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37268</th>\n",
       "      <td>61336.27</td>\n",
       "      <td>3.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37269</th>\n",
       "      <td>13937.60</td>\n",
       "      <td>9.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>8</td>\n",
       "      <td>12</td>\n",
       "      <td>0140105847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37270</th>\n",
       "      <td>445.79</td>\n",
       "      <td>3.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>0200726979</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>37251 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       total_open_amount  delay  doc_create_day  doc_create_month  due_day  \\\n",
       "0               25749.68    0.0              30                12       14   \n",
       "1               18116.87    1.0              30                12       14   \n",
       "2               32856.32    2.0              30                12       14   \n",
       "3               11982.59    4.0              30                12       14   \n",
       "4               32715.47    0.0              30                12       14   \n",
       "...                  ...    ...             ...               ...      ...   \n",
       "37266           31752.94    0.0              28                11       13   \n",
       "37267           45591.41    0.0              28                11       13   \n",
       "37268           61336.27    3.0              28                11       13   \n",
       "37269           13937.60    9.0              28                11        8   \n",
       "37270             445.79    3.0              28                11       13   \n",
       "\n",
       "       due_month          cn  \n",
       "0              1  0200769623  \n",
       "1              1  0200762301  \n",
       "2              1  0200955827  \n",
       "3              1  0200744019  \n",
       "4              1  0200794332  \n",
       "...          ...         ...  \n",
       "37266         12  0200769623  \n",
       "37267         12  0200744019  \n",
       "37268         12  0200769623  \n",
       "37269         12  0140105847  \n",
       "37270         12  0200726979  \n",
       "\n",
       "[37251 rows x 7 columns]"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data['cn'] = train_data['cust_number'].str.extract('(\\d+)')\n",
    "\n",
    "# then will drop that column\n",
    "train_data.drop('cust_number',axis=1,inplace=True)\n",
    "\n",
    "#lets see the dataframe\n",
    "train_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 37251 entries, 0 to 37270\n",
      "Data columns (total 7 columns):\n",
      " #   Column             Non-Null Count  Dtype  \n",
      "---  ------             --------------  -----  \n",
      " 0   total_open_amount  37251 non-null  float64\n",
      " 1   delay              37251 non-null  float64\n",
      " 2   doc_create_day     37251 non-null  int64  \n",
      " 3   doc_create_month   37251 non-null  int64  \n",
      " 4   due_day            37251 non-null  int64  \n",
      " 5   due_month          37251 non-null  int64  \n",
      " 6   cn                 37251 non-null  object \n",
      "dtypes: float64(2), int64(4), object(1)\n",
      "memory usage: 3.5+ MB\n"
     ]
    }
   ],
   "source": [
    "train_data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Converting the data-type of cn as float"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 37251 entries, 0 to 37270\n",
      "Data columns (total 7 columns):\n",
      " #   Column             Non-Null Count  Dtype  \n",
      "---  ------             --------------  -----  \n",
      " 0   total_open_amount  37251 non-null  float64\n",
      " 1   delay              37251 non-null  float64\n",
      " 2   doc_create_day     37251 non-null  int64  \n",
      " 3   doc_create_month   37251 non-null  int64  \n",
      " 4   due_day            37251 non-null  int64  \n",
      " 5   due_month          37251 non-null  int64  \n",
      " 6   cn                 37251 non-null  float64\n",
      "dtypes: float64(3), int64(4)\n",
      "memory usage: 3.5 MB\n"
     ]
    }
   ],
   "source": [
    "train_data['cn']=train_data['cn'].astype(str).astype(float)\n",
    "train_data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Splitting into x_train and y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        0.0\n",
       "1        1.0\n",
       "2        2.0\n",
       "3        4.0\n",
       "4        0.0\n",
       "        ... \n",
       "37266    0.0\n",
       "37267    0.0\n",
       "37268    3.0\n",
       "37269    9.0\n",
       "37270    3.0\n",
       "Name: delay, Length: 37251, dtype: float64"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train = train_data['delay']\n",
    "y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "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>total_open_amount</th>\n",
       "      <th>doc_create_day</th>\n",
       "      <th>doc_create_month</th>\n",
       "      <th>due_day</th>\n",
       "      <th>due_month</th>\n",
       "      <th>cn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>25749.68</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>18116.87</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>200762301.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>32856.32</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>200955827.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11982.59</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>200744019.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>32715.47</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>200794332.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37266</th>\n",
       "      <td>31752.94</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37267</th>\n",
       "      <td>45591.41</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>200744019.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37268</th>\n",
       "      <td>61336.27</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37269</th>\n",
       "      <td>13937.60</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>8</td>\n",
       "      <td>12</td>\n",
       "      <td>140105847.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37270</th>\n",
