01_Creditworthiness_Data_Cleaning_Exploration.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Project: Creditworthiness (Part1)\n",
    "\n",
    "### by Sooyeon Won\n",
    "### Keywords\n",
    "- Analytical Framework\n",
    "- Data Cleaning \n",
    "- Data Exploration \n",
    "- Data Visualization "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Table of Contents\n",
    "<ul>\n",
    "<li><a href=\"#Intro\">Introduction (Part1)</a></li>\n",
    "<li><a href=\"#Preparation\">Data Preparation (Part1)</a></li>\n",
    "<li><a href=\"#Exploration\">Data Exploration (Part1)</a></li>\n",
    "<li><a href=\"#Analysis\">Data Analysis (Part2)</a></li>\n",
    "<li><a href=\"#Conclusion\">Conclusion (Part3)</a></li>\n",
    "</ul>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='Intro'></a>\n",
    "## Introduction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It is the project of **Classification Models** from [Predictive Analytics for Business Nanodegree in Udacity](https://www.udacity.com/course/predictive-analytics-for-business-nanodegree--nd008). Classification algorithms are used to predict the probability of a categorical dependent variable(a.k.a. target variable). It is a powerful tool for business analysis. For this project, I approached the business issues using binary and non-binary classification models such as **Logistic Regression, Decision Tree, Random Forest, Boosted Model**. Based on the model comparisons, I investigate the optimal business insights. The business problem is as follows. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> At a small bank, I am tasked with determining if customers are creditworthy to give a loan to. Typically, my team gets 200 loan applications per week and approves them by hand. Due to a financial scandal that hit a competitive bank last week, the team suddenly have an influx of new people applying for loans for your bank instead of the other bank in your city. All of a sudden the team has nearly 500 loan applications to process this week. My task is to systematically evaluate the creditworthiness of these new loan applicants using classification modelling. <br>\n",
    "For this project, I analyse the business problem using the Problem Solving Framework and provide a list of creditworthy customers to the manager. The provided dataset contains the following information to work with:<br>\n",
    "- Data on all past applications\n",
    "- The list of customers that need to be processed in the next few days \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The analysis is separated into three parts. \n",
    "1. The first part of the analysis (Filename: 01_Creditworthiness_Data_Cleaning_Exploration.ipynb) contains the process of **Data Cleaning** and **Data Exploration with appropriate visualizations**. At the end of the Part1, I select the features associated with the creditworthiness of customers, which is the target variable of the analysis. \n",
    "2. In the second part of the analysis (Filename: 02_Creditworthiness_Data_Analysis.ipynb), I trained the dataset with various classification models: Logistic Regression, Decision Tree, Random Forest, Gradient Boosted Model, and AdaBoost Model. Then they are validated with the test dataset. Additionally, I evaluate the performance of the models with k-fold cross validation techniques.  \n",
    "3. After selecting the model with best performance, I conclude that how many new credit applicants are classified as creditworthy customers. \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='Preparation'></a>\n",
    "## Data Preparation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### - Data Assessment & Cleaning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import the relevant libraries \n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn import preprocessing\n",
    "import matplotlib.pyplot as plt \n",
    "plt.rc(\"font\", size=14)\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import train_test_split\n",
    "import seaborn as sns\n",
    "sns.set()\n",
    "sns.set(style=\"whitegrid\", color_codes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the data \n",
    "raw_data = pd.read_excel('credit-data-training.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Credit-Application-Result</th>\n",
       "      <th>Account-Balance</th>\n",
       "      <th>Duration-of-Credit-Month</th>\n",
       "      <th>Payment-Status-of-Previous-Credit</th>\n",
       "      <th>Purpose</th>\n",
       "      <th>Credit-Amount</th>\n",
       "      <th>Value-Savings-Stocks</th>\n",
       "      <th>Length-of-current-employment</th>\n",
       "      <th>Instalment-per-cent</th>\n",
       "      <th>Guarantors</th>\n",
       "      <th>Duration-in-Current-address</th>\n",
       "      <th>Most-valuable-available-asset</th>\n",
       "      <th>Age-years</th>\n",
       "      <th>Concurrent-Credits</th>\n",
       "      <th>Type-of-apartment</th>\n",
       "      <th>No-of-Credits-at-this-Bank</th>\n",
       "      <th>Occupation</th>\n",
       "      <th>No-of-dependents</th>\n",
       "      <th>Telephone</th>\n",
       "      <th>Foreign-Worker</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>1494</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>1</td>\n",
       "      <td>None</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Other Banks/Depts</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>1494</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>1</td>\n",
       "      <td>None</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>29.0</td>\n",
       "      <td>Other Banks/Depts</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>No Problems (in this bank)</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>1544</td>\n",
       "      <td>None</td>\n",
       "      <td>1-4 yrs</td>\n",
       "      <td>2</td>\n",
       "      <td>None</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>42.0</td>\n",
       "      <td>Other Banks/Depts</td>\n",
       "      <td>2</td>\n",
       "      <td>More than 1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>No Problems (in this bank)</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>3380</td>\n",
       "      <td>None</td>\n",
       "      <td>1-4 yrs</td>\n",
       "      <td>1</td>\n",
       "      <td>None</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>37.0</td>\n",
       "      <td>Other Banks/Depts</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>No Account</td>\n",
       "      <td>6</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>343</td>\n",
       "      <td>None</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>4</td>\n",
       "      <td>None</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>27.0</td>\n",
       "      <td>Other Banks/Depts</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Credit-Application-Result Account-Balance  Duration-of-Credit-Month  \\\n",
       "0              Creditworthy    Some Balance                         4   \n",
       "1              Creditworthy    Some Balance                         4   \n",
       "2              Creditworthy    Some Balance                         4   \n",
       "3              Creditworthy    Some Balance                         4   \n",
       "4              Creditworthy      No Account                         6   \n",
       "\n",
       "  Payment-Status-of-Previous-Credit       Purpose  Credit-Amount  \\\n",
       "0                           Paid Up         Other           1494   \n",
       "1                           Paid Up  Home Related           1494   \n",
       "2        No Problems (in this bank)  Home Related           1544   \n",
       "3        No Problems (in this bank)  Home Related           3380   \n",
       "4                           Paid Up  Home Related            343   \n",
       "\n",
       "  Value-Savings-Stocks Length-of-current-employment  Instalment-per-cent  \\\n",
       "0           £100-£1000                        < 1yr                    1   \n",
       "1           £100-£1000                        < 1yr                    1   \n",
       "2                 None                      1-4 yrs                    2   \n",
       "3                 None                      1-4 yrs                    1   \n",
       "4                 None                        < 1yr                    4   \n",
       "\n",
       "  Guarantors  Duration-in-Current-address  Most-valuable-available-asset  \\\n",
       "0       None                          2.0                              1   \n",
       "1       None                          2.0                              1   \n",
       "2       None                          1.0                              1   \n",
       "3       None                          1.0                              1   \n",
       "4       None                          1.0                              1   \n",
       "\n",
       "   Age-years Concurrent-Credits  Type-of-apartment No-of-Credits-at-this-Bank  \\\n",
       "0        NaN  Other Banks/Depts                  2                          1   \n",
       "1       29.0  Other Banks/Depts                  2                          1   \n",
       "2       42.0  Other Banks/Depts                  2                More than 1   \n",
       "3       37.0  Other Banks/Depts                  2                          1   \n",
       "4       27.0  Other Banks/Depts                  2                          1   \n",
       "\n",
       "   Occupation  No-of-dependents  Telephone  Foreign-Worker  \n",
       "0           1                 2          1               2  \n",
       "1           1                 2          1               2  \n",
       "2           1                 2          1               1  \n",
       "3           1                 2          1               1  \n",
       "4           1                 1          1               1  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_data.head() # Glance at the dataset(1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(500, 20)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_data.shape # Glance at the dataset(2). The dataset contains 500 datapoints with 20 fields."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 500 entries, 0 to 499\n",
      "Data columns (total 20 columns):\n",
      " #   Column                             Non-Null Count  Dtype  \n",
      "---  ------                             --------------  -----  \n",
      " 0   Credit-Application-Result          500 non-null    object \n",
      " 1   Account-Balance                    500 non-null    object \n",
      " 2   Duration-of-Credit-Month           500 non-null    int64  \n",
      " 3   Payment-Status-of-Previous-Credit  500 non-null    object \n",
      " 4   Purpose                            500 non-null    object \n",
      " 5   Credit-Amount                      500 non-null    int64  \n",
      " 6   Value-Savings-Stocks               500 non-null    object \n",
      " 7   Length-of-current-employment       500 non-null    object \n",
      " 8   Instalment-per-cent                500 non-null    int64  \n",
      " 9   Guarantors                         500 non-null    object \n",
      " 10  Duration-in-Current-address        156 non-null    float64\n",
      " 11  Most-valuable-available-asset      500 non-null    int64  \n",
      " 12  Age-years                          488 non-null    float64\n",
      " 13  Concurrent-Credits                 500 non-null    object \n",
      " 14  Type-of-apartment                  500 non-null    int64  \n",
      " 15  No-of-Credits-at-this-Bank         500 non-null    object \n",
      " 16  Occupation                         500 non-null    int64  \n",
      " 17  No-of-dependents                   500 non-null    int64  \n",
      " 18  Telephone                          500 non-null    int64  \n",
      " 19  Foreign-Worker                     500 non-null    int64  \n",
      "dtypes: float64(2), int64(9), object(9)\n",
      "memory usage: 78.2+ KB\n"
     ]
    }
   ],
   "source": [
    "raw_data.info() # Check the datatype of each field and missing values in each field."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> Brief Descriptions of Input variables: <br>\n",
    "1. **Account-Balance** (categorical: 'No Account', 'Some Balance'): Has an account with some balance?\n",
    "2. **Duration-of-Credit-Month** (numeric): Credit duration in month.\n",
    "3. **Payment-Status-of-Previous-Credit** (categorical: 'Paid Up', 'No Problems (in this bank)', 'Some Problems'): Has problems with previous credit?\n",
    "4. **Purpose** (categorical: 'Home Related', 'Used car', 'New car', 'Other' ): The reasons for taking credit.\n",
    "5. **Credit-Amount** (numeric): The amount of credit. \n",
    "6. **Value-Savings-Stocks** (categorical:'£100-£1000', 'None', '< £100'): The value of saving stocks for each customer. \n",
    "7. **Length-of-current-employment** (categorical: '< 1yr', '1-4 yrs', '4-7 yrs'): The number of years of employment. \n",
    "8. **Instalment-per-cent** (numeric): The percentage of the annual instalment. \n",
    "9. **Guarantors** (binary: \"None\", \"Yes\"): has a guarantor?\n",
    "10. **Duration-in-Current-address** (numeric)\n",
    "11. **Most-valuable-available-asset** (numeric)\n",
    "12. **Age-years** (numeric): The age of each customers in years. \n",
    "13. **Concurrent-Credits** (categorical: 'Other Banks/Depts'): has the customer concurrent credits? \n",
    "14. **Type-of-apartment** (categorical: '1', '2', '3', '4')\n",
    "15. **No-of-Credits-at-this-Bank**  (categorical: '1', 'More than 1')\n",
    "16. **Occupation** (binary: '1': yes, '0': no): are employed? \n",
    "17. **No-of-dependents** (numeric): The number of dependents.\n",
    "18. **Telephone** (numeric): The number of contacts with telephone\n",
