K UNITY SACCO EDA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a9332dee",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(149321, 8) \n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>OutstandingLoan</th>\n",
       "      <th>terms</th>\n",
       "      <th>disbursedamount</th>\n",
       "      <th>installmentstartdate</th>\n",
       "      <th>Lastinstallmentdate</th>\n",
       "      <th>LastInstallmentAmount</th>\n",
       "      <th>disbursementdate</th>\n",
       "      <th>Gender</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-18134.96</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2022-01-19</td>\n",
       "      <td>M</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-6769.98</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-01</td>\n",
       "      <td>F</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-10000.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-11-30</td>\n",
       "      <td>F</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-30000.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-11-26</td>\n",
       "      <td>F</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1269.98</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-11-26</td>\n",
       "      <td>M</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   OutstandingLoan  terms  disbursedamount installmentstartdate  \\\n",
       "0        -18134.96      1              0.0           2022-02-18   \n",
       "1         -6769.98      1              0.0           2021-12-31   \n",
       "2        -10000.00      1              0.0           2021-12-30   \n",
       "3        -30000.00      1              0.0           2021-12-26   \n",
       "4         -1269.98      1              0.0           2021-12-26   \n",
       "\n",
       "  Lastinstallmentdate  LastInstallmentAmount disbursementdate Gender  \n",
       "0          2022-02-18                    0.0       2022-01-19      M  \n",
       "1          2021-12-31                    0.0       2021-12-01      F  \n",
       "2          2021-12-30                    0.0       2021-11-30      F  \n",
       "3          2021-12-26                    0.0       2021-11-26      F  \n",
       "4          2021-12-26                    0.0       2021-11-26      M  "
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from datetime import timedelta, date \n",
    "import seaborn as sns\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler, MinMaxScaler #standardization and normalization\n",
    "scaler=StandardScaler()\n",
    "\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.cluster import KMeans, k_means\n",
    "\n",
    "#Load the data\n",
    "df=pd.read_excel(\"KCA Loan data.xlsx\")\n",
    "print(df.shape,\"\\n\")\n",
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "31b8e16e",
   "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>OutstandingLoan</th>\n",
       "      <th>terms</th>\n",
       "      <th>disbursedamount</th>\n",
       "      <th>installmentstartdate</th>\n",
       "      <th>Lastinstallmentdate</th>\n",
       "      <th>LastInstallmentAmount</th>\n",
       "      <th>disbursementdate</th>\n",
       "      <th>Gender</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-18134.96</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2022-01-19</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-6769.98</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-10000.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-11-30</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-30000.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-11-26</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1269.98</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2021-11-26</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   OutstandingLoan  terms  disbursedamount installmentstartdate  \\\n",
       "0        -18134.96      1              0.0           2022-02-18   \n",
       "1         -6769.98      1              0.0           2021-12-31   \n",
       "2        -10000.00      1              0.0           2021-12-30   \n",
       "3        -30000.00      1              0.0           2021-12-26   \n",
       "4         -1269.98      1              0.0           2021-12-26   \n",
       "\n",
       "  Lastinstallmentdate  LastInstallmentAmount disbursementdate Gender  \n",
       "0          2022-02-18                    0.0       2022-01-19      1  \n",
       "1          2021-12-31                    0.0       2021-12-01      0  \n",
       "2          2021-12-30                    0.0       2021-11-30      0  \n",
       "3          2021-12-26                    0.0       2021-11-26      0  \n",
       "4          2021-12-26                    0.0       2021-11-26      1  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#We format some the Gender Column to be more explanatory\n",
    "df.loc[df.Gender==\"M\", \"Gender\"]=1\n",
    "df.loc[df.Gender==\"F\", \"Gender\"]=0\n",
    "df.loc[df.Gender==\"U\", \"Gender\"]=2\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e934e8d3",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 149321 entries, 0 to 149320\n",
      "Data columns (total 8 columns):\n",
      " #   Column                 Non-Null Count   Dtype         \n",
      "---  ------                 --------------   -----         \n",
      " 0   OutstandingLoan        149321 non-null  float64       \n",
      " 1   terms                  149321 non-null  int64         \n",
      " 2   disbursedamount        149321 non-null  float64       \n",
      " 3   installmentstartdate   149285 non-null  datetime64[ns]\n",
      " 4   Lastinstallmentdate    149320 non-null  datetime64[ns]\n",
      " 5   LastInstallmentAmount  146885 non-null  float64       \n",
      " 6   disbursementdate       149274 non-null  datetime64[ns]\n",
      " 7   Gender                 149321 non-null  object        \n",
      "dtypes: datetime64[ns](3), float64(3), int64(1), object(1)\n",
      "memory usage: 9.1+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15053880",
   "metadata": {},
   "source": [
    "#### From the above data information, we can see that the data types are ok and may not need anyr reworking"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33cfbfcf",
   "metadata": {},
   "source": [
    "#### Handling missing values\n",
    "Let us see the columns with missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "56fa16a8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OutstandingLoan             0\n",
       "terms                       0\n",
       "disbursedamount             0\n",
       "installmentstartdate       36\n",
       "Lastinstallmentdate         1\n",
       "LastInstallmentAmount    2436\n",
       "disbursementdate           47\n",
       "Gender                      0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "97443249",
   "metadata": {},
   "source": [
    "#### installmentstartdate column has 36 missing values\n",
    "   we will impute this by adding 30 days to the **disbursmentdate** assuming that one starts paying one month after disbursment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "10645150",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df[\"installmentstartdate\"]=df.installmentstartdate.fillna(df.disbursementdate + timedelta(days=30))\n",
    "#df=df.astype({'installmentstartdate': 'datetime64[ns]'})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "207a9e0f",
   "metadata": {},
   "source": [
    "#### disbursmentdate column has 47 missing values\n",
    "   we will impute this by subtracting 30 days from  **installmentstartdate** assuming that one starts paying one month after disbursment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "f83dc845",
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"disbursementdate\"]=df.disbursementdate.fillna(df.installmentstartdate - timedelta(days=30))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "639d9a1a",
   "metadata": {},
   "source": [
    "#### Lastinstallmentdate column has 1 missing values\n",
    "   we will impute this  with **installmentstartdate**+30days since this was transaction was offered for 1 term only meaning one month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "4f23b7bc",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df[\"Lastinstallmentdate\"]=df.Lastinstallmentdate.fillna(df.installmentstartdate+timedelta(days=30))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1e7c53f",
   "metadata": {},
   "source": [
    "#### LastInstallmentAmount column has 2436 missing values\n",
    "This is quite alot of missing values. We are not willing to drop all those rows of data. So we will fill all those missing values with values zero assuming that it means there was NILL payment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ae073732",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OutstandingLoan          0\n",
       "terms                    0\n",
       "disbursedamount          0\n",
       "installmentstartdate     0\n",
       "Lastinstallmentdate      0\n",
       "LastInstallmentAmount    0\n",
       "disbursementdate         0\n",
       "Gender                   0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"LastInstallmentAmount\"]=df.LastInstallmentAmount.fillna(0)\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16e2d6d1",
   "metadata": {},
   "source": [
    "# Let us check the defaulters!\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ba076a69",
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime\n",
    "from datetime import timedelta as dt\n",
    "\n",
    "df[\"defaulter\"]=\"\"\n",
    "\n",
    "#date today minus when the persons started paying the installments\n",
    "df[\"terms_lapsed\"]=(datetime.datetime.now()) #this can be misleading if data was not extracted recently\n",
    "df[\"terms_lapsed\"]=df[\"terms_lapsed\"]-df.installmentstartdate\n",
    "\n",
    "#extract days from timedelta\n",
    "df[\"terms_lapsed\"]=pd.to_timedelta(df[\"terms_lapsed\"],errors=\"coerce\").dt.days \n",
    "\n",
    "#divide the days by 30 days of the month to get approximate months that have lapsed\n",
    "df[\"terms_lapsed\"]=df.terms_lapsed/30 \n",
    "\n",
    "#Get all those whose term has lapsed also includes..:\n",
    "#Those whose term lapsed, they paid last installment, yet they still have balances\n",
    "df.loc[((df.terms_lapsed)>(df.terms) | (((df.terms_lapsed)==(df.terms)) & (df.OutstandingLoan>0)  ) ,\"defaulter\")]= 1\n",
    "\n",
    "#Get all those whose terms are not yet over\n",
    "df.loc[(df.terms_lapsed)<(df.terms),\"defaulter\"]= 0 \n",
    "\n",
    "#get all those whose terms havent lapsed but they did not pay last installment\n",
    "df.loc[(df.terms_lapsed)<(df.terms) & (df.LastInstallmentAmount==0)  ,\"defaulter\"]= 2 #unknown"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "06b97e69",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Create a ratio of Disbursed amount and Outstanding loan\n",
    "df[\"disbursed_outstanding_ratio\"]=df[\"disbursedamount\"]/abs(df[\"OutstandingLoan\"])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c0cf8dd9",
   "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>OutstandingLoan</th>\n",
       "      <th>installmentstartdate</th>\n",
       "      <th>Lastinstallmentdate</th>\n",
       "      <th>disbursementdate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-18134.96</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>2022-01-19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-6769.98</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>2021-12-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-10000.00</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>2021-11-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-30000.00</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-11-26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1269.98</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-11-26</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   OutstandingLoan installmentstartdate Lastinstallmentdate disbursementdate\n",
       "0        -18134.96           2022-02-18          2022-02-18       2022-01-19\n",
       "1         -6769.98           2021-12-31          2021-12-31       2021-12-01\n",
       "2        -10000.00           2021-12-30          2021-12-30       2021-11-30\n",
       "3        -30000.00           2021-12-26          2021-12-26       2021-11-26\n",
       "4         -1269.98           2021-12-26          2021-12-26       2021-11-26"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#get date columns to attached later to the transformed df\n",
    "df_cols=df[['OutstandingLoan','installmentstartdate','Lastinstallmentdate','disbursementdate']]\n",