       "      <td>445.79</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>200726979.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>37251 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       total_open_amount  doc_create_day  doc_create_month  due_day  \\\n",
       "0               25749.68              30                12       14   \n",
       "1               18116.87              30                12       14   \n",
       "2               32856.32              30                12       14   \n",
       "3               11982.59              30                12       14   \n",
       "4               32715.47              30                12       14   \n",
       "...                  ...             ...               ...      ...   \n",
       "37266           31752.94              28                11       13   \n",
       "37267           45591.41              28                11       13   \n",
       "37268           61336.27              28                11       13   \n",
       "37269           13937.60              28                11        8   \n",
       "37270             445.79              28                11       13   \n",
       "\n",
       "       due_month           cn  \n",
       "0              1  200769623.0  \n",
       "1              1  200762301.0  \n",
       "2              1  200955827.0  \n",
       "3              1  200744019.0  \n",
       "4              1  200794332.0  \n",
       "...          ...          ...  \n",
       "37266         12  200769623.0  \n",
       "37267         12  200744019.0  \n",
       "37268         12  200769623.0  \n",
       "37269         12  140105847.0  \n",
       "37270         12  200726979.0  \n",
       "\n",
       "[37251 rows x 6 columns]"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train = train_data.drop([\"delay\"], axis=1).copy()\n",
    "x_train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Now we need to repeat the same process on val1, val2 and x_test."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 3246 entries, 0 to 3245\n",
      "Data columns (total 13 columns):\n",
      " #   Column                  Non-Null Count  Dtype         \n",
      "---  ------                  --------------  -----         \n",
      " 0   business_code           3246 non-null   object        \n",
      " 1   cust_number             3246 non-null   object        \n",
      " 2   name_customer           3246 non-null   object        \n",
      " 3   clear_date              3246 non-null   datetime64[ns]\n",
      " 4   buisness_year           3246 non-null   float64       \n",
      " 5   document_create_date.1  3246 non-null   datetime64[ns]\n",
      " 6   due_in_date             3246 non-null   datetime64[ns]\n",
      " 7   invoice_currency        3246 non-null   object        \n",
      " 8   total_open_amount       3246 non-null   float64       \n",
      " 9   baseline_create_date    3246 non-null   datetime64[ns]\n",
      " 10  cust_payment_terms      3246 non-null   object        \n",
      " 11  isOpen                  3246 non-null   int64         \n",
      " 12  delay                   3246 non-null   float64       \n",
      "dtypes: datetime64[ns](4), float64(3), int64(1), object(5)\n",
      "memory usage: 329.8+ KB\n"
     ]
    }
   ],
   "source": [
    "val1.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "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>total_open_amount</th>\n",
       "      <th>delay</th>\n",
       "      <th>doc_create_day</th>\n",
       "      <th>doc_create_month</th>\n",
       "      <th>due_day</th>\n",
       "      <th>due_month</th>\n",
       "      <th>cn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>21748.48</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>0200763152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>70698.34</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>897.46</td>\n",
       "      <td>5.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2367.26</td>\n",
       "      <td>3.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>16</td>\n",
       "      <td>12</td>\n",
       "      <td>0140106408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>34027.91</td>\n",
       "      <td>3.0</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>0200744019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3241</th>\n",
       "      <td>2020.42</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>0200726979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3242</th>\n",
       "      <td>256.91</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3243</th>\n",
       "      <td>256.91</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3244</th>\n",
       "      <td>1118.71</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3245</th>\n",
       "      <td>135465.20</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>0200707822</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3246 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      total_open_amount  delay  doc_create_day  doc_create_month  due_day  \\\n",
       "0              21748.48    0.0              28                11       13   \n",
       "1              70698.34    0.0              28                11       13   \n",
       "2                897.46    5.0              28                11       13   \n",
       "3               2367.26    3.0              28                11       16   \n",
       "4              34027.91    3.0              28                11       13   \n",
       "...                 ...    ...             ...               ...      ...   \n",
       "3241            2020.42    0.0              12                 1       27   \n",
       "3242             256.91    0.0              12                 1       27   \n",
       "3243             256.91    0.0              12                 1       27   \n",
       "3244            1118.71    0.0              12                 1       27   \n",
       "3245          135465.20    0.0              12                 1       27   \n",
       "\n",
       "      due_month          cn  \n",
       "0            12  0200763152  \n",
       "1            12  0200769623  \n",
       "2            12  0200769623  \n",
       "3            12  0140106408  \n",
       "4            12  0200744019  \n",
       "...         ...         ...  \n",
       "3241          1  0200726979  \n",
       "3242          1  0200769623  \n",
       "3243          1  0200769623  \n",
       "3244          1  0200769623  \n",