    "19. **Foreign-Worker** (binary: '1': yes, '0': no)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x000001692D7CFB48>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001692DFC2048>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001692DFF2D48>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x000001692E02AAC8>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001692E062848>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001692E094F88>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x000001692E0D40C8>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001692E10B108>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001692E10FCC8>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x000001692E148EC8>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001692E1B0448>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001692E1E8588>]],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6cAAAJPCAYAAACAfKIrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzde1wU9f4/8Ndy9YJlFKhfU0/HBA1UCi0pb5ByEVaUPIUX0EOp2cUgUxExSzNvJGne0uPDOuYpEUGSlDRN08Ab5YUe3krBGwFGotx2F/bz+8MH8wPd5brLDPB6Ph49YmZnPvv+zM68/bx3ZmdUQggBIiIiIiIiIhlZyB0AEREREREREYtTIiIiIiIikh2LUyIiIiIiIpIdi1MiIiIiIiKSHYtTIiIiIiIikh2LUyIiIiIiIpIdi1MySqfTYeDAgXjttdfkDoWIqMHKy8uxefNmBAUFITAwECNGjMDy5cuh1Wob1O7TTz+N69ev4+zZs5g+fToA4MyZM3j//ferXU+JOTYsLAz5+flyh0HUbFy/fh29evVCYGAgAgMDoVarERwcjN27d5v0fSofu5MnT8bvv/9usrYDAwNx586dOq/3xx9/4O2334ZarcbIkSMxYcIEnDx50mRx1dXq1avxww8/NKiN/Px8ODs7G3xt06ZNiIyMbFD7BFjJHQAp1759+9CzZ09kZGTgjz/+QPfu3eUOiYio3j744AMUFBTgyy+/RLt27VBcXIz33nsPc+fOxfLlyxvcfu/evbFq1SoAwO+//46cnJxql1dijv3555/lDoGo2WnVqhWSkpKk6Rs3bmDSpEmwtLSEj4+PSd6j8rG7ceNGk7RZoXLstXX58mVMnDgRixcvxqBBgwAAaWlpeP311/H111+jR48eJo2xNo4dO4Ynn3yy0d+X6obFKRn19ddfY8SIEejatSu+/PJLLFiwAACwYcMGxMfHo23btujXrx/279+PAwcOQKvVIiYmBidOnEB5eTmeeuopREdHw87Orkq76enpmDFjBg4cOAALCwuUlJTAy8sL3333HXQ6HRYsWIDs7GzodDr4+/vj9ddfBwCsX78e+/fvR2lpKUpKSjB79mwMHz4cn332GU6dOoXc3Fw4Oztj2rRpmDt3LrRaLYQQGDNmDMaPH9/o24+IlOP69evYtWsXjhw5IuWkNm3a4MMPP8Qvv/yCyMhI3L59G9euXcPQoUPxzjvvGM1nJ0+exMKFC6FSqdC7d2/o9XoA9wY+CxcuxMaNG7Fq1SrcvXsXc+bMweLFiw3GZCzHHjt2DCtWrECnTp1w5coVtG7dGlOmTMGWLVtw5coVeHt7IyoqCgCwbds2bNmyBRYWFnjssccwb948PPHEE4iMjESPHj3w6quvAkCVaS8vL4wePRppaWnIzs5GYGAgwsPDMWfOHADAxIkTsWHDBnTq1MmsnwlRS9W5c2dMnz4dmzZtwo8//ljtsdqnTx9cuHAB7777LqysrPD5559Dq9UiPz8fo0aNMnjsjh8/HitXrkTv3r2rzRF2dna4cOEC/vzzTzg7O2Pp0qVo27btA/E6OzsjLS0NBw8exL59+2BhYYGsrCy0atUKS5cuNfjF2saNG/HSSy9JhSkAeHh44JNPPkGrVq1w/fp1qNVq/PrrrwBQZTohIQHx8fEoKSmBnZ0dRo8eXWV6y5Yt2L59O77++mvo9Xq0b98e8+bNQ/fu3Y32a+fOncjIyMCyZctgaWmJ4cOHV4n3ypUrWLBgAYqKipCXl4eePXvi008/ha2tLfbu3YvY2Fi0bt0arq6u0jo6nQ4fffQRUlNT8eijj+LRRx9Fu3btAAAhISF4+OGHcfnyZYwdOxajRo3CokWLcPHiReh0Onh4eGDWrFmwsrLCqlWrsG/fPlhbW+ORRx7B4sWL4ejoaHR+syeIDLh06ZJwcXER+fn54vTp06JPnz4iPz9f/PTTT8LHx0cUFBQIvV4v5syZIzw9PYUQQnz22WdiyZIlQq/XCyGE+OSTT8T8+fMNtj9y5Ehx8OBBIYQQ27dvFxEREUIIIUJCQsT+/fuFEEKUlpaKkJAQ8d1334nr16+LkJAQUVJSIoQQIjk5WQQEBAghhFi1apXw8fEROp1OCCHEnDlzxOeffy6EECI3N1eEh4eL8vJyM2wlImoqUlJSxEsvvWT09dmzZ4uJEydK08bymUajEc8//7xITU0VQgixa9cu4eTkJK5duyaOHj0q/P39hRBC7NixQ0yZMsXo+xnLsUIIcfToUdGrVy/x22+/CSGEePXVV8Urr7wiNBqN+Ouvv4SLi4v4888/RWpqqhg2bJj466+/pPf08/MTer1ezJ49W/znP/+p0r+KaU9PT7FkyRIhhBB//vmn6N27t7h69aoQQggnJyepPSJquGvXrgk3N7cH5l+8eFH07du3xmN19erVQggh9Hq9mDBhgrhy5YoQ4t6x26tXL+l4rXzsenp6ijNnztSYIyryilarFaNGjRLx8fEG+1DR9o4dO4S7u7vIzs4WQgixYMECMWvWLIPrBAQESOO82myXytM7duwQ/fv3F3fv3jU4fezYMTFu3DhRXFwshBDi8OHDwtfXV9p+xvo1YcIEsWfPHoPxLFmyROzcuVMIIYRWqxUBAQEiJSVF5OXlCXd3d3Hp0iUhhBDr168XTk5OQgghvvjiCxEaGio0Go0oKioSo0ePFrNnz5bea86cOVL7kZGR4r///a8QQoiysjLx3nvviQ0bNoibN2+KZ555Rmg0GiGEEJs2bRL79u0zOr8l4JlTMujrr7+Gp6cnHnnkETzyyCN4/PHHERcXh7y8PPj6+uKhhx4CAIwfPx5Hjx4FABw8eBB3795FamoqgHvfKD366KMG2x8/fjzi4uIwZMgQbNu2DbNmzUJxcTFOnDiBgoICrFy5EgBQXFyM8+fPY8SIEVi2bBl27dqFrKwsnD59GkVFRVJ7bm5usLK6tzsPHz4cs2fPxpkzZ+Dh4YHo6GhYWPDn1UQtmYWFhXSG0xh3d3fpb2P57OLFi7CysoKHhwcAICAgoMbflhpiLMdOnToVAPD444/jqaeeAgB07doV7dq1g42NDezt7dG2bVsUFBTg8OHDGDFiBOzt7QEAQUFBWLRoEa5fv17j+7/44osAgA4dOuDRRx9FQUEBunTpUud+EFH9qFQqtGrVqsbl+vXrJy2/fv16HDx4EMnJyfjjjz8ghEBJSYnRdWvKEYMGDYKNjQ0AwMnJCQUFBTXG4+Ligo4dOwIAnnrqKezbt89o/2rKudVxdnaucuVd5emDBw8iKysLwcHB0ut37tzB7du3AdSvXzNnzsTPP/+MjRs3IjMzE7m5uSguLkZ6ejqcnJyky4FfeeUVrFixAsC9y5QDAgJgY2MDGxsbqNVqXLhwQWqz4rOriPns2bOIj48HAJSWlgK4l4N79uyJ0aNHY/DgwRg8eDA8PDyg1+sNzm8JWJzSA4qLi5GUlAQbGxt4eXkBAAoLC/HVV1/B398fQghpWUtLS+lvvV6PqKgoDBkyBABQVFQEjUaDs2fPIjo6WlouKSkJarUaK1aswNGjR1FcXIz+/fujsLAQQgh88803aN26NYB7Pzy3tbXFb7/9hjfeeAOTJk3CCy+8gP79++PDDz+U2mzTpo30t6enJ77//nukpqYiLS0Na9asQUJCgpRMiajl6dOnDy5fvozCwsIqA56cnBzMmzcPbdq0qZJHjOWzmzdvVsmBAKQvxoxZuXIlDhw4AADw8vLC5MmTjebYsLAwAJAGVtW9h6GBnxACZWVlUKlUVeLU6XRVlrO1tZX+vn9ZIjK/s2fPwsnJqcZjtSIvFRcXY/To0Rg2bBj69euHl156CT/88EO1x251OQJAleK4Io79+/dLv513dHR84PerhtbJycnBlClTpPkbNmyAm5sbTp06BU9Pzyrrr169Gl27doW7u3ut+m1oWq/XIzAwEDNnzpSmc3Nz8fDDDxuN8X6BgYHS3x999BE2bdqE8vJy+Pn5YejQocjOzpbWq7x+dfm+8pjYUMwrV66ULoG+c+cOVCoVLCws8NVXX+Hs2bNIS0vDxx9/jEGDBmHWrFlG5zd3PJ1ED9i1axfat2+Pw4cP48CBAzhw4AB++OEHFBcXw8XFBXv37sXdu3cBQPoGCAAGDhyIrVu3QqvVQq/XY968eVixYgV69+6NpKQk6T8AaN26NUaOHImoqCjpmy87Ozu4ublh8+bNAO4duGPHjsX+/ftx4sQJuLq64t///jeeffZZ7N+/H+Xl5QbjnzFjBnbv3g1/f3/Mnz8fdnZ2uHr1qjk3GREpXIcOHaBWqxEVFYXCwkIA9wrCDz74AO3bt3/gDIaxfObs7AwhBA4dOgQA2L9/v8Fv5S0tLaUB4DvvvCPlv3feeafaHJuSklLrPg0aNAi7d++W7tC5Y8cOtG/fHt26dcMjjzyCjIwMAPcK8OPHj9eqzcpxE5F5XLlyBWvXrkVYWFitj9WsrCwUFhYiPDwcXl5eOHbsmJSfAMPHbnU5wpgXX3xRyle1vbFShw4dqozzOnTogFdffRXbt2/HkSNHpOV++uknbNmyBT179sRDDz0EnU4n3VX4u+++q9V7Affy83fffYfc3FwA965EmThxYo3rVd5GlePt3bs3jhw5gjfffBMjRowAAJw+fRrl5eXo378/fv/9d5w/fx4AkJCQILU3aNAg7Ny5ExqNBhqNpto7MA8cOBBffPEFhBDQarWYNm0avvrqK5w/fx4BAQHo3r07pk6dikmTJuHs2bNG57cEPHNKD/j666/x73//u8o3QA899BBCQkLwxRdf4OWXX8Yrr7yCVq1aoUePHtJZzjfeeANLly7F6NGjUV5ejl69elV7S+2goCDExcVh1KhR0ryYmBgsXLgQarUaWq0WAQEBGDlyJG7duoW9e/fCz88Per0enp6eKCgokAaZlb3xxhuYO3cutm3bBktLSwwbNgz9+/c34RYioqZo/vz5WLt2LYKDg2FpaQmtVothw4bh7bfffuDSXGP5zNraGmvWrMEHH3yAFStWoFevXgZ/vuDm5oY1a9bgrbfewurVq6u8VlOOre034y+88AImTZqEiRMnQq/Xw97eHp9//jksLCwQEhKC9957Dz4+Pnj88ccxYMCAWrXp6+uLkJAQfPbZZ3BycqrVOkRUvdLSUulMnYWFBWxtbfHuu+9i6NChcHZ2rtWx6uzsjKFDh8LPzw82NjbSpaZZWVno2rVrlWO3QnU5wty6deuG9evX49NPP8XSpUul91+3bp2UW2bOnInJkyfD3t4evr6+tW574MCBmDx5MsLCwqBSqWBnZ4fVq1dDpVJVu56XlxdWrFgBnU6H0aNHV3ktIiICb775Jtq0aQM7Ozv0798fV69ehb29PWJiYvDee+/B2tq6yngyODgYV69eRUBAQI1F/9y5c7Fo0SKo1WrodDo8//zzeO2112BtbQ0/Pz+89NJLaNOmDVq1aoXo6Gj07NnT4PyWQCV4LQ/VwdmzZ/Hrr78iNDQUALB582acPn0an376aZ3aEUJg48aNuHHjRpXLc4mIiIiIqGVicUp1UlhYiKioKFy+fBkqlQqdOnXCwoUL0aFDhzq14+XlBUdHR6xdu1b6oT4REREREbVcLE6JiIiIiIhIdrwhEhEREREREcmOxSkRERERERHJTjF369Xr9SgqKoK1tXWNd9sioqZFCAGdToe2bds2yl0CmyrmQaLmi3mw9pgLiZqvmnKhYorToqIiXLx4Ue4wiMiMnJyc0K5dO7nDUCzmQaLmj3mwZsyFRM2fsVyomOLU2toawL1AbWxs6rx+RkYGXF1dTR2WSSk9RsbXMEqPD5AvRq1Wi4sXL0rHORlWmzzYFPYzc2ip/QZabt+bW7+ZB2uvrmNCpe4rSoxLiTEByoxLiTEBTT+umnKhYorTiss2bGxsYGtrW6826rteY1J6jIyvYZQeHyBvjLw8q3q1zYNNYT8zh5bab6Dl9r059pt5sGb1GRMqdV9RYlxKjAlQZlxKjAloHnEZy4X80QMRERERERHJjsUpERERERERyY7FqQFaXblZ2u3Zy8Us7RJRy2WufGWudomIiEge5vy33VR1jmJ+c6okNtaWUM9IMnm7uz4JNHmbRNSyMV8RERFRbZhrzACYbtzAM6dEREREREQkOxanREREREREJDsWp0RERERERCQ7FqdEREREREQkOxanREREREREJDsWp0RERERERCQ7FqdEREREREQkOxanRERERETNhFZXbvQ1d3d3s7RLZCpWcgdARERERESmYWNtCfWMJJO3u+uTQJO3SXQ/njklIqqF1atXw9/fH/7+/li2bBkAIDU1FWq1Gt7e3oiNjZWWPXfuHIKCguDj44O5c+eirKxMrrCJiIiImgwWp0RENUhNTcWRI0eQmJiInTt34rfffkNycjKioqKwdu1a7N69GxkZGTh06BAAYObMmXj//ffx/fffQwiBuLg4mXtAREREpHwsTomIauDg4IDIyEjY2NjA2toa3bt3R2ZmJrp164YuXbrAysoKarUaKSkpuHHjBkpLS+Hm5gYACAoKQkpKisw9ICIiIlI+/uaUiKgGPXr0kP7OzMzEnj17MGHCBDg4OEjzHR0dkZOTg9zc3CrzHRwckJOTU6f3y8jIqPb19PR06e+G3NyiJpXfRwmUFk9jaql9b6n9JiJqqVicEhHV0qVLlzB16lTMmjULlpaWyMzMlF4TQkClUkGv10OlUj0wvy5cXV1ha2tr8LX09HSzFqSVNdb71EZj9ltpWmrfm1u/NRpNjV88ERG1dLysl4ioFtLT0zFp0iTMmDEDo0ePRseOHZGXlye9npeXB0dHxwfm37p1C46OjnKETERERNSksDglIqpBdnY23nzzTcTExMDf3x8A0LdvX1y5cgVZWVkoLy9HcnIyBg8ejM6dO8PW1la6HDEpKQmDBw+WM3wiIiKiJoGX9RIR1WDTpk3QaDRYsmSJNC84OBhLlizB22+/DY1GgyFDhsDX1xcAEBMTg+joaBQWFsLFxQWhoaFyhU5ERETUZLA4JSKqQXR0NKKjow2+9u233z4wr2fPnoiPjzd3WERERETNCi/rJSIiIiIiItmxOCUiIiIiIiLZsTglIiIiIiIi2bE4JSIiIiIiItmxOCUiIiIiIiLZsTglIiIiIiIi2dWqOC0sLERAQACuX78OAEhNTYVarYa3tzdiY2Ol5c6dO4egoCD4+Phg7ty5KCsrM0/URERERERE1KzUWJyePn0aY8eORWZmJgCgtLQUUVFRWLt2LXbv3o2MjAwcOnQIADBz5ky8//77+P777yGEQFxcnFmDJyIiIiJlWLp0KSIjIwHwhAUR1U+NxWlcXBzmz58PR0dHAMCZM2fQrVs3dOnSBVZWVlCr1UhJScGNGzdQWloKNzc3AEBQUBBSUlLMGz0RERERyS4tLQ2JiYnSNE9YEFF9WNW0wKJFi6pM5+bmwsHBQZp2dHRETk7OA/MdHByQk5NT54AyMjLqvE6F9PT0eq9bmbu7u0naMcRUMZoL42sYpccHNI0YiYio6bh9+zZiY2Px+uuv4/z58wZPWKxatQrjxo2TOVIiUroai9P76fV6qFQqaVoIAZVKZXR+Xbm6usLW1rbO66Wnp5u1qDQVJceo9G3I+BpOrhg1Gk2DvngiIiLlev/99xEREYHs7GwAD57IaIwTFkr94lWOuJrqSRYlfoZKjAmof1zmHgOaYnvVuTjt2LEj8vLypOm8vDw4Ojo+MP/WrVvSpcBERERE1Pxs374dnTp1goeHBxISEgAYP5FRV7U9YaHUL4eVGldDmKs/StxWSowJUG5cQO32j5pOWNS5OO3bty+uXLmCrKwsPP7440hOTsZLL72Ezp07w9bWVtpgSUlJGDx4cF2bJyIiIqImYvfu3cjLy0NgYCAKCgpQXFwMlUrFExZEVC91Lk5tbW2xZMkSvP3229BoNBgyZAh8fX0BADExMYiOjkZhYSFcXFwQGhpq8oCJiIiISBk2b94s/Z2QkIDjx49j8eLFCAgI4AkLIqqzWhenBw4ckP728PDAt99++8AyPXv2RHx8vGkiIyJSmMLCQgQHB2P9+vV4/PHHMWfOHKSnp6N169YAgLfeegvDhw9HamoqFi9eDI1GAz8/P0RERMgcORFR4+IJCyKqjzqfOSUiaolOnz6N6Oho6ZnPwL2bdXz11VdVLlereBb0li1b0KlTJ0ydOhWHDh3CkCFDZIiaiKjxBAUFISgoCABPWBBR/dT4nFMiInrwmc8lJSW4efMmoqKioFarsWrVKuj1eqPPgiYiIiKi6vHMKRFRLdz/zOdbt25hwIABmD9/Ptq1a4epU6ciPj4ebdq0Mfgs6Lqo6fEJlW/V3lQfGVAfSounMbXUvrfUfhMRtVQsTomI6qFLly5Ys2aNNB0SEoKdO3fCx8enwY9QqO7xCY15C3kl3apeybfON7eW2vfm1m8+75mIqGa8rJeIqB4uXLiA77//XpoWQsDKysros6CJiIiIqHosTomI6kEIgY8//hgFBQXQ6XTYtm0bhg8fXuVZ0OXl5UhOTuYjFIiIiIhqgZf1NiKtrhw21pZNrm0ielDPnj0xZcoUjB07FmVlZfD29kZAQAAAGH0WdFPCfEVERESNjcVpI7KxtoR6RpJZ2t71SaBZ2iWiqio/83n8+PEYP378A8sYexZ0U8J8RURERI2Nl/USERERERGR7FicEhERERERkexYnBIREREREZHsWJwSERERERGR7FicNhNaXXmD2zD2sHNTtE1ERERERFQd3q23meCdNYmIiIiIqCnjmVMiIiIiIiKSHYtTIiIiIiIikh2LUyIiIiJqsnr2cjFb27zvBlHj4m9OiYiIiKjJatumFe+7QdRM8MwpERERERERyY7FKREREREREcmOxSkRERERERHJrskWp/f/QN3d3V2mSIiopSgsLERAQACuX78OAEhNTYVarYa3tzdiY2Ol5c6dO4egoCD4+Phg7ty5KCsrkytkIiIioiajyd4Qycbakj9+J6JGc/r0aURHRyMzMxMAUFpaiqioKGzZsgWdOnXC1KlTcejQIQwZMgQzZ87ERx99BDc3N0RFRSEuLg7jxo2TtwNERERECtdkz5wSETWmuLg4zJ8/H46OjgCAM2fOoFu3bujSpQusrKygVquRkpKCGzduoLS0FG5ubgCAoKAgpKSkyBk6ERERUZPQZM+cEhE1pkWLFlWZzs3NhYODgzTt6OiInJycB+Y7ODggJyen0eIkIiIiaqpYnBIR1YNer4dKpZKmhRBQqVRG59dFRkZGta+np6dLfzfV39tX7oM512kuWmrfW2q/iYhaKhanRET10LFjR+Tl5UnTeXl5cHR0fGD+rVu3pEuBa8vV1RW2trYGX0tPT2+yBWllde1Dc+l3fbTUvje3fms0mhq/eCIiaun4m1Mionro27cvrly5gqysLJSXlyM5ORmDBw9G586dYWtrK53xSUpKwuDBg2WOloiIiEj5eOaUiKgebG1tsWTJErz99tvQaDQYMmQIfH19AQAxMTGIjo5GYWEhXFxcEBoaKnO0RERERMrXoOI0JCQE+fn5sLK618yCBQtw9epVrFu3DmVlZZg4cSLGjx9vkkCJiJTgwIED0t8eHh749ttvH1imZ8+eiI+Pb8ywiIhks3r1auzZswcAMGTIEMyaNQupqalYvHgxNBoN/Pz8EBERIXOURNQU1Ls4FUIgMzMTP/74o1Sc5uTkICIiAgkJCbCxsUFwcDCee+45PPnkkyYLmIiIiIiUITU1FUeOHEFiYiJUKhVee+01JCcnIyYmxuBzoImIqlPv35xevnwZABAWFoaRI0fiq6++QmpqKgYMGID27dujTZs28PHx4fP9iIiIiJopBwcHREZGwsbGBtbW1ujevTsyMzMNPgeaiKgm9T5zeufOHXh4eGDevHnQ6XQIDQ2Fn5/fA8/9O3PmTJ3are2d7JrTHfyaAqXczl8pcRij9PiAphEjERE1DT169JD+zszMxJ49ezBhwgSDz4EmIqpJvYvTp59+Gk8//bQ0PWbMGCxevBjTpk2T5tXn+X7VPUKB5KOELwOU/lgBpccHyBcjH6FARNS8Xbp0CVOnTsWsWbNgaWmJzMxM6bX6jAcB5ZywaMiXunJ8IWzO7WHO/pij7Z69XNC2Tat6rVvTdiwqLsX5c7/Vq+2Gqu+2UvKxUqHexenJkyeh0+ng4eEB4F7i6dy5s8Hn/hERERFR85Seno7p06cjKioK/v7+OH78uEnGg0o5YVHfAX1T+NK6rszVH3NuK/WMJLO0u+uTQFk+XyXvV7WJq6YTFvX+zendu3exbNkyaDQaFBYWIjExEcuXL0daWhry8/NRUlKCvXv38vl+RERERM1UdnY23nzzTcTExMDf3x+A8edAExHVpN5nTj09PXH69GmMGjUKer0e48aNg7u7OyIiIhAaGgqdTocxY8agT58+poyXiIiIiBRi06ZN0Gg0WLJkiTQvODjY6HOgiYiq06DnnIaHhyM8PLzKPLVaDbVa3aCgiIiIiEj5oqOjER0dbfA1Q8+BJiKqTr0v66WWQ6srb1LtEhERERFR09OgM6fUMthYW5rlx+S7Pgk0eZtERERERNQ08cwpERERERERyY7FKREREREREcmOxSkRERERERHJjsUpERERERERyY43RCIiaoCQkBDk5+fDyupeOl2wYAGuXr2KdevWoaysDBMnTsT48eNljpKIiIhI+Vickmy0unLYWFvWenl3d3eztU1UH0IIZGZm4scff5SK05ycHERERCAhIQE2NjYIDg7Gc889hyeffFLmaImIiIiUjcUpycZcj6gB+JgaahyXL18GAISFheH27dt4+eWX0bZtWwwYMADt27cHAPj4+CAlJQVvvfWWnKESERERKR5/c0pEVE937tyBh4cH1qxZgy+++ALffPMNbt68CQcHB2kZR0dH5OTkyBglERERUdPAM6dERPX09NNP4+mnn5amx4wZg8WLF2PatGnSPCEEVCpVndrNyMio9vX09HTp77pc7q4U9bnsvjb9LCouxflzv9U3LEWr/Jm3JC2130RELRWLU7mlmBUAACAASURBVCKiejp58iR0Oh08PDwA3CtEO3fujLy8PGmZvLw8ODo61qldV1dX2NraGnwtPT29SRaklZnrkv5dnwQ2+W1jSHP4zOujufVbo9HU+MUTEVFLx8t6iYjq6e7du1i2bBk0Gg0KCwuRmJiI5cuXIy0tDfn5+SgpKcHevXsxePBguUMlIiIiUjyeOSUiqidPT0+cPn0ao0aNgl6vx7hx4+Du7o6IiAiEhoZCp9NhzJgx6NOnj9yhEhERESkei1MiogYIDw9HeHh4lXlqtRpqtVqmiFoucz5Cio+nIiIiMj8Wp0RE1Czw8VRERERNG39zSs2SVlfepNolIiIiImrpeOaUmiVz3g2UiIiIiIhMj2dOiYiIiIiISHYsTomIiIiIiEh2LE6JiIiIiIhIdixOiYiIiIiISHYsTomIiIiIiEh2LE6JiIiIiIhIdixOiRSCz2YlIiIiopaMzzklqgOtrhw21pYGX3N3d29Q23w2K5FyVXfsK7FdIiKipojFKVEdmKuABFhEEikZvzwiIiIyP17WS0RERERERLJjcUpERERERESyM0txumvXLowYMQLe3t7YunWrOd6CiGqp8g2RGvq72OrapqqYB0lu5jw+eexTbTEXElFdmPw3pzk5OYiNjUVCQgJsbGwQHByM5557Dk8++aSp34qIaoG/k218zINUW7W5IVJ9v1Qy57G/Y0mAWdrlDaKaF+ZCIqorkxenqampGDBgANq3bw8A8PHxQUpKCt566y1TvxURkSIxD1JtNdUvj3iDKKoN5kIiqiuTF6e5ublwcHCQph0dHXHmzJka1xNCAAC0Wm2t36t9W/N8u6rRaMzStrnabaptM+bGadvcMddGxXFdcZw3d+bOg/dv96a43zBm87bbVNsuLCqGtdW9dl1dXWudY2pDW1YOGyvTx6wrK5dirvb9W1geBJrPmFDO9evLXPnKnMzVvlL3DbneW+7tUVMuVAkTZ8l169ZBo9EgPDwcABAXF4eMjAwsWLCg2vXu3r2LixcvmjIUIlIYJycntGvXTu4wzI55kIiMaSl5EGAuJCLjjOVCk5857dixI06ePClN5+XlwdHRscb12rZtCycnJ1hbW0OlUpk6LCKSkRACOp0Obdu2lTuURsE8SET3a2l5EGAuJKIH1ZQLTV6cPv/88/jss8+Qn5+P1q1bY+/evVi4cGGN61lYWLSYbxKJWqJWrVrJHUKjYR4kIkNaUh4EmAuJyLDqcqHJi9MOHTogIiICoaGh0Ol0GDNmDPr06WPqtyEiUizmQSIi5kIiqjuT/+aUiIiIiIiIqK4s5A6AiIiIiIiIiMUpERERERERyY7FKREREREREcmOxSkRERERERHJjsUpERERERERya7JFqerV6+Gv78//P39sWzZMgBAamoq1Go1vL29ERsbK2t8K1euxIgRI+Dv74/NmzcrLr4KS5cuRWRkJADg3LlzCAoKgo+PD+bOnYuysjJZYwsJCYG/vz8CAwMRGBiI06dPY9euXRgxYgS8vb2xdetWWeM7cOAAgoKC4Ofnh48++giAcj7j7du3S9stMDAQ7u7uWLBggWLio/pR0v5vKnU5zo3tv0rLXcYUFhYiICAA169fB1D3/ty8eRPjx4+Hr68vpk2bhqKiIgDAnTt3MGXKFPj5+WH8+PHIy8tr/M7V4P6+z5kzB97e3tLnvm/fPgCm2ybUfN2/L1UmVy6oLqbVq1fD09NT2tcbK3cbGidXJte2qikuObaXoTF7ZXJtq5rikmvfAqrWD5WZJCeLJujnn38Wr7zyitBoNEKr1YrQ0FCxa9cuMWTIEHH16lWh0+lEWFiYOHjwoCzxHTt2TAQHBwudTidKSkqEp6enOHfunGLiq5Camiqee+45MXv2bCGEEP7+/uLXX38VQggxZ84csXXrVtli0+v1YuDAgUKn00nz/vzzT+Hp6Sn+/vtvUVRUJNRqtbh06ZIs8V29elUMHDhQZGdnC61WK8aOHSsOHjyouM9YCCEuXrwohg8fLm7evKnI+Kh2lLT/m0pdjvOSkhKj+6+Scpcxp06dEgEBAcLFxUVcu3atXv2ZMmWKSE5OFkIIsXr1arFs2TIhhBAffvih+Pzzz4UQQiQmJop33nmnsbtXrfv7LoQQAQEBIicnp8pyptwm1DwZ2pcqkyMX1BTT1KlTxS+//GL2OCozNE7eu3dvlWXk2Fa1iauxt5ehMfsff/xRZRk5tlVt4pJj3xLiwfqhMlPk5CZ55tTBwQGRkZGwsbGBtbU1unfvjszMTHTr1g1dunSBlZUV1Go1UlJSZInv2WefxX//+19YWVnhr7/+Qnl5Oe7cuaOY+ADg9u3biI2Nxeuvvw4AuHHjBkpLS+Hm5gYACAoKkjW+y5cvAwDCwsIwcuRIfPXVV0hNTcWAAQPQvn17tGnTBj4+PrLFuG/fPowYMQIdO3aEtbU1YmNj0bp1a0V9xhU++OADRERE4Nq1a4qMj2pHSfu/qdTlOD9z5ozB/VdpucuYuLg4zJ8/H46OjgBQ5/7odDqcOHECPj4+VeYDwMGDB6FWqwEAAQEB+Omnn6DT6WTopWH3972kpAQ3b95EVFQU1Go1Vq1aBb1eb9JtQs3T/ftSZXLlgupiAoCMjAx8/vnnUKvVWLBgATQajdljMjROvnnzpvS6XNuqpriAxt9ehsbsbdq0kV6Xa1vVFBcgz751f/1QmalycpMsTnv06CHtJJmZmdizZw9UKhUcHBykZRwdHZGTkyNXiLC2tsaqVavg7+8PDw8P5ObmKiq+999/HxEREXjooYcA4IH4HBwcZI3vzp078PDwwJo1a/DFF1/gm2++wc2bNxWzDbOyslBeXo7XX38dgYGB+N///qe4zxi4V9CUlpbCz89PkfFR7TXHz68ux7mx/istdxmzaNEi9OvXT5qua3/+/vtv2NnZwcrKqsr8+9uysrKCnZ0d8vPzG6NbtXJ/32/duoUBAwbg448/RlxcHE6ePIn4+HiTbhNqnu7flyqTKxdUF1NRURF69eqFmTNnIjExEXfu3MHatWvNHpOhcfKQIUOk1+XaVjXFJdf2un/M3qFDB+k1Of+NqS4uubbV/fVDZabKyU2yOK1w6dIlhIWFYdasWejSpQtUKpX0mhCiyrQcpk+fjrS0NGRnZyMzM1Mx8W3fvh2dOnWCh4eHNE+v1ysmPgB4+umnsWzZMrRr1w729vYYM2YMVq1apZgYy8vLkZaWho8//hjbtm3DmTNncO3aNcXEV+Gbb77Bv//9bwDK+4ypbprj51eX49xY/5vqdqlrfwz1y1g/hRCwsFDuP+9dunTBmjVr4OjoiNatWyMkJASHDh0y6zah5k+JuaBt27bYuHEjunfvDisrK4SFheHQoUON9v6Vx8n/+Mc/pPlybytjccm5vSqP2ePi4qT5cm8rY3HJsa0M1Q+VmSonK/dfrxqkp6dj0qRJmDFjBkaPHo2OHTtWuQlEXl6e0UsszO2PP/7AuXPnAACtW7eGt7c3jh07ppj4du/ejZ9//hmBgYFYtWoVDhw4gPj4+Crx3bp1S7b4AODkyZNIS0uTpoUQ6Ny5s2K24WOPPQYPDw/Y29ujVatWGDZsGFJTUxUTHwBotVqcOHECXl5eAKCoY4Tqrjl+fnU5zo31//75cueu2qprf+zt7XH37l2Ul5dXWR64d4bx1q1bAICysjIUFRWhffv2jdiburlw4QK+//57aVoIASsrK5NuE2p5lJgLbt68ifj4eGm6Yl9vDPePkyuTc1tVF5cc28vQmP3ChQvS63Jtq5rikmNbGaofPv74Y+l1U+XkJlmcZmdn480330RMTAz8/f0BAH379sWVK1ekyy2Tk5MxePBgWeK7fv06oqOjodVqodVqsX//fgQHBysmvs2bNyM5ORlJSUmYPn06vLy8sHjxYtja2iI9PR0AkJSUJFt8AHD37l0sW7YMGo0GhYWFSExMxPLly5GWlob8/HyUlJRg7969ssXo6emJI0eO4M6dOygvL8fhw4fh6+urmM8YuDcA/Mc//iH9RkFJxwjV3fPPP6+Y/d9U6nKcG9t/O3furKjcVVt17Y+1tTX69euH3bt3AwB27twp9XPIkCHYuXMngHuDh379+sHa2lqejtWCEAIff/wxCgoKoNPpsG3bNgwfPtyk24RaHiXmglatWmH58uW4du0ahBDYunUrhg8fbvb3NTROrkyubVVTXHJsL0Njdnd3d+l1ubZVTXHJsa0M1Q9RUVHS66bKyY3z9Y2Jbdq0CRqNBkuWLJHmBQcHY8mSJXj77beh0WgwZMgQ+Pr6yhLfkCFDcObMGYwaNQqWlpbw9vaGv78/7O3tFRGfMTExMYiOjkZhYSFcXFwQGhoqWyyenp44ffo0Ro0aBb1ej3HjxsHd3R0REREIDQ2FTqfDmDFj0KdPH1ni69u3L1577TWMGzcOOp0OL7zwAsaOHYt//vOfivmMr127ho4dO0rTtra2ijlGqO46dOigmP3fVOp6nBvbf5WUu2qruuPRWH/mz5+PyMhIrFu3Dp06dcKKFSsAAO+88w4iIyPh7++Pdu3aISYmRrZ+1UbPnj0xZcoUjB07FmVlZfD29kZAQACAun/GxrYJtRyTJ0/G9OnT0bt3b8XkgsoxLViwANOmTYNOp8Mzzzwj/dTGnIyNkw8cOCDrtqpNXI29vYyN2eXer2oTlxz7liFz586Fl5cXXnzxRZPkZJUQQpghTiIiIiIiIqJaa5KX9TYH169fR69evaQH56rVagQHB0unwk0lLCxMumvj5MmT8fvvv5u0fUPKy8sxbdo0+Pj44KuvvjK4zI8//oiQkBAEBgbC398f4eHhyM7ObvB7T506FQkJCQCAwMBA3LlzB3fv3q32Wy4vLy+4ubk98KDghIQEODs7N+iW4QcPHsTKlSul9qZOnVrvtoio7pydnaFWq6VcGxgYiLlz55qs/ZUrV0qX1DbUqVOn8Nxzz0Gv10vz3n33Xbi6uqKwsFCa98EHH2D58uV1ajskJISPWSFqZry8vHD27Nk6r1fTuKhCY41bahuPklUeb1PDNMnLepuLVq1aISkpSZq+ceMGJk2aBEtLS+kZQQ31888/S39v3LjRJG3WJCcnB0eOHMGpU6dgaWn5wOu7du3CunXrsG7dOnTr1g1CCGzYsAGhoaH47rvvYGNjY5I4Krbt9evXa0zejzzyCPbt24dRo0ZJ83bu3InHHnusQTGcPXsWBQUFDWqDiBrmyy+/hL29vVnafuedd0zWVsXlyxcuXECvXr1QVlaGY8eO4bnnnsPhw4fh5+cHADh69CgWLlxosvclopaloKCgXkWtuSgtnvqoPN6mhmFxqiCdO3fG9OnTsWnTJvz444/o0aMHXn31VQBAZGSkNO3l5YU+ffrgwoULePfdd2FlZYXPP/8cWq0W+fn5GDVqFMLDwzFnzhwAwMSJE7FhwwaMHz8eK1euRO/evbFt2zZs2bIFFhYWeOyxxzBv3jw88cQTiIyMhJ2dHS5cuIA///wTzs7OWLp0Kdq2bftAvCdPnsSyZctQUlICa2trhIeH45lnnsFrr72GsrIyBAUF4bPPPkPXrl2rrBcbG4uFCxeiW7duAO7dZnrKlCno1KkTtFotkpOTER8fj5KSEtjZ2WHLli3Yvn07vv76a+j1erRv3x7z5s1D9+7dkZOTg8jISOTm5uL//u//8Ndff0nv4+zsjLS0NMyZMwelpaUIDAxEQkKCwYJ55MiR+Pbbb6Xi9MaNGyguLsY///nPavs7ePBgJCQkYN++fbCwsEBWVhZatWqFpUuXorCwEN988w3Ky8vRrl07dOvWDXl5eZgyZQqys7NhaWmJTz75BN27d2/gnkNE9VHdMV3bHFQ5Nx86dAgxMTGwsLBAr169kJqaiv/97384fvy4wRxx/7FvYWGBgQMH4tixY+jVqxfS09Ph7OwMX19fHDhwAH5+fsjJycFff/2Fp59+GgCwZs0afPfdd7C0tMQTTzyBefPmwcHBASEhIXj44Ydx+fJljB07VnqPsrIyzJgxA1ZWVli6dClKSkqwaNEiXLx4ETqdDh4eHpg1axasrKzg6uqKF198EefPn0dMTAx69+7dqJ8PEdVO7969MWXKFPz888/Izc2V7omRl5eH2bNn4++//wZw73eMFePDyuOixMREbNu2DTqdDgUFBZg8eTLGjRtX5T1CQkLg4uKCU6dOIT8/Hy+//DJu3bqF48ePo6SkBJ9++imcnZ1x9+5doznFWJw1jdOeeuopTJ48GYcPH0ZxcTHeffddeHt7A0C1ufn27du4du0ahg4dipkzZ1Zps6ioCB999BF++eUXWFpaYtiwYYiIiIBOp0NMTAxOnDiB8vJyPPXUU4iOjoadnR28vLwwevRo6dEugYGBBsfbnTp1MtdH3TIIksW1a9eEm5vbA/MvXrwo+vbtK2bPni3+85//SPMrT3t6eorVq1cLIYTQ6/ViwoQJ4sqVK0IIIf7880/Rq1cv8ddffwkhhHBycpL+9vT0FGfOnBGpqali2LBh0vwdO3YIPz8/odfrxezZs8Urr7wiNBqN0Gq1YtSoUSI+Pv6BOPPz84WHh4c4deqUFPezzz4rrl69arRvFes5OTmJ4uJio9tmx44don///uLu3btCCCGOHTsmxo0bJ61z+PBh4evrK4QQ4o033hCxsbFCCCEyMzOFm5ub2LFjR5W+VxdPxXZJT08XHh4eIicnRwghxJo1a8SWLVvEhAkTxJ49e6rt744dO4S7u7vIzs4WQgixYMECMWvWLCGEEKtWrRIffvih1K9+/fqJzMxMIYQQCxcuFHPmzDEaFxE1nJOTkwgICBAjR46U/rt161aNx3Rtc1BFbs7PzxfPPvusOHfunBBCiISEBOHk5CSuXbtWbY64X2Jiopg2bZoQQojFixeLL7/8UuTk5Ihnn31WlJWVicTERBEeHi6EECI+Pl688soroqioSAhxL9+EhYUJIYSYMGFClfwyYcIE8e2334o33nhDfPjhh0Kv1wshhIiMjBT//e9/hRBClJWViffee09s2LBB2naJiYkN/gyIyDwqxnVOTk5iy5YtQgghzp49K1xdXUVpaalYvXq1mDdvnhBCiKKiIhEeHi7u3LlTZVxUWFgoXn75ZZGfny+EEOLXX3+VXtuxY4eYMmWKEOJeDnnrrbeEEEKcOnVKODk5if379wshhFi0aJGIjo4WQtScUwzFWdM4zcnJSaxbt04IIcS5c+eEu7u7+Ouvv2rMzRMnTjTa5scffywiIiJEWVmZ0Gg0Yvz48eLo0aPis88+E0uWLJFy5CeffCLmz58vbe8lS5YIIe6Nt3v37i2uXr0qxVgxrqaG4ZlThVGpVGjVqlWNy/Xr109afv369Th48CCSk5Pxxx9/QAiBkpISo+sePnwYI0aMkC5zCwoKwqJFi3D9+nUAwKBBg6RLa52cnAxelnrmzBl07doVffv2BQD06NEDzzzzDI4fP47nnnvO6HtXPBi+8m+qDHF2doadnR2Ae7/bzMrKQnBwsPT6nTt3cPv2baSmpmL27NkAgG7dulX73tWxtraGj48PkpOTERYWhj179mDLli3Ss/iq669KpYKLi4t0Z9ynnnoK+/btM/g+ffr0kc4Y9+rVy+hyRGQ6hi7rPXToULXHdG1zUIWTJ0+ie/fu6NmzJwBg9OjR+Oijj6TXa5sjBg8ejMWLF0Ov1+PHH3/Ef/7zHzg6OqJz587IyMjA0aNHMWTIEADATz/9hKCgIOlxUaGhoVi/fj20Wi2A///vRIWlS5eiqKgI+/btkx6MfvDgQZw9e1Z6Xl5paWmVde5vg4iU6cUXXwRwL9dotVoUFxdj0KBB0tVazz//PGbMmIF27dpVGde1bdsW69evx6FDh5CZmYnz58+juLjY4HtUPKqkS5cuAO6NFwGga9euOH78OICac4qhOGtjwoQJAO7d7dvJyQknTpzA6dOnq83NlR+9cr/U1FTMmTMHlpaWsLS0lO6Rsnz5cty9exepqakAAJ1Oh0cfffSB+Dt06IBHH30UBQUF0vYg02BxqjBnz56Fk5MTVCoVRKUbKet0uirLVQxGiouLMXr0aAwbNgz9+vXDSy+9hB9++KHKuvczVBgKIVBWVgYAVYrjijj279+PVatWAbj3wPexY8dKgxtDbVS4f72NGzfiH//4B06fPo3nn3++yrLvvPMOpk2bVqV/FfEGBgZKl2To9Xrk5ubi4YcffmA71fQAYkPxVBg1ahTmz58PNzc3PPHEE1UeYl9eXm60v9bW1ga3mSGV46tuOSIyr5qO6drmoAqWlpYPHM8VX8YBhvNqTk4OpkyZIs3fsGEDOnTogMcffxx79+6FpaWlNOgZOnQo0tPTcfz4ccyaNUuKo3If9Hp9lRxcuQ/AvZ8vCCEQHR2N9evXS+usXLlSusT4zp07Vdq8vw0iUiZbW1sAkI5fIQT69OmD/fv3Iy0tDUePHsW//vUvbNy4scr45s8//8Qrr7yCl19+Ge7u7vD19cWPP/5o8D3uvyeIoecp15RTDMVZmbFxWuVLffV6PSwtLWvMzZXzV2BgoPT3Rx99BCsrqypxZWdno1WrVtDr9YiKipK+BCwqKoJGo3kg/oo+cBxnerxbr4JcuXIFa9euRVhYGB555BFkZGQAuHeDoYpvpO6XlZWFwsJChIeHw8vLC8eOHYNWq5UKUEtLywcKxkGDBmH37t3SXcV27NiB9u3bS2f0DHnxxReRlJSEpKQkbNy4EW5ubrh8+TLOnDkDALh06RJOnDiBZ599ttr1AOCtt97CokWLkJWVBeDeIHHt2rU4f/58ld94Vhg4cCC+++475ObmAgC+/vprTJw4UerLtm3bAAA3b97EsWPHHljfysoK5eXlEEIYjKdC3759UVpaitjYWIwePbrKa7Xt7/0MbX8ikl9djunqclCFZ555RjrrAADff//9A4Oy+3Xo0EHKR0lJSejQoQOAe2dP165di6FDh0rLDh06FElJSXBwcJDOAg8aNAg7duyQzjxs2bIF/fv3N3pTuT59+iA8PBxXr15FXFyc1LcvvvgCQghotVpMmzbN6F3WiahpiYmJwdq1azFs2DDMnTsXTz75JC5dulRlXJSRkQF7e3u88cYbGDhwoFSYlpeX1+s965NTajNOq7gr+m+//YYrV66gf//+tcrNFSrn2t69e8PDwwOJiYnQ6/XQarWYPn06Tpw4gYEDB2Lr1q3SWHrevHm1elYnx3umwzOnMqr48Tdw7xt2W1tbvPvuuxg6dCicnZ3x3nvvwcfHB48//jgGDBhgsA1nZ2cMHToUfn5+sLGxgZOTE5588klkZWWha9eu8PX1RUhICD777DNpnRdeeAGTJk3CxIkTodfrYW9vj88//7zKt/w1sbe3x8qVK7Fw4UKUlpZCpVJh8eLFeOKJJ6TLg41Rq9UQQuDdd99FWVkZNBoNXFxc8OWXXxocVA0cOBCTJ09GWFgYVCoV7OzssHr1aqhUKsyfPx9z5syBn58fOnbsKF1SV5mDgwP69OkDf39/bN26FY888ojR2AIDA7F161bpUpXa9PfXX3812t6AAQPw3nvvYeHChXBxcal2uxBR46nLMV1dDqrQvn17rFixArNnz4aFhQVcXV1hZWWF1q1b1zm2iuJ03rx50rzevXvj1q1bVW5SMmbMGGRnZ+Nf//oX9Ho9unXrhpiYmGrbtrW1xZIlSxAWFoYBAwZg7ty5WLRoEdRqNXQ6HZ5//nm89tprdY6ZiJRn4sSJiIyMREBAAGxsbODs7Ax/f39YWlpK46LNmzejQ4cO8PX1hUqlwrPPPgt7e3vpBEJd1Sen1Gac9ssvvyAuLg56vR6xsbF4+OGHa5Wbjak4URIYGIjy8nKMGDEC3t7eGDx4MJYuXYrRo0ejvLwcvXr1QmRkZI3tVR5vOzk51bg8GacSPB9NRETUIIWFhVi7di3efvtttG7dGr/99humTp2Kw4cP12qgREREhlU8fcFcjwQjZeGZUyIiogays7ODtbU1xowZAysrK1hZWeHTTz9lYUpERFQHPHNKREREREREsuMNkYiIiIiIiEh2LE6JiIiIiIhIdor5zaler0dRURGsra35Gx2iZkYIAZ1Oh7Zt29bprtAtDfMgUfPFPFh7zIVEzVdNuVAxxWlRUREuXrwodxhEZEZOTk5o166d3GEoFvMgUfPHPFgz5kKi5s9YLlRMcWptbQ3gXqDGHiDe2DIyMuDq6ip3GEYxvoZRenyA8mOsbXxarRYXL16UjnMyrD55UOn7iCGMuXEw5sbBPGh6dc2FSt1vlBiXEmMClBmXEmMCmn5cNeVCxRSnFZdt2NjYwNbWVuZo/j8lxWII42sYpccHKD/GusTHy7OqV988qPR9xBDG3DgYc+NgHjSt+uRCpe43SoxLiTEByoxLiTEBzSMuY7mQP3ogIiIiIiIi2bE4JSIiIiIiItmxODVAqysHALi7u5ulXSIiU+nZy8Us7TJfERERNS/m/LfdVOMRxfzmVElsrC2hnpFk8nZ3fRJo8jaJqGVr26YV8xURERHVyFw1DmC6cQPPnBIREREREZHsWJwSERERERGR7FicEhERERERkexYnBIREREREZHsWJwSERERERGR7FicEhERERERkexYnBIREREREZHsWJwSERERERGR7FicNhNaXXmTbJuIiIiIiAgArOQOgEzDxtoS6hlJZml71yeBZmmXiIiIiIioAs+cEhERERERkexYnBIREREREZHsWJwSEdXB0qVLERkZCQA4d+4cgoKC4OPjg7lz56KsrAwAcPPmTYwfPx6+vr6YNm0aioqK5AyZiIiIqElgcUpEVEtpaWlITEyUpmfOnIn3338f33//PYQQiIuLAwB8+OGHGDduHFJSUuDq6oq1a9fKFTIRERFRk8HilIioFm7fvo3Y2Fi8/vrryczW5gAAIABJREFUAIAbN26gtLQUbm5uAICgoCCkpKRAp9PhxIkT8PHxqTKfiKipW716Nfz9/eHv749ly5YBAFJTU6FWq+Ht7Y3Y2FhpWWNXlhARVYfFKRFRLbz//vuIiIjAQw89BADIzc2Fg4OD9LqDgwNycnLw999/w87ODlZWVlXmExE1ZampqThy5AgSExOxc+dO/Pbbb0hOTkZUVBTWrl2L3bt3IyMjA4cOHQJg/MoSIqLq8FEyREQ12L59Ozp16gQPDw8kJCQAAPR6PVQqlbSMEAIqlUr6f2X3T9ckIyOj1su6u7vXqe26SE9Pb5JtmwtjbhyMWZkcHBwQGRkJGxsbAED37t2RmZmJbt26oUuXLgAAtVqNlJQUPPnkkw9cWbJq1SqMGzdOtviJqGlgcUpEVIPdu3cjLy8PgYGBKCgoQHFxMVQqFfLy8qRlbt26BUdHR9jb2+Pu3bsoLy+HpaUl8vLy4OjoWKf3c3V1ha2tram7UWfmKnzT09PNWlSbA2NuHM05Zo1GU6cvnpSmR48e0t+ZmZnYs2cPJkyYUOUKEkdHR+Tk5Bi9soSIqCYsTomIarB582bp74SEBBw/fhyLFy9GQECANDBNSkrC4MGDYW1tjX79+mH37t1Qq9XYuXMnBg8eLGP0RESmc+nSJUydOhWzZs2CpaUlMjMzpdcqrhwxdmVJXdWlmFfq2WslxqXEmABlxqXEmID6x2XuL/9Msb1YnBIR1VNMTAyio6NRWFgIFxcXhIaGAgDmz5+PyMhIrFu3Dp06dcKKFStkjpSIqOHS09Mxffp0REVFwd/fH8ePH69yBUnFlSIdO3Y0eGVJXdX2KhKlnnFXYlxKjAlQZlxKjAlQblxA7Yrfmq4iYXFKRFQHQUFBCAoKAgD07NkT8fHxDyzTuXNnbNmypbFDIyIym+zsbLz55puIjY2Fh4cHAKBv3764cuUKsv4fe/cfUGV5/3/8dQTEtJZjgzRzfpZmNlzSsJQ0GLoABcTQFeqkciu10unKNGU6TdPIRTqzXPNbm9kMLUWZoZtOV+IsqWSUlivR/AUo+QOUw6/r+4cfz0fkxwE8h/uAz8c/eu77Pvf9Ohf3uc553/d97uvgQd10001KT0/XsGHD1KlTJ/n6+la7sgQAnKE4BQAAQJ2WL18uu92uBQsWOKYlJCRowYIFmjBhgux2u8LCwhQVFSWp9itLAKAu9SpOi4qKlJCQoFdffVU33XSTMjMzNX/+fNntdg0aNEiTJ0+WdGFMqxkzZqi4uFi9e/fW7NmzHcMpAAAAoHlKSkpSUlJSjfPWr19fbVptV5YAQF2cjnO6Z88ejRgxwvGD95KSEsa0AgAAAAC4lNPiNDU1VbNmzXL8kD07O9sxppW3t7djTKsjR45UG9MqIyPDvekBAAAAAC2C02tu582bV+Xx5WNXMaYVAAAAAOBKNfgHobWNXWXFmFbu4s7bM7t6vKSL6/PUcYs8dXyoizw9n+T5GT09HwAAAJqHBhenl49dZdWYVs2VK4vIphznqDHb8eRxmCTPzyd5fsb65nM2phUAAADg9Denl7t0TKuKigqlp6crNDS0yphWkhjTCgAAAABQbw0+c+rr68uYVgAAAAAAl6p3cbp161bH/0NCQhjTCgAAAADgMg2+rBcAAAAAAFejOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlmm1xWlpWYXUEAAAAAICLeFsdoLFa+3gp9sk0t6x7w+/j3LJeAAAAAEDNmu2ZUwAAAABAy0FxCgAAAACwHMUpAAAAAMByFKcAAAAAAMtRnAIAAAAALEdxCgAAAACwHMUpAAAAAMByFKcAUA9LlixRdHS0oqOjlZycLEnKzMxUbGysIiIilJKS4lh27969io+PV2RkpGbMmKHy8nKrYgMAADQbFKcA4ERmZqY++OADrV27VuvWrdNnn32m9PR0TZ8+XUuXLtXGjRuVk5Oj7du3S5KmTJmimTNnatOmTTLGKDU11eJXAACuUVRUpJiYGB0+fFgSB+k8UWlZRa3zgoOD3bJewFW8rQ4AAJ7O399f06ZNU+vWrSVJXbt2VW5urrp06aLOnTtLkmJjY5WRkaFu3bqppKREQUFBkqT4+HgtXrxYI0eOtCw/ALjCnj17lJSUpNzcXElSSUmJpk+frhUrVqhjx44aO3astm/frrCwME2ZMkVz585VUFCQpk+frtTUVPrBJtLax0uxT6a5fL0bfh/n8nUCl+PMKQA4ccsttziKzdzcXL333nuy2Wzy9/d3LBMQEKC8vDzl5+dXme7v76+8vLwmzwwArpaamqpZs2YpICBAkpSdne04SOft7e04SHfkyJFqB+kyMjKsjA6gmeDMKQDU0/79+zV27Fg9/fTT8vLycpw9kCRjjGw2myorK2Wz2apNb4icnJx6L3sll2g5k5WV1SzX7S5kbhpk9lzz5s2r8vjyg3EcpANwpa6oOB09erQKCwvl7X1hNXPmzNGhQ4f0yiuvqLy8XA8++KBGjRrlkqAAYKWsrCxNnDhR06dPV3R0tD788EMVFBQ45hcUFCggIEAdOnSoMv3EiROOswz11bNnT/n6+rose2O5q/DNyspya1HtDmRuGi05s91ub9CBp+agtoNxrjhIJzXsQJ2nHiCwIhcHLV3HEzNJjc/l7v7VFe3V6OLUGKPc3Fz985//dBSneXl5mjx5st599121bt1aCQkJ6tOnj7p163bFQQHAKseOHdPjjz+ulJQUhYSESJJ69eqlAwcO6ODBg7rpppuUnp6uYcOGqVOnTvL19XV8YU1LS1NoaKjFrwAAXO/yg3GuPEgn1f9Anace1PDUXFfiajpo6YmZJM/NJdVv/3B2oK7RxenXX38tSRozZoxOnTql+++/X+3atVPfvn3Vvn17SVJkZKQyMjL0xBNPNHYzAGC55cuXy263a8GCBY5pCQkJWrBggSZMmCC73a6wsDBFRUVJkhYuXKikpCQVFRUpMDBQiYmJVkUHALfhIB0AV2t0cXrmzBmFhITot7/9rcrKypSYmKhBgwZV++1Bdna2S4ICgFWSkpKUlJRU47z169dXm9ajRw+tWbPG3bEAwFK+vr4cpAPgUo0uTu+44w7dcccdjsfDhw/X/PnzNX78eMc0d94IxFNPZ9eltKxCrX28XLa+pmyDxl5D7qnX6l/k6fkkz8/o6fkAAK61detWx/9DQkI4SAfAZRpdnO7evVtlZWWO318ZY9SpU6caf3vQEJ5yIxB3cNe4U5L7x55qTCHsydfES56fT/L8jFfzjUAAAADgWo0e5/Ts2bNKTk6W3W5XUVGR1q5dqxdeeEE7d+5UYWGhzp8/r82bN/MbAwAAAACAU40+cxoeHq49e/Zo6NChqqys1MiRIxUcHKzJkycrMTFRZWVlGj58uG6//XZX5gUAAAAAtEBXNM7ppEmTNGnSpCrTYmNjFRsbe0WhcHVw9W9wm2rdAAAAAFzviopT4Eo059/gAgAAwLNdyckKZ/fU4ESIe1CcAgAAAGhxOBHS/DT6hkgAAAAAALgKxSkAAAAAwHIUpwAAAAAAy1GcAgAAAAAsR3EKAACAZqvHbYFuW3dpWYXb1g2gOu7WCwAAgGarXds23JEVaCE4cwoAAAAAsBzFKdAA7ry8h0uHAAAAcDXjsl6gARjMGQAAAHAPzpwCAAAAACxHcQoAAAAAsBzFKQAAAADAchSncKqxN+oJDg52cRIAAAAALRU3RIJT7roJEDcAAgAAAHARZ04BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTtEilZRVOlwkODm6CJAAAAADqw9vqAIA7tPbxUuyTaS5f74bfx7l8ne5WWlah1j5ebll3j9sC3bJeAAAAXH0oToEWzl2FutQ8i3UAAAB4Jrdc1rthwwYNHjxYERERWrlypTs2AQAejX4Q8Cz1+blHY3AFSd3oCwE0hMvPnObl5SklJUXvvvuuWrdurYSEBPXp00fdunVz9aYAwCPRDwKeh597ND36QgAN5fIzp5mZmerbt6/at2+vtm3bKjIyUhkZGa7eDAB4LPpBAKAvBNBwLj9zmp+fL39/f8fjgIAAZWdnO32eMUaSVFpaWu9ttW/nnpu82O12t6zbXettrusmc/V1N2R6QzR15stdfF9ffJ+3dE3ZD0ru+fu6Yr+zcv3uQOam4c7MVr5XrrZ+UGo53wmtfH5j8bnwfzx137Bq21a3h7O+0GZc3Eu+8sorstvtmjRpkiQpNTVVOTk5mjNnTp3PO3v2rL788ktXRgHgYbp3767rrrvO6hhuRz8IoDZXSz8o0RcCqF1tfaHLz5x26NBBu3fvdjwuKChQQECA0+e1a9dO3bt3l4+Pj2w2m6tjAbCQMUZlZWVq166d1VGaBP0ggMtdbf2gRF8IoDpnfaHLi9O7775bf/jDH1RYWKhrrrlGmzdv1rPPPuv0ea1atbpqjiQCV6M2bdpYHaHJ0A8CqMnV1A9K9IUAalZXX+jy4vSGG27Q5MmTlZiYqLKyMg0fPly33367qzcDAB6LfhAA6AsBNJzLf3MKAAAAAEBDuXwoGQAAAAAAGoriFAAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7i9BJFRUWKiYnR4cOHJUmZmZmKjY1VRESEUlJSLM22ZMkSRUdHKzo6WsnJyR6Xb9GiRRo8eLCio6P1+uuve1y+i55//nlNmzZNkrR3717Fx8crMjJSM2bMUHl5uaXZRo8erejoaMXFxSkuLk579uzRhg0bNHjwYEVERGjlypWW5tu6davi4+M1aNAgzZ07V5Jn/o1bqsv7p0t52r58UV2Z//GPfyguLk5DhgzRY489ptOnT1uQsLq6Ml+0bds2DRgwoAlTOVdX7q+//lqjR4/WkCFD9Mtf/rJZtPVnn32mYcOGaciQIRo7dqzOnDljQcKqavocvpSnvg9bGk/sC+vKtGTJEoWHhzs+25vqs9xT91dnuaxor5q+w17KqrZylsuqfUuq+n36UkePHtWoUaMUFRWl8ePHq7i4uOErNzDGGPPpp5+amJgYExgYaL755htz/vx5ExYWZg4dOmTKysrMmDFjzLZt2yzJtmPHDvPAAw8Yu91uSktLTWJiotmwYYPH5Nu1a5dJSEgwZWVl5vz58yY8PNzs3bvXY/JdlJmZafr06WOmTp1qjDEmOjrafPLJJ8YYY5555hmzcuVKy7JVVlaa/v37m7KyMse048ePm/DwcPPtt9+a4uJiExsba/bv329JvkOHDpn+/fubY8eOmdLSUjNixAizbds2j/sbt1SX90+X86R9+aK6Mp89e9b069fPHD9+3BhjzEsvvWSeffZZK2JW4aydjTGmoKDAREVFmfDw8CZOV7u6cldWVpqIiAizfft2Y4wxL7zwgklOTrYiZhXO2vpiH2OMMfPnzzcvvvhiU0esoqbP4c2bN1dZxhPfhy2NJ/aFzjKNHTvWfPzxx27PcSlP3V/rk6up26um77BfffVVlWWsaKv65LJi3zKm+vfpSz366KMmPT3dGGPMkiVLGvV5w5nT/5WamqpZs2YpICBAkpSdna0uXbqoc+fO8vb2VmxsrDIyMizJ5u/vr2nTpql169by8fFR165dlZub6zH57rrrLv3lL3+Rt7e3Tp48qYqKCp05c8Zj8knSqVOnlJKSonHjxkmSjhw5opKSEgUFBUmS4uPjLc339ddfS5LGjBmjIUOG6M0331RmZqb69u2r9u3bq23btoqMjLQs49///ncNHjxYHTp0kI+Pj1JSUnTNNdd41N+4Jbu8f7qUp+3LF9WVuaysTLNmzdINN9wgSbr11lt17Nixpo5YTV2ZL0pKStITTzzRhKmcqyv3Z599prZt2yo0NFSSNG7cOI0aNaqpI1bjrK0rKysdR9zPnz+vNm3aNGW8amr6HD569Khjvqe+D1saT+wLne3LOTk5WrZsmWJjYzVnzhzZ7Xa3Z/LU/dVZLqnp26um77Bt27Z1zLeqrZzlkqzZty7/Pn2psrIyffTRR4qMjJTU+LaiOP1f8+bNU+/evR2P8/Pz5e/v73gcEBCgvLw8K6LplltucbwpcnNz9d5778lms3lMPkny8fHR4sWLFR0drZCQEI9qP0maOXOmJk+erO985zuSqv99/f39Lc135swZhYSE6OWXX9Ybb7yhVatW6ejRox7ThgcPHlRFRYXGjRunuLg4vfXWWx73N27JLu+fLuVp+/JFdWX+7ne/q3vvvVeSVFJSoj/+8Y/62c9+1pTxalRXZkn6y1/+oh/96Efq1atXE6Zyrq7chw4d0ve//31Nnz5d9913n2bNmlXtC44VnLX1tGnTlJSUpP79+yszM1MJCQlNmK66mj6Hw8LCHPM99X3Y0nhiX1hXpuLiYt12222aMmWK1q5dqzNnzmjp0qVuz+Sp+6uzXFa11+XfYS8eOJWsfW/Xlcuqtrr8+/Slvv32W1177bXy9vaW1Pi2ojitRWVlpWw2m+OxMabKYyvs379fY8aM0dNPP63OnTt7XL6JEydq586dOnbsmHJzcz0m3+rVq9WxY0eFhIQ4pnna3/eOO+5QcnKyrrvuOvn5+Wn48OFavHixx2SsqKjQzp079dxzz+ntt99Wdna2vvnmG4/JdzXztH25Ic6ePatHH31UPXr00H333Wd1nDp9+eWX2rx5sx577DGrozRIeXm5PvzwQ40YMUJr165V586dtWDBAqtj1amkpEQzZszQG2+8oQ8++EAjR47U1KlTrY4lqern8P/8z/84pjfn92FL4Yl/g3bt2um1115T165d5e3trTFjxmj79u1Ntn1P3V9ry2Vle136HTY1NdUx3eq2qi2XFW1V0/fpS9XUNo1pK4rTWnTo0EEFBQWOxwUFBXVe6uVuWVlZeuihh/Tkk0/qvvvu86h8X331lfbu3StJuuaaaxQREaFdu3Z5TL6NGzdqx44diouL0+LFi7V161atWbOmSr4TJ05Y+vfdvXu3du7c6XhsjFGnTp08pg2///3vKyQkRH5+fmrTpo1+9rOfKTMz02PyXc0u7wus3pfrKz8/XyNHjtStt96qefPmWR3HqYyMDBUUFGjYsGF69NFHHfk9nb+/v7p06aIf//jHkqSYmBhlZ2dbnKpuX375pXx9fXX77bdLkh544AF9+OGHFqeq/jl8qeb6PmxJPPFvcPToUa1Zs8bx2BjjOKvkbp66v9aVy4r2quk77BdffOGYb1VbOctlRVvV9H36ueeec8z38/PT2bNnVVFRIanx3wspTmvRq1cvHThwwHE5Y3p6uuM3O03t2LFjevzxx7Vw4UJFR0d7XL7Dhw8rKSlJpaWlKi0t1ZYtW5SQkOAx+V5//XWlp6crLS1NEydO1IABAzR//nz5+voqKytLkpSWlmZZPunCGaTk5GTZ7XYVFRVp7dq1euGFF7Rz504VFhbq/Pnz2rx5s2UZw8PD9cEHH+jMmTOqqKjQ+++/r6ioKI/5G1/NOnXq5FH7cn1cvER80KBBmjFjhuVnN+pj4sSJ2rRpk9LS0vTHP/5RAQEBeuutt6yO5dQdd9yhwsJC7du3T9KFu24HBgZanKpuXbp00fHjxx2/xd+yZYujuLZKTZ/Dl2qO78OWxhP/Bm3atNELL7ygb775RsYYrVy50vGTBnfy1P3VWS4r2qum77DBwcGO+Va1lbNcVrRVTd+np0+f7pjv4+Oj3r17a+PGjZKkdevWNaqtmubwTTPk6+urBQsWaMKECbLb7QoLC1NUVJQlWZYvXy673V7lUqyEhASPyRcWFqbs7GwNHTpUXl5eioiIUHR0tPz8/DwiX20WLlyopKQkFRUVKTAwUImJiZZlCQ8P1549ezR06FBVVlZq5MiRCg4O1uTJk5WYmKiysjINHz7ccSahqfXq1Uu/+tWvNHLkSJWVlalfv34aMWKEbr75Zo/+G7dkjzzyiCZOnKgf//jHHrUv1+Vi5uPHj+vzzz9XRUWFNm3aJEnq2bOnR55BvbSdm5NLc7/88stKSkrS+fPn1aFDhxqHb/AEl2aeP3++Jk2aJGOMvve971U5Om+F2j6Ht27d2uzehy2NJ/aFl2aaM2eOxo8fr7KyMv3kJz/Rww8/7Pbte+r+Wp9cTd1etX2HtXq/qk8uK/atmsyYMUMDBgzQwIEDNWvWLE2bNk2vvPKKOnbsqBdffLHB67MZY4wbcgIAAAAAUG9c1gsAAAAAsBzFqQscPnxYt956q37xi19Umzdt2jTdeuutKiwsbPB6s7OzNXPmTFdEdGRZvny5W57f2NfoKlu2bNHcuXMlSaNHj3Y6rtIf/vAHzZkzp8Z5AwYM0H/+8x+XZ3SFpKQk5eTkWB0D8CgX++DVq1dXmb58+XJNmzbNJduoqKjQ+PHjFRkZqTfffLPOZQsLC3Xrrbe6ZLtXYuzYsXr33Xcb/XxXfwYBaD7++te/asiQIRo8eLCio6M1ZcqUamOSWmX16tVauXKlpAs5//jHP1qcCK7Eb05dxNfXVwcOHNCRI0fUqVMnSdK5c+f08ccfN3qd//3vfxknrZ4GDhyogQMHWh3D7TIzM/XAAw9YHQPwOK1atdLzzz+v4OBg3XzzzS5ff15enj744AN9+umn8vLycvn6PRGfQcDV6fnnn9e+ffu0bNkydezYUZWVlVq/fr0eeOABrV69Wh06dLA0X1ZWlm655RZJ0ogRIyzNAtejOHURLy8vDRo0SBs2bNC4ceMkSZs3b9bAgQP1//7f/5Mkvf3221qxYoVatWql73//+/rtb3+rH/7wh9q9e7cWLFigyspKSReOdt9+++1avHixzp49q2eeeUbz58+vsr0nn3xSgYGBGjNmjCTprbfe0ocffqgXX3xRzz33nPbs2aPi4mIZYzR37twqd/iSLpzp3Llzp/z8/Ko8bt++fZ3Pz8rK0qZNm1RUVKR+/fpp6tSp1W5dvXr1av31r39VZWWl2rdvr9/+9rfq2rVrtTY7cOCA5syZo+LiYhUUFKhHjx566aWXtG7dOv3zn//Uq6++KunC7bQfeughbdu2TWvXrtXbb7+tsrIynT59Wo888ohGjhypd999V5s2bdKyZcuqbOPVV1/Vli1bVFJSovPnz2vq1KmOu5l99dVXGjVqlE6fPq3bbrtNs2bN0rXXXlvl+Vu3btUrr7yisrIytWnTRlOnTtUdd9xR7bWcOHFCM2fO1MmTJ1VQUKBOnTrppZde0ve+9z299dZbWrVqlXx8fOTr66s5c+aoW7dutU7Py8vTnDlzdOzYMZWVlSk6Olrjxo1TSkqK8vPz9dRTTyk5OVm9evWqcV8ErkZt2rTRww8/rKeeekqrVq1S69atHfPOnj2r2bNna9++fbLZbLrnnnv0m9/8psbb7u/evVvJyck6f/68fHx8NGnSJP3kJz/Rr371K5WXlys+Pl5/+MMf9IMf/KDK8zZv3qyUlBRdc8016tmzZ5V5tfWJ06ZNk6+vr/bt26eTJ0+qX79+SkpKko+Pj7766ivNmzdPp06dUkVFhUaPHq3hw4dr165dSklJUefOnbV//36Vl5dr9uzZCg4OVl5enqZNm6