    "df_cols.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "e9c6ee2c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'for col in cols:\\n    print(\"For column {}, the unique values are: {}\".format(col,df[col].unique()),\"\\n\",\"\\n\",\"***\"*40)'"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Function to get the ranges\n",
    "\"\"\"\n",
    "\n",
    "#create a copy dataframe before we transform our columns\n",
    "df_copy=df.copy()\n",
    "\n",
    "cols=['OutstandingLoan', 'disbursedamount','LastInstallmentAmount','disbursed_outstanding_ratio','terms','terms_lapsed']\n",
    "def create_bins(df):\n",
    "    for column in cols:\n",
    "        results=pd.qcut(df[column],q=3)#labels=['low','middle','upper']\n",
    "        df[column]=results\n",
    "    return df\n",
    "create_bins(df)\"\"\"\n",
    "\n",
    "\"\"\"for col in cols:\n",
    "    print(\"For column {}, the unique values are: {}\".format(col,df[col].unique()),\"\\n\",\"\\n\",\"***\"*40)\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c1cfb263",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "1c962c7f",
   "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>OutstandingLoan</th>\n",
       "      <th>terms</th>\n",
       "      <th>disbursedamount</th>\n",
       "      <th>installmentstartdate</th>\n",
       "      <th>Lastinstallmentdate</th>\n",
       "      <th>LastInstallmentAmount</th>\n",
       "      <th>disbursementdate</th>\n",
       "      <th>Gender</th>\n",
       "      <th>defaulter</th>\n",
       "      <th>terms_lapsed</th>\n",
       "      <th>disbursed_outstanding_ratio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>0</td>\n",
       "      <td>2022-01-19</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>0</td>\n",
       "      <td>2021-12-01</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>0</td>\n",
       "      <td>2021-11-30</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>0</td>\n",
       "      <td>2021-11-26</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>0</td>\n",
       "      <td>2021-11-26</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   OutstandingLoan  terms  disbursedamount installmentstartdate  \\\n",
       "0                2      0                0           2022-02-18   \n",
       "1                2      0                0           2021-12-31   \n",
       "2                2      0                0           2021-12-30   \n",
       "3                2      0                0           2021-12-26   \n",
       "4                2      0                0           2021-12-26   \n",
       "\n",
       "  Lastinstallmentdate  LastInstallmentAmount disbursementdate Gender  \\\n",
       "0          2022-02-18                      0       2022-01-19      1   \n",
       "1          2021-12-31                      0       2021-12-01      0   \n",
       "2          2021-12-30                      0       2021-11-30      0   \n",
       "3          2021-12-26                      0       2021-11-26      0   \n",
       "4          2021-12-26                      0       2021-11-26      1   \n",
       "\n",
       "  defaulter  terms_lapsed  disbursed_outstanding_ratio  \n",
       "0         1             0                            0  \n",
       "1         1             0                            0  \n",
       "2         1             0                            0  \n",
       "3         1             0                            0  \n",
       "4         1             0                            0  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols=['OutstandingLoan', 'disbursedamount','LastInstallmentAmount','disbursed_outstanding_ratio','terms','terms_lapsed']\n",
    "def create_bins(df):\n",
    "    for column in cols:\n",
    "        results=pd.qcut(df[column],q=3,labels=False)#labels=['low','middle','upper']\n",
    "        df[column]=results\n",
    "    return df\n",
    "create_bins(df)\n",
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "ac27fcf7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For column OutstandingLoan, the unique values are: [2 0 1] \n",
      " \n",
      " ************************************************************************************************************************\n",
      "For column disbursedamount, the unique values are: [0 1 2] \n",
      " \n",
      " ************************************************************************************************************************\n",
      "For column LastInstallmentAmount, the unique values are: [0 1 2] \n",
      " \n",
      " ************************************************************************************************************************\n",
      "For column disbursed_outstanding_ratio, the unique values are: [0 1 2] \n",
      " \n",
      " ************************************************************************************************************************\n",
      "For column terms, the unique values are: [0 2 1] \n",
      " \n",
      " ************************************************************************************************************************\n",
      "For column terms_lapsed, the unique values are: [0 1 2] \n",
      " \n",
      " ************************************************************************************************************************\n"
     ]
    }
   ],
   "source": [
    "for col in cols:\n",
    "    print(\"For column {}, the unique values are: {}\".format(col,df[col].unique()),\"\\n\",\"\\n\",\"***\"*40)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "85762ab9",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEICAYAAAB1f3LfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAXuElEQVR4nO3dfbRddZ3f8fdHghAfwvBwoZhAgyUzDrBEJGXiY8XYIbbTgbGgmeoQNctYZMaxna5ZMF2t42rTkdXOUJkKa6iMBHyALEYl1WJlBR1rB8EgKM8lBQtpEMKDCI6gid/+cX5XTy43lxN2zj0c7vu11llnn+/Zv32/m7vC5/723mefVBWSJD1bLxh1A5Kk8WaQSJI6MUgkSZ0YJJKkTgwSSVInBokkqZN5o25gth100EG1ePHiUbchSWPlhhtueKiqJqZ7b84FyeLFi9m0adOo25CksZLk/+7qPQ9tSZI6MUgkSZ0YJJKkTgwSSVInBokkqRODRJLUiUEiSerEIJEkdTLnPpC4Oy741tdH3cKccMbff+OoW5DUgTMSSVInBokkqRODRJLUiUEiSerEIJEkdWKQSJI6MUgkSZ0YJJKkTgwSSVInBokkqRODRJLUiUEiSerEIJEkdTLUIEnyS0muSHJHktuTvCbJAUmuTnJXe96/b/2zk2xOcmeSk/rqxye5ub13XpK0+j5JLm/165IsHub+SJKebtgzko8BX66qVwDHArcDZwEbq2oJsLG9JslRwErgaGAFcH6Svdp2LgDWAEvaY0WrrwYeraojgXOBc4a8P5KkKYYWJEkWAG8ELgKoqp9U1Q+Ak4F1bbV1wClt+WTgsqp6qqruATYDJyQ5FFhQVddWVQGXTBkzua0rgOWTsxVJ0uwY5ozk5cA24JNJbkzyiSQvBg6pqvsB2vPBbf2FwH1947e02sK2PLW+05iq2g48Bhw4tZEka5JsSrJp27Zte2r/JEkMN0jmAa8GLqiq44Af0Q5j7cJ0M4maoT7TmJ0LVRdW1dKqWjoxMTFz15Kk3TLMINkCbKmq69rrK+gFywPtcBXt+cG+9Q/rG78I2Nrqi6ap7zQmyTxgP+CRPb4nkqRdGlqQVNX3gfuS/EorLQduAzYAq1ptFXBlW94ArGxXYh1B76T69e3w1+NJlrXzH6dPGTO5rVOBa9p5FEnSLJk35O3/HvDpJC8E7gbeQy+81idZDdwLnAZQVbcmWU8vbLYDZ1bVjradM4CLgfnAVe0BvRP5lybZTG8msnLI+yNJmmKoQVJVNwFLp3lr+S7WXwusnaa+CThmmvqTtCCSJI2Gn2yXJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKmToQZJku8luTnJTUk2tdoBSa5Ocld73r9v/bOTbE5yZ5KT+urHt+1sTnJekrT6Pkkub/Xrkiwe5v5Ikp5uNmYkJ1bVq6pqaXt9FrCxqpYAG9trkhwFrASOBlYA5yfZq425AFgDLGmPFa2+Gni0qo4EzgXOmYX9kST1GcWhrZOBdW15HXBKX/2yqnqqqu4BNgMnJDkUWFBV11ZVAZdMGTO5rSuA5ZOzFUnS7Bh2kBTwlSQ3JFnTaodU1f0A7fngVl8I3Nc3dkurLWzLU+s7jamq7cBjwIFD2A9J0i7MG/L2X1dVW5McDFyd5I4Z1p1uJlEz1Gcas/OGeyG2BuDwww+fuWNJ0m4Z6oykqra25weBzwMnAA+0w1W05wfb6luAw/qGLwK2tvqiaeo7jUkyD9gPeGSaPi6sqqVVtXRiYmLP7JwkCRhikCR5cZKXTi4Dvw7cAmwAVrXVVgFXtuUNwMp2JdYR9E6qX98Ofz2eZFk7/3H6lDGT2zoVuKadR5EkzZJhHto6BPh8O/c9D/hMVX05ybeA9UlWA/cCpwFU1a1J1gO3AduBM6tqR9vWGcDFwHzgqvYAuAi4NMlmejORlUPcH0nSNIYWJFV1N3DsNPWHgeW7GLMWWDtNfRNwzDT1J2lBJEkaDT/ZLknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHUy7NvIS9Ju+/pXbx91C3PCG0/81T2yHWckkqRODBJJUicGiSSpE4NEktSJQSJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUicGiSSpE4NEktTJ0IMkyV5Jbkzyxfb6gCRXJ7mrPe/ft+7ZSTYnuTPJSX3145Pc3N47L0lafZ8kl7f6dUkWD3t/JEk7m40Zye8D/bfyPAvYWFVLgI3tNUmOAlYCRwMrgPOT7NXGXACsAZa0x4pWXw08WlVHAucC5wx3VyRJUw01SJIsAv4x8Im+8snAura8Djilr35ZVT1VVfcAm4ETkhwKLKiqa6uqgEumjJnc1hXA8snZiiRpdgx7RvKfgT8EftZXO6Sq7gdozwe3+kLgvr71trTawrY8tb7TmKraDjwGHLhH90CSNKOhBUmS3wAerKobBh0yTa1mqM80Zmova5JsSrJp27ZtA7YjSRrEMGckrwN+M8n3gMuANyf5FPBAO1xFe36wrb8FOKxv/CJga6svmqa+05gk84D9gEemNlJVF1bV0qpaOjExsWf2TpIEDDFIqursqlpUVYvpnUS/pqreBWwAVrXVVgFXtuUNwMp2JdYR9E6qX98Ofz2eZFk7/3H6lDGT2zq1/YynzUgkScMziu9s/yiwPslq4F7gNICqujXJeuA2YDtwZlXtaGPOAC4G5gNXtQfARcClSTbTm4msnK2dkCT1zEqQVNXXgK+15YeB5btYby2wdpr6JuCYaepP0oJImurHP9446hae9+bPn/afsuYYP9kuSepkoCBJ8rQ/7aarSZLmnhkPbSXZF3gRcFC7lcnk5bYLgJcNuTdJ0hh4pnMk7wc+RC80buAXQfJD4OPDa0uSNC5mDJKq+hjwsSS/V1V/Pks9SZLGyEBXbVXVnyd5LbC4f0xVXTKkviRJY2KgIElyKfD3gJuAyc92TN5AUZI0hw36OZKlwFF+alySNNWgnyO5Bfg7w2xEkjSeBp2RHATcluR64KnJYlX95lC6kiSNjUGD5I+H2YQkaXwNetXWXw+7EUnSeBr0qq3H+cUXRr0Q2Bv4UVUtGFZjkqTxMOiM5KX9r5OcApwwjIYkSePlWd39t6q+ALx5z7YiSRpHgx7aelvfyxfQ+1yJnymRJA181dY/6VveDnwPOHmPdyNJGjuDniN5z7AbkSSNp0G/2GpRks8neTDJA0n+KsmiYTcnSXruG/Rk+yeBDfS+l2Qh8N9aTZI0xw0aJBNV9cmq2t4eFwMTQ+xLkjQmBg2Sh5K8K8le7fEu4OFhNiZJGg+DBsl7gbcD3wfuB04FPAEvSRo4SP4dsKqqJqrqYHrB8sczDUiyb5Lrk3wnya1JPtLqByS5Osld7Xn/vjFnJ9mc5M4kJ/XVj09yc3vvvCRp9X2SXN7q1yVZvHu7L0nqatAgeWVVPTr5oqoeAY57hjFPAW+uqmOBVwErkiwDzgI2VtUSYGN7TZKjgJXA0cAK4Pwke7VtXQCsAZa0x4pWXw08WlVHAucC5wy4P5KkPWTQIHnBlJnDATzDZ1Cq54n2cu/2KHofZFzX6uuAU9ryycBlVfVUVd0DbAZOSHIosKCqrm3f0HjJlDGT27oCWD45W5EkzY5BP9n+p8DfJLmCXhi8HVj7TIPajOIG4Ejg41V1XZJDqup+gKq6P8nBbfWFwDf7hm9ptZ+25an1yTH3tW1tT/IYcCDw0JQ+1tCb0XD44YcPuMuSpEEMNCOpqkuAfwo8AGwD3lZVlw4wbkdVvQpYRG92ccwMq083k6gZ6jONmdrHhVW1tKqWTkx41bIk7UmDzkioqtuA257ND6mqHyT5Gr1zGw8kObTNRg4FHmyrbQEO6xu2CNja6oumqfeP2ZJkHrAf8Miz6VGS9Ow8q9vIDyLJRJJfasvzgbcAd9D7hPyqttoq4Mq2vAFY2a7EOoLeSfXr22Gwx5Msa+c/Tp8yZnJbpwLXtPMokqRZMvCM5Fk4FFjXzpO8AFhfVV9Mci2wPslq4F7gNICqujXJenqznu3AmVW1o23rDOBiYD5wVXsAXARcmmQzvZnIyiHujyRpGkMLkqr6LtNcIlxVDwPLdzFmLdOcxK+qTcDTzq9U1ZO0IJIkjcbQDm1JkuYGg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdDC1IkhyW5KtJbk9ya5Lfb/UDklyd5K72vH/fmLOTbE5yZ5KT+urHJ7m5vXdekrT6Pkkub/Xrkiwe1v5IkqY3zBnJduAPqupXgWXAmUmOAs4CNlbVEmBje017byVwNLACOD/JXm1bFwBrgCXtsaLVVwOPVtWRwLnAOUPcH0nSNIYWJFV1f1V9uy0/DtwOLAROBta11dYBp7Tlk4HLquqpqroH2AyckORQYEFVXVtVBVwyZczktq4Alk/OViRJs2NWzpG0Q07HAdcBh1TV/dALG+DgttpC4L6+YVtabWFbnlrfaUxVbQceAw4cyk5IkqY19CBJ8hLgr4APVdUPZ1p1mlrNUJ9pzNQe1iTZlGTTtm3bnqllSdJuGGqQJNmbXoh8uqo+18oPtMNVtOcHW30LcFjf8EXA1lZfNE19pzFJ5gH7AY9M7aOqLqyqpVW1dGJiYk/smiSpGeZVWwEuAm6vqj/re2sDsKotrwKu7KuvbFdiHUHvpPr17fDX40mWtW2ePmXM5LZOBa5p51EkSbNk3hC3/Trgd4Cbk9zUan8EfBRYn2Q1cC9wGkBV3ZpkPXAbvSu+zqyqHW3cGcDFwHzgqvaAXlBdmmQzvZnIyiHujyRpGkMLkqr6BtOfwwBYvosxa4G109Q3AcdMU3+SFkSSpNHwk+2SpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOhlakCT5yyQPJrmlr3ZAkquT3NWe9+977+wkm5PcmeSkvvrxSW5u752XJK2+T5LLW/26JIuHtS+SpF0b5ozkYmDFlNpZwMaqWgJsbK9JchSwEji6jTk/yV5tzAXAGmBJe0xuczXwaFUdCZwLnDO0PZEk7dLQgqSqvg48MqV8MrCuLa8DTumrX1ZVT1XVPcBm4IQkhwILquraqirgkiljJrd1BbB8crYiSZo9s32O5JCquh+gPR/c6guB+/rW29JqC9vy1PpOY6pqO/AYcODQOpckTeu5crJ9uplEzVCfaczTN56sSbIpyaZt27Y9yxYlSdOZ7SB5oB2uoj0/2OpbgMP61lsEbG31RdPUdxqTZB6wH08/lAZAVV1YVUuraunExMQe2hVJEsx+kGwAVrXlVcCVffWV7UqsI+idVL++Hf56PMmydv7j9CljJrd1KnBNO48iSZpF84a14SSfBd4EHJRkC/Bh4KPA+iSrgXuB0wCq6tYk64HbgO3AmVW1o23qDHpXgM0HrmoPgIuAS5NspjcTWTmsfZEk7drQgqSqfnsXby3fxfprgbXT1DcBx0xTf5IWRJKk0XmunGyXJI0pg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE7GPkiSrEhyZ5LNSc4adT+SNNeMdZAk2Qv4OPBW4Cjgt5McNdquJGluGesgAU4ANlfV3VX1E+Ay4OQR9yRJc8q8UTfQ0ULgvr7XW4Bfm7pSkjXAmvbyiSR3zkJvo3IQ8NCom9gdHxh1A88dY/e7006e77+/v7urN8Y9SDJNrZ5WqLoQuHD47Yxekk1VtXTUfWj3+bsbb3P59zfuh7a2AIf1vV4EbB1RL5I0J417kHwLWJLkiCQvBFYCG0bckyTNKWN9aKuqtif5XeB/AHsBf1lVt464rVGbE4fwnqf83Y23Ofv7S9XTTilIkjSwcT+0JUkaMYNEktSJQSJJ6sQgGXNJfiXJa5Ls3W4ZozHj7218JTkyydIk+4y6l1HyZPsYS/I24D8A/689NgEXV9UPR9qYBpLkl6vqf7flvapqx6h70uCS/Aa9f38PA98HPjz5+5xrnJGMqSR7A+8AVlfVcuBKeh/O/MMkC0banJ5R+5/QTUk+A1BVO5yZjI8krwX+E7Cqqk4EHgXm7N3HDZLxtgBY0pY/D3wReCHwz5JMd/sYPQckeTHwu8CHgJ8k+RQYJmPoo1V1Y1v+MHDAXD3EZZCMqar6KfBnwNuSvKGqfgZ8A7gJeP0oe9PMqupHwHuBzwD/Cti3P0xG2ZsGdh3wOfj5Oa596N3UcEGrHTi61mafQTLe/ifwFeB3kryxqnZU1WeAlwHHjrY1zaSqtlbVE1X1EPB+YP5kmCR5dZJXjLZDzaT9W5s8FxngB8AjVbUtyTuBf59k/sganGVjfYuUua6qnkzyaXp3PD67/c/nKeAQ4P6RNqeBVdXDSd4P/Mckd9C73c+JI25LA6qq7fS+nuK+JH8C/Drw7qr68YhbmzUGyZirqkeT/FfgNnp/2T4JvKuqHhhtZ9odVfVQku/S+7bPf1hVW0bdkwbTzkfuDbyhPS+vqrtG29Xs8vLf55F2rLba+RKNkST7A+uBP6iq7466H+2+JO8GvjUXbxxrkEjPEUn2raonR92Hnp0kqTn6P1SDRJLUiVdtSZI6MUgkSZ0YJJKkTgwSzQlJFiW5MsldSf5Pko8leeEzjPmjAbb7oSQv2oN9fi/JQW35bzps591J/sue6kuaiUGi5712nf/ngC9U1RLgl4GXAGufYegzBgm9+2XtsSDpV1WvHcZ2pT3NINFc8Gbgyar6JPz8flb/Anhvkg/0/+We5ItJ3pTko/RuW3JTkk8neXGSLyX5TpJbkrwjyQfp3Y7mq0m+2sZfkGRTkluTfKRvu99L8pEk305y8+QtUJIcmOQrSW5M8hf0brcxOeaJ9vymJF9LckWSO1o/ae/9o1b7RpLzknxxpv8QSf5l6/+WJB/qq38hyQ2t7zX9PSRZ2/b7m0kOeba/BD1/GSSaC44GbugvtPsk3csu7u5QVWcBP66qV1XVO4EVwNaqOraqjgG+XFXnAVuBE9utxAH+dVUtBV4J/IMkr+zb7ENV9WrgAno3a4TeXWO/UVXHARuAw3exD8fRm/0cBbwceF2SfYG/AN5aVa8HJmb6j5DkeOA9wK8By4D3JTmuvf3eqjoeWAp8sO+mgy8GvllVxwJfB94308/Q3GSQaC4IvfuRDVqfzs3AW5Kc0+62/Ngu1nt7km8DN9ILsKP63vtce74BWNyW3whM3vn3S/S+12I611fVlnbXgpva+FcAd1fVPW2dzz7DPrwe+HxV/aiqnmj9vKG998Ek3wG+Se97bSa/nuAn9L6eYGrf0s8ZJJoLbqX3l/bPtS//Ogx4jJ3/Hew73QbaN98dTy9Q/iTJv526TpIj6M00llfVK4EvTdneU+15BzvPhAYJs6f6lifH7+53zky7fpI3AW8BXtNmHjfyi75/2vdp7al9S4BBorlhI/CiJKfDz+9J9qfAxcDdwKuSvCDJYcAJfeN+mt43UZLkZcDfVtWn6H0z3qvbOo8DL23LC4AfAY+1cwlvHaC3rwPvbD/jrcD+u7FfdwAvT7K4vX7HAD/rlCQval+u9Vv0vopgP+DRqvrbdu5m2W70IPnXhZ7/qqqS/BZwfpJ/Q+8PqP9O76qsnwD30Jtp3AJ8u2/ohcB326GqS+jd5v1nwE+BM/rWuSrJ/VV1YpIb6c2A7gb+1wDtfQT4bPsZf03vvM2g+/XjJB8AvpzkIeD6Kau8O8kpfa+X0QvPyfU+UVU3JrkN+Oft7sN30ju8JQ3Me21JYyzJS6rqiXYV18eBu6rq3FH3pbnFQ1vSeHtfkpvozYL2o3cVlzSrnJFIkjpxRiJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUif/HxDn3i1LrirjAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEICAYAAAB1f3LfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAVn0lEQVR4nO3df9TedX3f8eeLRDHHCuNHYJjg4krsCq7iiBTrdMXYmp62gg5sXK1x5izK6NSddj2wnTNbVzY4czJ1wka1Emgt5GAtma1WjLXUFcGgVASlZIKQQSFCqrEVNPG9P67PXa/cXLlzh0+u++LO/Xycc53re72/38/n+nxzn9yv+/v9Xtfnm6pCkqQn67BJD0CSNL8ZJJKkLgaJJKmLQSJJ6mKQSJK6GCSSpC6LJz2AuXbsscfWihUrJj0MSZpXbr311m9U1dJR6xZckKxYsYKtW7dOehiSNK8k+fq+1nlqS5LUxSCRJHUxSCRJXQwSSVIXg0SS1MUgkSR1MUgkSV0MEklSl7F+ITHJvcAuYA+wu6pWJTkauBZYAdwLvLaqdrbtLwTWt+3fWlV/3OqnAVcCS4A/At5WVZXkcOAq4DTgEeAXquregzX+yz9/48HqSjM470Uvm/QQJHWYiyOSM6vq1Kpa1V5fAGypqpXAlvaaJCcDa4FTgDXAZUkWtTaXAxuAle2xptXXAzur6iTgUuCSOdgfSdKQSZzaOgvY2JY3AmcP1a+pqser6h5gG3B6khOAI6rqphrcF/iqaW2m+roOWJ0k498FSdKUcQdJAZ9McmuSDa12fFU9CNCej2v1ZcD9Q223t9qytjy9vlebqtoNfBM4Zgz7IUnah3FP2viSqnogyXHADUm+OsO2o44kaob6TG327ngQYhsAnvOc58w8YknSARnrEUlVPdCeHwY+CpwOPNROV9GeH26bbwdOHGq+HHig1ZePqO/VJsli4Ejg0RHjuKKqVlXVqqVLR86CLEl6ksZ2RJLkmcBhVbWrLf808E5gM7AOuLg9X9+abAY+nOTdwLMZXFS/par2JNmV5AzgZuANwPuG2qwDbgLOAT7drqNImsdu/JOvTHoIC8LLzvzRg9LPOE9tHQ98tF37Xgx8uKo+keTzwKYk64H7gHMBquqOJJuAO4HdwPlVtaf1dR4/+Pjvx9sD4IPA1Um2MTgSWTvG/ZEkjTC2IKmqrwEvGFF/BFi9jzYXAReNqG8Fnj+i/hgtiCRJk7Hg7pCoheM739ky6SEc8pYsGfk3oRYYp0iRJHUxSCRJXQwSSVIXg0SS1MUgkSR1MUgkSV0MEklSF4NEktTFIJEkdTFIJEldDBJJUheDRJLUxSCRJHUxSCRJXQwSSVIXg0SS1MUgkSR1MUgkSV0MEklSF4NEktTFIJEkdTFIJEldDBJJUheDRJLUxSCRJHUxSCRJXQwSSVIXg0SS1MUgkSR1MUgkSV0MEklSl7EHSZJFSb6Y5GPt9dFJbkhyd3s+amjbC5NsS3JXklcO1U9Lcntb994kafXDk1zb6jcnWTHu/ZEk7W0ujkjeBnxl6PUFwJaqWglsaa9JcjKwFjgFWANclmRRa3M5sAFY2R5rWn09sLOqTgIuBS4Z765IkqYba5AkWQ78LPCBofJZwMa2vBE4e6h+TVU9XlX3ANuA05OcABxRVTdVVQFXTWsz1dd1wOqpoxVJ0twY9xHJfwd+Dfj+UO34qnoQoD0f1+rLgPuHttveasva8vT6Xm2qajfwTeCYg7oHkqQZjS1Ikvwc8HBV3TrbJiNqNUN9pjbTx7IhydYkW3fs2DHL4UiSZmOcRyQvAV6V5F7gGuDlSX4HeKidrqI9P9y23w6cONR+OfBAqy8fUd+rTZLFwJHAo9MHUlVXVNWqqlq1dOnSg7N3kiRgjEFSVRdW1fKqWsHgIvqnq+r1wGZgXdtsHXB9W94MrG2fxHoug4vqt7TTX7uSnNGuf7xhWpupvs5p7/GEIxJJ0vgsnsB7XgxsSrIeuA84F6Cq7kiyCbgT2A2cX1V7WpvzgCuBJcDH2wPgg8DVSbYxOBJZO1c7IUkamJMgqarPAJ9py48Aq/ex3UXARSPqW4Hnj6g/RgsiSdJk+M12SVIXg0SS1MUgkSR1MUgkSV0MEklSF4NEktTFIJEkdTFIJEldDBJJUheDRJLUxSCRJHUxSCRJXQwSSVIXg0SS1MUgkSR1MUgkSV0MEklSF4NEktTFIJEkdTFIJEldDBJJUheDRJLUxSCRJHUxSCRJXQwSSVIXg0SS1MUgkSR1MUgkSV0MEklSF4NEktTFIJEkdTFIJEldxhYkSZ6R5JYkf5HkjiS/0epHJ7khyd3t+aihNhcm2ZbkriSvHKqfluT2tu69SdLqhye5ttVvTrJiXPsjSRptnEckjwMvr6oXAKcCa5KcAVwAbKmqlcCW9pokJwNrgVOANcBlSRa1vi4HNgAr22NNq68HdlbVScClwCVj3B9J0ghjC5Ia+HZ7+bT2KOAsYGOrbwTObstnAddU1eNVdQ+wDTg9yQnAEVV1U1UVcNW0NlN9XQesnjpakSTNjbFeI0myKMltwMPADVV1M3B8VT0I0J6Pa5svA+4far691Za15en1vdpU1W7gm8AxI8axIcnWJFt37NhxkPZOkgRjDpKq2lNVpwLLGRxdPH+GzUcdSdQM9ZnaTB/HFVW1qqpWLV26dD+jliQdiFkFSZIts6ntS1X9NfAZBtc2Hmqnq2jPD7fNtgMnDjVbDjzQ6stH1Pdqk2QxcCTw6GzHJUnqN2OQtE9eHQ0cm+So9omro9uno569n7ZLk/y9trwEeAXwVWAzsK5ttg64vi1vBta2T2I9l8FF9Vva6a9dSc5o1z/eMK3NVF/nAJ9u11EkSXNk8X7Wvxl4O4PQuJUfnEr6FvD+/bQ9AdjYPnl1GLCpqj6W5CZgU5L1wH3AuQBVdUeSTcCdwG7g/Kra0/o6D7gSWAJ8vD0APghcnWQbgyORtfvbYUnSwTVjkFTVe4D3JPk3VfW+A+m4qr4EvHBE/RFg9T7aXARcNKK+FXjC9ZWqeowWRJKkydjfEQkAVfW+JD8BrBhuU1VXjWlckqR5YlZBkuRq4IeB24Cp001T3+mQJC1gswoSYBVwsheyJUnTzfZ7JF8G/v44ByJJmp9me0RyLHBnklsYzKEFQFW9aiyjkiTNG7MNkl8f5yAkSfPXbD+19afjHogkaX6a7ae2dvGDOayezmAm37+pqiPGNTBJ0vww2yOSZw2/TnI2cPo4BiRJml+e1Oy/VfUHwMsP7lAkSfPRbE9tvWbo5WEMvlfid0okSbP+1NbPDy3vBu5lcHdCSdICN9trJP9y3AORJM1Ps72x1fIkH03ycJKHknwkyfL9t5QkHepme7H9QwxuIvVsBvdJ/9+tJkla4GYbJEur6kNVtbs9rgS8+bkkadZB8o0kr0+yqD1eDzwyzoFJkuaH2QbJm4DXAn8FPMjg/uhegJckzfrjv/8JWFdVOwGSHA28i0HASJIWsNkekfzYVIgAVNWjjLgfuyRp4ZltkByW5KipF+2IZLZHM5KkQ9hsw+C/AX+e5DoGU6O8FrhobKOSJM0bs/1m+1VJtjKYqDHAa6rqzrGOTJI0L8z69FQLDsNDkrSXJzWNvCRJUwwSSVIXg0SS1MUgkSR1MUgkSV0MEklSF4NEktRlbEGS5MQkf5LkK0nuSPK2Vj86yQ1J7m7Pw1OvXJhkW5K7krxyqH5aktvbuvcmSasfnuTaVr85yYpx7Y8kabRxHpHsBn6lqn4UOAM4P8nJwAXAlqpaCWxpr2nr1gKnAGuAy5Isan1dDmwAVrbHmlZfD+ysqpOAS4FLxrg/kqQRxhYkVfVgVX2hLe8CvsLgNr1nARvbZhuBs9vyWcA1VfV4Vd0DbANOT3ICcERV3VRVBVw1rc1UX9cBq6eOViRJc2NOrpG0U04vBG4Gjq+qB2EQNsBxbbNlwP1Dzba32rK2PL2+V5uq2g18EzhmLDshSRpp7EGS5IeAjwBvr6pvzbTpiFrNUJ+pzfQxbEiyNcnWHTt27G/IkqQDMNYgSfI0BiHyu1X1+638UDtdRXt+uNW3AycONV8OPNDqy0fU92qTZDFwJPDo9HFU1RVVtaqqVi1duvRg7JokqRnnp7YCfBD4SlW9e2jVZmBdW14HXD9UX9s+ifVcBhfVb2mnv3YlOaP1+YZpbab6Ogf4dLuOIkmaI+O8y+FLgF8Cbk9yW6v9e+BiYFOS9cB9wLkAVXVHkk0MpqrfDZxfVXtau/OAK4ElwMfbAwZBdXWSbQyORNaOcX8kSSOMLUiq6rOMvoYBsHofbS5ixJ0Xq2or8PwR9cdoQSRJmgy/2S5J6mKQSJK6GCSSpC4GiSSpi0EiSepikEiSuhgkkqQuBokkqYtBIknqYpBIkroYJJKkLgaJJKmLQSJJ6mKQSJK6GCSSpC4GiSSpi0EiSepikEiSuhgkkqQuBokkqYtBIknqYpBIkroYJJKkLgaJJKmLQSJJ6mKQSJK6GCSSpC4GiSSpi0EiSepikEiSuhgkkqQuBokkqcvYgiTJbyd5OMmXh2pHJ7khyd3t+aihdRcm2ZbkriSvHKqfluT2tu69SdLqhye5ttVvTrJiXPsiSdq3cR6RXAmsmVa7ANhSVSuBLe01SU4G1gKntDaXJVnU2lwObABWtsdUn+uBnVV1EnApcMnY9kSStE9jC5KquhF4dFr5LGBjW94InD1Uv6aqHq+qe4BtwOlJTgCOqKqbqqqAq6a1merrOmD11NGKJGnuzPU1kuOr6kGA9nxcqy8D7h/abnurLWvL0+t7tamq3cA3gWPGNnJJ0khPlYvto44kaob6TG2e2HmyIcnWJFt37NjxJIcoSRplroPkoXa6ivb8cKtvB04c2m458ECrLx9R36tNksXAkTzxVBoAVXVFVa2qqlVLly49SLsiSYK5D5LNwLq2vA64fqi+tn0S67kMLqrf0k5/7UpyRrv+8YZpbab6Ogf4dLuOIkmaQ4vH1XGS3wN+Ejg2yXbgHcDFwKYk64H7gHMBquqOJJuAO4HdwPlVtad1dR6DT4AtAT7eHgAfBK5Oso3Bkcjace2LJGnfxhYkVfW6faxavY/tLwIuGlHfCjx/RP0xWhBJkibnqXKxXZI0TxkkkqQuBokkqYtBIknqYpBIkroYJJKkLgaJJKmLQSJJ6mKQSJK6GCSSpC4GiSSpi0EiSepikEiSuhgkkqQuBokkqYtBIknqYpBIkroYJJKkLgaJJKmLQSJJ6mKQSJK6GCSSpC4GiSSpi0EiSepikEiSuhgkkqQuBokkqYtBIknqYpBIkroYJJKkLgaJJKmLQSJJ6jLvgyTJmiR3JdmW5IJJj0eSFpp5HSRJFgHvB34GOBl4XZKTJzsqSVpY5nWQAKcD26rqa1X1XeAa4KwJj0mSFpTFkx5Ap2XA/UOvtwM/Pn2jJBuADe3lt5PcNQdjm5RjgW9MehAH4l9PegBPHfPuZ6e9HOo/v3+wrxXzPUgyolZPKFRdAVwx/uFMXpKtVbVq0uPQgfNnN78t5J/ffD+1tR04cej1cuCBCY1Fkhak+R4knwdWJnlukqcDa4HNEx6TJC0o8/rUVlXtTvLLwB8Di4Dfrqo7JjysSVsQp/AOUf7s5rcF+/NL1RMuKUiSNGvz/dSWJGnCDBJJUheDRJLUxSCZ55L8SJIXJ3lamzJG84w/t/kryUlJViU5fNJjmSQvts9jSV4D/Gfg/7XHVuDKqvrWRAemWUnyvKr6y7a8qKr2THpMmr0kP8fg/98jwF8B75j6eS40HpHMU0meBvwCsL6qVgPXM/hy5q8lOWKig9N+tV9CtyX5MEBV7fHIZP5I8hPAu4B1VXUmsBNYsLOPGyTz2xHAyrb8UeBjwNOBf5Fk1PQxegpI8kzgl4G3A99N8jtgmMxDF1fVF9vyO4CjF+opLoNknqqq7wHvBl6T5KVV9X3gs8BtwD+d5Ng0s6r6G+BNwIeBXwWeMRwmkxybZu1m4Pfh765xHc5gUsMjWu2YyQ1t7hkk89ufAZ8EfinJy6pqT1V9GHg28ILJDk0zqaoHqurbVfUN4M3AkqkwSfJPkvyjyY5QM2n/16auRQb4a+DRqtqR5BeB30yyZGIDnGPzeoqUha6qHkvyuwxmPL6w/fJ5HDgeeHCig9OsVdUjSd4M/NckX2Uw3c+ZEx6WZqmqdjO4PcX9Sf4L8NPAG6vqOxMe2pwxSOa5qtqZ5LeAOxn8ZfsY8PqqemiyI9OBqKpvJPkSg7t9/lRVbZ/0mDQ77Xrk04CXtufVVXX3ZEc1t/z47yGknautdr1E80iSo4BNwK9U1ZcmPR4duCRvBD6/ECeONUikp4gkz6iqxyY9Dj05SVIL9BeqQSJJ6uKntiRJXQwSSVIXg0SS1MUg0YKS5NeT/GqSdyZ5xQzbXZnknLkc27T3vzfJsZN6/1GSnJ3k5EmPQ089BokWpKr6j1X1qXH1f4jOmXU2YJDoCQwSHfKS/IckdyX5FPAjrfZ3RxxJLk5yZ5IvJXnXUNNXJPmzJH/ZZuslyRuT/I+hvj+W5Cfb8rfbkc7NwItH9ZtkaZKPJPl8e7yk1Y9J8skkX0zyvxhMuzH1Hn+Q5NYkdyTZMFT/dpJL2rpPJTk9yWeSfC3Jq9o2z0jyoSS3t77PnOV+XJTkL5J8LsnxbbbbVzH49v1tSX744Px0dCgwSHRIS3IasBZ4IfAa4EXT1h8NvBo4pap+DPjNodUrgH8G/CzwP5M8Yz9v90zgy1X14wxmGhjV73uAS6vqRcA/Bz7Q6u8APltVLwQ2A88Z6vdNVXUasAp469CEgM8EPtPW7Wrv8VPtfd/ZtjkfoKr+MfA6YOMs9+NzVfUC4EbgX1XVn7dx/buqOrWq/u9++tAC4hQpOtS9FPhoVf0tQJLN09Z/i8G0Mh9I8ocMpuKfsqnNEnB3kq8B+5tIcQ/wkf30+wrg5KFZ/o9I8izgZQyCjqr6wyQ7h/p9a5JXt+UTGdw64BHgu8AnWv124PGq+l6S2xmEIAxmgn5f6/erSb4OPG8/+/HdofHeyiCcpH3yiEQLwT6/ddsm3DudQQCczQ9+MY9qV8Bu9v5/M/zX/WNT08DP0O9hwIvbX/WnVtWyqtq1r3G2002vaG1eAHxx6D2/N/RN6u8zmLCTFn5TfyTu6740M+3HcL978A9O7YdBokPdjcCrkyxpf/n//PDKJD8EHFlVf8TgRlOnDq0+N8lh7XrAPwTuAu4FTm31ExmExRPM0O8nGdzUamq7qfqNwC+22s8AR7X6kcDOqvrbNrvzGQe2+3v1+zwGp8xmvR/T7AKedYDvrwXAvzR0SKuqLyS5lsENv77O4B4uw54FXN+uGwT4t0Pr7gL+lMG0/G9p0/b/H+AeBqeSvgx8YR9vva9+3wq8v830u5jBL/q3AL8B/F6SL7T3vK9t/wngLW37u4DPHeA/wWUMru/czuAo5I1V9fgB7Mewa4DfSvJW4Byvk2iKc21Jkrp4akuS1MUgkSR1MUgkSV0MEklSF4NEktTFIJEkdTFIJEldDBJJUpf/D+mFPSUL//zLAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "features=['OutstandingLoan', 'terms', 'disbursedamount', 'LastInstallmentAmount','Gender', \n",
    "          'terms_lapsed', 'defaulter','disbursed_outstanding_ratio']\n",
    "\n",
    "for f in features:\n",
    "    sns.countplot(x=f,data=df,palette=\"Set3\") \n",
    "    plt.xticks(rotation=45)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51d2f33f",
   "metadata": {},
   "source": [
    "**Males were the most while unknown gender were the least**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "30ccfda3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 149321 entries, 0 to 149320\n",
      "Data columns (total 11 columns):\n",
      " #   Column                       Non-Null Count   Dtype         \n",
      "---  ------                       --------------   -----         \n",
      " 0   OutstandingLoan              149321 non-null  int64         \n",
      " 1   terms                        149321 non-null  int64         \n",
      " 2   disbursedamount              149321 non-null  int64         \n",
      " 3   installmentstartdate         149321 non-null  datetime64[ns]\n",
      " 4   Lastinstallmentdate          149321 non-null  datetime64[ns]\n",
      " 5   LastInstallmentAmount        149321 non-null  int64         \n",
      " 6   disbursementdate             149321 non-null  datetime64[ns]\n",
      " 7   Gender                       149321 non-null  object        \n",
      " 8   defaulter                    149321 non-null  object        \n",
      " 9   terms_lapsed                 149321 non-null  int64         \n",
      " 10  disbursed_outstanding_ratio  149321 non-null  int64         \n",
      "dtypes: datetime64[ns](3), int64(6), object(2)\n",
      "memory usage: 12.5+ MB\n"
     ]
    }
   ],
   "source": [
    "\"\"\"features = ['OutstandingLoan','terms','disbursedamount','LastInstallmentAmount','terms_lapsed',\n",
    "           \"disbursed_outstanding_ratio\",\"defaulter\"]\n",
    "for col in features:\n",
    "    df.boxplot(column=col, by='defaulter', figsize=(6,6))\n",
    "    plt.title(col)\n",
    "plt.show()\"\"\"\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "bd00bc0a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'#this took so long\\nsns.pairplot(data=df,palette=\"Set3\")'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"#this took so long\n",
    "sns.pairplot(data=df,palette=\"Set3\")\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "da5f8e56",
   "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>OutstandingLoan</th>\n",
       "      <th>terms</th>\n",
       "      <th>disbursedamount</th>\n",
       "      <th>LastInstallmentAmount</th>\n",
       "      <th>Gender</th>\n",
       "      <th>defaulter</th>\n",
       "      <th>terms_lapsed</th>\n",
       "      <th>disbursed_outstanding_ratio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   OutstandingLoan  terms  disbursedamount  LastInstallmentAmount Gender  \\\n",
       "0                2      0                0                      0      1   \n",
       "1                2      0                0                      0      0   \n",
       "2                2      0                0                      0      0   \n",
       "3                2      0                0                      0      0   \n",
       "4                2      0                0                      0      1   \n",
       "\n",
       "  defaulter  terms_lapsed  disbursed_outstanding_ratio  \n",
       "0         1             0                            0  \n",
       "1         1             0                            0  \n",
       "2         1             0                            0  \n",
       "3         1             0                            0  \n",
       "4         1             0                            0  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The dataset: we drop date columns since they contain no important information for our clustering activity.\n",
    "# We have actually done some feature engineering from the dates and created a new column; defaulter\n",
    "\n",
    "#we will drop defaulter = unkonwn and defaulter = no rows as it seems df is for defaulter\n",
    "#df=df[~((df.defaulter==0)|(df.defaulter==2))]\n",
    "df.drop(['installmentstartdate','Lastinstallmentdate','disbursementdate'], axis=1,inplace=True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b922784d",
   "metadata": {},
   "source": [
    "# PREDICTION AND CREDIT RATING\n",
    "## The problem\n",
    "\n",
    "1. Segmentation of customers with similar credit behaviours\n",
    "2. identify the common characteristics in each of the clusters\n",
    "3. Develop a system that can give credit rating for a customer. This rating can be used to judge whether to lend to them and if yes,  the amounts that are ideal for lending to them. This will help the business reduce non-performing loans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "5e39696a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of values of features : OutstandingLoan is : 3 which are -- [0 1 2]\n",