       "3245          1  0200707822  \n",
       "\n",
       "[3246 rows x 7 columns]"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val1['delay']=val1.delay.apply(lambda x: 0 if x<0 else x)\n",
    "val1.drop(['business_code', 'name_customer', 'clear_date', 'buisness_year', 'invoice_currency', 'baseline_create_date', 'cust_payment_terms', 'isOpen'], axis=1, inplace=True)\n",
    "\n",
    "val1[\"doc_create_day\"] = val1[\"document_create_date.1\"].dt.day\n",
    "val1[\"doc_create_month\"] = val1[\"document_create_date.1\"].dt.month\n",
    "\n",
    "val1[\"due_day\"] = val1[\"due_in_date\"].dt.day\n",
    "val1[\"due_month\"] = val1[\"due_in_date\"].dt.month\n",
    "\n",
    "val1['cn'] = val1['cust_number'].str.extract('(\\d+)')\n",
    "val1.drop('cust_number',axis=1,inplace=True)\n",
    "\n",
    "val1.drop(columns=['document_create_date.1', 'due_in_date'], inplace=True)\n",
    "val1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 3246 entries, 0 to 3245\n",
      "Data columns (total 7 columns):\n",
      " #   Column             Non-Null Count  Dtype  \n",
      "---  ------             --------------  -----  \n",
      " 0   total_open_amount  3246 non-null   float64\n",
      " 1   delay              3246 non-null   float64\n",
      " 2   doc_create_day     3246 non-null   int64  \n",
      " 3   doc_create_month   3246 non-null   int64  \n",
      " 4   due_day            3246 non-null   int64  \n",
      " 5   due_month          3246 non-null   int64  \n",
      " 6   cn                 3246 non-null   float64\n",
      "dtypes: float64(3), int64(4)\n",
      "memory usage: 177.6 KB\n"
     ]
    }
   ],
   "source": [
    "val1['cn']=val1['cn'].astype(str).astype(float)\n",
    "val1.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Splitting into x_val and y_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       0.0\n",
       "1       0.0\n",
       "2       5.0\n",
       "3       3.0\n",
       "4       3.0\n",
       "       ... \n",
       "3241    0.0\n",
       "3242    0.0\n",
       "3243    0.0\n",
       "3244    0.0\n",
       "3245    0.0\n",
       "Name: delay, Length: 3246, dtype: float64"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_val1 = val1['delay']\n",
    "y_val1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "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>total_open_amount</th>\n",
       "      <th>doc_create_day</th>\n",
       "      <th>doc_create_month</th>\n",
       "      <th>due_day</th>\n",
       "      <th>due_month</th>\n",
       "      <th>cn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>21748.48</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>200763152.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>70698.34</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>897.46</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2367.26</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>16</td>\n",
       "      <td>12</td>\n",
       "      <td>140106408.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>34027.91</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>12</td>\n",
       "      <td>200744019.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3241</th>\n",
       "      <td>2020.42</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>200726979.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3242</th>\n",
       "      <td>256.91</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3243</th>\n",
       "      <td>256.91</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3244</th>\n",
       "      <td>1118.71</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3245</th>\n",
       "      <td>135465.20</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>200707822.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3246 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      total_open_amount  doc_create_day  doc_create_month  due_day  due_month  \\\n",
       "0              21748.48              28                11       13         12   \n",
       "1              70698.34              28                11       13         12   \n",
       "2                897.46              28                11       13         12   \n",
       "3               2367.26              28                11       16         12   \n",
       "4              34027.91              28                11       13         12   \n",
       "...                 ...             ...               ...      ...        ...   \n",
       "3241            2020.42              12                 1       27          1   \n",
       "3242             256.91              12                 1       27          1   \n",
       "3243             256.91              12                 1       27          1   \n",
       "3244            1118.71              12                 1       27          1   \n",
       "3245          135465.20              12                 1       27          1   \n",
       "\n",
       "               cn  \n",
       "0     200763152.0  \n",
       "1     200769623.0  \n",
       "2     200769623.0  \n",
       "3     140106408.0  \n",
       "4     200744019.0  \n",
       "...           ...  \n",
       "3241  200726979.0  \n",
       "3242  200769623.0  \n",
       "3243  200769623.0  \n",
       "3244  200769623.0  \n",
       "3245  200707822.0  \n",
       "\n",
       "[3246 rows x 6 columns]"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_val1 = val1.drop([\"delay\"], axis=1).copy()\n",
    "x_val1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "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>total_open_amount</th>\n",
       "      <th>delay</th>\n",
       "      <th>doc_create_day</th>\n",
       "      <th>doc_create_month</th>\n",
       "      <th>due_day</th>\n",
       "      <th>due_month</th>\n",
       "      <th>cn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32547.72</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>0200789077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4388.75</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6040.68</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>71525.23</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>0200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>231.24</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>0200117492</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4654</th>\n",
       "      <td>2878.29</td>\n",