b8/HzdeOONOnnypCNDQ9d34403VvkMSkpK0jPPPKODBw+qVatWCgwM1Jw5c9SqFRdgAS3J8ePHtWrVKm3btk3XX3+9pAsH/4YOHaqcnBwtW7ZMiYmJmjlzpgoLC9WqVSuNHz9egwcP1oEDB2qcPmDAAC1atMhxY7mLj7/73e9q9OjRuueee7Rnzx4ZYzRz5kz17t271u9VH3/8sbZu3aodO3aoTZs2Kiws1LfffquZM2dq//79mjNnjk6dOiWbzaYxY8Zo6NChdfab8EAGV+ybb74xQUFB5j//+Y+JiopyTH/wwQfNF198Ybp3727+9re/mZ/97Gfm5MmTxhhj3nnnHTNo0CBTWVlpEhMTTXp6ujHGmL1795rf/e53jmUeffTRGre5c+dOExMT43g8fPhws2PHDvPxxx+bCRMmmIqKCmOMMcuWLTNjx441xhgzdepU86c//ckYY0z37t0dWS597Oz59913nykuLjZ2u9384he/MCtXrqzy/F27dpmRI0eac+fOGWOMef/996u0yaUWLFhg1q1bZ4wxprS01MTExJiMjAxz9uxZ07t3b5Ofn2+MMSY5Odm8+OKLpqioyNx///2msLDQGGPMJ598YoKCgqq11S9+8Qvz3nvvmcOHD5vRo0eb8+fPG2OMSU9Pd7TZ4sWLzU9/+lNz8uRJU1lZaZ588kmTnJxsjDEmPDzcZGdnmwMHDpiYmBjH9r788kvTr18/U1xcXO21vPHGG2bZsmXGGGMqKyvNr371K7N8+XJTXl5uAgMDTV5enjHGmLVr15pVq1bVOt0YY0aPHm22bNlijDGmpKTEjB492vztb3+rkg3A/7nYB1dUVJhRo0aZBQsWGGOM+dOf/mSmTp1qnn76afPss8+ayspKY7fbzZgxYxzv10sVFhaakJAQ8+mnnxpjLrzn77rrLnPo0CHHNmpSUFBggoODzf79+40xxrz66qume/fuxhhTZ584depUM3ToUFNUVGTsdrsZNWqUWbFihSkrKzODBw82OTk5xhhjzpw5YwYNGmQ++eQT8+9//9vcdttt5vPPPzfGGLN8+XIzatQoY4wxjz32mElJSTHGGJObm2uCgoLMO++80+j1Xdqvrl271owZM8YYY0x5ebmZMWOGyc3NbcRfC4Any8jIMPHx8TWe/NToAAAgAElEQVTO27Jli4mNjTVDhw41b775pjHGmKNHj5qBAweas2fP1jr98u8uFx9/8803pnv37mb9+vXGGGO2bdtm+vXrZ0pLS2v9XmVM1e+zixcvNrNnzzZlZWVm4MCBZtOmTcYYY44fP27uuece8/HHH9fZz8HzcObUhXr27CkvLy/l5OToe9/7noqLi9W9e3dJ0vvvv6/Bgwc7zlTGx8dr3rx5Onz4sAYNGqQ5c+Zo69atuvvuu/Wb3/zG6bb69Okju92u//znP7rmmmtUWFiokJAQ2Ww2XX/99Vq1apW++eYb7dq1S+3atav3a7jjjjvqfH5cXJzatm0rSRoyZIi2b99eZSD6bdu26eDBg0pISHBMO3PmjE6dOqX27dtX2daUKVO0Y8cOvfbaa8rNzVV+fr7OnTuna6+9Vvfee6/Wr1+vhx56SBs2bNDKlSvVrl07vfrqq9q+fbtyc3O1b98+nTt3rtbX0qlTJyUnJ2vDhg06ePCg42zwRffee6/j7zFs2LBqwyvs2LFD+fn5euihhxzTbDabDh06pB49elRZ9sEHH9Tu3bv1+uuvKzc3V/v371evXr3k5eWlqKgoJSQk6Kc//an69++vsLCwWqefO3dOH330kU6fPq1FixZJunB5+L59+zR48OA6/3bA1a5Vq1Z64YUXNHToUPXv398x/V//+pf++te/ymazqXXr1kpISNCf//xnPfroo1Wen52drR/84AeOqxJuueUW/eQnP9GHH36oPn361LrdrKwsde/eXd26dZMkPfDAA47b59fVJ0rSfffd5+hj4+LitGXLFvXt21eHDh2qMn5cSUmJPv/8c3Xt2lU33nijbrvtNknSj370I61du1bShcv+p06dKunCWKEXM+fm5jZqfZcKDg5WSkqKRo8erbvvvlsPPvigunTpUmubAGi+ysvLa5xeWloqY4z27dunn//855Kkjh076h//+IdOnTpV43Rnrr/+esXGxkqS4/vRF198Uev3qtrk5ubKbrcrIiJCknTDDTcoIiJC77//vvr06VOvfg6egeLUxYYMGaL169fLz89PcXFxjuk1DTJvjFF5ebkSEhIUHh6uHTt26P3339eSJUuq3dBn0aJF2rp1q6QLl0P8+te/1vDhw5WWliYfHx8NHz5cNptN27Zt07x58/Twww9r4MCBuvnmm7V+/fo6M5eWljr+7+z5l/7WyhhT7bK4yspKxcXFacqUKY7H+fn5uv7666u0x9y5c7V8+XJVVFRo0KBB+ulPf6pjx47J/O/IRvfff7/j0reuXbuqc+fOOn78uB544AHdf//9Cg4OVlRUlP75z3/W+ro+++wzPfbYY3rooYfUr18/3XnnnZo9e3aNr6WysrLG1xISEqKXXnrJMe3YsWMKCAjQI488ovz8fEnSxIkT9fHHHys7O1vDhg1Tnz59VF5e7ngtCxcu1JdffqnMzEz98Y9/VFpamhYtWlTj9Hnz5skYo1WrVumaa66RdOHmKr6+vrW+TgD/p2PHjpo9e7amTp2qoUOHSrrwXr60D66srFR5ebm2bNmixYsXS5ICAgI0YsSIan31xX76Upc/b/jw4Y73u6QqfUldfaJUvU9t1aqVKioqdN111yktLc0x78SJE7ruuuv06aefqk2bNo7pNpvNse1L/39pjsau71KdO3fW3//+d+3atUv//ve/9fDDD2vOnDkaMGBAtWUBNF9BQUE6ePCgCgoK5O/vX2Xerl271LNnT/33v/+t0ld+/fXXjmUvn37jjTdKUpV+5dLvnZf/hr+yslJeXl564YUXav1eVZOKioo6++/69HPwDPxYxMXi4uKUkZGhjRs3KiYmxjH9nnvu0caNGx13tH3nnXfUvn17denSRQkJCdq7d6/i4+P17LPP6syZMyooKJCXl5fjTfXrX/9aaWlpSktL069//WtJF464b926VZs2bVJ8fLykC2f7wsPDNXLkSPXs2VP/+Mc/VFFRUS2nn5+f44606enpjunOnv+3v/1NpaWlstvtWrt2rUJDQ6ust3///vrb3/7mKNz++te/6sEHH5QkR/60tDT9+Mc/1gcffKDHH3/ccUZwz549jm0FBQVJkl5++WXHUbicnBz5+fnpscceU//+/R2FaU2vT5I++ugj9ezZUw8//LDuuusubdmypcqyW7du1enTp1VRUaHU1NRqryUkJEQ7duzQV199JUnavn27hgwZopKSEr322muO1zJw4EB98MEHevDBBzV06FB973vfU2ZmpioqKlRYWKiwsDC1b99eDz30kCZNmqT//Oc/tU6/9tprFRQUpNdff13ShTMsI0aM0JYtWySpyj4BoGZRUVEKDQ3Vn//8Z0kX+qU333xTxhiVlpYqNTVVd999twYOHOh4H7/22msKCgrS119/rezsbEnS/v379dFHH+muu+6qsv7Ln3fnnXfqv//9r/bt2ydJVe6QW1efKEnvvfdelT41PDxcP/zhD9WmTRtHMXns2DHFxMQ4vVP3Pffco7fffluSdPToUe3atUuSGr2+S/ubt956S88884z69++vKVOmqH///vr888/rfD6A5ueGG27Q6NGj9Zvf/KbKDdHeeecdbd68WU888YQCAwO1bt06SRf6kxEjRqikpKTG6WfPnpWfn5+jv9m1a5cKCgoc6y0sLNS//vUvSRe+l/n4+Kh79+61fq+Sav4udPPNN8vb21ubN2+WdOEmdps2bdLdd9/tppaCu3Dm1MVuuOEGde3aVdddd12Vy1j79Omjhx56SA8++KAqKyvl5+enZcuWqVWrVnrqqaf03HPP6aWXXpLNZtMTTzyhm266SRUVFXr55Zf1xBNPaMmSJdW25e/vrx/96EcqLy/XDTfcIElKSEjQk08+qdjYWJWXl6tfv37avHmz42ZLFyUlJWnOnDn6zne+o7vvvttxxMvZ82+66SaNHDlSxcXFuvfee3XfffdVWW///v31yCOPaMyYMbLZbLr22mu1ZMmSGs8cT548WY8//rjatm2ra6+9VnfeeacOHTrkmP/zn/9cS5cu1c9+9jNJUr9+/bRmzRpFRUXJZrPprrvukp+fnw4ePFjj3yImJkabN2/WoEGDVFlZqfDwcJ0+fVpFRUWSpK5du2rs2LE6c+aMgoODq13i161bN82ZM0e/+c1vHGeJX3nllRovk3788ceVnJysRYsWycfHRz/5yU906NAh+fn5afz48XrooYfUpk0beXl5ae7cubVOly6caX322WcVGxur0tJSxcTEaMiQIZIuXIo8ZcoU/e53v6ty2SKAqpKSkpSVleX4/9y5cxUbG6uysjLdc889jhvXXcrPz0+LFi3Ss88+q5KSEtlsNs2fP18//OEPdfjw4Vq35efnp4ULF+qpp56Sj4+P7rzzTsc8Z31imzZtNHLkSJ05c0aRkZEaNmyYWrVqpaVLl2revHn605/+pPLycv36179WcHCwo+CsyaxZs/TMM89o0KBB6tChg+PnB61bt27U+oKCghyfQcnJyfrwww81ePBgXXPNNerYsaNGjx5d9x8BQLP05JNPavXq1Ro/frxKS0tVWlqqH//4x1q1apU6deqk3//+95o9e7ZWrFghm82mefPmyd/fv9bpTz31lH73u9/p7bffVmBgoAIDAx3b8vX1VVpamhYuXKg2bdro5ZdflpeXV63fqyQpNDRUCxYsqJLZx8dHS5cu1dy5c/WHP/xBFRUVevzxx9W3b986+zl4HpvhvDYAAE1u2rRpuuWWW/TLX/7S6igA0OQOHz6s2NhYffLJJ1ZHgQfhsl4AAAAAgOU4cwoAAAAAsBxnTgEAAAAAlvOYGyJVVlaquLhYPj4+Nd48B0DzZYxRWVmZ2rVrp1atOCZWG/pBoOWiH6w/+kKg5XLWF3pMcVpcXKwvv/zS6hgA3Kh79+667rrrrI7hsegHgZaPftA5+kKg5autL/SY4tTHx0fShaCtW7d2unxOTo569uzp7lgN5om5PDGTRK6G8MRMUv1zlZaW6ssvv3S8z1GzhvaDkufuG3Uhc9Mgc9OgH3S9xvSF7tKS90lPQuam4QmZnfWFHlOcXrxso3Xr1vL19a3Xc+q7XFPzxFyemEkiV0N4YiapYbm4PKtujekHJc/dN+pC5qZB5qZBP+haje0L3cUTMjQUmZsGmRuvtr6QHz0AAAAAACxHcQoAAAAAsBzFKVqk0rKKRj83ODjYLesF3KHHbYFuWS/7OgCgqfBZhos85jengCu19vFS7JNpLl/vht/HuXydwJVo17YN+zoAoFnjswwXceYUAAAAAGA5ilMAAAAAgOUoTgGgAZ5//nlNmzZNkrR3717Fx8crMjJSM2bMUHl5uSTp6NGjGjVqlKKiojR+/HgVFxdbGRkAAKBZoDgFgHrauXOn1q5d63g8ZcoUzZw5U5s2bZIxRqmpqZKk2bNna+TIkcrIyFDPnj21dOlSqyIDAAA0GxSnAFAPp06dUkpKisaNGydJOnLkiEpKShQUFCRJio+PV0ZGhsrKyvTRRx8pMjKyynQAAADUjbv1AkA9zJw5U5MnT9axY8ckSfn5+fL393fM9/f3V15enr799ltde+218vb2rjK9IXJycuq9bF1DH12prKysZrludyFz0yAzAFy9KE4BwInVq1erY8eOCgkJ0bvvvitJqqyslM1mcyxjjJHNZnP8e6nLHzvTs2dP+fr6XnnwK+SuwjcrK8utRbU7kLlptOTMdru9QQeeAOBqRHEKAE5s3LhRBQUFiouL0+nTp3Xu3DnZbDYVFBQ4ljlx4oQCAgLk5+ens2fPqqKiQl5eXiooKFBAQICF6QEAAJoHfnMKAE68/vrrSk9PV1pamiZOnKgBAwZo/vz58vX1dVzOl5aWptDQUPn4+Kh3797auHGjJGndunUKDQ21Mj4AAECzUO/itKioSDExMTp8+LAk6ZlnnlFERITi4uIUFxenv//975KkzMxMxcbGKiIiQikpKe5JDQAeYOHChZo/f76ioqJ07tw5JSYmSpJmzZql1NRUDR48WLt379akSZMsTgoAAOD56nVZ7549e5SUlKTc3FzHtJycHL355ptVLlcrKSnR9OnTtWLFCnXs2FFjx47V9u3bFRYW5vLgAGCF+Ph4xcfHS5J69OihNWvWVFumU6dOWrFiRVNHAwAAaNbqdeY0NTVVs2bNchSi58+f19GjRzV9+nTFxsZq8eLFqqysVHZ2trp06aLOnTvL29tbsbGxDKEAAAAAAHCqXmdO582bV+XxiRMn1LdvX82aNUvXXXedxo4dqzVr1qht27ZVhlYICAho8BAKAAAAAICrT6Pu1tu5c2e9/PLLjsejR4/WunXrFBkZWePQCg3RkNuse+q4Yp6YyxMzSe7L1VzHfvTE7TrjqbkAAADQvDSqOP3iiy+Um5uryMhISReKUG9vb3Xo0KHK0AqNGUKhvuP7FZ8rUbu2bRoWvJ5KyyrU2serUc/1xDHaPDGT5Lm5nLEis6e2FeP7AQAAwFUaVZwaY/Tcc8+pb9++atu2rd5++23dd9996tWrlw4cOKCDBw/qpptuUnp6uoYNG+bqzJKkdm3bKPbJNLese8Pv49yyXgAAAABAzRpVnPbo0UOPPvqoRowYofLyckVERCgmJkaStGDBAk2YMEF2u11hYWGKiopyaWAAAABY5/nnn9e3336rBQsWaO/evZoxY4aKi4vVu3dvzZ49W97e3jp69KimTJmikydP6oc//KEWLlyodu3aWR0dgIer9zinkrR161bddNNNkqRRo0Zp48aN2rx5s5566inHMiEhIVq/fr02bdqk6dOnN/g3pwAAAPBMO3fu1Nq1ax2Pp0yZopkzZ2rTpk0yxig1NVWSNHv2bI0cOVIZGRnq2bOnli5dalVkAM1Ig4pTAAAAXJ1OnTqllJQUjRs3TpJ05MgRlZSUKCgoSNKFcaAzMjJUVlamjz76yHFvkovTAcAZilMAAAA4NXPmTE2ePFnf+c53JEn5+flVhhD09/dXXl6evv32W1177bXy9vauMh0AnGnUb04BAABw9Vi9erU6duyokJAQvfvuu5KkysrKGocQrGkowcb8zMtT7vLeHIdMa26Zm+sQgM2tnSXPz0xxCgAAgDpt3LhRBQUFiouL0+nTp3Xu3DnZbLYqQwieOHFCAQEB8vPz09mzZ1VRUSEvL69GDS0o1X94QXfy1KHc6tIcM7uTu9qiObazJ2R2Nrwgl/UCAACgTq+//rrS09OVlpamiRMnasCAAZo/f758fX0dZ2LS0tIUGhoqHx8f9e7dWxs3bpQkrVu3TqGhoVbGB9BMUJwCAACgURYuXKj58+crKipK586dU2JioiRp1qxZSk1N1eDBg7V7925NmjTJ4qQAmgMu6wUAAEC9xcfHKz4+XpLUo0cPrVmzptoynTp10ooVK5o6GoBmjjOnAAAAAADLUZwCQD0sWrRIgwcPVnR0tF5//XVJUmZmpmJjYxUREaGUlBTHsnv37lV8fLwiIyM1Y8YMlZeXWxUbAACg2aA4BQAnPvzwQ/373//W+vXr9c4772jFihXat2+fpk+frqVLl2rjxo3KycnR9u3bJUlTpkzRzJkztWnTJhljlJqaavErAAAA8HwUpwDgxF133aW//OUv8vb21smTJ1VRUaEzZ86oS5cu6ty5s7y9vRUbG6uMjAwdOXJEJSUlCgoKknTht1kZGRkWvwIAAADPR3EKAPXg4+OjxYsXKzo6WiEhIcrPz5e/v79jfkBAgPLy8qpN9/f3V15enhWRAQAAmhXu1gsA9TRx4kQ98sgjGjdunHJzc2Wz2RzzjDGy2WyqrKyscXpD1DU49eXcOZj2xbELm9u63YXMTYPMAHD1ojgFACe++uorlZaW6rbbbtM111yjiIgIZWRkyMvLy7FMQUGBAgIC1KFDBxUUFDimnzhxQgEBAQ3aXs+ePeXr6+uy/I3lrsI3KyvLrUW1O5C5abTkzHa7vUEHngDgasRlvQDgxOHDh5WUlKTS0lKVlpZqy5YtSkhI0IEDB3Tw4EFVVFQoPT1doaGh6tSpk3x9fR1nUtLS0hQaGmrxKwAAAPB8nDkFACfCwsKUnZ2toUOHysvLSxEREYqOjpafn58mTJggu92usLAwRUVFSZIWLlyopKQkFRUVKTAwUImJiRa/AgAAAM9HcQoA9TBhwgRNmDChyrSQkBCtX7++2rI9evTQmjVrmioaAABAi8BlvQAAAAAAy1GcAgAAAAAsR3EKAAAAALAcxSkAAAAAwHIUpwAAAAAAy9WrOC0qKlJMTIwOHz4sScrMzFRsbKwiIiKUkpLiWG7v3r2Kj49XZGSkZsyYofLycvekBgAAAAC0KE6L0z179mjEiBHKzc2VJJWUlGj69OlaunSpNm7cqJycHG3fvl2SNGXKFM2cOVObNm2SMUapqaluDQ8AAAAAaBmcFqepqamaNWuWAgICJEnZ2dnq0qWLOnfuLG9vb8XGxiojI0NHjhxRSUmJgoKCJEnx8fHKyMhwb3oAAAAAQIvg7WyBefPmVXmcn58vf39/x+OAgADl5eVVm+7v76+8vLwGB8rJyanXcsHBwQ1ed0NkZWVZ8tzG6nFboNq1bVPjvCttq+JzJdq397MrWkdt3NVW7tw/rPj7WrldZzw1FwAAAJoXp8Xp5SorK2Wz2RyPjTGy2Wy1Tm+onj17ytfXt8HPc7XGFjdZWVluL5xrE/tkmlvWu+H3cW55TVa21ZWwIrOntlV9c9nt9nofeAIAAMDVqcF36+3QoYMKCgocjwsKChQQEFBt+okTJxyXAgMAAAAAUJcGF6e9evXSgQMHdPDgQVVUVCg9PV2hoaHq1KmTfH19HZf4paWlKTQ01OWBAQAAAAAtT4Mv6/X19dWCBQs0YcIE2e12hYWFKSoqSpK0cOFCJSUlqaioSIGBgUpMTHR5YAAAAABAy1Pv4nTr1q2O/4eEhGj9+vXVlunRo4fWrFnjmmQAAAAAgKtGgy/rBQAAAADA1ShOAQAAAACWozgFgHpYsmSJoqOjFR0dreTkZElSZmamYmNjFRERoZSUFMeye/fuVXx8vCIjIzVjxgyVl5dbFRsAAKDZoDgFACcyMzP1wQcfaO3atVq3bp0+++wzpaena/r06Vq6dKk2btyonJwcbd++XZI0ZcoUzZw5U5s2bZIxRqmpqRa/AgAAAM9HcQoATvj7+2vatGlq3bq1fHx81LVrV+Xm5qpLly7q3LmzvL29FRsbq4yMDB05ckQlJSUKCgqSJMXHxysjI8PiVwAAAOD5KE4BwIlbbrnFUWzm5ubqvffek81mk7+/v2OZgIAA5eXlKT8/v8p0f39/5eXlNXlmAACA5qbB45wCwNVq//79Gjt2rJ5++ml5eXkpNzfXMc8YI5vNpsrKStlstmrTGyInJ6feywYHBzdo3Q2RlZXVLNftLmRuGmQGgKsXxSkA1ENWVpYmTpyo6dOnKzo6Wh9++KEKCgoc8wsKChQQEKAOHTpUmX7ixAkFBAQ0aFs9e/aUr6+vy7I3lrsK36ysLLcW1e5A5qbRkjPb7fYGHXgCgKsRl/UCLVxpWYXb1t3jtkC3rduTHDt2TI8//rgWLlyo6OhoSVKvXr104MABHTx4UBUVFUpPT1doaKg6deokX19fx5mUtLQ0hYaGWhkfAACgWeDMKdDCtfbxUuyTaW5Z94bfx7llvZ5m+fLlstvtWrBggWNaQkKCFixYoAkTJshutyssLExRUVGSpIULFyopKUlFRUUKDAxUYmKiVdEBwGWWLFmi9957T5IUFhamp59+WpmZmZo/f77sdrsGDRqkyZMnS7owpNaMGTNUXFys3r17a/bs2fL25msngLrRSwCAE0lJSUpKSqpx3vr166tN69Gjh9asWePuWADQZC4dUstms+lXv/qV0tPTtXDhQq1YsUIdO3bU2LFjtX37doWFhWnKlCmaO3eugoKCNH36dKWmpmrkyJFWvwwAHo7LegEAAFAnhtQC0BQoTgEAAFAnhtQC0BS4rBcAAAD10lRDakkNG1bLnZrjUEHNLTPDojUdT89McQoAAACnmnJILckzhtVqycMbXS0YFu3/eEJmZ8NqcVkvAAAA6sSQWgCaAmdOAQAAUCeG1ALQFChOAQAAUCeG1ALQFLisFwAAAABgOYpTAAAAAIDlKE4BAAAAAJa7ot+cjh49WoWFhfL2vrCaOXPm6NChQ3rllVdUXl6uBx98UKNGjXJJUAAAAABAy9Xo4tQYo9zcXP3zn/90FKd5eXmaPHmy3n33XbVu3VoJCQnq06ePunXr5rLAAAAAAICWp9HF6ddffy1JGjNmjE6dOqX7779f7dq1U9++fdW+fXtJUmRkpDIyMvTEE0+4Ji0AAAAAoEVq9G9Oz5w5o5CQEL388st64403tGrVKh09elT+/v6OZQICApSXl+eSoAAAAACAlqvRZ07vuOMO3XHHHY7Hw4cP1/z58zV+/HjHNGOMbDZbg9abk5NTr+WCg4MbtN6GysrKsuS5jeXJ7WHFet3ZHlb8fa9ku8113wAAAMDVpdHF6e7du1VWVqaQkBBJFwrRTp06qaCgwLFMQUGBAgICGrTenj17ytfXt7GxXKaxX+izsrLcXgxYwR2vqbm2lRWZPbmt6pPLbrfX+8ATAAAArk6Nvqz37NmzSk5Olt1uV1FRkdauXasXXnhBO3fuVGFhoc6fP6/NmzcrNDTUlXkBwDJFRUWKiYnR4cOHJUmZmZmKjY1VRESEUlJSHMvt3btX8fHxioyM1IwZM1ReXm5VZAAAgGaj0cVpeHi4wsLCNHToUA0bNkzDhg1TcHCwJk+erMTERA0dOlQxMTG6/fbbXZkXACyxZ88ejRgxQrm5uZKkkpISTZ8+XUuXLtXGjRuVk5Oj7du3S5KmTJmimTNnatOmTTLGKDU11cLkAAAAzcMVjXM6adIkTZo0qcq02NhYxcbGXlEoAPA0qampmjVrlp5++mlJUnZ2trp06aLOnTtLutD3ZWRkqFu3biopKVFQUJAkKT4+XosXL9bIkSMtyw4AANAcXFFxCgBXi3nz5lV5nJ+fX+PdyS+f7u/v3+C7ljfk97nN9eZfzfFGWmRuGmQGgKsXxSkANEJlZWWVu5FfvDt5bdMbornfGM4ZT77BV23I3DRacmZuDAcAzjX6N6cAcDXr0KFDjXcnv3z6iRMnGnzXcgAAgKsRxSkANEKvXr104MABHTx4UBUVFUpPT1doaKg6deokX19fx2V+aWlp3LUcAACgHrisFwAawdfXVwsWLNCECRNkt9sVFhamqKgoSdLChQuVlJSkoqIiBQYGKjEx0eK0AAAAno/iFAAaYOvWrY7/h4SEaP369dWW6dGjh9asWdOUsQAAAJo9LusFAAAAAFiO4hQAALR4pWUVbllvj9sC3bJeALgacVkvAABo8Vr7eCn2yTSXr3fD7+Ncvk4AuFpx5hQAAAAAYDmKUwAAAACA5ShOAQAAAACWozgFAAAAAFiO4hQAAAAAYDmKUwAAAACA5ShOAQAAAACWozgFAAAAAFiO4hQAAAAAYDmKUwAAAACA5ShOAQAAAACWozgFAAAAAFiO4hQAAAAAYDm3FKcbNmzQ4MGDFRERoZUrV7pjEwDg0egHAYC+EEDDeLt6hXl5eUpJSdG7776r1q1bKyEhQX369FG3bt1cvSkA8Ej0gwBAXwig4VxenGZmZqpv375q3769JCkyMlIZGRl64okn6nyeMUaSVFpaWu9ttW/n1figdbDb7ZY+v7E8tT2sWrc72sOqv+2VbtvqfePi+/ri+7yla8p+UGqe+7qV76XGInPTaKmfC1dbPyg1fV/oLryPmgafZU3D6szO+kKbcXEvuWzZMp07d06TJ0+WJK1evVrZ2dl69tln63ze2bNn9eWXX7oyCgAP0717d1133XVWx3A7+kEAtbla+kGJvhBA7WrrC11+5rSyslnw1SoAACAASURBVFI2m83x2BhT5XFt2rVrp+7du8vHx6deywNoPowxKisrU7t27ayO0iToBwFc7mrrByX6QgDVOesLXV6cdujQQbt373Y8LigoUEBAgNPntWrV6qo5kghcjdq0aWN1hCZDPwigJldTPyjRFwKoWV19ocvv1nv33Xdr586dKiws1Pnz57V582aFhoa6ejMA4LHoBwGAvhBAw7n8zOkNN9ygyZMnKzExUWVlZRo+fLhuv/12V28GADwW/SAA0BcCaDiX3xAJAAAAAICGcvllvQAAAAAANBTFKQAAAADAchSnAAAAAADLUZwCAAAAACzXLIrToqIixcTE6PDhw9Xm7d27V/Hx8YqMjNSMGTNUXl5ueaYlS5YoPDxccXFxiouL08qVK5sk05IlSxQdHa3o6GglJydXm29VWznLZUV7LVq0SIMHD1Z0dLRef/31avOtaitnuazatyTp+eef17Rp06pNP3r0qEaNGqWoqCiNHz9excXFTZbpauKJ/aAzdWX+xz/+obi4OA0ZMkSPPfaYTp8+bUHC6urKfNG2bds0YMCAJkzlXF25v/76a40ePVpDhgzRL3/5y2bR1p999pmGDRumIUOGaOzYsTpz5owFCavy1M9YuMaGDRs0ePBgRURE1PjZ6onvI2eZPfF9xGdZ02iun2WSJOPhPv30UxMTE2MCAwPNN998U21+dHS0+eSTT4wxxjzzzDNm5cqVlmcaO3as+fjjj92e41I7duwwDzzwgLHb7aa0tNQkJiaazZs3V1nGiraqT66mbq9du3aZhIQEU1ZWZs6fP2/Cw8PNV199VWUZK9qqPrms2LeMMSYzM9P06dPHTJ06tdq8Rx991KSnpxtjjFmyZIlJTk5u6ngtnif2g87Ulfns2bOmX79+5vjx48YYY1566SXz7LPPWhGzCmftbIwxBQUFJioqyoSHhzdxutrVlbuystJERESY7du3G2OMeeGFFzziPeqsrUeMGGG2bdtmjDFm/vz55sUXX2zqiFV46mcsXOP48eMmPDzcfPvtt6a4uNjExsaa/fv3O+Z74vvIWWZjPO99xGdZ02iun2UXefyZ09TUVM2aNUsBAQHV5h05ckQlJSUKCgqSJMXHxysjI8PSTJKUk5OjZcuWKTY2VnPmzJHdbnd7Jn9/f02bNk2tW7eWj4+PunbtqqNHjzrmW9VWznJJTd9ed911l/7yl7/I29tbJ0+eVEVFhdq2beuYb1VbOcslWbNvnTp1SikpKRo3bly1eWVlZfroo48UGRkpqena6mrjif2gM3VlLisr06xZs3TDDTdIkm699VYdO3asqSNW46xvl6SkpCQ98cQTTZjKubpyf/b/27v/6KjqO//jrzEJ4ZdKaTPG0py0IhAMCAqlopAsdBsCQwSiLQE1xfTwo9LQIgvSZIQV5IeYirIUV11qFwpqxEKQQsJZOLBCULrpljQ0CAcJyq9kUCsEyEyY+Xz/8MssMZBJMJM7kzwf53gOc++de1+fe/P5XN9z79w5eFAdO3ZUUlKSJGnatGl65JFHWjpiPYH2tc/n89+FcenSJbVv374l49UTqudYNI/i4mLdd9996tKlizp27KgRI0bUOX6h2I8CZZZCrx9xLmsZ4XouuyLki9NFixZp4MCB15xXVVWlmJgY/+uYmBhVVlZamunChQvq3bu3Zs+erY0bN+rcuXNatWpV0DP16NHD36ErKiq0bds2JScn++dbta8C5bJqf0VFRWnFihVyOBwaPHiwf2CRrNtXgXJZta/mzZunmTNn6pZbbqk37/PPP1fnzp0VGRkpqWX3VVsSiuNgIA1l/sY3vqEf/ehHkqSamhq9+uqr+ud//ueWjHdNDWWWpDVr1uiuu+5Sv379WjBVYA3l/vjjj/Wtb31LOTk5GjdunObPn1/vQy8rBNrXc+fOldPp1JAhQ1RcXKyMjIwWTFdfqJ5j0Ty+evzsdnud4xeK/ShQZin0+hHnspYRrueyK0K+OG2Iz+eTzWbzvzbG1HlthU6dOum1115T9+7dFRkZqaysLO3evbvFtn/kyBFlZWVpzpw5+u53v+ufbvW+ul4uK/fXjBkztG/fPp0+fVr5+fn+6Vbvq+vlsmJfvf3227r99ts1ePDga86/1r6xug+2NVb/vX4d58+f15QpU5SQkKBx48ZZHadBhw8f1vbt2/XEE09YHaVJLl++rP3792vChAnauHGj4uLitHTpUqtjNaimpka5ubn6/e9/rz179mjixIl66qmnrI4lKXTPsfh6Ah2/UOxHgTKHcj+6lnDuQ5zLmldYF6exsbFyuVz+12fPnm3wEnZLOHXqlDZs2OB/bYzxX1UKtpKSEk2aNEmzZs2q1zms3FcN5bJifx09elTl5eWSpA4dOiglJUUffvihf75V+ypQLiv21datW7V3716NGTNGK1as0M6dO7V48WL//K5du+r8+fPyer2SJJfLZXkfbGtCcRxsjKqqKk2cOFG9evXSokWLrI4TUGFhoVwulx566CFNmTLFnz/UxcTEKD4+Xn379pUkjR49WqWlpRanatjhw4cVHR2tu+++W5I0fvx47d+/3+JUoXuOxdf31eP31XNZKPajQJlDtR9dT7j2Ic5lzS+si9Nu3bopOjpaJSUlkqSCggL/9wGs0r59ez3//PP65JNPZIzRunXr/Jf8g+n06dOaPn268vLy5HA46s23al8FymXF/jpx4oScTqc8Ho88Ho927NihAQMG+Odbta8C5bJiX73++uvasmWLCgoKNGPGDA0fPlw5OTn++VFRURo4cKC2bt0qSdq0aZPlfbCtCcVxMBCv16tp06Zp5MiRys3NDYtPx2fMmKGioiIVFBTo1Vdfld1u1/r1662OFdA999yjzz77TIcOHZIk7dy5U4mJiRanalh8fLzOnDmjjz76SJK0Y8cOf1FglVA9x6J53H///dq3b58+++wzXbp0Sdu3b69z/EKxHwXKHIr9qCHh2Ic4lwVHy1zSa2aTJ0/WjBkz1LdvX+Xl5cnpdKq6ulqJiYnKzMy0PNOCBQv085//XLW1tbr33nv1+OOPB337q1evltvtrnObSUZGhnbu3GnpvmpMrpbeX8nJySotLdXYsWMVERGhlJQUORwOy/+uGpPLir+ta8nNzdXw4cP1wx/+UPPnz9fcuXP18ssv6/bbb9cLL7xgSaa2xuq/1xtxJfOZM2f097//XV6vV0VFRZKkPn36hOSnzlfv53Byde7f/va3cjqdunTpkmJjY6/5Myih4OrMS5Ys0a9+9SsZY/TNb36zzh0bVgjVcyyax2233aaZM2cqMzNTtbW1evjhh3X33XeHdD9qTOZQ60fXwrmsZYTTucxmjDFWhwAAAAAAtG1hfVsvAAAAAKB1CMvbehEczz