      "The number of values of features : terms is : 3 which are -- [0 1 2]\n",
      "The number of values of features : disbursedamount is : 3 which are -- [0 1 2]\n",
      "The number of values of features : LastInstallmentAmount is : 3 which are -- [0 1 2]\n",
      "The number of values of features : Gender is : 3 which are -- [0 1 2]\n",
      "The number of values of features : defaulter is : 3 which are -- [0 1 2]\n",
      "The number of values of features : terms_lapsed is : 3 which are -- [0 1 2]\n",
      "The number of values of features : disbursed_outstanding_ratio is : 3 which are -- [0 1 2]\n"
     ]
    }
   ],
   "source": [
    "for column in df:\n",
    "    unique_vals=np.unique(df[column])\n",
    "    len_unique=len(unique_vals)\n",
    "    if len_unique<10:\n",
    "        print(\"The number of values of features : {} is : {} which are -- {}\".format(column,len_unique,unique_vals))\n",
    "        \n",
    "    else:\n",
    "        print(\"The number of values of features : {} is : {}\".format(column,len_unique))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8d50d934",
   "metadata": {},
   "outputs": [],
   "source": [
    "#predict_df.groupby(\"defaulter\").size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "f71f7b54",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The clusters are: [4 4 4 ... 0 0 0]\n",
      "The inertia is: 280760.632793947\n"
     ]
    }
   ],
   "source": [
    "#Get the train data\n",
    "X_train=df.values\n",
    "\n",
    "#running kmeans \n",
    "kmeans=KMeans(n_clusters=5,random_state=540)\n",
    "kmeans=kmeans.fit(X_train)\n",
    "\n",
    "#print the cluster assigned to each observation\n",
    "print(\"The clusters are:\",kmeans.labels_)\n",
    "\n",
    "#print the inertia\n",
    "print(\"The inertia is:\",kmeans.inertia_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "cea91fce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The inertia for 2 is 435066.9935203006\n",
      "The inertia for 3 is 357625.94311398733\n",
      "The inertia for 4 is 317522.18947382516\n",
      "The inertia for 5 is 280354.1948962724\n",
      "The inertia for 6 is 253101.66467658774\n",
      "The inertia for 7 is 232723.03309276834\n",
      "The inertia for 8 is 219198.9309080343\n",
      "The inertia for 9 is 203887.3388284323\n",
      "The inertia for 10 is 197035.59071036766\n",
      "The inertia for 11 is 183996.43565360186\n",
      "The inertia for 12 is 178315.81694432066\n",
      "The inertia for 13 is 171210.03130563186\n",
      "The inertia for 14 is 163255.96581045387\n",
      "The inertia for 15 is 159819.87729463988\n",
      "The inertia for 16 is 156303.47355456388\n",
      "The inertia for 17 is 151545.05838556477\n",
      "The inertia for 18 is 144336.54294503538\n",
      "The inertia for 19 is 144563.4071789343\n"
     ]
    }
   ],
   "source": [
    "# finding best inertia\n",
    "\n",
    "nclusters=range(2,20)\n",
    "inertia=[]\n",
    "for n in nclusters:\n",
    "    kmeans=KMeans(n_clusters=n,random_state=2)\n",
    "    kmeans=kmeans.fit(X_train)\n",
    "    u=kmeans.inertia_\n",
    "    inertia.append(u)\n",
    "    print('The inertia for {} is {}'.format(n,u))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "7e8cbe91",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Inertia Plot Per Cluster')"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#elbow method for inertia\n",
    "fig,(ax1)=plt.subplots(1,figsize=(16,6))\n",
    "xx=np.arange(len(nclusters))\n",
    "ax1.plot(xx,inertia)\n",
    "ax1.set_xticks(xx)\n",
    "ax1.set_xticklabels(nclusters,rotation=45)\n",
    "plt.xlabel(\"Number of Clusters\")\n",
    "plt.ylabel(\"Inertia Score\")\n",
    "plt.title(\"Inertia Plot Per Cluster\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "5f06f36a",
   "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>Cluster 1</th>\n",
       "      <th>Cluster 2</th>\n",
       "      <th>Cluster 3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>39530</td>\n",
       "      <td>54451</td>\n",
       "      <td>55340</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Cluster 1  Cluster 2  Cluster 3\n",
       "0      39530      54451      55340"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Get the train data\n",
    "X_train=df.values\n",
    "\n",
    "#running kmeans \n",
    "kmeans=KMeans(n_clusters=3,random_state=540)\n",
    "kmeans=kmeans.fit(X_train)\n",
    "\n",
    "#prediction for new data\n",
    "predictions=kmeans.predict(X_train)\n",
    "\n",
    "#calculating the count of the clusters\n",
    "unique,counts=np.unique(predictions, return_counts=True)\n",
    "counts=counts.reshape(1,3)\n",
    "\n",
    "clcounts_df=pd.DataFrame(counts,columns=[\"Cluster 1\",\"Cluster 2\",\"Cluster 3\"])\n",
    "clcounts_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cb674b86",
   "metadata": {},
   "source": [
    " # PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "d15e8102",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variation (first two componens): [0.44200519 0.21319806]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.decomposition import PCA\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "\n",
    "X = X_train\n",
    "y_num = predictions\n",
    "\n",
    "target_names = [\"Cluster 1\",\"Cluster 2\",\"Cluster 3\"]\n",
    "pca=PCA(n_components=2, random_state =453)\n",
    "X_r=pca.fit(X).transform(X)\n",
    "\n",
    "# % OF VARIANCE explained by each component\n",
    "print(\"Explained Variation (first two componens): %s\" %str(pca.explained_variance_ratio_))\n",
    "\n",
    "#plotting the data\n",
    "plt.figure()\n",
    "plt.figure(figsize=(12,8))\n",
    "colors=[\"navy\",\"turquoise\",\"darkorange\",\"red\",\"black\"]\n",
    "lw=2\n",
    "\n",
    "for color, i, target_name in zip(colors, [0,1,2],target_names):\n",
    "    plt.scatter(X_r[y_num==i,0],X_r[y_num == i,1],color=color,alpha=0.8,lw=lw,label=target_name)\n",
    "    \n",
    "plt.legend(loc=\"best\",shadow=False,scatterpoints=1)\n",
    "plt.legend(bbox_to_anchor=(1.05,1),loc=2,borderaxespad=0.6)\n",
    "plt.title(\"PCA of 2 Items\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "bebdbedf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Variance in our dataset is :  4.327577788307964\n",
      "The 95% variance we weant to have is :  4.111198898892566\n",
      "\n",
      "Variance Explained With 2 n_components: 2.8354430237176897\n",
      "Variance Explained With 3 n_components: 3.4392153262570426\n",
      "Variance Explained With 4 n_components: 3.891037935701698\n",
      "Variance Explained With 5 n_components: 4.091478849091813\n",
      "Variance Explained With 6 n_components: 4.235431827065212\n",
      "Variance Explained With 7 n_components: 4.321708213652514\n",
      "Variance Explained With 8 n_components: 4.327577788307964\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABGQAAAHgCAYAAADjbbusAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAB3Q0lEQVR4nO3dd3hUVf7H8fehJ1SlSVNEVERF1KiIDbvYsJe1LLoWXLHjir23tbefvfey9i7YO6gIghUEQSyASkvo5/fHHUjAAAGS3Jnk/Xqe7zN35tyZ+SSTXck355wbYoxIkiRJkiSp8tRIO4AkSZIkSVJ1Y0NGkiRJkiSpktmQkSRJkiRJqmQ2ZCRJkiRJkiqZDRlJkiRJkqRKZkNGkiRJkiSpktVKO8DyaNasWWzfvn3aMSRJkpSCceOS27Zt082h7PbZZ59NjDE2TzuHJC1OTjZk2rdvz+DBg9OOIUmSJClLhRDGpJ1BkpbEJUuSJEmSJEmVzIaMJEmScsoxxyQlSVIuy8klS5IkSaq+mjZNO4EkSSvOhowkSZJyyuWXp51AkqQV55IlSZIkSZKkSmZDRpIkSTnliCOSkiQpl7lkSZIkSTmlXbu0E0iStOJsyEiSJCmnXHRR2gkkSVpxLlmSJEmSJEmqZDZkJEmSlFMOPTQpSZJymUuWJEmSlFPWXjvtBJIkrTgbMpIkScop556bdgJJklacS5YkSZIkSZIqmQ0ZSZIk5ZSDDkpKkqRc5pIlSZIk5ZSuXdNOIEnSirMhs6JiTG5DSDeHJElSNdG/f9oJJElacS5ZWhFvvw2bbw6PP552EkmSJEmSlENsyKyI77+HTz6Biy+GefPSTiNJklQt7LtvUpIk5TIbMivin/+EVVeFESPgf/9LO40kSVK1sPnmSUmSlMtsyKyIOnXgzDOT44sucpaMJElSJejXLylJknKZDZkVdcQR0LYtfPUVPPts2mkkSZIkSVIOsCGzourWLd7q31kykiRJFW7PPZOSJCmX2ZApD//6F7RuDV9+Cc8/n3YaSZKkKm377ZOSJCmX2ZApD/XqwRlnJMcXXQQxpptHkiSpCjvppKQkScplNmTKy9FHwyqrwBdfwIsvpp1GkiRJkiRlMRsy5SUvD/7zn+TYWTKSJEkVpmfPpCRJymU2ZMrTscdCixYweDC88kraaSRJkqqkPfZISpKkXGZDpjzl58PppyfHzpKRJEmqEP/+d1KSJOUyGzLl7bjjoFkz+OQTeP31tNNIkiRJkqQsZEOmvNWvD/36JccXXugsGUmSpHK2ww5JSZKUy2zIVITjj4emTeGjj2DgwLTTSJIkVSkHHpiUJEm5zIZMRWjQAE49NTl2lowkSVK5OvropCRJymU2ZCpK376w0krw/vvw9ttpp5EkSZIkSVnEhkxFadQITjklOb7oonSzSJIkVSE9eiQlSVIusyFTkU48EZo0SWbIvPtu2mkkSZKqhN69k5IkKZfZkKlIjRvDyScnx86SkSRJKhc2ZCRJVYENmYp24onJ8qWBA+GDD9JOI0mSlPNmz05KkqRcZkOmoq20Epx0UnLsLBlJkqQVtuOOSUmSlMtsyFSGk0+Ghg3h9dfh44/TTiNJkpTTjjoqKUmScpkNmcqw8spwwgnJsbNkJEmSVsihhyYlSVIusyFTWU45BerXh1degU8/TTuNJElSziosTEqSpFxmQ6ayNGsGffsmxxdfnG4WSZKkHLbrrklJkpTLbMhUptNOg/x8ePFF+OyztNNIkiTlpOOOS0qSpFxmQ6YyNW8O//53cuwsGUmSpOVy4IFJSZKUy2zIVLZ+/SAvD557DoYMSTuNJElSzpk8OSlJknKZDZnK1rIl9OmTHDtLRpIkaZn16pWUJEm5zIZMGk4/HerVg6efhqFD004jSZKUU048MSlJknKZDZk0tGoFxxyTHF9ySbpZJEmScsw++yQlSVIusyGTlv/8B+rUgaeeguHD004jSZKUMyZOTEqSpFxmQyYtbdrA0UdDjM6SkSRJWgb77ZeUJEm5zIZMms44A2rXhscfh6+/TjuNJElSTjjttKQkScplNmTS1K4d/OtfzpKRJElaBnvskZQkSbmswhsyIYRdQgjfhhB+CCH0X8J5m4QQ5oYQqtcE1P79k1kyjz0G336bdhpJkqSs9+uvSUmSlMsqtCETQqgJ3AL0BDoDB4cQOi/mvCuB1yoyT1ZabTXo3RvmzYNLL007jSRJUtY76KCkJEnKZRU9Q2ZT4IcY46gY4yzgMaBXKeedAPwP+L2C82SnM8+EWrXg4Yfhhx/STiNJkpTV+vdPSpKkXFbRDZk2wNgS98dlHlsghNAG2Bu4bUkvFEI4JoQwOIQweMKECeUeNFWrrw6HH+4sGUmSpDLYZZekJEnKZRXdkAmlPBYXuX89cEaMce6SXijGeEeMsSDGWNC8efPyypc9zjoLataEBx+EUaPSTiNJkpS1xo5NSpKkXFbRDZlxQLsS99sC4xc5pwB4LIQwGtgP+L8Qwl4VnCv7rLEGHHoozJ0Ll12WdhpJkqSsddhhSUmSlMtCjItOWCnHFw+hFvAdsD3wMzAI+EeMcfhizr8PeDHG+NSSXregoCAOHjy4nNNmge+/h06doEaN5Lh9+7QTSZIkZZ0BA5LbHXZIN4eyWwjhsxhjQdo5JGlxKnSGTIxxDtCX5OpJXwNPxBiHhxD6hBD6VOR756Q114R//APmzIHLL087jSRJUlbaYQebMZKk3FehM2QqSpWdIQPwzTfQuXNy1aUffoBVV007kSRJUlaZv91ehw7p5lB2c4aMpGxX0XvIaFl16gQHHQSzZ8MVV6SdRpIkKesceWRSkiTlMmfIZKMRI2C99aB2bRg5Etq2TTuRJElS1njnneR2m23SzaHs5gwZSdnOGTLZqHNn2H9/mDULrrwy7TSSJElZZZttbMZIknKfDZlsdc45ye2dd8L4Ra8ULkmSVH19+21SkiTlMhsy2Wr99WHffWHmTPjvf9NOI0mSlDWOPTYpSZJymXvIZLMvv4SuXaFePfjxR1hllbQTSZIkpe7DD5Pb7t3TzaHs5h4ykrKdM2Sy2QYbwF57wYwZcNVVaaeRJEnKCt2724yRJOU+GzLZ7rzzkttbb4Xff083iyRJUhb46qukJEnKZTZkst2GG8Iee0BREVx9ddppJEmSUte3b1KSJOUy95DJBYMHwyabQH4+jB4NzZunnUiSJCk1gwYlt5tskm4OZTf3kJGU7ZwhkwsKCmDXXaGwEK69Nu00kiRJqdpkE5sxkqTcZ0MmV8zfS+bmm2HSpHSzSJIkpWjIkKQkScplNmRyxWabwc47w7RpcN11aaeRJElKzcknJyVJUi6zIZNLzj8/ub3xRvjzz3SzSJIkpeT665OSJCmX2ZDJJZtvDjvsAFOn+q8QSZJUbXXtmpQkSbnMhkyumT9L5oYb4K+/Uo0iSZKUhkGDiq+0JElSrrIhk2u23BK23RYmT06WLkmSJFUzp5+elCRJucyGTC6aP0vmuutgypR0s0iSJFWym29OSpKkXGZDJhdtsw1svXWyZOmmm9JOI0mSVKnWWy8pSZJymQ2ZXDV/lsy11yab/EqSJFUTH36YlCRJucyGTK7adlvYYgv44w+45Za000iSJFWas85KSpKkXGZDJleFUDxL5uqrYdq0dPNIkiRVkttvT0qSpFxmQyaX7bADdOsGkybBrbemnUaSJKlSrL12UpIk5TIbMrms5CyZq66C6dPTzSNJklQJ3nknKUmScpkNmVy3886wySYwYYJzdyVJUrVw/vnFf5OSJClX2ZDJdSVnyfz3v1BYmG4eSZKkCnbPPUlJkpTLbMhUBbvuChtvDL/9BnfemXYaSZKkCtWhQ1KSJOUyGzJVQQhw3nnJ8ZVXwowZ6eaRJEmqQAMGJCVJUi6zIVNV7LEHdO0Kv/wCd92VdhpJkqQKc8klSUmSlMtsyFQVJWfJXHEFzJyZbh5JkqQK8uCDSUmSlMtsyFQlvXrB+uvDzz+7050kSaqy2rVLSpKkXGZDpiqpUaN4lszllztLRpIkVUmvvpqUJEm5zIZMVbPPPrDuujB2LNx/f9ppJEmSyt0VVyQlSVIusyFT1dSoAeeemxxfdhnMmpVuHkmSpHL22GNJSZKUy2zIVEX77QedOsGYMe54J0mSqpxVVklKkqRcZkOmKqpZs3iWzKWXwuzZ6eaRJEkqRy+8kJQkSbnMhkxVdeCBsNZa8OOP8PDDaaeRJEkqN9dck5QkSbnMhkxVVbMmnHNOcnzJJTBnTrp5JEmSyslTTyUlSVIusyFTlR18MHTsCCNHwqOPpp1GkiSpXDRrlpQkSbnMhkxVVqsWnH12cnzJJTB3brp5JEmSysHTTyclSVIusyFT1R1yCHToAN99B48/nnYaSZKkFXbjjUlJkpTLbMhUdbVrw1lnJccXX+wsGUmSlPOeey4pSZJymQ2Z6uDww2G11eCbb9wBT5Ik5bzGjZOSJCmX2ZCpDhadJTNvXrp5JEmSVsDjj7sSW5KU+2zIVBe9e0O7djB8uLvgSZKknHbrrUlJkpTLbMhUF3XqwJlnJscXXeQsGUmSlLNefjkpSZJymQ2Z6uTII6FNGxg2zJ3wJElSzsrPT0qSpFxmQ6Y6qVsX+vdPji+6CGJMN48kSdJyeOihpCRJymU2ZKqbo46CVq1gyBB44YW000iSJC2zu+5KSpKkXGZDprqpVw/OOCM5vvBCZ8lIkqSc88YbSUmSlMtsyFRHxxwDLVvC55+7I54kSco5tWsnJUlSLrMhUx3l5cF//pMcO0tGkiTlmPvuS0qSpFxmQ6a66tMHWrSAQYPg1VfTTiNJklRmNmQkSVWBDZnqKj8f+vVLjp0lI0mScsjbbyclSVIusyFTnR13HDRrBp984s54kiRJkiRVIhsy1VmDBnDaacmxs2QkSVKOuPPOpCRJymU2ZKq744+HlVeGDz+EN99MO40kSdJSPf54UpIk5TIbMtVdw4Zw6qnJ8UUXpZtFkiSpDAYMSEqSpFxmQ0ZwwgnQpAm8+6475EmSJEmSVAlsyAgaNYJTTkmOnSUjSZKy3P/9X1KSJOUyGzJKnHgiNG4Mb70F772XdhpJkqTFeuGFpCRJymU2ZJRo0gROOik5dpaMJEnKYq+8kpQkSbnMhoyKnXxyssnvgAHJVZckSZIkSVKFsCGjYiutlCxdAmfJSJKkrHXDDUlJkpTLbMhoYaecAg0awGuvwSefpJ1GkiTpbwYOTEqSpFxmQ0YLa9oU+vZNjp0lI0mSstDzzyclSVIusyGjvzvtNKhfH15+GQYNSjuNJEmSJElVjg0Z/V2zZnD88cnxxRenm0WSJGkRV1+dlCRJucyGjEp32mmQnw8vvACff552GkmSpAU++igpSZJymQ0Zla5FCzjuuOTYWTKSJCmL/O9/SUmSlMtsyGjx+vWDevXg2Wfhyy/TTiNJkiRJUpVhQ0aLt8oq0KdPcuwsGUmSlCWuuCIpSZJymQ0ZLdnpp0Pdusm84GHD0k4jSZLEkCFJSZKUy2zIaMlat4ZjjkmOL7kk3SySJEnAY48lJUlSLrMho6X7z3+gTh148kkYMSLtNJIkSZIk5TwbMlq6tm3hqKMgRmfJSJKk1F18sdvbSZJynw0Zlc0ZZ0Dt2sn84G++STuNJEmqxr79NilJknKZDRmVzaqrwpFHJrNkLr007TSSJKkae+ihpCRJymU2ZFR2/ftDrVrwyCPw3Xdpp5EkSZIkKWfZkFHZtW8PvXvDvHlw2WVpp5EkSdXUeeclJUlSLqvwhkwIYZcQwrchhB9CCP1LGe8VQhgaQhgSQhgcQtiyojNpBZx5JtSsmcwTHjky7TSSJKkaGjs2KUmSclmFNmRCCDWBW4CeQGfg4BBC50VOGwhsEGPsChwJ3FWRmbSCOnSAww+HuXPdS0aSJKXi3nuTkiQpl1X0DJlNgR9ijKNijLOAx4BeJU+IMU6LMcbM3fpARNntrLOSWTIPPAA//ph2GkmSJEmSck5FN2TaACUnlI7LPLaQEMLeIYRvgJdIZskom3XsCIccksyScS8ZSZJUyc48MylJknJZRTdkQimP/W0GTIzxmRhjJ2Av4OJSXyiEYzJ7zAyeMGFC+abUsjv7bKhRA+67D8aMSTuNJEmqRiZNSkqSpFxW0Q2ZcUC7EvfbAuMXd3KM8V1gjRBCs1LG7ogxFsQYC5o3b17+SbVs1loLDj4Y5syByy9PO40kSapG7rgjKUmScllFN2QGAWuGEFYPIdQBDgKeL3lCCKFjCCFkjjcC6gD+zSMXnH02hAD33OOlDiRJkiRJWgYV2pCJMc4B+gKvAV8DT8QYh4cQ+oQQ+mRO2xf4KoQwhOSKTAeW2ORX2WyddeDAA2H2bLjiirTTSJKkaqJfv6QkScplIRd7HwUFBXHw4MFpxxDA8OGw/vpQuzaMGgVt/rZnsyRJUrk6/vjk9pZb0s2h7BZC+CzGWJB2DklanIpesqSqbt11Yb/9YNYsuPLKtNNIkqRq4JZbbMZIknKfDRmtuHPPTW7vuAN++SXdLJIkSZIk5QAbMlpx668P++wDM2fCf/+bdhpJklTFnXxyUpIk5TIbMiof82fJ3HYb/PprulkkSZIkScpyNmRUPrp2hV69YMYMuPrqtNNIkqQq7Prrk5IkKZfZkFH5mT9L5tZb4fff080iSZIkSVIWsyGj8rPxxrD77lBYCNdck3YaSZJURR1/fPGlryVJylU2ZFS+zjsvub3lFpg4Md0skiSpSsrLS0qSpFxmQ0bla5NNoGdPmD4drr027TSSJKkKuvpqt6yTJOU+GzIqf/Nnydx0E/zxR7pZJEmSJEnKQjZkVP66dYOddoJp0+C669JOI0mSqphjjklKkqRcZkNGFeP885PbG2+EP/9MN4skSapSmjZNSpKkXGZDRhWje3fYfnuYMgVuuCHtNJIkqQq5/PKkJEnKZTZkVHHmz5K5/nqYPDnVKJIkSZIkZRMbMqo4W20FPXokzZgbb0w7jSRJqiKOOCIpSZJymQ0ZVaz5s2Suuy5ZviRJkrSC2rVLSpKkXGZDRhVrm22SmTJ//gk335x2GkmSVAVcdFFSkiTlMhsyqlghFM+SueYamDo13TySJEmSJGUBGzKqeNttl1x16Y8/4P/+L+00kiQpxx16aFKSJOUyGzKqeCVnyVx9NUyfnm4eSZKU09ZeOylJknKZDRlVjh13hM02g4kT4dZb004jSZJy2LnnJiVJUi6zIaPKUXKWzFVXQWFhunkkSZIkSUqRDRlVnl12gYIC+P13uP32tNNIkqQcddBBSUmSlMtsyKjylJwl89//QlFRunkkSVJO6to1KUmScpkNGVWu3XaDjTaCX3+FO+9MO40kScpB/fsnJUlSLrMho8oVApx3XnJ85ZUwY0a6eSRJkiRJSoENGVW+PfeEDTaA8ePh7rvTTiNJknLMvvsmJUlSLrMho8pXcpbMFVfAzJnp5pEkSTll882TkiQpl9mQUTr22gvWWw/GjYN77007jSRJyiH9+iUlSVIuW6aGTAihZQhh90y1qKhQqgZq1CieJXP55TBrVrp5JEmSJEmqRGVuyIQQDgA+BfYHDgA+CSHsV1HBVA3suy907gw//QT33592GkmSlCP23DMpSZJy2bLMkDkb2CTG+M8Y4+HApsC5FRNL1UKNGnBu5kfosstg9ux080iSpJyw/fZJSZKUy5alIVMjxvh7ifuTlvH50t/tvz906gSjR8ODD6adRpIk5YCTTkpKkqRctiwNlVdDCK+FEHqHEHoDLwEvV0wsVRs1a8I55yTHl17qLBlJkiRJUrVQ5oZMjPF04A6gC7ABcEeM8YyKCqZq5MADYc01YdQoeOSRtNNIkqQs17NnUpIk5bJay3JyjPF/wP8qKIuqq1q1klky//wnXHIJHHJI8pgkSVIp9tgj7QSSJK24pc6QCSG8n7mdGkKYUqKmhhCmVHxEVQv/+AessQb88AM89ljaaSRJUhb797+TkiQply21IRNj3DJz2zDG2KhENYwxNqr4iKoWatWCs89Oji+5BObOTTePJEmSJEkVqMx7yIQQ/nYJnNIek5bboYfC6qvDt9/CE0+knUaSJGWpHXZISpKkXLYsV1lat+SdEEItYOPyjaNqrXZtOOus5Pjii50lI0mSSnXggUlJkpTLyrKHzJkhhKlAl5L7xwC/Ac9VeEJVL4cfDquuCl9/Df9z/2hJkvR3Rx+dlCRJuawse8hcHmNsCFy1yP4xTWOMZ1ZCRlUndeoUz5K56CKYNy/dPJIkSZIkVYAyL1mKMZ4ZQlgphLBpCGHr+VWR4VRN9e4N7drB8OHwzDNpp5EkSVmmR4+kJEnKZcuyqe9RwLvAa8CFmdsLKiaWqrW6daF//+TYWTKSJGkRvXsnJUlSLluWTX1PAjYBxsQYtwU2BCZUSCrpX/+CNm1g6FB4/vm000iSpCxiQ0aSVBUsS0NmRoxxBkAIoW6M8Rtg7YqJpWqvbl0444zk+KKLIMZ080iSpKwxe3ZSkiTlsmVpyIwLITQBngXeCCE8B4yviFASkFw+oVUr+OILePHFtNNIkqQsseOOSUmSlMuWZVPfvWOMf8UYLwDOBe4G9qqgXBLUqwf/+U9yfOGFzpKRJEkAHHVUUpIk5bIQy/BLbgihBjA0xrhexUdauoKCgjh48OC0Y6gyFBZChw7w22/w0kuw665pJ5IkSVIOCCF8FmMsSDuHJC1OmWbIxBjnAV+GEFat4DzSwvLz4fTTk2NnyUiSJJK/1xQWpp1CkqQVsyx7yLQChocQBoYQnp9fFRVMWqBPH2jeHD79FF5/Pe00kiQpZbvu6qRZSVLuq7UM515YYSmkJalfH/r1S666dOGFsNNOEELaqSRJUkqOOy7tBJIkrbgy7SFTphcK4aMY4+bl8mJL4R4y1dC0adC+PUyalMyS8dIKkiRJWgL3kJGU7ZZlydLS1CvH15IW1qABnHZacuxeMpIkVWuTJyclSVIuK8+GjL8hq2L17QsrrwwffABvvZV2GkmSlJJevZKSJCmXlWdDRqpYDRvCKackxxddlG4WSZKUmhNPTEqSpFxWng0Zd1lVxTvhBGjSBN55JylJklTt7LNPUpIk5bLybMgcVo6vJZWucWM4+eTk2FkykiRVSxMnJiVJUi5bakMmhDA1hDBlcTX/vBjjVxUbVco46SRo1AjefBPefz/tNJIkqZLtt19SkiTlslpLOyHG2BAghHAR8CvwIMnypEOAhhWaTipNkyZJU+bii5NZMq+/nnYiSZJUieZfeFGSpFwWYhkvHxxC+CTGuNnSHqsMBQUFcfDgwZX9tsomf/wB7dvD1Knw4Yew+eZpJ5IkSVIWCSF8FmMsSDuHJC3OsuwhMzeEcEgIoWYIoUYI4RBgbkUFk5Zo5ZWTDX7BvWQkSapmfv01KUmSctmyNGT+ARwA/Jap/TOPSek49VSoXx9efRU+/TTtNJIkqZIcdFBSkiTlsqXuITNfjHE00KviokjLqGlT6NsXrrwymSXz4otpJ5IkSZWgf/+0E0iStOLKPEMmhLBWCGFgCOGrzP0uIYRzKi6aVAannQb5+fDSS+C+QpIkVQu77JKUJEm5bFmWLN0JnAnMBogxDgWcLKp0NW8O//53cnzxxelmkSRJlWLs2KQklU0IoUkI4akQwjchhK9DCGW+IkYIYc8QQlbOSwshtA8h5Mw2GiGEC0II/Up5fK8QQueUMvUOIdy8lHOW+WcghHBSCOGrEMLwEMLJJR6/IITwcwhhSKZ2zTy+RQhhaAhhUAihY+axJiGE10IIYTm+tLJkvCiEsENFvHZZLUtDJj/GuOhGHXPKM4y0XPr1g7w8eP55+OKLtNNIkqQKdthhSUkqsxuAV2OMnYANgK/L+sQY4/MxxisqLNmKac8y7msaQijzth2VaC8glYZMWSzrz0AIYT3gaGBTkp+33UMIa5Y45boYY9dMvZx57DRgX+As4LjMY+cCl8WyXhp6GYQQasYYz4sxDijv114Wy9KQmRhCWAOIACGE/YBfKiSVtCxatoQ+fZJjZ8lIklTlnXNOUpKWLoTQCNgauBsgxjgrxvhXKec1DyH8LzNDYVAIYYvM4wtmUIQQ1gghfJwZvyiEMK3E80/PPD40hHBh5rH2mRk5d2ZmSrweQsgLIawTQvi0xHPbhxCGZo43DiG8E0L4LDM7olXm8Y4hhAEhhC9DCJ9nfje9AtgqM9PilBBCvRDCvSGEYSGEL0II25b4Gp4MIbwAvB5CaBVCeDfzvK9CCFuV8v0YHUJoljkuCCG8nTnepsTsji9CCA0X9/VnHj87hPBtCGEAsHYp79Md2BO4KvOaa4QQuma+z0NDCM+EEFZahs/ruRDC4ZnjY0MID2eO3w4hXB9C+DDzNW9aymvuEUL4JPN1DQghtCzlZ+C+EMKNmdcZlekLLGod4OMYY2GMcQ7wDrB3KeeVNBvIA/KB2ZnPt02M8Z3STg4h9AwhPFHifo/M50sI4dYQwuDMz1zJz2J0COG8EML7wP6Zr2W/zNh5me/jVyGEO0JIZuVkvm9XhhA+DSF8N/9nJSRXn74687M2NIRwQubxUn9+F2dZGjLHA7cDnUIIPwMnU9y5ktJ1+ulQrx488wwMHZp2GkmSVIF22CEpSWXSAZgA3Jv5RfuuEEL9Us67gWTmwiYkMxXuWsw5N2TOGT//wRDCTsCaJDMiugIbhxC2zgyvCdwSY1wX+AvYN8b4NVAnhNAhc86BwBMhhNrATcB+McaNgXuASzPnPJx5nQ2A7iSTA/oD72VmWlxH8jsrMcb1gYOB+0MI9TLP3xz4Z4xxO5JZNa/FGLuSzOAYspTvYUn9gOMzz90KKFrc1x9C2Jhkm48NgX2ATRZ9sRjjh8DzwOmZr2Mk8ABwRoyxCzAMOL+UHIv7vI4Bzss0Dk4DTijxnPoxxu7Av0m+t4t6H+gWY9wQeAz4z2K+B62ALYHdSZpii/oK2DqE0DSEkA/sCrQrMd4308S4p0Sz6XLgDpI+w80kn/u5i3l/gDeAbiV+lg8EHs8cnx1jLAC6ANuEELqUeN6MGOOWMcbHFnm9m2OMm8QY1yNpDO1eYqxWjHHTTLb5n8UxwOrAhpnP6eGl/PyWqswNmRjjqBjjDkBzoFPmixhd1udLFapVKzjmmOTYWTKSJFVpo0YlJalMagEbAbdmftGeTtLIWNQOwM0hhCEkDYJG82d/lLA58GTm+JESj++UqS+Az4FOJA0KgB9jjEMyx5+RLDMCeAI4IHM8/5fptYH1gDcyOc4B2mZytIkxPgMQY5wRYyws5WvYEngwc843wBhgrczYGzHGPzLHg4AjQggXAOvHGKeW8lqL8wFwbQjhRKBJZgbI4r7+rYBnMjNFppB8X5cohNA487rzZ4bcTzLDaVGlfl4xxt+A84C3gNNKfM0AjwLEGN/NnN9kkddsC7wWQhgGnA6su5iYz8YY58UYRwAtFx3MNNyuJGmavAp8SfF2J7cCa5A0rn4Brsk8Z0iMsVuMcVuSJuL45NsRHg8hPDR/tk6J95iTee09QrIMbTfguczwASGEz0k+j3VZeDnY45Ru28zsoGHAdot87U9nbkv+/O4A3JbJQeb7XOrP72LeD1iGy16HEOqSdN7aA7UyM3iIMV5U1teQKtQZZ8Dtt8NTT8FXX8F666WdSJIkVYAjj0xu33471RhSrhgHjIsxfpK5/xSlN2RqAJvHGItKPhjKtp9qAC6PMd6+yHPbAzNLPDSXZPYBJL8YPxlCeBqIMcbvQwjrA8NjjJsv8jqNyhIik2Nxps8/iDG+m5nBsxvwYAjhqhjjA4ucP4fiCQz1Sjz3ihDCSySzPj4Oyaawi/v6Tyaz5UcFKPXzylgfmAS0XuTxRbMsev8m4NoY4/MhhB7ABYt575Kfaanf8xjj3WSWyYUQLiP5OSTTMCLz+J3AiyWfl1kqdA5Jk+5mkhkp7YETgbMXeZvHSWZF/QEMijFODSGsTjKLaZMY458hhPso8flR4uegxHvWA/4PKIgxjs006ko+Z/7XO5fiHkrg79+/QCk/v0uyLEuWngN6kfxgTi9RUnZo3RqOOio5vvBCKP+9nyRJUha48MKkJC1djPFXYGwIYf7+JdsDI0o59XWg7/w7IYSupZzzMckf6WHhK+6+BhwZQmiQeW6bEEKLpeQaSfIL7rkUz1r4FmgeMleBCiHUDiGsm5ldMi6EsFfm8bqZpTBTgZKzeN4FDsmcsxawauY1FxJCWA34PcZ4J0nTYKNSIo4GNs4cz/+aCSGsEWMcFmO8EhhMMhtmcV//u8DeIdk3pyGwx2K+HQu+jhjjZODPULyvzWEke7AsqtTPKyR7w/QkWSbVL9OgmO/AzDlbApMz71VSY+DnzPE/F5O1TOZ//iGEVUmWaz2auV9yT5W9SZY3lfRP4KUY458k+8nMy1R+KW/zNslndzTFP0ONSPoUkzOzanqWIe785svEzGdY2r44i3od6JOZnUMIYWUW8/O7pBdZlh2m28YYd1mG86XK178/3HlnMktmzz3httugTZu0U0mSpHK0zTZpJ5Byzgkke1zUAUYBR5RyzonALSHZXLcWSTOhzyLnnAw8FEI4DXgJmAwQY3w9hLAO8FFmRs004FCShsuSPA5cRbIXBzHGWSHZZPXGzNKdWsD1wHCSxsTtIYSLSDaA3R8YCswJIXwJ3Ecyy+G2zLKTOUDvGOPMUmb59ABODyHMzmQ9vJRsFwJ3hxDOAj4p8fjJIdkseC5JY+uVzHv87euPMX4eQnicZI+aMcB7i/k+PAbcmVkGtR9JU+K2TNOpzJ9XCOEk4E7giBjj+MzndE8IYbvMc/4MIXxI0rQ4spTXvIBk1tLPJM231Us5p6z+F0JoSvJZHZ9psAD8N9M8iiRNr2PnPyHz9f6TZPkXwLXA/4BZJHsCLSTGODeE8CLQO/M8YoxfhhC+IPmZGUWyxGyJYox/ZWbrDMtkGlSGr+8ukuVwQzM/R3fGGG9ews9vqUJZryAVQrgDuCnGOKxMT6hABQUFcfDgwWnHULZ6/HE49liYPBkaNYKrr05mzlTM5eslSVIl+zbz9+61/3a9EqlYCOGzzMaeKieZX5iLYowxhHAQcHCMsVfaubR0IblKVL8Yo79IZ5FlWbK0JfBZSC7ZNTRkLu9UUcGk5XbggTB8eDJDZsqUZLPf7beHkSPTTiZJksrBsccmJanSbQwMyfwe+G+Sq/hIWk7LMkNmtdIejzGOKddEZeAMGZVJjMlsmRNOgIkTIS8PLr0UTjwRatZMO50kSVpOH36Y3Hbvnm4OZTdnyEjKdkudIVNiR+upiykpO4UABx0EI0bAP/4BRUVw6qmw5ZbJY5IkKSd1724zRpKU+8qyZGn+9eU/I9lF+rMS5TQVZb/mzeHhh+H555MNfj/+GDbcEC6+GGbPTjudJElaRl99lZQkSbmszEuWsolLlrTcJk+G009PrsQE0KUL3HMPbLzxkp8nSZKyRo8eye3bb6eZQtnOJUuSst2ybOpLCGGlEMKmIYSt51dFBZMqROPGcMcdMHAgdOgAQ4fCZpsll8suKko7nSRJKoOrrkpKkqRcVuaGTAjhKJJr0b9Gck3210iuUy7lnu22S5oxp5wC8+bBlVdC167w3ntpJ5MkSUuxySZJSZKUy5ZlhsxJwCbAmBjjtsCGwISlPSmEsEvmUtk/hBD6lzJ+SOYy2kNDCB+GEDZYhkzS8qtfH669NrlUwzrrwHffwdZbQ9++MNX9qiVJylZDhiQlSVIuW5aGzIwY4wyAEELdGOM3wNpLekIIoSZwC9AT6AwcHELovMhpPwLbxBi7ABcDdyxDJmnFdesGX3wB55wDtWrBLbfAeuvB66+nnUySJJXi5JOTkqqTEEKLEELttHNIKj/L0pAZF0JoAjwLvBFCeA4Yv5TnbAr8EGMcFWOcBTwG9Cp5Qozxwxjjn5m7HwNtlyGTVD7q1k2uujR4MGy0Efz0E+y8MxxxBPz559KfL0mSKs311yclVRch1NwR6oyGhk+EEELaeSSVjzI3ZGKMe8cY/4oxXgCcC9wN7LWUp7UBxpa4Py7z2OL8C3iltIEQwjEhhMEhhMETJix1pZS0fDbYAD75BK64ImnS3HcfdO4MTz+ddjJJkpTRtWtSUnUQQt3jocFz8HwetN8J8i9KO5Ok8rHUhkwIYeVFCxgGvA80WNrTS3ms1OtshxC2JWnInFHaeIzxjhhjQYyxoHnz5kuLLS2/WrXgjDPgyy9hyy3h119h331h//3ht9/STidJUrU3aFBSUlUWQqgZQsNbYZX/wmd5sDPwej7UPy2EWgennU/SiivLDJnPgMGZ20Vr8FKeOw5oV+J+W0pZ5hRC6ALcBfSKMU4qQyap4q29NrzzDtx8c7IB8FNPJbNlHnwQYql9RUmSVAlOPz0pqaoKITSCRgNgg8NhSD50zIw0B3asBXmnpJlPUvkIsQJ/sQwh1AK+A7YHfgYGAf+IMQ4vcc6qwJvA4THGD8vyugUFBXHw4KX1gqRyNGYMHHNM8Ua/PXvCbbfBqqumm0uSpGroq6+S2/XWSzeHslsI4bMYY0HaOZZVCGE1aPAmHNAGbqsL8/fxnQbsWwgffwVTesYY/0gzp6QVtyyb+hJC2CeEcG0I4ZoQwl5LOz/GOAfoC7wGfA08EWMcHkLoE0LokzntPKAp8H8hhCEhBDstyj6rrQavvgr33gtNmsArr8C668Ktt8K8eWmnkySpWllvPZsxqppCCN0gfwhcuBrcVaIZMxbYeDp89AxM2cpmjFQ1lHmGTAjh/0jmyj2aeehAYGSM8fgKyrZYzpBRqn79FY4/vnij3623hrvugjXXTDeXJEnVxIeZOdXdu6ebQ9kt12bIJPvC1LsbHsuD3UuMfArsUgQzLoSi/8aKXOIgqVItS0NmOLDe/P8DCCHUAIbFGNetwHylsiGjrPDUU0lj5vffoV695LLZJ5+cbAosSZIqTI8eye3bb6eZQtkuVxoyyWWs8y+G+qfAgHzoUmL0SaB3Icz4R4xzn0sro6SKsSxLlr4FSm6Y0Q4YWr5xpByy334wYgQcdhjMmJHsLti9OwwblnYySZKqtNtvT0rKdSGEetDwaVjjZBhaohkTgYvnQO9JULilzRipalqWhkxT4OsQwtshhLeBEUCLEMLzIYTnKySdlO2aNoUHHoCXXoK2bZNrcG68MVxwAcyalXY6SZKqpLXXTkrKZSGEltDwE9h+Z/ikPqySGZkJHDQDrvoeCtePMX6RZk5JFWdZlixts6TxGOM75ZKoDFyypKw0ZQr0759s9AvJboP33AObbJJuLkmSqph3Mv/q3GaJ/zpVdZfNS5ZCCOtB/YFw0kpwce3iv5NPAHaZDj+8DVMOiDEWphhTUgVbloZM5xjjiEUe6xFjfLsigi2JDRlltXfegaOOgh9+gBo14JRT4KKLID8/7WSSJFUJ7iGjssjWhkwINXeFek/A7flwaCgeGQHsUAhTboHp/WOMXspTquKWZcnSEyGE/4REXgjhJuDyigom5axttoEvv4R+/ZL711wDG2xQ/Oc8SZK0Qu65JykplyS/RtU7BRo+BW/UX7gZ8xqwWRFM/HeM0/5jM0aqHpalIbMZyaa+HwKDgPHAFhURSsp5+flw1VXw8cfJ0qUffkj+nHfcccnSJkmStNw6dEhKyhUhhFrQ4E5ofQl8kQclr9l+yzzYZwpM2znGWfenFlJSpVuWhsxsoAjIA+oBP9q5lZZik03gs8+STX5r14bbboN114WXX047mSRJOWvAgKSkXBBCaAKN3oKCg+GLfFg9MzIH+Pcs6D8OCjeKMb6XYkxJKViWhswgkoZMAbAlcHAI4akKSSVVJXXqwPnnw+efJw2aceNgt93g8MNh0qS000mSlHMuuSQpKduFEDpAgyHwj03gjXxonBmZAuxcCA8PgmkbxBhHphhTUkqWpSFzNPA9cFaM8VfgBGBIRYSSqqT11oMPP0yWMtWrBw8+CJ07w5NPQhk315YkScl/Qh98MO0U0pKFELaE/M/hsnZwa12olRkZDWxYCIOegCnbxhj/Si+lpDQtS0PmCKAbcHDm/lSgV7knkqqyWrWSzX6HDoWtt4bff4cDDoB994Vffkk7nSRJOaFdu6SkbBVCncOhwevwv8ZwQonfuT4CNiyC8efA1CNjjLNTCykpdcu0qW+M8XhgBkCM8U+gdoWkkqq6NdeEt96CW2+Fhg3hmWeS2TL33utsGUmSluLVV5OSsk0IoUYI9a+ElW+Fj/NglxKjjwA7TIcp+8dYdF2M/qNPqu6WaVPfEEJNIAKEEJrPP5a0HGrUgD59YPhw6NkT/voLjjwSdtkFRo9OO50kSVnriiuSkrJJCCEfGj4Hax8PQ/Nh3cxIBM6dA8dMgMLNY5z7Upo5JWWPZWnI3Ag8A7QIIVwKvA9cViGppOqkXTt46aVkMfzKK8Prryf7zdx8M8zzQmaSJC3qsceSkrJFCKEVNBwEu+wAH9aHFpmRImC/IrjhG5i+foxxWJo5JWWXsCwz5UIInYDtgQAMjDF+XVHBlqSgoCAOHjw4jbeWKtZvv8EJJyQb/QJssQXcfTesvXa6uSRJknJMCOGzGGNBJbzPBlB/AJzeBM6rlfyqBPAbsPN0GDUAph4cYyyq6CyScsuyzJAhxvhNjPGWGOPNaTVjpCqtZUt44gl4+mlYZRX44APYYINkXvacOWmnkyQpK7zwQlJS2kKouQfkfwh3N4PzSzRjhgFdCuH7G2Dq3jZjJJVmmRoykirJ3nvDiBFwxBEwcyaceSZsthl8+WXaySRJSt011yQlpSWEEELIOx0aPw5v5sOBJUZfBroXwqSjY5x+tpv3SlocGzJStlppJbjnHnjtNVhtNfj8cygogHPPTZo0kiRVU089lZSUhhBCbWh4P7Q7H4bkwWaZkQjcMA/2/wum7RDjnEdSjCkpB9iQkbLdTjvBsGHQt2+ybOmSS2DDDeHjj9NOJklSKpo1S0qqbCGElaDRu9BtX/isPqyaGZkNHD0TzhkDhRvGGD9KM6ek3GBDRsoFDRvCTTfBe+/BWmvB119D9+5wyikwfXra6SRJqlRPP52UVJlCCGtC/aHwz43glXxomBn5C9i+EJ78GKZ1jTGOTi+lpFxiQ0bKJVtuCUOGQP/+UKMGXH89dOkCb76ZdjJJkirNjTcmJVWWEMI2kD8Yrm4NN9aBmpmRkUDXQhjyIEzZIcY4Jc2cknKLDRkp1+TlweWXwyefJFdgGjUKtt8ejj4aJk9OO50kSRXuueeSkipDCHX/BQ1fgecaQZ8Svz+9B2xcBL+cEeOUPjFGL4kpaZnYkJFy1cYbw6BBcPHFUKcO3HUXdO7sdUAlSVVe48ZJSRUphFAjhAbXQbMb4dM82KHE6P0RdpmWXNJ65s2phZSU02zISLmsdm045xz44gvo1g3Gj4c994R//AMmTEg7nSRJFeLxx5OSKkoIoX4yK6bzMfBlPnTKjMwD+s+G43+Hws1inPtamjkl5TYbMlJV0LkzvP8+XHcd5OfDo48mjz32GMSYdjpJksrVrbcmJVWEEEIbaPg57LE1vJcP8y/pVQjsVQT/Nxymrx9jHJFmTkm5z4aMVFXUrAknn5xcInu77WDiRDj4YNhrL/j557TTSZJUbl5+OSmpvIUQNob8odC/AzxUD+pmRn4BNi2Et1+Gqd1ijE5FlrTCbMhIVU2HDjBgANx5JzRqBM8/D+uum+wx42wZSVIVkJ+flFSeQqi5N+S/Cw+uDGfVgpAZGQJ0KYIfr4Kp+8cYZ6YYU1IVYkNGqopCgKOOguHDYffdk6svHX007LhjclUmSZJy2EMPJSWVhxBCCCHvbGjyMLybD/uUGH0e2LIQ/uwd4/QLYvSvW5LKjw0ZqSpr2zaZIfPII9CsGQwcCOuvD9dfD3Pnpp1OkqTlctddSUkrKoRQBxo+DO3PhC/zYOPMSAT+OxcO/hOmbxvjnCfSzCmparIhI1V1ISR7yYwYAQcdBIWFcMopsNVW8PXXaaeTJGmZvfFGUtKKCCE0hYYfwpZ7waD60DYzMgs4YgZc8iMUbhBj/DTFmJKqMBsyUnXRvHly9aXnnoPWreGjj6BrV7j0Upg9O+10kiSVWe3aSUnLK4SwNtQfCsesDy/mQYPMyB9Aj0J45gOYumGMcWyaOSVVbTZkpOpmzz2TvWWOOgpmzYJzzoFNNoHPP087mSRJZXLffUlJyyOEmttD/iC4oRVcXaf4V6LvgA0KYdg9MGXnGOO0NHNKqvpsyEjVUZMmyVWYBgyA1VeHL7+ETTeFM8+EGTPSTidJ0hLZkNHyCqHecVD/BXipIfwrFI+8BRQUwW+nxjj1hBijm+1JqnA2ZKTqbPvtYdgwOPlkmDcPrrgiWcb0wQdpJ5MkabHefjspqaxCCDVDaHgLtLgaPsuDHiVG74qw+1SYvnuMs25PK6Ok6seGjFTd1a8P112XNGHWWQe+/TbZ8PfEE2GaM3UlSVJuCyE0hIavw/q9YUg+rJkZmQucOgtO/gUKN4lx7pspxpRUDdmQkZTYfHP44gs4+2yoUQNuugnWW8/LWEiSss6ddyYlLU0IYVVo8AXsswW8kw8rZ0amAbsXwd1DYXqXGOO3aeaUVD3ZkJFUrG5duOQSGDwYNtwQxoyBnXaCI4+EP/9MO50kSQA8/nhS0lLkQ/4QOK893FsX5l+aaxxQUAgfPAtTtogxTkotoaRqzYaMpL/r2hU++QQuvzxp0tx7L3TuDM8+m3YySZIYMCApaXEee+wJIH9teHQlOL0mzN+/dzCwQRGMvRSmHhJjnJViTEnVnA0ZSaWrXRv694chQ2CLLeDXX2HvveGAA+C339JOJ0mS9DcxRs4//xKOPLIf8EEN2LPE6P+AbQrhr0NinH5ZjDGmFFOSABsykpamUyd491248cZkA+Ann0xmyzz0EPjvGElSCv7v/5KSSpo5cyb77384V1/9HEVFnwBdMyMRuHQOHP4HFG4V49xn0kspScVsyEhauho14IQT4KuvYMcd4Y8/4LDDYPfdYezYtNNJkqqZF15ISppvwoQJdOu2PS+/PIPCwneAVpmRmcChM+DKkVDYJcb4eYoxJWkhNmQklV379vDaa3DPPdCkCbz8Mqy7Ltx+O8ybl3Y6SVI18corSUkAI0aMYP31N2P48B4UFT0O5GdGJgJbFcILb8PUjWOMP6cWUpJKYUNG0rIJAY44AkaMgL32gqlToU8f2H57+OGHtNNJkqRq5NVXX2PTTXvw228XMnv2JRT/evMN0GUejLgNpu4WY5yeYkxJKpUNGUnLp1UrePppeOIJaNEC3n4bunSBa66BuXPTTidJqsJuuCEpVW833vh/7LNPb6ZPfxo4rMTIAPLytgF++ynGaafFGJ3GKykr2ZCRtPxCgP33T2bLHHooFBVBv37QvXuy34wkSRVg4MCkVD3NmTOHY445kTPPvJmiog+ALReMhXA7jRodyquvPgnMm5RaSEkqg5CLV3srKCiIgwcPTjuGpEW99FKyfGncuOSy2WefDWeeCXXqpJ1MkiRVAZMnT2aPPQ7is8/mUlj4BNAkMzKXOnX60aLFK7z11ot07NiREMJnMcaCFONK0hI5Q0ZS+dltNxg+HI49FmbPhgsugIICGDQo7WSSJCnH/fjjj3TtugWffro6hYUvU9yMmUpeXi822mgYQ4d+RMeOHVNMKUllZ0NGUvlq1Ahuuw3eegvWWAOGDYNu3eA//4Hff087nSSpCrj66qRUfXz44YdsuGF3fvrpWGbOvAWolRkZQ37+Fuy3XxveffcVVlpppTRjStIysSEjqWL06AFDh8JppyX3r7oKVlkFttwy+Vf099+nGk+SlLs++igpVQ8PPfQIO+64F5Mn3828eScAITPyCXl5m3PBBUdw//23Ubt27TRjStIycw8ZSRXv00/hwgthwACYNav48c6doVev5PLZBQVQwx6xJElKxBg566wLuOGG+ykqegFYv8To4+Tn9+XRR+9hzz33KPX57iEjKdvZkJFUeaZOhVdfheeegxdfhMmTi8datUqaM716wbbbQt266eWUJEmpKioq4sADj2DgwDEUFj4LtMyMRGrVupgmTe5mwIDn2WCDDRb7GjZkJGU7GzKS0jF7Nrz7Ljz7bNKgGTu2eKxhQ+jZM5k507MnNGmSUkhJUja64orktn//dHOoYvz666/suONejBy5OkVF9wB5mZEZ1Kv3Lzp0+J4BA56jVatWS3wdGzKSsp3rAySlo3Zt2H57uOkmGDMGPv8czjsPunRJZtI88QT84x/QvDnstBPccsvCTRtJUrU1ZEhSqnqGDh1Kly7d+OabnhQVPUJxM+Z38vO3Y/vtZzN48DtLbcZIUi5whoyk7PPjj8msmeeeS2bRzJtXPLbxxsnMmV69YL31IITFvowkScodL774Egcd1Jvp028EDi4x8hX5+XvQt++hXH75hdQo455zzpCRlO1syEjKbpMmwUsvJUubXnsNCguLxzp0KN53ZostoFatxb6MJEnKTjFGrr32Rs4990qKiv4HbF5i9FXy8g7nttuu5fDDD12m17UhIynb2ZCRlDuKimDgwKQ58/zzMGFC8VjTprDHHklzZqedID8/tZiSpIp18cXJ7bnnpptDK2727Nkcc8yJPPHE+xQWvgistmCsRo2badjwUl566Sm22GKLZX5tGzKSsp1/TpaUO/LyYPfdk5o7Fz7+OFnW9Oyz8P33cN99SeXlwY47Js2ZPfZI9qGRJFUZ336bdgKVh7/++ovddjuAIUNqUVj4AdAoMzKHOnVOplWrt3jzzQ/o0KFDmjElqcI4Q0ZS7osRvvmm+IpNn3xSPFajBnTvXrzvTMeOaaWUJEkZI0eOZNttd+e333Zm1qyrKf478WTy8w9ko43gxRcfp3Hjxsv9Hs6QkZTtbMhIqnrGj4cXXkgaNG++CbNmFY+tu25xc2bjjZOGjSRJqjTvvfceu+++P1Onnk+Mx5UY+ZH8/N05+OBtue2266m1gnvD2ZCRlO1syEiq2qZMgVdfTWbOvPQSTJ5cPNamDey5Z9Kc2XZbqFMnvZySpDI777zk9qKL0s2hZXfvvfdz/PGnU1T0ELBTiZEPyMvbj8suO5uTT+5bLu9lQ0ZStrMhI6n6mDUruYz2/KVN48YVjzVqBD17JrNnevaEFZgiLUmqWEcckdzee2+6OVR28+bN4z//OYdbb32cwsIXgM4lRh+ifv1TefLJB+jZc5dye08bMpKynQ0ZSdVTjPD558WbAg8bVjxWu3YyY6ZXr2QGTdu2qcWUJCnXFRYWsu++h/Huu79TWPg0MH+z/XnUqnU+K6/8EAMHvsB6661Xru9rQ0ZStrMhI0kAo0YlzZnnnoP33oN584rHCgqK951Zd10IIbWYkiTlkvHjx7P99nsyenRnZsy4E6ibGSmiXr3erLXWWN5441latGhR7u9tQ0ZStrMhI0mLmjgx2W/m2WfhtdegqKh4bI01ksbMXnslV2+qWTOtlJJUbZ15ZnJ7+eXp5tCSffHFF+y4Yy8mT+7DnDlnAvP/oPEr+fm92Gmnjjz66N3Uq1evQt7fhoykbOflRSRpUc2awT//Cc88kzRnnnsOjjwyeXzkSLj2Wth6a1hlleTx556DwsK0U0tStTFpUlLKXs8++xxbbrkTkyZdy5w5Z1HcjPmSvLzNOOWU3Xj66YcqrBkjSbnAGTKSVFZz58JHHyUzZ559NmnOzJeXBzvtlMyc2X33pHkjSVI1E2Pkiiuu4eKLr6Oo6FlgkxKjL5KXdwR33XUT//jHQRWexRkykrKdDRlJWh4xwogRxZsCDxpUPFajBmy5ZbK0qVevZJmTJElV3KxZszjiiH/z7LODM1dSapcZidSocT2NGl3FK688Tbdu3Soljw0ZSdnOhowklYeff4bnn08aNG++CbNnF4+tt17xpsAbb+ymwJK0gvr1S26vvjrdHCr2xx9/sMsu+/LVV40oKnoYaJAZmU3dun1p0+Yj3nzzBVZbbbVKy2RDRlK2cw8ZSSoPbdrAccfBq6/ChAnw2GNw0EHQqBF89RVccglssgmsuiocfzy88QbMmpV2aknKSUVFC++3rnR99913dOnSjSFDCigqepriZsyf5Of3pFu3cXzxxfuV2oyRpFzgDBlJqkizZsHbbxdfUvvnn4vHGjWC3XZLZs707JnclyQph7z11lv06nUQ06ZdSoxHlRj5gfz83Tn88J7cfPPV1EzhqoTOkJGU7WzISFJliRE++6x435mvvioeq10bttsuWdq0557QunVaKSVJKpM77ribk08+i6KiR4HtSoy8S17eAVx11QUcf3yftOLZkJGU9WzISFJaRo4snjnz/vswb17x2CabFO8707mz+85IUgknn5zcXn99mimqr7lz53LqqWdy113PUFj4ErBWidH7aNDgDJ5++mF23HGHtCICNmQkZT8bMpKUDSZMgJdeSmbOvP76wpsjdOyYNGb22gs23xxSmPYtSdnEhkx6pk2bxj77HMoHH/xFYeH/gKaZkXnUrn02zZo9yZtvvkinTp3SjAnYkJGU/WzISFK2KSxMNv197jl44QWYOLF4rHlz2GOPpEGz446Ql5deTklStTJu3Di2224Pxo7diBkzbgXqZEamk5d3OOusM4HXXnuaZs2apRlzARsykrKdDRlJymZz58KHHyYzZ559FkaNKh7Lz4eddkpmzuy+OzRtupgXkSRpxQwePJiddtqLKVNOYu7cfsD8pbQ/k5+/J7vvvj4PPHA7devWTTPmQmzISMp2NmQkKVfECMOHF28KXPL/B2vUgK22SmbO9OoFHTqkFlOSKtrxxye3t9ySbo7q4oknnuKII46jsPBOYK8SI5+Tl9eLM888nnPOOYOQZfud2ZCRlO1qpB1AklRGIcB668HZZ8OgQTB2bPLbyE47JQ2Zd96BU0+FNdaALl3gvPOSqzrlYONdkpYkL88Vm5UhxshFF11O796nUFj4Ggs3Y54lP39n7r//es49t3/WNWMkKRc4Q0aSqoLJk+GVV5KZMy+/DFOnFo+1a5dcSnuvvWCbbZJLbEuStAQzZ87ksMOO4aWXvqKw8HmgTWYkUrPmVTRqdCOvv/4sBQXZOwHFGTKSsp0NGUmqambOhLffLr6k9vjxxWONG8NuuyXLmnbZBRo1Si2mJCk7TZw4kZ122ptvvmlBUdEDQP3MyCzq1etDu3Zf8OabL9C2bds0Yy6VDRlJ2c4lS5JU1dStCzvvDP/3f8mypk8/hbPOgnXXTWbSPPIIHHhgcsWm/faDoUPTTixJy+SYY5JS+fv666/p0qUbX321FUVFT1LcjJlEfv5ObLnlJD7//L2sb8ZIUi6wISNJVVmNGrDJJnDppfDVV/D993D11bDlljB7Nvzvf7DBBkmD5uuv004rSWXStKkXlqsIr7/+Bptu2oNffjmX2bMvo/hXhW/Jz+/GUUdtyquvPk2DBg3SjClJVUaFN2RCCLuEEL4NIfwQQuhfyninEMJHIYSZIYR+FZ1Hkqq1jh3htNPgvfdg3Dg46SSoUweeeCLZMPjww2HkyLRTStISXX55Uio/t9xyG3vtdRjTpj0J/LPEyJvk5W3Nddf154Yb/kvNmjXTiihJVU6FNmRCCDWBW4CeQGfg4BBC50VO+wM4Ebi6IrNIkhbRujVcfz388AMce2wym+bBB2HtteHoo2HMmLQTSpIq2Ny5cznuuFP4z3+up6jofWDrBWMh3EXDhgfz0kuPccwx/0ovpCRVURU9Q2ZT4IcY46gY4yzgMaBXyRNijL/HGAcBsys4iySpNO3awW23wXffQe/eyWWy77oL1lwT+vZdeFNgScoCRxyRlFbM1KlT2WGHXjzwwDAKCz8COmZG5lKnTj9at/4vgwe/x7bbbptmTEmqsiq6IdMGGFvi/jiKr5m3TEIIx4QQBocQBk+YMKFcwkmSSlh9dbj3XhgxAg4+GObMgVtugTXWSJY5/f572gklCUj6yO3apZ0it40ZM4auXbfgo4/aUlj4CrBSZmQaeXn70LXr5wwd+jFrrbVWmjElqUqr6IZMKOWx5brOdozxjhhjQYyxoHnz5isYS5K0WGuvnVyJaehQ2GcfmDEDrr0WOnRIrtb0xx9pJ5RUzV10UVJaPp988gldu3Zn9OgjmTnzVqB2ZmQs+flbsvfezXnvvVdZeeWV04wpSVVeRTdkxgEl/37RFnDuuyTlgvXWS67C9NlnsNtuMH16sovm6qvDhRcml9CWJOWURx55jO2224O//rqdefNOpvjvp4PIy9ucc889jIceupM6deqkmFKSqoeKbsgMAtYMIaweQqgDHAQ8X8HvKUkqTxttBC++CB99BDvsAFOmwAUXJI2ZK65IGjWSVIkOPTQplV2MkbPPvpCjjjqDwsIBwO4lRp8kP39XHnnkFvr3P40QSpvkLkkqbxXakIkxzgH6Aq8BXwNPxBiHhxD6hBD6AIQQVgkhjANOBc4JIYwLITSqyFySpOXQrRu88Qa8/TZstRX8+SeceWbSmLnuOigqSjuhpGpi7bWTUtnMmDGDffY5hOuvf5miok+ALpmRSK1al9K06Wm8//7r7LVXryW9jCSpnIUYl2tLl1QVFBTEwYMHpx1DkqqvGGHAADjnHPj00+Sx1q3h7LPhX/+CunXTzSdJAuC3335jp5325vvv21FUdB+QlxmZSb16R7H66t8wYMBztG7dOsWUFSOE8FmMsSDtHJK0OBW9ZEmSVBWFADvuCB9/DC+8AF27JpfHPv745M/Wd98Ns2ennVKSqrWvvvqKLl26MWLEDhQVPUpxM2YC+fnbs+22RQwe/E6VbMZIUi6wISNJWn4hwO67Jxv/PvUUdO4MY8bAUUclxw89BHPnpp1SUhVz0EFJafFefvkVunXbjt9/v4Q5cy6i+J/9I8jP78a//70NL774BPn5+WnGlKRqzYaMJGnF1agB++6bXCr74YdhzTXhhx/gsMNg/fXhySdh3ry0U0qqIrp2TUqlu+66m9hvvyOZPv0Z4JASI6+Tl9eDW245n6uuupQaNfxVQJLS5B4ykqTyN2cOPPggXHQRjB6dPLbBBsn9PfZIZtZIksrVnDlz6NPnZB599C0KC18EVl8wFsKtNGx4IS+++CRbbbVVeiErkXvISMp2tsUlSeWvVi044gj49lu49VZo0wa+/BJ69YLNNoPXXks2BpYklYvJkyfTo8duPProSAoLP6S4GTOHOnVOom3bG/n88w+qTTNGknKBDRlJUsWpUwf69EmWL11/PbRsCYMGwS67JJfOfvvttBNKykH77puUEqNGjWKDDbozaNBaFBa+ADTOjEwhP39PNt54BEOHfsQaa6yRZkxJ0iJsyEiSKl69enDSSTByJFx5Jay8MnzwAWy7LeywA3z0UdoJJeWQzTdPSvDBBx+w0UZbMHbsccyadRNQKzMymvz8Ldh//1V5552XadKkSYopJUmlcQ8ZSVLlmzIFbrgBrr46OQbYdddkj5mNN043myTliAceeIg+fU6lqOgBYJcSIx+Rl7cvF198BqeeeiKhmu7b5R4ykrKdM2QkSZWvUSM499xkw9+zz4b69eHll6GgAPbZB4YNSzuhJGWtefPmccYZ53LccedSVPQWCzdjHqV+/V488cSdnHbaSdW2GSNJucCGjCQpPSutBJdcAj/+CP36JUubnnkmuSLTwQcnmwJL0iL23DOp6qiwsJA99zyIm29+k8LCT4B1MyORWrUupHnzM/noo4HsvvtuacaUJJWBDRlJUvqaN4erroJRo+CEE6B2bXjsMejcGXr3Th6XpIztt0+quvnll1/YZJMeDBxYh8LCgUCLzEgR9eodQqdOrzBs2Cesv/76acaUJJWRDRlJUvZo1QpuvDG5KtMxx0CNGnD//bD22nDssTB2bNoJJWWBk05Kqjr58ssv6dKlG999twczZjwI1MuM/EZ+/nbstFPk00/fomXLlmnGlCQtAxsykqTs064d3H57smTpn/+EefPgjjugY0c48UT45Ze0E0pSpXn++Rfo3n0HJk78L3PmnAvM3xdmGHl5m3HSSTvz7LOPkJeXl2ZMSdIysiEjScpeHTrAfffB8OFw0EEwezbcdBOssQacfjpMmJB2Qkkp6Nkzqaouxsh//3stBx3Uh8LCF4EDS4y+TH7+9txxx2VcdtkFbt4rSTnIhowkKft16gSPPgpffgl77w1FRckls1dfHc45B/78M+2EkirRHnskVZXNnj2bf/6zDxdeeD9FRR8Bm2VGIjVq3Ejjxv/ijTee5dBD/5FmTEnSCggxxrQzLLOCgoI4ePDgtGNIktLy2Wdw3nnJpbIBGjeG005LNpVo1CjdbJK0gv7880969tyPoUPzKSp6BGiYGZlN3bon0arVu7z11ou0b98+xZTZL4TwWYyxIO0ckrQ4zpCRJOWejTeGl16CDz9MLrUyeXLSoFl9dfjvf2H69LQTStJy+eGHH+jSpRtffLEBRUXPUtyM+Yv8/N3YdNMf+fLLD23GSFIVYENGkpS7Nt8cBgyAt96CLbeEP/6AM85I9p654QaYMSPthJIqwA47JFXVvPPOO2y88Zb8/POpzJp1LVAzMzKK/PzuHHJIJ9588wUaORNQkqoEGzKSpNzXowe8+y68+ipssgn8/jucfHJyVabbboNZs9JOKKkcHXhgUlXJXXfdS8+eBzBlykPEeGyJkffJy+vO5Zcfzx133EitWrVSyyhJKl/uISNJqlpihBdfhHPPTTYBBmjfPlnSdNhh4C8zkrLIvHnzOO20s7jjjqcyV1LqVGL0AerX78dTTz3ILrvsnFbEnOUeMpKynTNkJElVSwjJ5Vc+/xyeeALWWQdGj4Yjj4TOneGRR2Du3LRTShLTp09n11334447PqSw8GOKmzHzqF37HFq2vIBPPnnLZowkVVE2ZCRJVVONGrD//jBsGDz4YLJ86fvv4ZBDoEsX+N//YN68tFNKWg49eiSVy37++Wc23nhr3nmnEYWFbwDNMiOF5OUdyLrrvs2wYZ+w7rrrphlTklSBbMhIkqq2mjXh0EPh66/h7rth1VVhxAjYb7/kak0vvpgsc5KUM3r3TipXff7553Tp0o0fftifGTPuBepmRn4hP38bdt21Hh9/PJDmzZunGVOSVMFsyEiSqodatZJlS999B7fcAq1bw5AhyfKmzTeHN96wMSPliFxuyDz99DNstdXO/PHHDcyd2x8ImZEh5OVtxumn78WTTz5A3bp1l/QykqQqwIaMJKl6qVsX/v1v+OEHuO46aNECPvkEdtqp+GpNkrLa7NlJ5ZIYI5dcciWHHnoChYWvAvuUGH2e/PyduPfea7jggrMJISzuZSRJVYgNGUlS9ZSXl1wae9QouOIKWGmlpBmzzTaw447w8cdpJ5S0GDvumFSumDVrFv/4x7+4/PLHKSr6GNg4MxKpUeNqmjQ5jrfeepEDD9w/zZiSpEpmQ0aSVL3Vrw9nnAE//ggXXACNGsGAAckypt13hy++SDuhpEUcdVRSuWDSpEl0774jzz33B4WF7wJtMyOzqFv3GFZf/UG+/PIjNt100zRjSpJSYENGkiSAxo3h/POTxsxZZyWNmpdego02SjYAHj487YSSMg49NKls9+2339KlSzeGDu1GUdHTQIPMyB/k5+/CFlv8yhdfvM+qq66aZkxJUkpsyEiSVNLKK8OllyZLmU47DerVSy6Rvf768I9/JJsCS0pVYWFS2WzAgIEUFGzNL7+cyezZV1L8z+7vyc/fnCOO2JDXX3+Whg0bphlTkpQiGzKSJJWmRQu4+moYORL69k2u0vToo7DOOnDEEclMGkmp2HXXpLLVbbfdyZ57HsK0aY8T45ElRt4mL29Lrr76NG6++Rpq1qyZWkZJUvpsyEiStCStW8NNNyVXZTr6aAgB7rsP1loLjjsOxo1LO6FU7Rx3XFLZZu7cufTtexqnnXY1RUXvAT0WjIVwDw0bHsjzzz/Ccccdk1pGSVL2CDHGtDMss4KCgjh48OC0Y0iSqqORI+Gii+Chh2DevOQy2n36QP/+sMoqaaeTlJJp06ax117/4KOPplFY+BSwcmZkHrVr96dZs6d5880X6dSpU5oxq5UQwmcxxoK0c0jS4jhDRpKkZbHGGnD//ckmvwceCDNnwg03QIcOydWaJk5MO6FU5U2enFS2GDt2LBtuuCXvv9+SwsLXKG7GTCcvb1+6dPmEYcM+sRkjSVqIDRlJkpZHp07w2GPw5ZfQqxcUFcF//wurrw7nnQd//ZV2QqnK6tUrqWzw6aef0qVLN3788TBmzrwDqJ0ZGUd+/lbsuWcTPvzwDZo2bZpmTElSFrIhI0nSiujSBZ59FgYNgp49Ydo0uPjipDFz6aUwdWraCaUq58QTk0rb448/ybbb7s5ff93K3LmnASEz8hl5ed0488wDefTRe6hTp06aMSVJWco9ZCRJKk8ffgjnngtvvpncb9o02V/m3/+G/Px0s0kqFzFGLrzwMv7739spKnoe6Fpi9Gny84/lwQfvYJ999k4pocA9ZCRlP2fISJJUnrp3h4EDk4bMFlvApElw+unJ3jM33ZTsOSNphUycmN52TTNnzmT//Q/nqquepajoE4qbMZGaNa9g5ZVP4r33XrMZI0laKhsykiRVhG23hffeg1degYIC+PXXZI1Fx45wxx0we3baCaWctd9+SVW2CRMm0K3b9rz88gwKC98BWmVGZlGv3pF07PgkQ4d+zEYbbVT54SRJOceGjCRJFSUE2GUX+PTTZJ+ZLl1g3Dg49lhYe+3kak1z5qSdUso5p52WVGUaMWIE66+/GcOH96Co6HFg/hLEieTn78g220zms8/epU2bNpUbTJKUs2zISJJU0UJILgnzxRfw+OPJFZp+/BF694Z1102u1jRvXtoppZyxxx5JVZZXX32NTTftwW+/Xcjs2ZdQ/E/ob8jP78Yxx2zOyy8/Rf369SsvlCQp59mQkSSpstSoAQccAF99BQ88kOwr8913cPDBsMEG8MwzkIOb7UuV7ddfk6oMN974f+yzT2+mT38aOKzEyADy8rbmhhvO5rrrrqBGDf9ZLUlaNl5lSZKktMyenSxbuvhi+Omn5LEuXaBbN1h1VWjXrvi2bVuoWzfdvFKW6NEjuX377Yp7jzlz5vDvf5/Kww8PoLDwRaDDgrEQbqdhw/N5/vnH2WabbSouhFaIV1mSlO1syEiSlLaZM+Guu+DSS+GXXxZ/XsuWSYNm0WbN/NuWLZNZOFIV9+qrye0uu1TM60+ZMoU99jiIwYPnUFj4BNAkMzKXOnX60aLFy7z55ousueaaFRNA5cKGjKRsZ0NGkqRsUVSUXC579GgYOzaZNTP/9uefYe7cJT+/du1kJs2SmjaNG1fKlyLlqtGjR7Pttrvzyy9bM3PmjUCtzMhU8vIOpkuXIl555SlWWmmlNGOqDGzISMp2tZZ+iiRJqhR5ebDbbqWPzZ2bzJ6Z36Ap2ayZfztxYrJZ8I8/Lv49GjX6e6Om5HGbNi6NUtYbOza5bddu+Z7/4Ycfsuaaa9K8efOFHv/oo4/o2XNfpk49k3nz+gIhM/IT+fm7s+++3bj77luoXbv2cmeXJGk+Z8hIklRVFBYml9VeUtOmsHDpr7PKKktu2rRo4dIopWpF9pApKiqiRYt2tG/fkU8/fYu8vDwAHnroEY499mQKC+8Ddi3xjE/Iy9uHCy44jdNPP4UQQmkvqyzkDBlJ2c4ZMpIkVRX5+bDWWkmVJkb488/SGzXzj3/+ufgSNoMGlf46deokS6OW1LRp1Kjivk5Ve+ecs/zPfeSRR4lxE374YSUOPPAInn32Ec455yKuv/4+iooGAuuXOPsJ8vOP59FH72HPPSvxOtuSpGrBGTKSJKnYnDlJM2ZJTZuJE5f+Oo0aLX4fm1VXTRo6depU/NcjlRBjpGPHDRk16kpgG/Lzt6V16ymMH9+IwsJngZbzz6RWrUto0uQuBgx4ng022CC90FpuzpCRlO2cISNJkorVqpU0S9q2Xfw585dGLalpM2UKfPVVUqUJofiqUYubaePSKC3GqFHJbYcOSz5vUe+99x6//TYD2BGoQWHhs4wceQcx9gPyMmfNoF69o+jQ4TsGDPiYVq1alV9wSZJKcIaMJEkqXzHCH3/8vVlT8rYsV42avzRqSTNtXBpVLS3vHjK77LIfr722LXD8Ys74nfz8vdl22zY8+eT9C/aXUW5yhoykbOcMGUmSVL5CgKZNk+ratfRz5sxZ+KpRpd1OnJhMhZg/HaI0jRsvvlnTrp1Lo6qoCy9c9uf89NNPvPPOW8C9iznjW/Lze9K37yFcfvmF1HB2liSpgtmQkSRJla9WraRh0q4ddO9e+jmFhUljZklNm8mTk1rS0qjSrhpV8talUTlnm22W/Tk33HAr8+YdBjRczBmDaNBgDmeffbrNGElSpXDJkiRJyk2LLo0qrWkzfnzZlkbNbw4trmnj0qis8u23ye3aa5ft/ORS16sybdpHQMfFnBWpW/d4NttsNAMHPk+tWv7dMte5ZElStvO/NJIkKTctz9Ko0po2kybByJFJLU7jxrDaasWbDpes1VaDVq2gZs0K+TL1d8cem9yWdQ+Zhx56hBi7sfhmDEBg5swbGDRoF44//jRuv/2GFUwpSdKS2ZCRJElV14osjSp5PHkyDB2aVGlq1izegLhko6bk/YaLWyqjZXXZZWU/N8bI5ZffyPTpVy/mjInAe9Su/S55ee8yY8a3jBrVgBgjIYTyiCtJUqlsyEiSpOotPz9Z+7K49S8xJrNo5jdpStaYMcntr78mx2PGLP59mjQpvVEzv5xlU2aL662V5t1332XChNnADplHxgPvUrfuu9Sp8y6zZ49lww27s9tuW9Ojx40UFBRQt27dCkgtSdLC3ENGkiRpRc2cCePG/b1RU7KKipb8GrVq/X2WzaINnAYNKufryXLz93Beb72ln7vzzvvy+utF5OWtQs2a7xLjn2y22dbsttvWbL31VnTt2tX9Yqoo95CRlO38r48kSdKKqlsX1lgjqdLMn2VTWqNmfgPnt99g9OikFmellRa/JGrVVZMrSlWDWTZ9+ya3ZdlD5q+/JrHbbqvQs2cB22xzKp07d/YqSpKkrOAMGUmSpGwwY8bCs2xKm2kzY8aSX2P+LJvSmjWrrZbspVMFZtkMGpTcbrJJujmU3ZwhIynbOUNGkiQpG9SrBx07JlWaGGHixMUvifrpp7LNsll55SUvi1plFcjyGSQ2YiRJVYENGUmSpFwQAjRvntTGG5d+TlHR0mfZ/PFHUkOGlP4atWsX72WzuA2I69evsC+zLOZHX9zVziVJygU2ZCRJkqqKvDxYc82kShMjTJiw5Fk2v/8OP/6Y1OI0bVp6o2Z+A6dlywqdZXPyycltWfaQkSQpW9mQkSRJqi5CgBYtkipYzNYaRUUwduziL/H900/JBsWTJsEXX5T+GrVrJ/vVLG4D4nbtVmiWzfXXL/dTJUnKGjZkJEmSVCwvD9ZaK6nSzJtXPMtmcTNtJkyAUaOSWpymTRe/JGq11ZKm0WJm2bhUSZJUFdiQkSRJUtnVqJEsSWrZcvG76xYWlj7LZn4DZ+zY4lk2n39e+mvUqbPwLJsSM20GTV4LWrZkk63qVdzXKUlSBbMhI0mSpPKVnw9rr51UaebNS/aqWdLmwxMnwsiRSS3idN4CfuTt9kfAuutC587Ftc460LBhxX59kiSVgxBjTDvDMisoKIiDBw9OO4YkSZIqyvxZNqUsifrq+7rwyy+sN/fL0p/brt3CTZr51aRJpX4JSlcI4bMY42I2S5Kk9NmQkSRJUu6ZPTuZPTNixML1zTcwc2bpz2nVauEGzfzZNU2bVm52VQobMpKynQ0ZSZIk5ZQPP0xuu3cvZXDu3GQz4UUbNV9/nVxBqjQtWpQ+o6ZFi+TKVMpJNmQkZTsbMpIkScopPXokt2+/vQxPmjcvWf60aKNmxAiYNq3056y8cumNmtatbdTkABsykrKdDRlJkiTllG+/TW4Xt2fwMokRxo1buEEzfHhyO3ly6c9p1GjhJU/zq107GzVZxIaMpGxnQ0aSJElaVIzwyy9/n00zfDj88Ufpz2nQILnK06Izatq3Ty4XrkplQ0ZStrMhI0mSpJzyzjvJ7TbbpPDmMcKECaUvffrtt9Kfk5cHnTr9vVHToQPUqlW5+asRGzKSsp0NGUmSJOWU5dpDpjJMmpRsHjx/ydP8Gj++9PPr1EnWXS269KljR6hdu3KzV0E2ZCRluyrVkJk3bx4TJ07kr7/+Yu7cuSkkkwRQs2ZNmjRpQrNmzajhFG1JUjkbNSq57dAh3Rxl9tdfSaNm0Rk1P/1U+vm1asFaa/19Rs1aa0HdupUaPZfZkJGU7apUQ+ann34ihEDLli2pXbs2wU3VpEoXY2T27Nn89ttvxBhZddVV044kSVJ2mjoVvvnm73vUjB6dLI1aVM2asMYaf2/UdOqULIvSQmzISMp2Vaoh8/XXX7P22mv7F3kpC8ybN49vv/2WddZZJ+0okqQqZsCA5HaHHdLNUWGmT08uJbXojJqRI5PLdy8qhGS6UGmNmgYNKj9/lrAhIynbVbldxGzGSNnB/y1KkirKJZckt1W2IVO/Pmy0UVIlzZgB333390bN998nzZqRI+GFFxZ+zmqr/b1Rs8460Lhx5X09kqRSVbmGjCRJkqq2Bx9MO0FK6tWDLl2SKmnWLPjhh+IlT/MbNd9+C2PGJPXKKws/p02bpDlTckPhddaBlVeuvK9Hkqo5GzJSjrrgggv44YcfeOihh5Z6bp8+fWjTpg3nnntuuecIIfD999/TsWPHcn9tSZJK065d2gmyTJ06xU2V/fYrfnzOnGTWzKIzar75Bn7+Oak33lj4tVZZ5e8zajp3hubNK/drkqRqwDUFylpvv/02bdu2LbfX+/rrr9luu+1o3LgxHTt25JlnnlkwNnr0aEIINGjQYEFdfPHFC8YfeeQRWrVqxeqrr87bJa6xOXLkSLp37571V/W67bbbKqQZI0lSGl59NSktRa1ayWW1994bzj4bHn4YvvgCpk1LZtQ8/zxccQUcfjgUFEB+Pvz6K7z5Jtx8M/z738k1xlu0SBoy22wDxx0HN90EAwfCL7+UvvmwJKlMKnyGTAhhF+AGoCZwV4zxikXGQ2Z8V6AQ6B1j/Lyic6l6mTNnDr169aJPnz688cYbvPPOO+yxxx588cUXrLXWWgvO++uvv6hVq9bfntu/f38+//xzPvvsM/r27ctXX30FwIknnsi1115LzZo1K/XrkSSpOrsi86/JXXZJN0fOmn+1pjXWgD32KH583rzkUtyLzqgZMQImToR3302qpJVWWngmzfrrw/bbV+7XI0k5qkJnyIQQagK3AD2BzsDBIYTOi5zWE1gzU8cAt1ZkpjS1b9+eq6++mi5dutC4cWMOPPBAZsyYsdTnPffcc3Tt2pVGjRqxxhpr8GrmT0Ljx49nzz33ZOWVV6Zjx47ceeedC55zwQUXsP/++3PooYfSsGFD1l9/fb777jsuv/xyWrRoQbt27Xj99dcXnN+jRw/OPPNMNt10Uxo3bkyvXr34448/Fow///zzrLvuujRp0oQePXrw9ddfl/nrevHFF+natStNmjShe/fuDB06dKnPnT59Oj179mT8+PELZqyMHz+eTz/9lIKCAho1akTLli059dRTy/S9/+abbxg/fjynnHIKNWvWZLvttmOLLbbgwTIsQp80aRJt2rShVatW7LDDDowaNQqAp556ijZt2tCtW7elvsY999zDOuusw0orrcTOO+/MmDFjALjyyivp1q0bc+bMAeDWW29l3XXXZcaMGQtm7dxxxx20bt2aVq1acc011yz2Pfbff39WWWUVGjduzNZbb83w4cMXjPXu3ZtzzjkHKJ55dM0119CiRQtatWrFvffeu+DcmTNn0q9fP1ZddVVatmxJnz59KCoqWjB+1VVX0apVK1q3bs0999yz1K9dkqTy9thjSamc1agB7dvDrrtCv35wzz3w8ccweTKMHQuvvQbXXQdHHw1bbAFNmsCff8IHH8Cdd8Ipp8BJJ6X9VUhSzqjoJUubAj/EGEfFGGcBjwG9FjmnF/BATHwMNAkhtCqvAD16wH33JcezZyf352+5UViY3H/88eT+5MnJ/aefTu5PnJjcn79Z/a+/JvfnT5EdO3bZ8zzxxBO8+uqr/PjjjwwdOpT75odbjE8//ZTDDz+cq666ir/++ot3332X9u3bA3DwwQfTtm1bxo8fz1NPPcVZZ53FwIEDFzz3hRde4LDDDuPPP/9kww03ZOedd2bevHn8/PPPnHfeeRx77LELvdcDDzzAPffcw/jx46lVqxYnnngiAN999x0HH3ww119/PRMmTGDXXXdljz32YNasWUv9uj7//HOOPPJIbr/9diZNmsSxxx7LnnvuycyZM5f43Pr16/PKK6/QunVrpk2bxrRp02jdujUnnXQSJ510ElOmTGHkyJEccMABZfq+l3Z59xjjgpku86222mq0bduWI444gokTJwLQvHlzJk2axLhx43jjjTdYd911mTZtGpdccgmXX375Ut/72Wef5bLLLuPpp59mwoQJbLXVVhx88MEAnH766dSpU4dLLrmE77//nrPOOouHHnqIevXqLXj+W2+9xffff8/rr7/OFVdcwYD51/pcRM+ePfn+++/5/fff2WijjTjkkEMWm+nXX39l8uTJ/Pzzz9x9990cf/zx/PnnnwCcccYZfPfddwwZMoQffviBn3/+mYsuugiAV199lauvvpo33niD77//frFZJEmqSKuskpQqSQjQti3stBOcfDLccQe8/z788UeybGngwGQZ03HHwf77p51WknJGRTdk2gAl2xbjMo8t6zmEEI4JIQwOIQyeMGFCuQetLCeeeCKtW7dm5ZVXZo899mDIkCFLPP/uu+/myCOPZMcdd6RGjRq0adOGTp06MXbsWN5//32uvPJK6tWrR9euXTnqqKMWmvGx1VZbsfPOO1OrVi32339/JkyYQP/+/alduzYHHXQQo0eP5q+//lpw/mGHHcZ6661H/fr1ufjii3niiSeYO3cujz/+OLvtths77rgjtWvXpl+/fhQVFfHhhx8u9eu68847OfbYY9lss82oWbMm//znP6lbty4ff/zxcn1PateuzQ8//MDEiRNp0KBBmWanAHTq1IkWLVpw1VVXMXv2bF5//XXeeecdCgsLAWjWrBmDBg1izJgxfPbZZ0ydOnVBQ6NGjRrceuut7Lffflx99dXceeednHfeeZxwwgkMGzaMbbfdlp133vlvzZ35br/9ds4880zWWWcdatWqxVlnncWQIUMYM2YMNWrU4IEHHuDGG29kzz335D//+Q8bbrjhQs8///zzqV+/Puuvvz5HHHEEjz76aKnvc+SRR9KwYUPq1q3LBRdcwJdffsnkyZMX+30877zzqF27NrvuuisNGjTg22+/JcbInXfeyXXXXcfKK69Mw4YNOeuss3gs82fIJ554giOOOGLBz8kFF1xQpu+/JEnl6YUX/n51Z6UghKQztt120Lcv/N//wfnnp51KknJGRTdkQimPLTpVoSznEGO8I8ZYEGMsaL4Mu7y//Tb07p0c166d3D/00OR+fn5y/8ADk/uNGyf399knud+sWXJ//tLaVVZJ7s9fr7w8O/yvUuLPOfn5+UybNm2J548dO5Y11ljjb4+PHz9+wS/M86222mr8/PPPC+63bNlywXFeXh7NmjVbsNdJXl4ewELv367EF7Taaqsxe/ZsJk6cyPjx41lttdUWjNWoUYN27dot9F6L+7rGjBnDNddcQ5MmTRbU2LFjGT9+/HJ9T+6++26+++47OnXqxCabbMKLL75Y6nk9e/ZcsNTp4Ycfpnbt2jz77LO89NJLrLLKKlxzzTUccMABCzYNbtCgAQUFBdSqVYuWLVty88038/rrrzNlyhQAtt9+ez7++GPeeecdatSoweDBg+nduzeHHXYY9913H+eeey5HHXVUqVnGjBnDSSedtODrX3nllYkxLvj+tW/fnm233ZbRo0dz/PHH/+35i34uJb93882dO5f+/fuzxhpr0KhRowWzqObP8llU06ZNF9orZ/73fcKECRQWFrLxxhsvyLvLLrswvwk6fvz4v+WRJKmyXXNNUpIk5bKK3tR3HFCybdEWWPS3ybKcU221a9eOkSNH/u3x1q1b88cffzB16tQFTZmffvqJNm3+NrmozMaWWIP1008/Ubt2bZo1a0br1q0ZNmzYgrEYI2PHji3Te7Vr146zzz6bs88+e5nzJPs9L2zNNdfk0UcfZd68eTz99NPst99+TJo0ifr16y903iuvvPK353bp0oV33nlnwf3u3bvzz3/+c4nvvehSpxgjffv25cYbb2TixInMnTuX1VZbjVVWWWWhvXFKmv89WNwSopdffpmPPvqI7bffntNPP53bb799ofGxY8fSqVMnIPlcWrdu/bfXeOSRR3juuecYMGAA7du3Z/Lkyay00kqlLtVakmbNmpGXl8fw4cNL/XxbtWr1t58TSZIq21NPpZ1AkqQVV9EzZAYBa4YQVg8h1AEOAp5f5JzngcNDohswOcb4SwXnyhn/+te/uPfeexk4cOCC/V+++eYb2rVrR/fu3TnzzDOZMWMGQ4cO5e67717iviFL89BDDzFixAgKCws577zz2G+//ahZsyYHHHAAL730EgMHDmT27Nlcc8011K1bl+7duy/1NY8++mhuu+02PvnkE2KMTJ8+nZdeeompU6cu9bktW7Zk0qRJCy27eeihh5gwYQI1atSgSZMmAGW+wtHQoUOZMWMGhYWFXH311fzyyy/0zkyf+uSTT/j222+ZN28ekyZN4sQTT6RHjx40btx4ode466672HDDDenatStNmzalqKiIESNG8NZbb9GhQ4dS37dPnz5cfvnlCzbZnTx5Mk8++SSQzGD517/+xV133cX999/PCy+8wMsvv7zQ8y+++GIKCwsZPnw49957LwfOn9JVwtSpU6lbty5NmzalsLCQs846q0zfk0XVqFGDo48+mlNOOYXff/8dgJ9//pnXXnsNgAMOOID77rtvwc/JhRdeuFzvI0nSimjWLClJknJZhTZkYoxzgL7Aa8DXwBMxxuEhhD4hhD6Z014GRgE/AHcC/67ITLlm00035d577+WUU06hcePGbLPNNguu0PPoo48yevRoWrduzd57782FF17IjjvuuNzvddhhh9G7d29WWWUVZsyYwY033gjA2muvzUMPPcQJJ5xAs2bNeOGFF3jhhReoU6fOUl+zoKCAO++8k759+7LSSivRsWPHpW5kPF+nTp04+OCD6dChA02aNGH8+PG8+uqrrLvuujRo0ICTTjqJxx57bKENcJfkwQcfpFWrVrRo0YKBAwfyxhtvULduXQBGjRrFLrvsQsOGDVlvvfWoW7fu3/ZqmThxIjfccAMXX3wxALVq1eLmm29mu+22o0+fPtx0002lvu/ee+/NGWecwUEHHUSjRo1Yb731FszgOeaYY+jVqxe77rorTZs25e677+aoo45i0qRJC56/zTbb0LFjR7bffnv69evHTjvt9Lf3OPzww1lttdVo06YNnTt3LvPeOqW58sor6dixI926daNRo0bssMMOfPvtt0CyFOzkk09mu+22o2PHjmy33XbL/T6SJC2vp58uvgiDJEm5KizrkoZsUFBQEAcPHvy3x7/++mvWWWedFBLlvh49enDooYcudh8UVb7Ro0ez+uqrM3v27IX2e8kl/m9SklQRevRIbt9+O80UynYhhM9ijAVp55CkxcnN3/IkSZJUbT33XNoJJElacRW9h4zK4LLLLltwRaCS1bNnz7SjSZIkZZ3GjZOSJCmXuWRJUoXxf5OSpIrw+OPJbSn73EsLuGRJUrZzyZIkSZJyyq23Jrc2ZCRJuazKNWTmzZtHjRquxJLSNm/evLQjSJKqqJdfTjuBJEkrrkp1LurXr8/PP//MrFmzyMWlWFJVEGNk1qxZ/Pzzz9SvXz/tOJKkKig/PylJknJZlZoh07ZtWyZOnMiYMWOYM2dO2nGkaqtWrVo0btyYZs2apR1FklQFPfRQcnvooenmkCRpRVSphkyNGjVo0aIFLVq0SDuKJEmSKshddyW3NmQkSbmsSjVkJEmSVPW98UbaCSRJWnE2ZCRJkpRTatdOO4EkSSuuSm3qK0mSpKrvvvuSkiQpl9mQkSRJUk6xISNJqgpCLl4eOoQwARiTdo4SmgET0w6hhfiZZCc/l+zjZ5J9/Eyyk59L9vEzyU7Z9LmsFmNsnnYISVqcnGzIZJsQwuAYY0HaOVTMzyQ7+blkHz+T7ONnkp38XLKPn0l28nORpLJzyZIkSZIkSVIlsyEjSZIkSZJUyWzIlI870g6gv/EzyU5+LtnHzyT7+JlkJz+X7ONnkp38XCSpjNxDRpIkSZIkqZI5Q0aSJEmSJKmS2ZCRJEmSJEmqZDZkVkAIYZcQwrchhB9CCP3TziMIIdwTQvg9hPBV2lmUCCG0CyG8FUL4OoQwPIRwUtqZBCGEeiGET0MIX2Y+lwvTzqRECKFmCOGLEMKLaWdRIoQwOoQwLIQwJIQwOO08ghBCkxDCUyGEbzL/fdk87UzVWQhh7cz/PubXlBDCyWnnkqRs5x4yyymEUBP4DtgRGAcMAg6OMY5INVg1F0LYGpgGPBBjXC/tPIIQQiugVYzx8xBCQ+AzYC//t5KuEEIA6scYp4UQagPvAyfFGD9OOVq1F0I4FSgAGsUYd087j5KGDFAQY5yYdhYlQgj3A+/FGO8KIdQB8mOMf6UcSyz4N/LPwGYxxjFp55GkbOYMmeW3KfBDjHFUjHEW8BjQK+VM1V6M8V3gj7RzqFiM8ZcY4+eZ46nA10CbdFMpJqZl7tbOlB36lIUQ2gK7AXelnUXKViGERsDWwN0AMcZZNmOyyvbASJsxkrR0NmSWXxtgbIn74/CXTGmJQgjtgQ2BT1KOIhYsjRkC/A68EWP0c0nf9cB/gHkp59DCIvB6COGzEMIxaYcRHYAJwL2Z5X13hRDqpx1KCxwEPJp2CEnKBTZkll8o5TH/uiwtRgihAfA/4OQY45S08whijHNjjF2BtsCmIQSX+aUohLA78HuM8bO0s+hvtogxbgT0BI7PLI9VemoBGwG3xhg3BKYD7uWXBTLLx/YEnkw7iyTlAhsyy28c0K7E/bbA+JSySFkts0fJ/4CHY4xPp51HC8tM9X8b2CXdJNXeFsCemf1KHgO2CyE8lG4kAcQYx2dufweeIVm2rPSMA8aVmNX3FEmDRunrCXweY/wt7SCSlAtsyCy/QcCaIYTVM38NOAh4PuVMUtbJbB57N/B1jPHatPMoEUJoHkJokjnOA3YAvkk1VDUXYzwzxtg2xtie5L8pb8YYD005VrUXQqif2ZCczLKYnQCv5JeiGOOvwNgQwtqZh7YH3Cg+OxyMy5UkqcxqpR0gV8UY54QQ+gKvATWBe2KMw1OOVe2FEB4FegDNQgjjgPNjjHenm6ra2wI4DBiW2a8E4KwY48vpRRLQCrg/czWMGsATMUYvsyz9XUvgmaS3TC3gkRjjq+lGEnAC8HDmj2KjgCNSzlPthRDySa4+emzaWSQpV3jZa0mSJEmSpErmkiVJkiRJkqRKZkNGkiRJkiSpktmQkSRJkiRJqmQ2ZCRJkiRJkiqZDRlJkspBCKF9COEfaeeQJElSbrAhI0lS+WgP2JCRJElSmdiQkSQtl8yMkK9DCHeGEIaHEF4PIeQt5tyOIYQBIYQvQwifhxDWCImrQghfhRCGhRAOzJzbI4TwTgjhiRDCdyGEK0IIh4QQPs2ct0bmvPtCCLeFEN7LnLd75vF6IYR7M+d+EULYNvN47xDC0yGEV0MI34cQ/lsi304hhI8y2Z4MITTIPD46hHBh5vFhIYROmce3CSEMydQXIYSGwBXAVpnHTqnI770kSZJynw0ZSdKKWBO4Jca4LvAXsO9izns4c94GQHfgF2AfoCuwAbADcFUIoVXm/A2Ak4D1gcOAtWKMmwJ3ASeUeN32wDbAbsBtIYR6wPEAMcb1gYOB+zOPk3m/AzOve2AIoV0IoRlwDrBDjHEjYDBwaon3mJh5/FagX+axfsDxMcauwFZAEdAfeC/G2DXGeN3SvnGSJEmq3mzISJJWxI8xxiGZ489IGiQLycweaRNjfAYgxjgjxlgIbAk8GmOcG2P8DXgH2CTztEExxl9ijDOBkcDrmceHLfIeT8QY58UYvwdGAZ0yr/tg5r2+AcYAa2XOHxhjnBxjnAGMAFYDugGdgQ9CCEOAf2Yen+/pUr6+D4BrQwgnAk1ijHOW/q2SJEmSitVKO4AkKafNLHE8FyhtyVJYzHMX9/iirzuvxP15LPzfrrjI8+IyvO7czGsF4I0Y48FLec7884kxXhFCeAnYFfg4hLDDEt5TkiRJ+htnyEiSKlSMcQowLoSwF0AIoW4IIR94l2TZUM0QQnNga+DTZXz5/UMINTL7ynQAvs287iGZ91oLWDXz+OJ8DGwRQuiYeU5+5nmLFUJYI8Y4LMZ4JckSp07AVKDhMuaXJElSNWVDRpJUGQ4DTgwhDAU+BFYBngGGAl8CbwL/iTH+uoyv+y3JUqdXgD6ZpUj/B9QMIQwDHgd6Z5Y+lSrGOAHoDTyayfcxSYNlSU7ObEb8Jcn+Ma9kvpY5mY2L3dRXkiRJSxRiXHS2tyRJ2S+EcB/wYozxqbSzSJIkScvKGTKSJEmSJEmVzBkykqRyE0K4BdhikYdviDHem0YeSZIkKVvZkJEkSZIkSapkLlmSJEmSJEmqZDZkJEmSJEmSKpkNGUmSJEmSpEpmQ0aSJEmSJKmS2ZCRJEmSJEmqZDZkJEmSJEmSKtn/Ayra9w7j8WSTAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#trying with dimensionality reduction and then kmeans\n",
    "n_components=X.shape[1]\n",
    "#running PCA with all components\n",
    "pca=PCA(n_components=n_components, random_state =453)\n",
    "X_r=pca.fit(X).transform(X)\n",
    "\n",
    "#calculatig the 95% variance\n",
    "total_variance=sum(pca.explained_variance_)\n",
    "print(\"Total Variance in our dataset is : \",total_variance)\n",
    "var_95=total_variance*0.95\n",
    "print(\"The 95% variance we weant to have is : \", var_95)\n",
    "print(\"\")\n",
    "\n",
    "#creating of a df with the components and explained variance\n",
    "a=zip(range(0,n_components),pca.explained_variance_)\n",
    "a=pd.DataFrame(a,columns=[\"PCA Comp\",'Explained Variance'])\n",
    "\n",
    "#Trying to hit 95%\n",
    "print(\"Variance Explained With 2 n_components:\", sum(a[\"Explained Variance\"][0:2]))\n",
    "print(\"Variance Explained With 3 n_components:\", sum(a[\"Explained Variance\"][0:3]))\n",
    "print(\"Variance Explained With 4 n_components:\", sum(a[\"Explained Variance\"][0:4]))\n",
    "print(\"Variance Explained With 5 n_components:\", sum(a[\"Explained Variance\"][0:5]))\n",
    "print(\"Variance Explained With 6 n_components:\", sum(a[\"Explained Variance\"][0:6]))\n",
    "print(\"Variance Explained With 7 n_components:\", sum(a[\"Explained Variance\"][0:7]))\n",
    "print(\"Variance Explained With 8 n_components:\", sum(a[\"Explained Variance\"][0:8]))\n",
    "\n",
    "#plotting the data\n",
    "plt.figure(1,figsize=(14,8))\n",
    "plt.plot(pca.explained_variance_ratio_,linewidth=2,c='r')\n",
    "plt.xlabel('n_componenst')\n",
    "plt.ylabel('explained_ratio')\n",
    "\n",
    "#plotiing line with 95%\n",
    "plt.axvline(6,linestyle=':',label='n_components -95% explained',c='blue')\n",
    "plt.legend(prop=dict(size=12))\n",
    "\n",
    "#adding arrow\n",
    "plt.annotate('6 eigenvectors used to explain 95% variance',xy=(6,pca.explained_variance_ratio_[6]),\n",
    "                                                               xytext=(8,pca.explained_variance_ratio_[1]),\n",
    "                                                              arrowprops=dict(facecolor='blue',shrink=0.05))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "dbdd4cfa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The inertia for 2 is 435066.9935203006\n",
      "The inertia for 3 is 357625.94311398733\n",
      "The inertia for 4 is 317522.18947382516\n",
      "The inertia for 5 is 280354.1948962724\n",
      "The inertia for 6 is 253101.66467658774\n",
      "The inertia for 7 is 232723.03309276834\n",
      "The inertia for 8 is 219198.9309080343\n",
      "The inertia for 9 is 203887.33882843226\n",
      "The inertia for 10 is 197035.59071036766\n",
      "The inertia for 11 is 183996.43565360186\n",
      "The inertia for 12 is 178315.81694432066\n",
      "The inertia for 13 is 171210.03130563186\n",
      "The inertia for 14 is 163255.96581045387\n",
      "The inertia for 15 is 159819.87729463994\n",
      "The inertia for 16 is 156303.47355456388\n",
      "The inertia for 17 is 151545.05838556477\n",
      "The inertia for 18 is 144336.54294503538\n",
      "The inertia for 19 is 144563.4071789343\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Inertia Plot Per Cluster')"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#running PCA with 6 n_components\n",
    "pca=PCA(n_components=6, random_state =453)\n",
    "X_r=pca.fit(X).transform(X)\n",
    "\n",
    "\n",
    "inertia=[]\n",
    "for n in nclusters:\n",
    "    kmeans=KMeans(n_clusters=n,random_state=2)\n",
    "    kmeans=kmeans.fit(X_train)\n",
    "    u=kmeans.inertia_\n",
    "    inertia.append(u)\n",
    "    print('The inertia for {} is {}'.format(n,u))\n",
    "    \n",
    "#elbow method for inertia\n",
    "fig,(ax1)=plt.subplots(1,figsize=(16,6))\n",
    "xx=np.arange(len(nclusters))\n",
    "ax1.plot(xx,inertia)\n",
    "ax1.set_xticks(xx)\n",
    "ax1.set_xticklabels(nclusters,rotation=45)\n",