       "      <td>11.0</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>0140106408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4655</th>\n",
       "      <td>10770.76</td>\n",
       "      <td>0.0</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0200721222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4656</th>\n",
       "      <td>39408.22</td>\n",
       "      <td>3.0</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>3</td>\n",
       "      <td>0200744707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4657</th>\n",
       "      <td>70504.55</td>\n",
       "      <td>0.0</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>3</td>\n",
       "      <td>0200759878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4658</th>\n",
       "      <td>90195.02</td>\n",
       "      <td>2.0</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>18</td>\n",
       "      <td>3</td>\n",
       "      <td>0200700919</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4659 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      total_open_amount  delay  doc_create_day  doc_create_month  due_day  \\\n",
       "0              32547.72    0.0              13                 1       28   \n",
       "1               4388.75    0.0              13                 1       28   \n",
       "2               6040.68    0.0              13                 1       28   \n",
       "3              71525.23    0.0              13                 1       28   \n",
       "4                231.24    0.0              13                 1       28   \n",
       "...                 ...    ...             ...               ...      ...   \n",
       "4654            2878.29   11.0              27                 2        9   \n",
       "4655           10770.76    0.0              27                 2        2   \n",
       "4656           39408.22    3.0              27                 2       13   \n",
       "4657           70504.55    0.0              27                 2       13   \n",
       "4658           90195.02    2.0              27                 2       18   \n",
       "\n",
       "      due_month          cn  \n",
       "0             1  0200789077  \n",
       "1             1  0200769623  \n",
       "2             1  0200769623  \n",
       "3             1  0200769623  \n",
       "4             1  0200117492  \n",
       "...         ...         ...  \n",
       "4654          3  0140106408  \n",
       "4655          5  0200721222  \n",
       "4656          3  0200744707  \n",
       "4657          3  0200759878  \n",
       "4658          3  0200700919  \n",
       "\n",
       "[4659 rows x 7 columns]"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "val2['delay']=val2.delay.apply(lambda x: 0 if x<0 else x)\n",
    "val2.drop(['business_code', 'name_customer', 'clear_date', 'buisness_year', 'invoice_currency', 'baseline_create_date', 'cust_payment_terms', 'isOpen'], axis=1, inplace=True)\n",
    "\n",
    "val2[\"doc_create_day\"] = val2[\"document_create_date.1\"].dt.day\n",
    "val2[\"doc_create_month\"] = val2[\"document_create_date.1\"].dt.month\n",
    "\n",
    "val2[\"due_day\"] = val2[\"due_in_date\"].dt.day\n",
    "val2[\"due_month\"] = val2[\"due_in_date\"].dt.month\n",
    "\n",
    "val2['cn'] = val2['cust_number'].str.extract('(\\d+)')\n",
    "val2.drop('cust_number',axis=1,inplace=True)\n",
    "\n",
    "val2.drop(columns=['document_create_date.1', 'due_in_date'], inplace=True)\n",
    "val2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4659 entries, 0 to 4658\n",
      "Data columns (total 7 columns):\n",
      " #   Column             Non-Null Count  Dtype  \n",
      "---  ------             --------------  -----  \n",
      " 0   total_open_amount  4659 non-null   float64\n",
      " 1   delay              4659 non-null   float64\n",
      " 2   doc_create_day     4659 non-null   int64  \n",
      " 3   doc_create_month   4659 non-null   int64  \n",
      " 4   due_day            4659 non-null   int64  \n",
      " 5   due_month          4659 non-null   int64  \n",
      " 6   cn                 4659 non-null   float64\n",
      "dtypes: float64(3), int64(4)\n",
      "memory usage: 254.9 KB\n"
     ]
    }
   ],
   "source": [
    "val2['cn']=val2['cn'].astype(str).astype(float)\n",
    "val2.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        0.0\n",
       "1        0.0\n",
       "2        0.0\n",
       "3        0.0\n",
       "4        0.0\n",
       "        ... \n",
       "4654    11.0\n",
       "4655     0.0\n",
       "4656     3.0\n",
       "4657     0.0\n",
       "4658     2.0\n",
       "Name: delay, Length: 4659, dtype: float64"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_val2 = val2['delay']\n",
    "y_val2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "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>total_open_amount</th>\n",
       "      <th>doc_create_day</th>\n",
       "      <th>doc_create_month</th>\n",
       "      <th>due_day</th>\n",
       "      <th>due_month</th>\n",
       "      <th>cn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32547.72</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>200789077.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4388.75</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6040.68</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>71525.23</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>200769623.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>231.24</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>200117492.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4654</th>\n",
       "      <td>2878.29</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>140106408.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4655</th>\n",
       "      <td>10770.76</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>200721222.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4656</th>\n",
       "      <td>39408.22</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>3</td>\n",
       "      <td>200744707.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4657</th>\n",
       "      <td>70504.55</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>3</td>\n",
       "      <td>200759878.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4658</th>\n",
       "      <td>90195.02</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>18</td>\n",