77rP785z9L+vLhPN26dfP/JtZbb711zd/HKi4u1nPPPaeCgoIb2ubXfT+AtutGxqxg++Mf/6gVK1aoe/fuWr16dYtue9euXTpw4IB++ctfBn1bK1euVEJCQkj8bAIAoPWgOIWf0+n0/3v48OHKy8sLi3vTAbRNoThmbdq0STNnztSYMWNafNt/+9vf9MUXX7TItj744APdeeedLbItAEDbQXGKRjly5IgWLVqkc+fOyev1atKkSfUepe/xeLRs2TKVlJTI6/UqMTFRubm56ty5s5KSkjR27Fjt2bNH1dXV+tnPfqbx48dLkqqrq/XLX/5SFRUV8ng8WrRoke69916dO3dOzzzzjP/nVIYNG+a/IjBw4EBNmjRJ77//vqqqqjR16lT/+t566y299dZb8vl86tq1q55++ml973vfa8G9BcBKmzdv1vr16/Xmm29K+vJnmH7yk59o586d6t+/vyZPnqz33ntPFy9e1JNPPqmUlBRJX/6u7xtvvCGfz6cuXbro6aefVvfu3eut//z583rmmWd06NAh2Ww2DR06VE8++aSWLVumv/3tbzpx4oQ+//xzTZo0qc77jh07pgULFujChQtyuVxKSEjQiy++qOjoaN11113XzHXx4kX967/+q44fP65//OMf6tSpk/Ly8nTHHXfoscce06233qqPPvpIo0aN0ptvvimv16ubb75Z8fHx2r59u3w+n06dOqXbbrtNP/nJT/SHP/xBFRUVevzxx5WVldVgu+fOnavOnTvrww8/1JkzZ9SrVy8999xz2rRpk8rKyrRs2TJFRES0yBPpAQBthAGuYdiwYaa0tNQYY4zH4zEjR4405eXlxhhjvvjiCzNixAhTWlpq9u7dax588EFjjDEvvviief75543P5zPGGPPcc8+ZhQsXGmOMGTp0qJk/f74xxpiTJ0+aQYMGmSNHjpi9e/eaxMREc+DAAWOMMa+99prJysoyxhjz5JNPmiVLlhhjjKmpqTE//elPzX/8x3+Y2tpa07NnT7N+/XpjjDF//etfTd++fY3H4zHFxcXm0UcfNZcuXTLGGLNr1y4zevToYO8uABa7esxyu91m8ODB5vDhw8aYL8emvLw8Y4wxPXv2NC+//LIxxpjy8nIzYMAA8+mnn5oPPvjATJw40Vy8eNEYY8x7771nUlNTr7mtOXPmmIULFxqfz2fcbrfJysoyr7zyijHGmEcffdRs27btmu9bunSp2bRpkzHmy3F19OjRprCwsMFc27Zt84+jxhjz9NNPmwULFvi39etf/9o/b8WKFeaZZ54xxhjzzjvvmAEDBphTp04Zr9drRo0aZbKzs43X6zXl5eWmb9++xuv1Ntjup556yowfP9643W7j8XjM2LFjzYYNGwK2EwCAG8WVUwR09OhRffLJJ3rqqaf80zwej8rLy/Wd73zHP23Xrl26ePGi3nvvPUlSbW1tnR9QfvTRRyVJ3/72t/XAAw9o79696tGjh+Lj4/0/Et27d29t2bJFkrRnzx5t2LBBkhQdHa3x48frzTff1E9/+lNJ0g9/+ENJ0l133SW3262amhrt2rVLx44d819FlaTPP/9c58+f180339zs+wZA6GnXrp1+/OMf6+2339ZTTz2ljRs3au3atf75V8aihIQE9ezZU3/+85914MABHT9+XBkZGf7lzp07p3/84x/q0qVLnfX/93//t9544w3ZbDa1a9dOGRkZ+s///E9NmTKlwVyzZ8/W3r179dprr6miokJVVVW6ePFig7lSU1MVFxentWvX6vjx49q/f7/uuece/3sGDhx43e317dtXt99+uyTpO9/5joYMGaKbbrpJcXFxcrvdunTpknbt2nXddkvS0KFD1a5dO0lSz549W+y2YQBA20RxioCu3Op19UOLXC6XbrnlFv+PJUtf/hjxvHnz9MADD0j68nbd2tpa//yIiIg667zyOjLy//4MbTabzP//dSOv11vnB42NMbp8+bL/9ZWHnVxZxhgjr9erhx56SDNnzvSvw+VyUZgCbUxGRoYefvhhDRo0SD169FBcXJx/3rXGIp/PpzFjxmj27Nn+6VVVVbr11lvrfH/02Weflc/nqzM2+Xy+OmPTFV993+rVq+X1ejVy5Ej90z/9k06fPu0f766Xa/369crPz9cjjzyitLQ0denSRSdOnPAv17Fjx+vugytF5RVXj7VXb+d67ZZU56FSV4/PAAAEAz8lg4DuvPNO3XTTTfrTn/4kSTp58qRGjx6tQ4cO1VluyJAhWrt2rWpra+X1epWTk6MXX3zRP3/Tpk2SpBMnTmjfvn0aOnRog9sdMmSI/vCHP0iS3G638vPzdf/99zf4nqFDh+rdd9/V2bNnJUnr1q3zf68KQNtx++23q3///lq8eLEmTJhQZ96VsejgwYM6duyYvv/972vIkCH605/+pKqqKknSG2+84b9Lo6CgwP9f3759/WOTMUYej+e6Y9NX37dnzx5Nnz5do0aNkiQdOHBAXq+3wVx79uzRuHHj9OMf/1jf+973tHPnzjrvuVpERMQ1i+SGNNTuhtzItgAACIQrpwioXbt2evnll7V48WL9+7//uy5fvqxZs2apX79+Ki4u9i+XnZ2t5557TmPHjvU/EGnOnDn++R9//LHGjRsnj8ej+fPnKz4+XidPnrzudufNm6eFCxdq9OjRqq2tVVJSkiZPntxg1uTkZE2aNEmTJk2SzWbTLbfcon/7t3/7+jsBQNhJT0/XwoULlZycXGf6X/7yF+Xn58vn82n58uW69dZbNWTIEE2ePFlZWVmy2Wzq3LmzVq5cWecK6RVOp1PPPvus0tLSVFtbq6FDh2ratGkB88ycOVPTp09Xx44d1blzZ33/+9/Xxx9/3GCurKwszZs3z/8Vh/79++vw4cPXXP99992nf/mXf9HChQuVmJjYqH3UlHZfbfjw4XrhhRdUW1tb7+F4AADcKJvhHh20gKSkJL3yyivq3bu31VEAtAE+n08LFizQt7/97TrfBe3Vq5f27dunrl27WpiuvlDNBQBAS+K2XgBAq1JdXa0f/OAHOn36tDIzM62OAwAAGokrpwAAAAAAy3HlFAAAAABguZB5IJLP59OFCxcUFRUV8EEMAMKLMUa1tbXq1KmTbrqJz8Suh3EQaL0YBwEgsJApTi9cuHDdJxACaB169uzJb842gHEQaP0YBwHg+kKmOI2KipL05aD91R8Ov5aysjL16dMn2LGaVThmlsIzN5lbRmMzezweHT582N/PcW1NHQel8Py7aQraF/5aexsZBwGg+TSqOH3ppZdUVFQkm82mhx9+WI8//riKi4u1ZMkSud1ujRw5UjNnzpQklZeXKzc3VxcuXNDAgQP1zDPPKDIy8Gau3MLWrl07RUdHNyp8Y5cLJeGYWQrP3GRuGU3JzK2qDbuRcVAKz7+bpqB94a+1t5FxEACaR8AvPezfv1/vv/++Nm/erHfeeUdr167VoUOHlJOTo1WrVmnr1q0qKyvT7t27JUmzZ8/WvHnzVFRUJGOM8vPzg94IAAAAAEB4C1icDho0SGvWrFFkZKQ+/fRTeb1enTt3TvHx8YqLi1NkZKTS0tJUWFiokydPqqamRv3795ckpaenq7CwMOiNAAAAAACEt0Y9Li4qKkorVqyQw+HQ4MGDVVVVpZiYGP98u92uysrKetNjYmJUWVnZ/KkBAAAAAK1Kox+INGPGDE2ePFnTpk1TRUVFne9MGGNks9nk8/muOb0pysrKGrVcQu/EJq23KS5crNGh8oNBWXdJSUlQ1hts4ZibzC0jHDMDoSKY5zJPrVftoiKCtn4AAJpbwOL06NGj8ng86t27tzp06KCUlBQVFhYqIuL/Tngul0t2u12xsbFyuVz+6WfPnpXdbm9SoD59+jT6wQJpswqatO7Gevc3YzRgwIBmX29JSUlQ1hts4ZibzC2jsZndbnejP3gC2pJOHdsH9VwGAEA4CXhb74kTJ+R0OuXxeOTxeLRjxw5lZGTo2LFjOn78uLxer7Zs2aKkpCR169ZN0dHR/ispBQUFSkpKCnojAAAAAADhLeCV0+TkZJWWlmrs2LGKiIhQSkqKHA6HunbtquzsbLndbiUnJys1NVWSlJeXJ6fTqerqaiUmJiozMzPojQAAAAAAhLdGfec0Oztb2dnZdaYNHjxYmzdvrrdsQkKCNmzY0DzpAAAAAABtQqOe1gsAAAAAQDBRnAIAAAAALEdxCgAAAACwHMUpAAAAAMByFKcAAAAAAMtRnAIAAAAALEdxCgAAAACwHMUpADTCSy+9pFGjRsnhcOj111+XJBUXFystLU0pKSlavny5f9ny8nKlp6drxIgRys3N1eXLl62KDQAAEDYoTgEggP379+v999/X5s2b9c4772jt2rU6dOiQcnJytGrVKm3dulVlZWXavXu3JGn27NmaN2+eioqKZIxRfn6+xS0AAAAIfRSnABDAoEGDtGbNGkVGRurTTz+V1+vVuXPnFB8fr7i4OEVGRiotLU2FhYU6efKkampq1L9/f0lSenq6CgsLLW4BAABA6Iu0OgAAhIOoqCitWLFCv/vd75SamqqqqirFxMT459vtdlVWVtabHhMTo8rKyiZtq6ysrEnLl5SUNGn5cNOa2zdgwICgrj9U9l2o5AiW1t4+AGgpFKcA0EgzZszQ5MmTNW3aNFVUVMhms/nnGWNks9nk8/muOb0p+vTpo+jo6EYtW1JSEvQCx0qtvX3BFgr7rrUfw8a2z+12N/mDJwBoa7itFwACOHr0qMrLyyVJHTp0UEpKij744AO5XC7/Mi6XS3a7XbGxsXWmnz17Vna7vcUzAwAAhBuKUwAI4MSJE3I6nfJ4PPJ4PNqxY4cyMjJ07NgxHT9+XF6vV1u2bFFSUpK6deum6Oho/21+BQUFSkpKsrgFAAAAoY/begEggOTkZJWWlmrs2LGKiIhQSkqKHA6HunbtquzsbLndbiUnJys1NVWSlJeXJ6fTqerqaiUmJiozM9PiFgAAAIQ+ilMAaITs7GxlZ2fXmTZ48GBt3ry53rIJCQnasGFDS0UDAABoFbitFwAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlGlWcrly5Ug6HQw6HQ8uWLZMkFRcXKy0tTSkpKVq+fLl/2fLycqWnp2vEiBHKzc3V5cuXg5McAAAAANBqBCxOi4uLtWfPHm3cuFGbNm3SwYMHtWXLFuXk5GjVqlXaunWrysrKtHv3bknS7NmzNW/ePBUVFckYo/z8/KA3AgAAAAAQ3gIWpzExMZo7d67atWunqKgode/eXRUVFYqPj1dcXJwiIyOVlpamwsJCnTx5UjU1Nerfv78kKT09XYWFhUFvBAAAAAAgvAUsTnv06OEvNisqKrRt2zbZbDbFxMT4l7Hb7aqsrFRVVVWd6TExMaqsrAxCbAAAAABAaxLZ2AWPHDmiqVOnas6cOYqIiFBFRYV/njFGNptNPp9PNput3vSmKCsra9RyAwYMaNJ6m6qkpCSs1hts4ZibzF9K6J2oTh3bN/t6r6w7HPczAAAAQk+jitOSkhLNmDFDOTk5cjgc2r9/v1wul3++y+WS3W5XbGxsnelnz56V3W5vUqA+ffooOjq6Se8JhmAUvyUlJUEvqoMhHHOTua60WQVBWe+7vxnTqMxut7vRHzwBAACgbQp4W+/p06c1ffp05eXlyeFwSJL69eunY8eO6fjx4/J6vdqyZYuSkpLUrVs3RUdH+6+kFBQUKCkpKbgtAAAAAACEvYBXTlevXi23262lS5f6p2VkZGjp0qXKzs6W2+1WcnKyUlNTJUl5eXlyOp2qrq5WYmKiMjMzg5ceAAAAANAqBCxOnU6nnE7nNedt3ry53rSEhARt2LDh6ycDAAAAALQZAW/rBQAAAAAg2ChOAQAAAACWozgFgEZYuXKlHA6HHA6Hli1bJkkqLi5WWlqaUlJStHz5cv+y5eXlSk9P14gRI5Sbm6vLly9bFRsAACBsUJwCQADFxcXas2ePNm7cqE2bNungwYPasmWLcnJytGrVKm3dulVlZWXavXu3JGn27NmaN2+eioqKZIxRfn6+xS0AAAAIfRSnABBATEyM5s6dq3bt2ikqKkrdu3dXRUWF4uPjFRcXp8jISKWlpamwsFAnT55UTU2N+vfvL0lKT09XYWGhxS0AAAAIfQGf1gsAbV2PHj38/66oqNC2bdv06KOPKiYmxj/dbrersrJSVVVVdabHxMSosrKySdsrKytr0vJXflu6tWrN7RswYEBQ1x8q+y5UcgRLa28fALQUilMAaKQjR45o6tSpmjNnjiIiIlRRUeGfZ4yRzWaTz+eTzWarN70p+vTpo+jo6EYtW1JSEvQCx0qtvX3BFgr7rrUfw8a2z+12N/mDJwBoa7itFwAaoaSkRJMmTdKsWbM0btw4xcbGyuVy+ee7XC7Z7fZ608+ePSu73W5FZAAAgLBCcQoAAZw+fVrTp09XXl6eHA6HJKlfv346duyYjh8/Lq/Xqy1btigpKUndunVTdHS0/za/goICJSUlWRkfAAAgLHBbLwAEsHr1arndbi1dutQ/LSMjQ0uXLlV2drbcbreSk5OVmpoqScrLy5PT6VR1dbUSExOVmZlpVXQAAICwQXEKAAE4nU45nc5rztu8eXO9aQkJCdqwYUOwYwEAALQq3NYLAAAAALAcxSkAAAAAwHIUpwAAAAAAy1GcAgAAAAAsR3EKAAAAALAcxSkAAAAAwHIUpwAAAAAAy1GcAgAAAAAsR3EKAAAAALAcxSkAAAAAwHIUpwAAAAAAy1GcAgAAAAAsR3EKAAAAALAcxSkAAAAAwHIUpwAAAAAAy1GcAgAAAAAs16jitLq6WqNHj9aJEyckScXFxUpLS1NKSoqWL1/uX668vFzp6ekaMWKEcnNzdfny5eCkBgAAAAC0KgGL0wMHDmjChAmqqKiQJNXU1CgnJ0erVq3S1q1bVVZWpt27d0uSZs+erXnz5qmoqEjGGOXn5wc1PAAAAACgdQhYnObn52v+/Pmy2+2SpNLSUsXHxysuLk6RkZFKS0tTYWGhTp48qZqaGvXv31+SlJ6ersLCwuCmBwAAAAC0CpGBFli0aFGd11VVVYqJifG/ttvtqqysrDc9JiZGlZWVTQ5UVlbWqOUGDBjQ5HU3RUlJSVitN9jCMTeZvxSufQUAAABtS8Di9Kt8Pp9sNpv/tTFGNpvtutObqk+fPoqOjm7y+5pbMP6HvqSkJOiFQjCEY24yt5zGZHa73Y3+4AkAAABtU5Of1hsbGyuXy+V/7XK5ZLfb600/e/as/1ZgAGgNeDgcAABA8DS5OO3Xr5+OHTum48ePy+v1asuWLUpKSlK3bt0UHR3tv8WvoKBASUlJzR4YAKzAw+EAAACCq8nFaXR0tJYuXars7GyNGjVKd9xxh1JTUyVJeXl5WrJkiVJTU3Xx4kVlZmY2e2AAsAIPhwMAAAiuRn/ndOfOnf5/Dx48JhEdvAAACh5JREFUWJs3b663TEJCgjZs2NA8yQAghLTkw+Ga+v3c1v5QqtbcvrbywLJQyREsrb19ANBSmvxAJABAcB8O15QHw4Xrg7Qaq7W3L9hCYd+19mPY2PbxYDgACKzJt/UCAHg4HAAAQHOjOAWAGxAqD4dL6J0YlPV6ar1BWS8AAMD1cFsvANyAqx8O53a7lZycXOfhcE6nU9XV1UpMTAzqw+E6dWyvtFkFzb7ed38zptnXCQAA0BCKUwBoAh4OBwAAEBzc1gsAAAAAsBzFKQAAAADAchSnAAAAAADLUZwCAAAAACxHcQoAAAAAsBzFKQAAAADAchSnAAAAAADLUZwCAIBWz1PrDcp6E3onBmW9ANAWRVodAAAAINjaRUUobVZBs6/33d+MafZ1AkBbxZVTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgOYpTAAAAAIDlKE4BAAAAAJajOAUAAAAAWI7iFAAAAABguaAUp++++65GjRqllJQUrVu3LhibAICQxjgIAADQNJHNvcLKykotX75cf/zjH9WuXTtlZGToBz/4ge68887m3hQAhCTGQQAAgKZr9uK0uLhY9913n7p06SJJGjFihAoLC/WLX/yiwfcZYyRJHo+n0dvq0inixoM2wO12B2W9wV53MIVjbjL/H6v7ypV+faWft3YtOQ5KwTm+odR/QilLMFjdP1tCqGSxsq+0tXEQAG6EzTTzKPnKK6/o4sWLmjlzpiTp7bffVmlpqRYuXNjg+86fP6/Dhw83ZxQAIaZnz566+eabrY4RdIyDAK6nrYyDAHAjmv3Kqc/nk81m8782xtR5fT2dOnVSz549FRUV1ajlAYQPY4xqa2vVqVMnq6O0CMZBAF/V1sZBALgRzV6cxsbG6n/+53/8r10ul+x2e8D33XTTTXySCLRi7du3tzpCi2EcBHAtbWkcBIAb0exP673//vu1b98+ffbZZ7p06ZK2b9+upKSk5t4MAIQsxkEAAICma/Yrp7fddptmzpypzMxM1dbW6uGHH9bdd9/d3JsBgJDFOAgAANB0zf5AJAAAAAAAmqrZb+sFAAAAAKCpKE4BAAAAAJajOAUAAAAAWI7iFAAAAABgubAoTqurqzV69GidOHGi3rzy8nKlp6drxIgRys3N1eXLly1IWF9Dmf/rv/5LY8aM0YMPPqgnnnhCX3zxhQUJ62so8xW7du3S8OHDWzBVwxrK/NFHH+mxxx7Tgw8+qJ/97GdhsZ8PHjyohx56SA8++KCmTp2qc+fOWZCwrpUrV8rhcMjhcGjZsmX15odqH2xtwnEcbKqG2rhy5UoNGzZMY8aM0ZgxY7Ru3ToLEt64ttCPArUx3I/hSy+9pFGjRsnhcOj111+vN781HEMAsJwJcX/961/N6NGjTWJiovnkk0/qzXc4HOZ///d/jTHG/PrXvzbr1q1r6Yj1NJT5/Pnz5oEHHjBnzpwxxhjz4osvmoULF1oRs45A+9kYY1wul0lNTTXDhg1r4XTX1lBmn89nUlJSzO7du40xxjz//PNm2bJlVsSsI9B+njBhgtm1a5cxxpglS5aYF154oaUj1rF3714zfvx443a7jcfjMZmZmWb79u11lgnFPtjahOM42FSB2jh16lTzl7/8xYJkX19b6EeNaWM4H8MPPvjAZGRkmNraWnPp0iUzbNgwc/To0TrLhPsxBIBQEPJXTvPz8zV//nzZ7fZ6806ePKmamhr1799fkpSenq7CwsKWjlhPQ5lra2s1f/583XbbbZKkXr166fTp0y0dsZ6GMl/hdDr1i1/8ogVTNayhzAcPHlTHjh2VlJQkSZo2bZoeeeSRlo5YT6D97PP5dOHCBUnSpUuX1L59+5aMV09MTIzmzp2rdu3aKSoqSt27d9epU6f880O1D7Y24TgONlWgvlFWVqZXXnlFaWlpWrBggdxudwsnvHFtoR8FaqMU3sdw0KBBWrNmjSIjI/Xpp5/K6/WqY8eO/vmt4RgCQCiItDpAIIsWLbruvKqqKsXExPhfx8TEqLKysiViNaihzN/4xjf0ox/9SJJUU1OjV199VY899lhLRbuuhjJL0po1a3TXXXepX79+LZQosIYyf/zxx/rWt76lnJwclZeX64477tDTTz/dgumuLdB+njt3rrKysrR48WJ16NBB+fn5LZTs2nr06OH/d0VFhbZt26Y33njDPy1U+2BrE47jYFM11MYLFy6od+/emj17tuLj4zV37lytWrVKM2fObMGEN64t9KNAbQz3YyhJUVFRWrFihX73u98pNTXV/yGz1DqOIQCEgpC/ctoQn88nm83mf22MqfM6lJ0/f15TpkxRQkKCxo0bZ3WcBh0+fFjbt2/XE088YXWURrt8+bL279+vCRMmaOPGjYqLi9PSpUutjtWgmpoa5ebm6ve//7327NmjiRMn6qmnnrI6liTpyJEjysrK0pw5c/Td737XPz2c+2Br0RaOQadOnfTaa6+pe/fuioyMVFZWlnbv3m11rCZrC/3oem1sLcdwxowZ2rdvn06fPl3nw8PWdAwBwEphXZzGxsbK5XL5X589e7bB21JDRVVVlSZOnKhevXoFvJIWCgoLC+VyufTQQw9pypQp/vyhLCYmRvHx8erbt68kafTo0SotLbU4VcMOHz6s6Oho3X333ZKk8ePHa//+/RankkpKSjRp0iTNmjWr3gcp4doHW5O2cAxOnTqlDRs2+F8bYxQZGfI3/tTRFvpRQ20M92N49OhRlZeXS5I6dOiglJQUffjhh/75reUYAoDVwro47datm6Kjo1VSUiJJKigo8H/HMFR5vV5NmzZNI0eOVG5ublh8sjpjxgwVFRWpoKBAr776qux2u9avX291rAbdc889+uyzz3To0CFJ0s6dO5WYmGhxqobFx8frzJkz+uijjyRJO3bs8BfXVjl9+rSmT5+uvLw8ORyOevPDsQ+2Nm3hGLRv317PP/+8PvnkExljtG7dOv/XI8JBW+hHgdoY7sfwxIkTcjqd8ng88ng82rFjhwYMGOCf3xqOIQCEgvD52PIqkydP1owZM9S3b1/l5eXJ6XSqurpaiYmJyszMtDreNV3JfObMGf3973+X1+tVUVGRJKlPnz4heQX16v0cLq7O/Nvf/lZOp1OXLl1SbGzsNX/aIBRcnXnJkiX61a9+JWOMvvnNb2rx4sWWZlu9erXcbnedW6IzMjK0c+fOsOuDrU04joNNdXUbFyxYoJ///Oeqra3Vvffeq8cff9zqeI3WFvpRY9oYzscwOTlZpaWlGjt2rCIiIpSSkiKHw9Em+iEAtCSbMcZYHQIAAAAA0LaF9W29AAAAAIDWgeIUAAAAAGA5ilMAAAAAgOUoTgEAAAAAlqM4BQAAAABYjuIUAAAAAGA5ilMAAAAAgOUoTgEAAAAAlvt/iDXbTOSZLn4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1152x720 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "raw_data.hist(column= raw_data.columns,  figsize = (16,10)) # Check the distribution of each columns "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> First, I built  the training set given the data provided. It has 500 datapoints and 20 features. Two features contain missing values. We've also checked the datatypes of each field. I made the following notes to clean the data. \n",
    "- (1) All column names include 'dash(-)' between the words. It is replaced with 'underscore' for convenient usage of python.\n",
    "- (2) The column **\"Type-of-apartment\"** should have an \"object\" type, rather than \"int64\". \n",
    "- (3) **'Telephone'** is the contact communication type. It is dropped due to irrelevancy. \n",
    "- (4) **'Occupation'**, **'Concurrent-Credits'** have only one value for the entire samples. Although the columns are associated to creditworthiness, it doesn't make sense to be included in the model. \n",
    "- (5) The fields: **'Guarantors'**, **'Foreign-Worker'**, **'No-of-dependents'** highly skew towards one type of data. Such low variabilities hinder appropriate prediction. They are dropped.\n",
    "- (6) **'Duration-in-Current-address'** is removed due the high percentage of missing values. ca. 69% of datapoints are missing. It is dropped.\n",
    "- (7) **'Age-years'** has only 2% of missing values. I imputed them with the median value. Unlike the mean value, median is the already existing value in the dataset, and it is less susceptible to outliers comparing to mean imputation. I prefer the median here because the distribution of the variable is skewed to the left. And when we have skewed distribution the median value is a more accurate measure of centre.\n",
    "\n",
    "> Therefore, the initial  data is modified as below. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make a copy of the original dataset. It works as a checkpoint.\n",
    "data = raw_data.copy() \n",
    "# (1) Change the column names with underscore\n",
    "data.columns = data.columns.str.replace(\"-\", \"_\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# (2) Change the datatype of the column: \"Type_of_apartment\"\n",
    "data.Type_of_apartment=data.Type_of_apartment.apply(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# (3) - (6) Dropped the irrelevant or uncessary columns \n",
    "data.drop(['Telephone'], axis = 1, inplace= True) # Irrelevant feature is dropped.\n",
    "data.drop(['Occupation','Concurrent_Credits'], axis = 1, inplace= True) # Same value for all datapoints are removed.\n",
    "data.drop(['Guarantors', 'Foreign_Worker','No_of_dependents'], axis = 1, inplace= True) # Columns with low variabilities are excluded.\n",
    "data.drop(['Duration_in_Current_address'], axis = 1, inplace= True) # The field with huge percentage of missing values is dropped."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>Credit_Application_Result</th>\n",
       "      <th>Account_Balance</th>\n",
       "      <th>Duration_of_Credit_Month</th>\n",
       "      <th>Payment_Status_of_Previous_Credit</th>\n",
       "      <th>Purpose</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Value_Savings_Stocks</th>\n",
       "      <th>Length_of_current_employment</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "      <th>Type_of_apartment</th>\n",
       "      <th>No_of_Credits_at_this_Bank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>1494</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>No Account</td>\n",
       "      <td>9</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>1199</td>\n",
       "      <td>None</td>\n",
       "      <td>1-4 yrs</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>More than 1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>No Account</td>\n",
       "      <td>9</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>1549</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>12</td>\n",
       "      <td>No Problems (in this bank)</td>\n",
       "      <td>Other</td>\n",
       "      <td>930</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>4-7 yrs</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>More than 1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>12</td>\n",
       "      <td>No Problems (in this bank)</td>\n",
       "      <td>Other</td>\n",