    "plt.xlabel(\"Number of Clusters\")\n",
    "plt.ylabel(\"Inertia Score\")\n",
    "plt.title(\"Inertia Plot Per Cluster\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "ed210db7",
   "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>OutstandingLoan</th>\n",
       "      <th>terms</th>\n",
       "      <th>disbursedamount</th>\n",
       "      <th>LastInstallmentAmount</th>\n",
       "      <th>Gender</th>\n",
       "      <th>defaulter</th>\n",
       "      <th>terms_lapsed</th>\n",
       "      <th>disbursed_outstanding_ratio</th>\n",
       "      <th>clusters</th>\n",
       "      <th>cluster_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   OutstandingLoan  terms  disbursedamount  LastInstallmentAmount Gender  \\\n",
       "0                2      0                0                      0      1   \n",
       "1                2      0                0                      0      0   \n",
       "2                2      0                0                      0      0   \n",
       "3                2      0                0                      0      0   \n",
       "4                2      0                0                      0      1   \n",
       "\n",
       "  defaulter  terms_lapsed  disbursed_outstanding_ratio  clusters  \\\n",
       "0         1             0                            0         1   \n",
       "1         1             0                            0         1   \n",
       "2         1             0                            0         1   \n",
       "3         1             0                            0         1   \n",
       "4         1             0                            0         1   \n",
       "\n",
       "  cluster_category  \n",
       "0        cluster 2  \n",
       "1        cluster 2  \n",
       "2        cluster 2  \n",
       "3        cluster 2  \n",
       "4        cluster 2  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#running PCA with 6 n_components\n",
    "pca=PCA(n_components=6, random_state =453)\n",
    "X_r=pca.fit(X).transform(X)\n",
    "\n",
    "\n",
    "#running kmeans  with 3 clusters\n",
    "kmeans=KMeans(n_clusters=3,random_state=540)\n",
    "kmeans=kmeans.fit(X_train)\n",
    "\n",
    "clusters=kmeans.labels_\n",
    "\n",
    "#attach clusters to original dataset\n",
    "df['clusters']=kmeans.labels_\n",
    "\n",
    "#create a cluster category\n",
    "df[\"cluster_category\"]=\"No Data\"\n",
    "df.loc[df.clusters==0,'cluster_category']='cluster 1'\n",
    "df.loc[df.clusters==1,'cluster_category']='cluster 2'\n",
    "df.loc[df.clusters==2,'cluster_category']='cluster 3'\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "5e3a7b12",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['OutstandingLoan', 'terms', 'disbursedamount', 'LastInstallmentAmount',\n",
       "       'Gender', 'defaulter', 'terms_lapsed', 'disbursed_outstanding_ratio',\n",
       "       'clusters', 'cluster_category'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "a2fb63fa",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['OutstandingLoan'].loc[df['OutstandingLoan']==1]=\"Betwn 0 & 48,000\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:5: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['terms'].loc[df['terms']==0]=\"Betwn 0 & 3\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:9: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['disbursedamount'].loc[df['disbursedamount']==0]=\"Betwn 0 & 9,000\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:13: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['LastInstallmentAmount'].loc[df['LastInstallmentAmount']==0]=\"Betwn 0 & 2,500\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:17: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['disbursed_outstanding_ratio'].loc[df['disbursed_outstanding_ratio']==0]=\"Betwn 0 & 0.2\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:21: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['terms_lapsed'].loc[df['terms_lapsed']==0]=\"Betwn 0 & 70\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:25: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['Gender'].loc[df['Gender']==0]=\"Female\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:26: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['Gender'].loc[df['Gender']==1]=\"Male\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:27: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['Gender'].loc[df['Gender']==2]=\"Undisclosed\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:29: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['defaulter'].loc[df['defaulter']==0]=\"No\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:30: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['defaulter'].loc[df['defaulter']==1]=\"Yes\"\n",
      "C:\\Users\\julmn\\AppData\\Local\\Temp\\ipykernel_11268\\3796329904.py:31: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['defaulter'].loc[df['defaulter']==2]=\"Unknown\"\n"
     ]
    }
   ],
   "source": [
    "df['OutstandingLoan'].loc[df['OutstandingLoan']==1]=\"Betwn 0 & 48,000\"\n",
    "df['OutstandingLoan'].loc[df['OutstandingLoan']==0]=\"Betwn 49,000 & 60,000\"\n",
    "df['OutstandingLoan'].loc[df['OutstandingLoan']==2]=\"Above 60,000\"\n",
    "\n",
    "df['terms'].loc[df['terms']==0]=\"Betwn 0 & 3\"\n",
    "df['terms'].loc[df['terms']==2]=\"Betwn 3 & 12\"\n",
    "df['terms'].loc[df['terms']==1]=\"Above 12\"\n",
    "\n",
    "df['disbursedamount'].loc[df['disbursedamount']==0]=\"Betwn 0 & 9,000\"\n",
    "df['disbursedamount'].loc[df['disbursedamount']==1]=\"Betwn 9,000 & 40,000\"\n",
    "df['disbursedamount'].loc[df['disbursedamount']==2]=\"Above 40,000\"\n",
    "\n",
    "df['LastInstallmentAmount'].loc[df['LastInstallmentAmount']==0]=\"Betwn 0 & 2,500\"\n",
    "df['LastInstallmentAmount'].loc[df['LastInstallmentAmount']==1]=\"Betwn 2,500 & 5800\"\n",
    "df['LastInstallmentAmount'].loc[df['LastInstallmentAmount']==2]=\"Above 5,800\"\n",
    "\n",
    "df['disbursed_outstanding_ratio'].loc[df['disbursed_outstanding_ratio']==0]=\"Betwn 0 & 0.2\"\n",
    "df['disbursed_outstanding_ratio'].loc[df['disbursed_outstanding_ratio']==1]=\"Betwn 0.2 & 1\"\n",
    "df['disbursed_outstanding_ratio'].loc[df['disbursed_outstanding_ratio']==2]=\"Above 1\"\n",
    "\n",
    "df['terms_lapsed'].loc[df['terms_lapsed']==0]=\"Betwn 0 & 70\"\n",
    "df['terms_lapsed'].loc[df['terms_lapsed']==1]=\"Betwn 70 & 121\"\n",
    "df['terms_lapsed'].loc[df['terms_lapsed']==2]=\"Above 121\"\n",
    "\n",
    "df['Gender'].loc[df['Gender']==0]=\"Female\"\n",
    "df['Gender'].loc[df['Gender']==1]=\"Male\"\n",
    "df['Gender'].loc[df['Gender']==2]=\"Undisclosed\"\n",
    "\n",
    "df['defaulter'].loc[df['defaulter']==0]=\"No\"\n",
    "df['defaulter'].loc[df['defaulter']==1]=\"Yes\"\n",
    "df['defaulter'].loc[df['defaulter']==2]=\"Unknown\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "a45096cb",
   "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>OutstandingLoan</th>\n",
       "      <th>terms</th>\n",
       "      <th>disbursedamount</th>\n",
       "      <th>LastInstallmentAmount</th>\n",
       "      <th>Gender</th>\n",
       "      <th>defaulter</th>\n",
       "      <th>terms_lapsed</th>\n",
       "      <th>disbursed_outstanding_ratio</th>\n",
       "      <th>clusters</th>\n",
       "      <th>cluster_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Female</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Female</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Female</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  OutstandingLoan        terms  disbursedamount LastInstallmentAmount  Gender  \\\n",
       "0    Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500    Male   \n",
       "1    Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500  Female   \n",
       "2    Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500  Female   \n",
       "3    Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500  Female   \n",
       "4    Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500    Male   \n",
       "\n",
       "  defaulter  terms_lapsed disbursed_outstanding_ratio  clusters  \\\n",
       "0       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "1       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "2       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "3       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "4       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "\n",
       "  cluster_category  \n",
       "0        cluster 2  \n",
       "1        cluster 2  \n",
       "2        cluster 2  \n",
       "3        cluster 2  \n",
       "4        cluster 2  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "a8df1289",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_=pd.concat([df,df_cols],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "75dd2f8f",
   "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>OutstandingLoan</th>\n",
       "      <th>terms</th>\n",
       "      <th>disbursedamount</th>\n",
       "      <th>LastInstallmentAmount</th>\n",
       "      <th>Gender</th>\n",
       "      <th>defaulter</th>\n",
       "      <th>terms_lapsed</th>\n",
       "      <th>disbursed_outstanding_ratio</th>\n",
       "      <th>clusters</th>\n",
       "      <th>cluster_category</th>\n",
       "      <th>OutstandingLoan</th>\n",
       "      <th>installmentstartdate</th>\n",
       "      <th>Lastinstallmentdate</th>\n",
       "      <th>disbursementdate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "      <td>-18134.96</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>2022-02-18</td>\n",
       "      <td>2022-01-19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Female</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "      <td>-6769.98</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>2021-12-31</td>\n",
       "      <td>2021-12-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Female</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "      <td>-10000.00</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>2021-12-30</td>\n",
       "      <td>2021-11-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Female</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "      <td>-30000.00</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-11-26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Above 60,000</td>\n",
       "      <td>Betwn 0 &amp; 3</td>\n",
       "      <td>Betwn 0 &amp; 9,000</td>\n",
       "      <td>Betwn 0 &amp; 2,500</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Betwn 0 &amp; 70</td>\n",
       "      <td>Betwn 0 &amp; 0.2</td>\n",
       "      <td>1</td>\n",
       "      <td>cluster 2</td>\n",
       "      <td>-1269.98</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-12-26</td>\n",
       "      <td>2021-11-26</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   OutstandingLoan        terms  disbursedamount LastInstallmentAmount  \\\n",
       "0     Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500   \n",
       "1     Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500   \n",
       "2     Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500   \n",
       "3     Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500   \n",
       "4     Above 60,000  Betwn 0 & 3  Betwn 0 & 9,000       Betwn 0 & 2,500   \n",
       "\n",
       "   Gender defaulter  terms_lapsed disbursed_outstanding_ratio  clusters  \\\n",
       "0    Male       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "1  Female       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "2  Female       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "3  Female       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "4    Male       Yes  Betwn 0 & 70               Betwn 0 & 0.2         1   \n",
       "\n",
       "  cluster_category  OutstandingLoan installmentstartdate Lastinstallmentdate  \\\n",
       "0        cluster 2        -18134.96           2022-02-18          2022-02-18   \n",
       "1        cluster 2         -6769.98           2021-12-31          2021-12-31   \n",
       "2        cluster 2        -10000.00           2021-12-30          2021-12-30   \n",
       "3        cluster 2        -30000.00           2021-12-26          2021-12-26   \n",
       "4        cluster 2         -1269.98           2021-12-26          2021-12-26   \n",
       "\n",
       "  disbursementdate  \n",
       "0       2022-01-19  \n",
       "1       2021-12-01  \n",
       "2       2021-11-30  \n",
       "3       2021-11-26  \n",
       "4       2021-11-26  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "8c38c580",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_.to_csv(\"cleaned_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "378ca659",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}