       "      <td>3</td>\n",
       "      <td>200700919.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4659 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      total_open_amount  doc_create_day  doc_create_month  due_day  due_month  \\\n",
       "0              32547.72              13                 1       28          1   \n",
       "1               4388.75              13                 1       28          1   \n",
       "2               6040.68              13                 1       28          1   \n",
       "3              71525.23              13                 1       28          1   \n",
       "4                231.24              13                 1       28          1   \n",
       "...                 ...             ...               ...      ...        ...   \n",
       "4654            2878.29              27                 2        9          3   \n",
       "4655           10770.76              27                 2        2          5   \n",
       "4656           39408.22              27                 2       13          3   \n",
       "4657           70504.55              27                 2       13          3   \n",
       "4658           90195.02              27                 2       18          3   \n",
       "\n",
       "               cn  \n",
       "0     200789077.0  \n",
       "1     200769623.0  \n",
       "2     200769623.0  \n",
       "3     200769623.0  \n",
       "4     200117492.0  \n",
       "...           ...  \n",
       "4654  140106408.0  \n",
       "4655  200721222.0  \n",
       "4656  200744707.0  \n",
       "4657  200759878.0  \n",
       "4658  200700919.0  \n",
       "\n",
       "[4659 rows x 6 columns]"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_val2 = val2.drop([\"delay\"], axis=1).copy()\n",
    "x_val2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_open_amount</th>\n",
       "      <th>doc_create_day</th>\n",
       "      <th>doc_create_month</th>\n",
       "      <th>due_day</th>\n",
       "      <th>due_month</th>\n",
       "      <th>cn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>16157.16</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>3</td>\n",
       "      <td>0200772595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>722.50</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>0140105847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4206.28</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>3</td>\n",
       "      <td>0200744019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>51877.23</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>3</td>\n",
       "      <td>0200943275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>365.88</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>3</td>\n",
       "      <td>0200726979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4814</th>\n",
       "      <td>1226.00</td>\n",
       "      <td>12</td>\n",
       "      <td>5</td>\n",
       "      <td>24</td>\n",
       "      <td>5</td>\n",
       "      <td>0200592182</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4815</th>\n",
       "      <td>36513.74</td>\n",
       "      <td>12</td>\n",
       "      <td>5</td>\n",
       "      <td>27</td>\n",
       "      <td>5</td>\n",
       "      <td>200769623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4816</th>\n",
       "      <td>15220.00</td>\n",
       "      <td>13</td>\n",
       "      <td>5</td>\n",
       "      <td>28</td>\n",
       "      <td>5</td>\n",
       "      <td>0200915438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4817</th>\n",
       "      <td>17893.25</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>5</td>\n",
       "      <td>100030964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4818</th>\n",
       "      <td>8326.87</td>\n",
       "      <td>19</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>0200915438</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4819 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      total_open_amount  doc_create_day  doc_create_month  due_day  due_month  \\\n",
       "0              16157.16              27                 2       13          3   \n",
       "1                722.50              27                 2       10          3   \n",
       "2               4206.28              27                 2       13          3   \n",
       "3              51877.23              27                 2       13          3   \n",
       "4                365.88              27                 2       13          3   \n",
       "...                 ...             ...               ...      ...        ...   \n",
       "4814            1226.00              12                 5       24          5   \n",
       "4815           36513.74              12                 5       27          5   \n",
       "4816           15220.00              13                 5       28          5   \n",
       "4817           17893.25              16                 5       31          5   \n",
       "4818            8326.87              19                 5        3          6   \n",
       "\n",
       "              cn  \n",
       "0     0200772595  \n",
       "1     0140105847  \n",
       "2     0200744019  \n",
       "3     0200943275  \n",
       "4     0200726979  \n",
       "...          ...  \n",
       "4814  0200592182  \n",
       "4815   200769623  \n",
       "4816  0200915438  \n",
       "4817   100030964  \n",
       "4818  0200915438  \n",
       "\n",
       "[4819 rows x 6 columns]"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test.drop(['business_code', 'name_customer', 'clear_date', 'buisness_year', 'invoice_currency', 'baseline_create_date', 'cust_payment_terms', 'isOpen'], axis=1, inplace=True)\n",
    "\n",
    "x_test[\"doc_create_day\"] = x_test[\"document_create_date.1\"].dt.day\n",
    "x_test[\"doc_create_month\"] = x_test[\"document_create_date.1\"].dt.month\n",
    "\n",
    "x_test[\"due_day\"] = x_test[\"due_in_date\"].dt.day\n",
    "x_test[\"due_month\"] = x_test[\"due_in_date\"].dt.month\n",
    "\n",
    "x_test['cn'] = x_test['cust_number'].str.extract('(\\d+)')\n",
    "x_test.drop('cust_number',axis=1,inplace=True)\n",
    "\n",
    "x_test.drop(columns=['document_create_date.1', 'due_in_date'], inplace=True)\n",
    "x_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4819 entries, 0 to 4818\n",
      "Data columns (total 5 columns):\n",