       "      <td>1402</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>1-4 yrs</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>151</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>12</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>1736</td>\n",
       "      <td>None</td>\n",
       "      <td>1-4 yrs</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>172</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>No Account</td>\n",
       "      <td>12</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>3017</td>\n",
       "      <td>None</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>321</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>No Account</td>\n",
       "      <td>24</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>2384</td>\n",
       "      <td>None</td>\n",
       "      <td>4-7 yrs</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>345</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>24</td>\n",
       "      <td>No Problems (in this bank)</td>\n",
       "      <td>Other</td>\n",
       "      <td>3617</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>4-7 yrs</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>More than 1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>363</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>24</td>\n",
       "      <td>No Problems (in this bank)</td>\n",
       "      <td>Other</td>\n",
       "      <td>5103</td>\n",
       "      <td>None</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>More than 1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>389</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>30</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>1867</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>4-7 yrs</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>495</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>No Account</td>\n",
       "      <td>60</td>\n",
       "      <td>No Problems (in this bank)</td>\n",
       "      <td>Other</td>\n",
       "      <td>7418</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Credit_Application_Result Account_Balance  Duration_of_Credit_Month  \\\n",
       "0                Creditworthy    Some Balance                         4   \n",
       "52               Creditworthy      No Account                         9   \n",
       "63               Creditworthy      No Account                         9   \n",
       "115              Creditworthy    Some Balance                        12   \n",
       "139              Creditworthy    Some Balance                        12   \n",
       "151              Creditworthy    Some Balance                        12   \n",
       "172              Creditworthy      No Account                        12   \n",
       "321              Creditworthy      No Account                        24   \n",
       "345              Creditworthy    Some Balance                        24   \n",
       "363              Creditworthy    Some Balance                        24   \n",
       "389              Creditworthy    Some Balance                        30   \n",
       "495              Creditworthy      No Account                        60   \n",
       "\n",
       "    Payment_Status_of_Previous_Credit Purpose  Credit_Amount  \\\n",
       "0                             Paid Up   Other           1494   \n",
       "52                            Paid Up   Other           1199   \n",
       "63                            Paid Up   Other           1549   \n",
       "115        No Problems (in this bank)   Other            930   \n",
       "139        No Problems (in this bank)   Other           1402   \n",
       "151                           Paid Up   Other           1736   \n",
       "172                           Paid Up   Other           3017   \n",
       "321                           Paid Up   Other           2384   \n",
       "345        No Problems (in this bank)   Other           3617   \n",
       "363        No Problems (in this bank)   Other           5103   \n",
       "389                           Paid Up   Other           1867   \n",
       "495        No Problems (in this bank)   Other           7418   \n",
       "\n",
       "    Value_Savings_Stocks Length_of_current_employment  Instalment_per_cent  \\\n",
       "0             £100-£1000                        < 1yr                    1   \n",
       "52                  None                      1-4 yrs                    4   \n",
       "63            £100-£1000                        < 1yr                    4   \n",
       "115           £100-£1000                      4-7 yrs                    4   \n",
       "139           £100-£1000                      1-4 yrs                    3   \n",
       "151                 None                      1-4 yrs                    3   \n",
       "172                 None                        < 1yr                    3   \n",
       "321                 None                      4-7 yrs                    4   \n",
       "345           £100-£1000                      4-7 yrs                    4   \n",
       "363                 None                        < 1yr                    3   \n",
       "389           £100-£1000                      4-7 yrs                    4   \n",
       "495           £100-£1000                        < 1yr                    1   \n",
       "\n",
       "     Most_valuable_available_asset  Age_years Type_of_apartment  \\\n",
       "0                                1        NaN                 2   \n",
       "52                               2        NaN                 2   \n",
       "63                               1        NaN                 2   \n",
       "115                              1        NaN                 2   \n",
       "139                              3        NaN                 1   \n",
       "151                              1        NaN                 2   \n",
       "172                              1        NaN                 1   \n",
       "321                              1        NaN                 1   \n",
       "345                              4        NaN                 1   \n",
       "363                              4        NaN                 3   \n",
       "389                              3        NaN                 2   \n",
       "495                              1        NaN                 2   \n",
       "\n",
       "    No_of_Credits_at_this_Bank  \n",
       "0                            1  \n",
       "52                 More than 1  \n",
       "63                           1  \n",
       "115                More than 1  \n",
       "139                          1  \n",
       "151                          1  \n",
       "172                          1  \n",
       "321                          1  \n",
       "345                More than 1  \n",
       "363                More than 1  \n",
       "389                          1  \n",
       "495                          1  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# (7) Impute the missing values in Age_years column \n",
    "# Find the samples whose 'Age_years' column is omitted.\n",
    "null_data = data[data.isnull().any(axis=1)]\n",
    "null_data "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33.0"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "median_age = data['Age_years'].median()\n",
    "median_age # Calculate the median value of age in years"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fill up the missing datapoints with its median values \n",
    "values = {'Age_years': median_age} \n",
    "data.fillna(value=values, inplace = True); "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> All cleaning processes earlier mentioned, are done. Now, using the methods .head() and .info(), check whether the changes are well applied. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Credit_Application_Result</th>\n",
       "      <th>Account_Balance</th>\n",
       "      <th>Duration_of_Credit_Month</th>\n",
       "      <th>Payment_Status_of_Previous_Credit</th>\n",
       "      <th>Purpose</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Value_Savings_Stocks</th>\n",
       "      <th>Length_of_current_employment</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "      <th>Type_of_apartment</th>\n",
       "      <th>No_of_Credits_at_this_Bank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>1494</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>33.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>1494</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>29.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>No Problems (in this bank)</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>1544</td>\n",
       "      <td>None</td>\n",
       "      <td>1-4 yrs</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>42.0</td>\n",
       "      <td>2</td>\n",
       "      <td>More than 1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>No Problems (in this bank)</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>3380</td>\n",
       "      <td>None</td>\n",
       "      <td>1-4 yrs</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>37.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>No Account</td>\n",
       "      <td>6</td>\n",
       "      <td>Paid Up</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>343</td>\n",
       "      <td>None</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>27.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Credit_Application_Result Account_Balance  Duration_of_Credit_Month  \\\n",
       "0              Creditworthy    Some Balance                         4   \n",
       "1              Creditworthy    Some Balance                         4   \n",
       "2              Creditworthy    Some Balance                         4   \n",
       "3              Creditworthy    Some Balance                         4   \n",
       "4              Creditworthy      No Account                         6   \n",
       "\n",
       "  Payment_Status_of_Previous_Credit       Purpose  Credit_Amount  \\\n",
       "0                           Paid Up         Other           1494   \n",
       "1                           Paid Up  Home Related           1494   \n",
       "2        No Problems (in this bank)  Home Related           1544   \n",
       "3        No Problems (in this bank)  Home Related           3380   \n",
       "4                           Paid Up  Home Related            343   \n",
       "\n",
       "  Value_Savings_Stocks Length_of_current_employment  Instalment_per_cent  \\\n",
       "0           £100-£1000                        < 1yr                    1   \n",
       "1           £100-£1000                        < 1yr                    1   \n",
       "2                 None                      1-4 yrs                    2   \n",
       "3                 None                      1-4 yrs                    1   \n",
       "4                 None                        < 1yr                    4   \n",
       "\n",
       "   Most_valuable_available_asset  Age_years Type_of_apartment  \\\n",
       "0                              1       33.0                 2   \n",
       "1                              1       29.0                 2   \n",
       "2                              1       42.0                 2   \n",
       "3                              1       37.0                 2   \n",
       "4                              1       27.0                 2   \n",
       "\n",
       "  No_of_Credits_at_this_Bank  \n",
       "0                          1  \n",
       "1                          1  \n",
       "2                More than 1  \n",
       "3                          1  \n",
       "4                          1  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 500 entries, 0 to 499\n",
      "Data columns (total 13 columns):\n",
      " #   Column                             Non-Null Count  Dtype  \n",
      "---  ------                             --------------  -----  \n",
      " 0   Credit_Application_Result          500 non-null    object \n",
      " 1   Account_Balance                    500 non-null    object \n",
      " 2   Duration_of_Credit_Month           500 non-null    int64  \n",
      " 3   Payment_Status_of_Previous_Credit  500 non-null    object \n",
      " 4   Purpose                            500 non-null    object \n",
      " 5   Credit_Amount                      500 non-null    int64  \n",
      " 6   Value_Savings_Stocks               500 non-null    object \n",
      " 7   Length_of_current_employment       500 non-null    object \n",
      " 8   Instalment_per_cent                500 non-null    int64  \n",
      " 9   Most_valuable_available_asset      500 non-null    int64  \n",
      " 10  Age_years                          500 non-null    float64\n",
      " 11  Type_of_apartment                  500 non-null    object \n",
      " 12  No_of_Credits_at_this_Bank         500 non-null    object \n",
      "dtypes: float64(1), int64(4), object(8)\n",
      "memory usage: 50.9+ KB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> Column names are changed with underscores between the words. All fields in the dataset do not contain missing values. Looks Nice. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='Exploration'></a>\n",
    "## Data Exploration with Visualizations\n",
    "\n",
    "> In this section, I explore each column individually deeper. To begin with, I compare the creditworthy customers with the non-creditworthy. Furthermore, for numeric input variables, I analysed their group means, and whether they are statistically significant. Several visualizations will help you to understand whether  the individual fields are associated with the target variables. <br>\n",
    "\n",
    "> **Contents of the Section**<br>\n",
    "    <ul>\n",
    "    <li><a href=\"#target\">1. Credit Application Result (Desired Target)</a></li>\n",
    "    <li><a href=\"#correlation\">2. Correlation between numeric variables</a></li>\n",
    "    <li><a href=\"#balance\">3. Account Balance</a></li>\n",
    "    <li><a href=\"#status\">4. Payment Status of Previous Credit</a></li>\n",
    "    <li><a href=\"#purpose\">5. Purpose</a></li>\n",
    "    <li><a href=\"#value\">6. Value Savings Stocks</a></li>\n",
    "    <li><a href=\"#length\">7. Length of current employment</a></li>\n",
    "    <li><a href=\"#apartment\">8. Types of Apartment</a></li>\n",
    "    <li><a href=\"#no_Credits\">9. Number of Credits at this Bank</a></li>\n",
    "    </ul>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='target'></a>\n",
    "### 1. Credit Application Result (Desired Target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Creditworthy        358\n",
       "Non-Creditworthy    142\n",
       "Name: Credit_Application_Result, dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Credit_Application_Result'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Counts')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "base_color =sns.color_palette()[0]\n",
    "sns.countplot(x='Credit_Application_Result',data = data, color =base_color)\n",
    "plt.xlabel('Credit Application Result')\n",
    "plt.ylabel('Counts')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- Proportion of Non-Creditworthy: 0.284\n",
      "- Proportion of Creditworthy: 0.716\n"
     ]
    }
   ],
   "source": [