      " #   Column             Non-Null Count  Dtype  \n",
      "---  ------             --------------  -----  \n",
      " 0   total_open_amount  4819 non-null   float64\n",
      " 1   doc_create_day     4819 non-null   int64  \n",
      " 2   doc_create_month   4819 non-null   int64  \n",
      " 3   due_day            4819 non-null   int64  \n",
      " 4   cn                 4819 non-null   float64\n",
      "dtypes: float64(2), int64(3)\n",
      "memory usage: 188.4 KB\n"
     ]
    }
   ],
   "source": [
    "x_test['cn']=x_test['cn'].astype(str).astype(float)\n",
    "x_test.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# FEATURE SELECTION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(14,10))\n",
    "cor = train_data.corr()\n",
    "sns.heatmap(cor,cmap = 'viridis',annot=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see from the above graph due_month and doc_create_month are highly correlated and thus they will have a negative impact on our model. So drop remove due_month. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = x_train.drop([\"due_month\"], axis=1).copy()\n",
    "x_val1 = x_val1.drop([\"due_month\"], axis=1).copy()\n",
    "x_val2 = x_val2.drop([\"due_month\"], axis=1).copy()\n",
    "x_test = x_test.drop([\"due_month\"], axis=1).copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MODELING"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As this is a regression model we will use the following models:-\n",
    "- Linear Regressor\n",
    "- SVR\n",
    "- Decision Tree\n",
    "- Random Forest\n",
    "- XgBoost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "#fitting the model \n",
    "\n",
    "clf = xgb.XGBRegressor()\n",
    "clf.fit(x_train, y_train)\n",
    "\n",
    "# Predicting the Validation Set Results\n",
    "predicted = clf.predict(x_val1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error = 1.82\n",
      "Mean squared error = 18.13\n",
      "Median absolute error = 0.85\n",
      "Explain variance score = 0.51\n",
      "R2 score = 0.51\n",
      "Accuracy=  0.5067037516860886\n"
     ]
    }
   ],
   "source": [
    "# predicting all types of error and accuracy\n",
    "\n",
    "print(\"Mean absolute error =\", round(sm.mean_absolute_error(y_val1, predicted), 2)) \n",
    "print(\"Mean squared error =\", round(sm.mean_squared_error(y_val1, predicted), 2)) \n",
    "print(\"Median absolute error =\", round(sm.median_absolute_error(y_val1, predicted), 2)) \n",
    "print(\"Explain variance score =\", round(sm.explained_variance_score(y_val1, predicted), 2)) \n",
    "print(\"R2 score =\", round(sm.r2_score(y_val1, predicted), 2))\n",
    "print(\"Accuracy= \", clf.score(x_val1,y_val1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf2 = RandomForestRegressor()\n",
    "clf2.fit(x_train, y_train)\n",
    "\n",
    "# Predicting the Validation Set Results\n",
    "predicted2 = clf2.predict(x_val1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error = 2.08\n",
      "Mean squared error = 19.05\n",
      "Median absolute error = 1.03\n",
      "Explain variance score = 0.51\n",
      "R2 score = 0.48\n",
      "Accuracy=  0.481575048144291\n"
     ]
    }
   ],
   "source": [
    "# predicting all types of error and accuracy\n",
    "\n",
    "print(\"Mean absolute error =\", round(sm.mean_absolute_error(y_val1, predicted2), 2)) \n",
    "print(\"Mean squared error =\", round(sm.mean_squared_error(y_val1, predicted2), 2)) \n",
    "print(\"Median absolute error =\", round(sm.median_absolute_error(y_val1, predicted2), 2)) \n",
    "print(\"Explain variance score =\", round(sm.explained_variance_score(y_val1, predicted2), 2)) \n",
    "print(\"R2 score =\", round(sm.r2_score(y_val1, predicted2), 2))\n",
    "print(\"Accuracy= \", clf2.score(x_val1,y_val1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### DecisionTreeRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fitting Decision Tree Regressor to the Training Set\n",
    "clf3 = DecisionTreeRegressor()\n",
    "clf3.fit(x_train, y_train)\n",
    "\n",
    "# Predicting the Validation Set Results\n",
    "predicted3 = clf3.predict(x_val1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error = 2.24\n",
      "Mean squared error = 53.16\n",
      "Median absolute error = 1.0\n",
      "Explain variance score = -0.43\n",
      "R2 score = -0.45\n",
      "Accuracy=  -0.44636587721378373\n"
     ]
    }
   ],
   "source": [
    "# predicting all types of error and accuracy\n",
    "\n",
    "print(\"Mean absolute error =\", round(sm.mean_absolute_error(y_val1, predicted3), 2)) \n",
    "print(\"Mean squared error =\", round(sm.mean_squared_error(y_val1, predicted3), 2)) \n",
    "print(\"Median absolute error =\", round(sm.median_absolute_error(y_val1, predicted3), 2)) \n",
    "print(\"Explain variance score =\", round(sm.explained_variance_score(y_val1, predicted3), 2)) \n",
    "print(\"R2 score =\", round(sm.r2_score(y_val1, predicted3), 2))\n",
    "print(\"Accuracy= \", clf3.score(x_val1,y_val1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### LinearRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fitting Simple Linear Regression to the Training Set\n",
    "clf5 = LinearRegression()\n",
    "clf5.fit(x_train, y_train)\n",
    "\n",
    "# Predicting the Test Set Results\n",
    "predicted5 = clf.predict(x_val1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error = 1.82\n",
      "Mean squared error = 18.13\n",
      "Median absolute error = 0.85\n",
      "Explain variance score = 0.51\n",
      "R2 score = 0.51\n",
      "Accuracy=  0.32265869429822236\n"
     ]
    }
   ],
   "source": [
    "# predicting all types of error and accuracy\n",
    "\n",
    "print(\"Mean absolute error =\", round(sm.mean_absolute_error(y_val1, predicted5), 2)) \n",
    "print(\"Mean squared error =\", round(sm.mean_squared_error(y_val1, predicted5), 2)) \n",
    "print(\"Median absolute error =\", round(sm.median_absolute_error(y_val1, predicted5), 2)) \n",
    "print(\"Explain variance score =\", round(sm.explained_variance_score(y_val1, predicted5), 2)) \n",
    "print(\"R2 score =\", round(sm.r2_score(y_val1, predicted5), 2))\n",
    "print(\"Accuracy= \", clf5.score(x_val1,y_val1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We'll continue with XgBoost as it gives us the best result from all the other models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### HYPER-PARAMETER TUNING"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "# now we will hypertune our parameters for better and accurate results and avoiding overfitting\n",
    "\n",
    "# we will set some parameter\n",
    "params={\n",
    " \"learning_rate\"    : [0.05, 0.10, 0.15, 0.20, 0.25, 0.30 ] ,\n",
    " \"max_depth\"        : [ 3, 4, 5, 6, 8, 10, 12, 15],\n",
    " \"min_child_weight\" : [ 1, 3, 5, 7 ],\n",
    " \"gamma\"            : [ 0.0, 0.1, 0.2 , 0.3, 0.4 ],\n",
    " \"colsample_bytree\" : [ 0.3, 0.4, 0.5 , 0.7 ]\n",
    "    \n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this we will fit and check for best parameters for our model using RandomizedSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "reg=xgb.XGBRegressor()\n",
    "random_search=RandomizedSearchCV(reg,param_distributions=params,n_iter=5,n_jobs=-1,cv=5,verbose=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 5 candidates, totalling 25 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  19 out of  25 | elapsed:   16.8s remaining:    5.2s\n",
      "[Parallel(n_jobs=-1)]: Done  25 out of  25 | elapsed:   18.7s finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomizedSearchCV(cv=5,\n",
       "                   estimator=XGBRegressor(base_score=None, booster=None,\n",
       "                                          colsample_bylevel=None,\n",
       "                                          colsample_bynode=None,\n",
       "                                          colsample_bytree=None, gamma=None,\n",
       "                                          gpu_id=None, importance_type='gain',\n",
       "                                          interaction_constraints=None,\n",
       "                                          learning_rate=None,\n",
       "                                          max_delta_step=None, max_depth=None,\n",
       "                                          min_child_weight=None, missing=nan,\n",
       "                                          monotone_constraints=None,\n",
       "                                          n_estimators=100, n...\n",
       "                                          random_state=None, reg_alpha=None,\n",
       "                                          reg_lambda=None,\n",
       "                                          scale_pos_weight=None, subsample=None,\n",
       "                                          tree_method=None,\n",
       "                                          validate_parameters=None,\n",
       "                                          verbosity=None),\n",
       "                   n_iter=5, n_jobs=-1,\n",
       "                   param_distributions={'colsample_bytree': [0.3, 0.4, 0.5,\n",
       "                                                             0.7],\n",
       "                                        'gamma': [0.0, 0.1, 0.2, 0.3, 0.4],\n",
       "                                        'learning_rate': [0.05, 0.1, 0.15, 0.2,\n",
       "                                                          0.25, 0.3],\n",
       "                                        'max_depth': [3, 4, 5, 6, 8, 10, 12,\n",
       "                                                      15],\n",
       "                                        'min_child_weight': [1, 3, 5, 7]},\n",
       "                   verbose=3)"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_search.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking for best parameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBRegressor(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "             colsample_bynode=1, colsample_bytree=0.7, gamma=0.1, gpu_id=-1,\n",
       "             importance_type='gain', interaction_constraints='',\n",
       "             learning_rate=0.05, max_delta_step=0, max_depth=5,\n",
       "             min_child_weight=7, missing=nan, monotone_constraints='()',\n",
       "             n_estimators=100, n_jobs=8, num_parallel_tree=1, random_state=0,\n",
       "             reg_alpha=0, reg_lambda=1, scale_pos_weight=1, subsample=1,\n",
       "             tree_method='exact', validate_parameters=1, verbosity=None)"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_search.best_estimator_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using these parameters and checking against val2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "cl = xgb.XGBRegressor(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
    "             colsample_bynode=1, colsample_bytree=0.7, gamma=0.2, gpu_id=-1,\n",
    "             importance_type='gain', interaction_constraints='',\n",
    "             learning_rate=0.2, max_delta_step=0, max_depth=8,\n",
    "             min_child_weight=3, monotone_constraints='()',\n",
    "             n_estimators=100, n_jobs=8, num_parallel_tree=1, random_state=0,\n",
    "             reg_alpha=0, reg_lambda=1, scale_pos_weight=1, subsample=1,\n",
    "             tree_method='exact', validate_parameters=1, verbosity=None)\n",
    "cl.fit(x_train, y_train)\n",
    "# Predicting the Test Set Results\n",
    "predicted = cl.predict(x_val2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error = 2.34\n",
      "Mean squared error = 30.33\n",
      "Median absolute error = 1.16\n",
      "Explain variance score = 0.49\n",
      "R2 score = 0.49\n",
      "Accuracy=  0.4925351710531721\n"
     ]
    }
   ],
   "source": [
    "# predicting all types of error and accuracy\n",
    "\n",
    "print(\"Mean absolute error =\", round(sm.mean_absolute_error(y_val2, predicted), 2)) \n",
    "print(\"Mean squared error =\", round(sm.mean_squared_error(y_val2, predicted), 2)) \n",
    "print(\"Median absolute error =\", round(sm.median_absolute_error(y_val2, predicted), 2)) \n",
    "print(\"Explain variance score =\", round(sm.explained_variance_score(y_val2, predicted), 2)) \n",
    "print(\"R2 score =\", round(sm.r2_score(y_val2, predicted), 2))\n",
    "print(\"Accuracy= \", cl.score(x_val2,y_val2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TEST"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "predicted1 = cl.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 7, 3, ..., 3, 9, 4])"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predicted1=np.around(predicted1)\n",
    "predicted1.astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4819"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(predicted1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We got the delay on test set."