    "non_creditworthy = len(data[data['Credit_Application_Result']=='Non-Creditworthy'])\n",
    "creditworthy = len(data[data['Credit_Application_Result']=='Creditworthy'])\n",
    "pct_of_non_creditworthy = non_creditworthy/(non_creditworthy+creditworthy)\n",
    "print(\"- Proportion of Non-Creditworthy:\", pct_of_non_creditworthy)\n",
    "pct_of_creditworthy = creditworthy/(non_creditworthy+creditworthy)\n",
    "print(\"- Proportion of Creditworthy:\", pct_of_creditworthy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> The classes are imbalanced, and the ratio of \"Creditworthy\" to \"Non-Creditworthy\" shows is **72:28**. Before I balanced the classes, some more explorations are conducted. I calculated the group means of numeric variables (Duration_of_Credit_Month, Credit_Amount, Instalment_per_cent, Most_valuable_available_asset, Age_years) for both groups and tested whether the group means are statistically different."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Group Means Comparisons"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Duration_of_Credit_Month</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Credit_Application_Result</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Creditworthy</th>\n",
       "      <td>19.865922</td>\n",
       "      <td>2840.229050</td>\n",
       "      <td>2.966480</td>\n",
       "      <td>2.265363</td>\n",
       "      <td>35.952514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Non-Creditworthy</th>\n",
       "      <td>25.387324</td>\n",
       "      <td>4106.957746</td>\n",
       "      <td>3.119718</td>\n",
       "      <td>2.598592</td>\n",
       "      <td>34.619718</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           Duration_of_Credit_Month  Credit_Amount  \\\n",
       "Credit_Application_Result                                            \n",
       "Creditworthy                              19.865922    2840.229050   \n",
       "Non-Creditworthy                          25.387324    4106.957746   \n",
       "\n",
       "                           Instalment_per_cent  Most_valuable_available_asset  \\\n",
       "Credit_Application_Result                                                       \n",
       "Creditworthy                          2.966480                       2.265363   \n",
       "Non-Creditworthy                      3.119718                       2.598592   \n",
       "\n",
       "                           Age_years  \n",
       "Credit_Application_Result             \n",
       "Creditworthy               35.952514  \n",
       "Non-Creditworthy           34.619718  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calcuate the group means.\n",
    "data.groupby('Credit_Application_Result').mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> - The average credit duration of \"Creditworthy\" customers are shorter than that of \"Non_Creditworthy\" customers. The difference is about 5.52 months. \n",
    "- The average credit amount of \"Creditworthy\" individuals are smaller than that of the \"Non_Creditworthy\". \"Non_Creditworthy\" customers take USD 1,266.73 more than the Creditworthy on average. \n",
    "- The variable \"Most_valuable_available_asset\" is numeric, but in discrete value. The average number of available assets that creditworthy customers own is ca. 2.27. Similarly. the Non_Creditworthy own ca. 2.60.\n",
    "- \"Age_years\" is in the same situation with \"Most_valuable_available_asset\". The average age of Creditworthy group is a little bit higher than that of Non_Creditworthy group.  \n",
    "- Since the differences of the two variables seem quite trivial, I test the mean difference using two samples independent t-test below. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Two Sample Independent T-tests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy import stats\n",
    "data_test=data.copy()\n",
    "worthy= data_test.query('Credit_Application_Result ==\"Creditworthy\"') # Filter out the creditworthy. \n",
    "non_worthy = data_test.query('Credit_Application_Result != \"Creditworthy\"') # Filter out the non-creditworthy. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=-4.347773395109759, pvalue=2.0650349072181e-05)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# T-test for \"Duration_of_Credit_Month\"\n",
    "stats.ttest_ind(worthy['Duration_of_Credit_Month'],non_worthy['Duration_of_Credit_Month'], equal_var = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=-3.72534707078571, pvalue=0.0002602378463940242)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# T-test for \"Credit_Amount\"\n",
    "stats.ttest_ind(worthy['Credit_Amount'],non_worthy['Credit_Amount'], equal_var = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=-1.392518904544035, pvalue=0.16495295478821406)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# T-test for \"Instalment_per_cent\"\n",
    "stats.ttest_ind(worthy['Instalment_per_cent'],non_worthy['Instalment_per_cent'], equal_var = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=-3.1659536259830956, pvalue=0.0017332335688768836)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# T-test for \"Most_valuable_available_asset\"\n",
    "stats.ttest_ind(worthy['Most_valuable_available_asset'],non_worthy['Most_valuable_available_asset'], equal_var = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=1.1613893683418994, pvalue=0.2465931246944281)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# T-test for \"Age_years\"\n",
    "stats.ttest_ind(worthy['Age_years'],non_worthy['Age_years'], equal_var = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> The T-tests results show that the mean differences for the columns: \"Duration_of_Credit_Month\", \"Credit_Amount\", \"Most_valuable_available_asset\" are statistically significant, with the alpha level 0.05. On the other hand, the average of \"Instalment_per_cent\", \"Age_years\" are not significantly different between the two groups. As the next step of data exploration, I checked the correlation between the input variables. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='correlation'></a>\n",
    "### 2. Correlation between Numeric Variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1692e419c88>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,6))\n",
    "sns.heatmap(data.corr(), annot = True,cmap=\"RdBu\", center=0, vmin=-1, vmax=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> It seems that \"Credit Amount\" and \"Duration of Credit Month\" are moderately (but not severely correlated) and positively correlated. [Correlation Reference.](https://www.dummies.com/education/math/statistics/how-to-interpret-a-correlation-coefficient-r/) Other than the two, none of remaining numerical fields have strong correlations (larger than 0.7) each other. <br>\n",
    "> Next, I calculated group means for other categorical variables in the dataset to get a more detailed sense of our data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='balance'></a>\n",
    "### 3. Account Balance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "No Account      262\n",
       "Some Balance    238\n",
       "Name: Account_Balance, dtype: int64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Account_Balance'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Duration_of_Credit_Month</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Account_Balance</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>No Account</th>\n",
       "      <td>22.618321</td>\n",
       "      <td>3560.492366</td>\n",
       "      <td>2.935115</td>\n",
       "      <td>2.366412</td>\n",
       "      <td>34.770992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Some Balance</th>\n",
       "      <td>20.130252</td>\n",
       "      <td>2803.113445</td>\n",
       "      <td>3.092437</td>\n",
       "      <td>2.352941</td>\n",
       "      <td>36.457983</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Duration_of_Credit_Month  Credit_Amount  Instalment_per_cent  \\\n",
       "Account_Balance                                                                 \n",
       "No Account                      22.618321    3560.492366             2.935115   \n",
       "Some Balance                    20.130252    2803.113445             3.092437   \n",
       "\n",
       "                 Most_valuable_available_asset  Age_years  \n",
       "Account_Balance                                            \n",
       "No Account                            2.366412  34.770992  \n",
       "Some Balance                          2.352941  36.457983  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('Account_Balance').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1]), <a list of 2 Text xticklabel objects>)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "pd.crosstab(data['Account_Balance'],data['Credit_Application_Result']).plot(kind='bar')\n",
    "plt.title('Credit Application Frequency for Account Balance')\n",
    "plt.xlabel('Account Balance')\n",
    "plt.ylabel('Frequency of Credit Application')\n",
    "plt.xticks(rotation =360)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> The frequency of Credit Application Results depends on whether each customer has some balance on this bank. Thus, the column 'Account_Balance' is a good predictor of the outcome variable."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='status'></a>\n",
    "### 4. Payment Status of Previous Credit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Paid Up                       260\n",
       "No Problems (in this bank)    204\n",
       "Some Problems                  36\n",
       "Name: Payment_Status_of_Previous_Credit, dtype: int64"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Payment_Status_of_Previous_Credit'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Duration_of_Credit_Month</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Payment_Status_of_Previous_Credit</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>No Problems (in this bank)</th>\n",
       "      <td>22.088235</td>\n",
       "      <td>3291.720588</td>\n",
       "      <td>3.058824</td>\n",
       "      <td>2.411765</td>\n",
       "      <td>38.147059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Paid Up</th>\n",
       "      <td>20.134615</td>\n",
       "      <td>3004.515385</td>\n",
       "      <td>2.965385</td>\n",
       "      <td>2.257692</td>\n",
       "      <td>33.757692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Some Problems</th>\n",
       "      <td>27.111111</td>\n",
       "      <td>4091.805556</td>\n",
       "      <td>3.055556</td>\n",
       "      <td>2.805556</td>\n",
       "      <td>34.111111</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   Duration_of_Credit_Month  Credit_Amount  \\\n",
       "Payment_Status_of_Previous_Credit                                            \n",
       "No Problems (in this bank)                        22.088235    3291.720588   \n",
       "Paid Up                                           20.134615    3004.515385   \n",
       "Some Problems                                     27.111111    4091.805556   \n",
       "\n",
       "                                   Instalment_per_cent  \\\n",
       "Payment_Status_of_Previous_Credit                        \n",
       "No Problems (in this bank)                    3.058824   \n",
       "Paid Up                                       2.965385   \n",
       "Some Problems                                 3.055556   \n",
       "\n",
       "                                   Most_valuable_available_asset  Age_years  \n",
       "Payment_Status_of_Previous_Credit                                            \n",
       "No Problems (in this bank)                              2.411765  38.147059  \n",
       "Paid Up                                                 2.257692  33.757692  \n",
       "Some Problems                                           2.805556  34.111111  "
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('Payment_Status_of_Previous_Credit').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1692e558b48>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "table=pd.crosstab(data['Payment_Status_of_Previous_Credit'],data['Credit_Application_Result'])\n",
    "table.div(table.sum(1).astype(float), axis=0).plot(kind='bar', stacked=True)\n",
    "plt.title('Stacked Bar Chart of Previous Payment Status vs Credit Approval')\n",
    "plt.xlabel('Payment Status of Previous Credit')\n",
    "plt.ylabel('Proportion of Credit Application Result')\n",
    "plt.xticks(rotation =360)\n",
    "plt.legend( bbox_to_anchor = (1, 0.5), loc = 6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> The proportion of Credit Application Results is different whether each credit applicant has some problems or not. The group of people either in 'No Problems (in this bank)' or in 'Paid up' shows little difference. Thus, it would be better that after integrating both groups then it should be used as a predictor of the outcome variable. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "data.Payment_Status_of_Previous_Credit.replace('No Problems (in this bank)', 'No Problems (in this bank) or Paid Up', inplace = True)\n",
    "data.Payment_Status_of_Previous_Credit.replace('Paid Up', 'No Problems (in this bank) or Paid Up', inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1692e5ead08>"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "table=pd.crosstab(data['Payment_Status_of_Previous_Credit'],data['Credit_Application_Result'])\n",
    "table.div(table.sum(1).astype(float), axis=0).plot(kind='bar', stacked=True)\n",
    "plt.title('Stacked Bar Chart of Previous Payment Status vs Credit Approval')\n",
    "plt.xlabel('Payment Status of Previous Credit')\n",
    "plt.ylabel('Proportion of Credit Application Result')\n",
    "plt.xticks(rotation =360)\n",
    "plt.legend( bbox_to_anchor = (1, 0.5), loc = 6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='purpose'></a>\n",
    "### 5. Purpose"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Home Related    355\n",
       "Used car         83\n",
       "New car          47\n",
       "Other            15\n",
       "Name: Purpose, dtype: int64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Purpose'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>Duration_of_Credit_Month</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Purpose</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Home Related</th>\n",
       "      <td>21.425352</td>\n",
       "      <td>3024.952113</td>\n",
       "      <td>3.090141</td>\n",
       "      <td>2.323944</td>\n",
       "      <td>35.963380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>New car</th>\n",
       "      <td>26.191489</td>\n",
       "      <td>5200.255319</td>\n",
       "      <td>2.680851</td>\n",
       "      <td>2.851064</td>\n",
       "      <td>37.936170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>20.000000</td>\n",
       "      <td>3271.066667</td>\n",
       "      <td>3.333333</td>\n",
       "      <td>2.066667</td>\n",
       "      <td>34.600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Used car</th>\n",
       "      <td>19.036145</td>\n",
       "      <td>2803.060241</td>\n",
       "      <td>2.795181</td>\n",
       "      <td>2.289157</td>\n",