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "Creating a new dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "Col=pd.DataFrame()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Appending the delay predicted column in it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "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>Delay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4814</th>\n",
       "      <td>1 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4815</th>\n",
       "      <td>0 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4816</th>\n",
       "      <td>3 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4817</th>\n",
       "      <td>9 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4818</th>\n",
       "      <td>4 days</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4819 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Delay\n",
       "0    1 days\n",
       "1    7 days\n",
       "2    3 days\n",
       "3    1 days\n",
       "4    1 days\n",
       "...     ...\n",
       "4814 1 days\n",
       "4815 0 days\n",
       "4816 3 days\n",
       "4817 9 days\n",
       "4818 4 days\n",
       "\n",
       "[4819 rows x 1 columns]"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Delay=[]\n",
    "for x in predicted1:\n",
    "    Delay.append(pd.Timedelta(days=x))\n",
    "Col['Delay'] = Delay\n",
    "Col"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Predicting the clear date"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "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>Delay</th>\n",
       "      <th>clear_date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1 days</td>\n",
       "      <td>2020-03-14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7 days</td>\n",
       "      <td>2020-03-17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3 days</td>\n",
       "      <td>2020-03-16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1 days</td>\n",
       "      <td>2020-03-14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1 days</td>\n",
       "      <td>2020-03-14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4814</th>\n",
       "      <td>1 days</td>\n",
       "      <td>2020-05-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4815</th>\n",
       "      <td>0 days</td>\n",
       "      <td>2020-05-27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4816</th>\n",
       "      <td>3 days</td>\n",
       "      <td>2020-05-31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4817</th>\n",
       "      <td>9 days</td>\n",
       "      <td>2020-06-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4818</th>\n",
       "      <td>4 days</td>\n",
       "      <td>2020-06-07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4819 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Delay clear_date\n",
       "0    1 days 2020-03-14\n",
       "1    7 days 2020-03-17\n",
       "2    3 days 2020-03-16\n",
       "3    1 days 2020-03-14\n",
       "4    1 days 2020-03-14\n",
       "...     ...        ...\n",
       "4814 1 days 2020-05-25\n",
       "4815 0 days 2020-05-27\n",
       "4816 3 days 2020-05-31\n",
       "4817 9 days 2020-06-09\n",
       "4818 4 days 2020-06-07\n",
       "\n",
       "[4819 rows x 2 columns]"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Col['clear_date'] = ts['due_in_date']+Col['Delay']\n",
    "Col"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creating the Aging Bucket"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "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>clear_date</th>\n",
       "      <th>Aging Bucket</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2020-03-14</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2020-03-17</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2020-03-16</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2020-03-14</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2020-03-14</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4814</th>\n",
       "      <td>2020-05-25</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4815</th>\n",
       "      <td>2020-05-27</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4816</th>\n",
       "      <td>2020-05-31</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4817</th>\n",
       "      <td>2020-06-09</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4818</th>\n",
       "      <td>2020-06-07</td>\n",
       "      <td>0-15days</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4819 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     clear_date Aging Bucket\n",
       "0    2020-03-14     0-15days\n",
       "1    2020-03-17     0-15days\n",
       "2    2020-03-16     0-15days\n",
       "3    2020-03-14     0-15days\n",
       "4    2020-03-14     0-15days\n",
       "...         ...          ...\n",
       "4814 2020-05-25     0-15days\n",
       "4815 2020-05-27     0-15days\n",
       "4816 2020-05-31     0-15days\n",
       "4817 2020-06-09     0-15days\n",
       "4818 2020-06-07     0-15days\n",
       "\n",
       "[4819 rows x 2 columns]"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "aging_bucket = []\n",
    "for x in predicted1:\n",
    "    if x<=15:\n",
    "        aging_bucket.append(\"0-15days\")\n",
    "    elif x<=30:\n",
    "        aging_bucket.append(\"16-30days\")\n",
    "    elif x<=45:\n",
    "        aging_bucket.append(\"31-45days\")\n",
    "    elif x<=60:\n",
    "        aging_bucket.append(\"46-60days\")\n",
    "    else:\n",
    "        aging_bucket.append(\"Greater than 60 days\")\n",
    "Col['Aging Bucket']= aging_bucket\n",
    "Col.drop(['Delay'],axis=1,inplace=True)\n",
    "Col"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}