       "      <td>32.746988</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Duration_of_Credit_Month  Credit_Amount  Instalment_per_cent  \\\n",
       "Purpose                                                                      \n",
       "Home Related                 21.425352    3024.952113             3.090141   \n",
       "New car                      26.191489    5200.255319             2.680851   \n",
       "Other                        20.000000    3271.066667             3.333333   \n",
       "Used car                     19.036145    2803.060241             2.795181   \n",
       "\n",
       "              Most_valuable_available_asset  Age_years  \n",
       "Purpose                                                 \n",
       "Home Related                       2.323944  35.963380  \n",
       "New car                            2.851064  37.936170  \n",
       "Other                              2.066667  34.600000  \n",
       "Used car                           2.289157  32.746988  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('Purpose').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Frequency of Credit Application')"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "purpose_order=['Home Related', 'Used car','New car', 'Other']\n",
    "sns.countplot(data = data, x = 'Purpose', hue = 'Credit_Application_Result',order = purpose_order)\n",
    "plt.title('Credit Application Frequency for Purpose')\n",
    "plt.xlabel('Purpose')\n",
    "plt.ylabel('Frequency of Credit Application')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> The frequency of Credit Application Results depends on the purpose of credits. Thus, the column 'Purpose' is a good predictor of the outcome variable."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='value'></a>\n",
    "### 6. Value Savings Stocks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "None          298\n",
       "£100-£1000    154\n",
       "< £100         48\n",
       "Name: Value_Savings_Stocks, dtype: int64"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Value_Savings_Stocks'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>Duration_of_Credit_Month</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Value_Savings_Stocks</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>&lt; £100</th>\n",
       "      <td>23.208333</td>\n",
       "      <td>3272.187500</td>\n",
       "      <td>3.104167</td>\n",
       "      <td>2.729167</td>\n",
       "      <td>33.895833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>None</th>\n",
       "      <td>21.265101</td>\n",
       "      <td>3214.137584</td>\n",
       "      <td>2.986577</td>\n",
       "      <td>2.328859</td>\n",
       "      <td>35.338926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>£100-£1000</th>\n",
       "      <td>21.207792</td>\n",
       "      <td>3150.077922</td>\n",
       "      <td>3.025974</td>\n",
       "      <td>2.305195</td>\n",
       "      <td>36.551948</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Duration_of_Credit_Month  Credit_Amount  \\\n",
       "Value_Savings_Stocks                                            \n",
       "< £100                               23.208333    3272.187500   \n",
       "None                                 21.265101    3214.137584   \n",
       "£100-£1000                           21.207792    3150.077922   \n",
       "\n",
       "                      Instalment_per_cent  Most_valuable_available_asset  \\\n",
       "Value_Savings_Stocks                                                       \n",
       "< £100                           3.104167                       2.729167   \n",
       "None                             2.986577                       2.328859   \n",
       "£100-£1000                       3.025974                       2.305195   \n",
       "\n",
       "                      Age_years  \n",
       "Value_Savings_Stocks             \n",
       "< £100                33.895833  \n",
       "None                  35.338926  \n",
       "£100-£1000            36.551948  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('Value_Savings_Stocks').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['£100-£1000', 'None', '< £100'], dtype=object)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Value_Savings_Stocks'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Proportion of Credit Application Results')"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "stocks_order = ['None', '< £100', '£100-£1000' ] # The order of Value_Savings_Stocks\n",
    "credit_order = ['Creditworthy', 'Non-Creditworthy'] # The order of Credit_Application_Result\n",
    "\n",
    "artists = [] # For storing references to plot elements\n",
    "baselines = np.zeros(len(stocks_order))\n",
    "stocks_counts = data['Value_Savings_Stocks'].value_counts()\n",
    "\n",
    "# For each second-variable category:\n",
    "for i in range(len(credit_order)):\n",
    "    # isolate the counts of the first category,\n",
    "    cat_credit= credit_order[i]\n",
    "    inner_counts = data[data['Credit_Application_Result'] == cat_credit]['Value_Savings_Stocks'].value_counts()\n",
    "    inner_props = inner_counts / stocks_counts\n",
    "    # then plot those counts on top of the accumulated baseline\n",
    "    bars = plt.bar(x = np.arange(len(stocks_order)),\n",
    "                   height = inner_props[stocks_order],\n",
    "                   bottom = baselines)\n",
    "    artists.append(bars)\n",
    "    baselines += inner_props[stocks_order]\n",
    "\n",
    "plt.xticks(np.arange(len(stocks_order)), stocks_order)\n",
    "plt.legend(reversed(artists), reversed(credit_order), framealpha = 1,\n",
    "           bbox_to_anchor = (1, 0.5), loc = 6);\n",
    "\n",
    "plt.title('Credit Application Frequency for Savings Stocks')\n",
    "plt.xlabel('Value Savings Stocks')\n",
    "plt.ylabel('Proportion of Credit Application Results')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> We should note that the column \"Value_Savings_Stocks\" is a categorical variable, and also the ordinal variable. The proportion of Creditworthy gets increasing as the value of saving stocks larger. The column is a good predictor of the outcome variable. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='length'></a>\n",
    "### 7. Length of current employment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "< 1yr      279\n",
       "4-7 yrs    124\n",
       "1-4 yrs     97\n",
       "Name: Length_of_current_employment, dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Length_of_current_employment'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Duration_of_Credit_Month</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Length_of_current_employment</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1-4 yrs</th>\n",
       "      <td>21.402062</td>\n",
       "      <td>2835.649485</td>\n",
       "      <td>3.051546</td>\n",
       "      <td>2.402062</td>\n",
       "      <td>32.639175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4-7 yrs</th>\n",
       "      <td>21.669355</td>\n",
       "      <td>3168.129032</td>\n",
       "      <td>3.306452</td>\n",
       "      <td>2.604839</td>\n",
       "      <td>43.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>&lt; 1yr</th>\n",
       "      <td>21.340502</td>\n",
       "      <td>3340.802867</td>\n",
       "      <td>2.863799</td>\n",
       "      <td>2.236559</td>\n",
       "      <td>33.293907</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              Duration_of_Credit_Month  Credit_Amount  \\\n",
       "Length_of_current_employment                                            \n",
       "1-4 yrs                                      21.402062    2835.649485   \n",
       "4-7 yrs                                      21.669355    3168.129032   \n",
       "< 1yr                                        21.340502    3340.802867   \n",
       "\n",
       "                              Instalment_per_cent  \\\n",
       "Length_of_current_employment                        \n",
       "1-4 yrs                                  3.051546   \n",
       "4-7 yrs                                  3.306452   \n",
       "< 1yr                                    2.863799   \n",
       "\n",
       "                              Most_valuable_available_asset  Age_years  \n",
       "Length_of_current_employment                                            \n",
       "1-4 yrs                                            2.402062  32.639175  \n",
       "4-7 yrs                                            2.604839  43.000000  \n",
       "< 1yr                                              2.236559  33.293907  "
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('Length_of_current_employment').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['< 1yr', '1-4 yrs', '4-7 yrs'], dtype=object)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Length_of_current_employment'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Proportion of Credit Application Results')"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "length_order = ['< 1yr', '1-4 yrs', '4-7 yrs'] # The order of Length_of_current_employment\n",
    "credit_order = ['Creditworthy', 'Non-Creditworthy'] # The order of Credit_Application_Result\n",
    "\n",
    "artists = [] # for storing references to plot elements\n",
    "baselines = np.zeros(len(length_order))\n",
    "length_counts = data['Length_of_current_employment'].value_counts()\n",
    "\n",
    "# for each second-variable category:\n",
    "for i in range(len(credit_order)):\n",
    "    # isolate the counts of the first category,\n",
    "    credit = credit_order[i]\n",
    "    inner_counts = data[data['Credit_Application_Result'] == credit]['Length_of_current_employment'].value_counts()\n",
    "    inner_props = inner_counts / length_counts\n",
    "    # then plot those counts on top of the accumulated baseline\n",
    "    bars = plt.bar(x = np.arange(len(length_order)),\n",
    "                   height = inner_props[length_order],\n",
    "                   bottom = baselines)\n",
    "    artists.append(bars)\n",
    "    baselines += inner_props[length_order]\n",
    "\n",
    "plt.xticks(np.arange(len(length_order)), length_order)\n",
    "plt.legend(reversed(artists), reversed(credit_order), framealpha = 1,\n",
    "           bbox_to_anchor = (1, 0.5), loc = 6);\n",
    "\n",
    "plt.title('Credit Application Frequency for Employment Length')\n",
    "plt.xlabel('Length of Current Employment')\n",
    "plt.ylabel('Proportion of Credit Application Results')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> The column \"Length_of_current_employment\" is a categorical variable, and also the ordinal variable. Although any particular tendency can be recognized from the visualizations, we can notice that the percentage of being creditworthy for employment less than 1 year is weigh lower than the customers with longer employment. Thus, it would be better using  the input as a predictor after integrating  both groups (i.e. '1-4 yrs' and '4-7 yrs')."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.Length_of_current_employment.replace('1-4 yrs', 'longer than 1 yrs employment', inplace = True)\n",
    "data.Length_of_current_employment.replace('4-7 yrs', 'longer than 1 yrs employment', inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1692ec5d788>"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "table=pd.crosstab(data['Length_of_current_employment'],data['Credit_Application_Result'])\n",
    "table.div(table.sum(1).astype(float), axis=0).plot(kind='bar', stacked=True)\n",
    "plt.title('Credit Application Frequency for Employment Length')\n",
    "plt.xlabel('Length of Current Employment')\n",
    "plt.ylabel('Proportion of Credit Application Result')\n",
    "plt.xticks(rotation =360)\n",
    "plt.legend( bbox_to_anchor = (1, 0.5), loc = 6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='apartment'></a>\n",
    "### 8. Types of Apartment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2    352\n",
       "1     92\n",
       "3     56\n",
       "Name: Type_of_apartment, dtype: int64"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Type_of_apartment'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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>Duration_of_Credit_Month</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Type_of_apartment</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>20.293478</td>\n",
       "      <td>3072.934783</td>\n",
       "      <td>2.869565</td>\n",
       "      <td>2.173913</td>\n",
       "      <td>30.434783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>20.593750</td>\n",
       "      <td>2897.894886</td>\n",
       "      <td>3.019886</td>\n",
       "      <td>2.153409</td>\n",
       "      <td>35.394886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>28.589286</td>\n",
       "      <td>5307.517857</td>\n",
       "      <td>3.178571</td>\n",
       "      <td>3.964286</td>\n",
       "      <td>45.142857</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Duration_of_Credit_Month  Credit_Amount  \\\n",
       "Type_of_apartment                                            \n",
       "1                                 20.293478    3072.934783   \n",
       "2                                 20.593750    2897.894886   \n",
       "3                                 28.589286    5307.517857   \n",
       "\n",
       "                   Instalment_per_cent  Most_valuable_available_asset  \\\n",
       "Type_of_apartment                                                       \n",
       "1                             2.869565                       2.173913   \n",
       "2                             3.019886                       2.153409   \n",
       "3                             3.178571                       3.964286   \n",
       "\n",
       "                   Age_years  \n",
       "Type_of_apartment             \n",
       "1                  30.434783  \n",
       "2                  35.394886  \n",
       "3                  45.142857  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('Type_of_apartment').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1692edb4848>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "table=pd.crosstab(data['Type_of_apartment'],data['Credit_Application_Result'])\n",
    "table.div(table.sum(1).astype(float), axis=0).plot(kind='bar', stacked=True)\n",
    "plt.title('Credit Application Frequency for Apartment Types')\n",
    "plt.xlabel('Apartment Type')\n",
    "plt.ylabel('Proportion of Credit Application Result')\n",
    "plt.xticks(rotation =360)\n",
    "plt.legend( bbox_to_anchor = (1, 0.5), loc = 6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> The detail information about each type of apartment type is unknown. However according to the data, we can speculate that Type3 is more expensive apartment comparing to others. The average credit durations of Type3 is longer, the average credit amount of the type is much larger, and therefore its instalment percentage is higher. Average age of customers whose apartment type is '3' is 10-15 years older than the customers in Apartment Type 2 or 3."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='no_Credits'></a>\n",
    "### 9. Number of Credits at this Bank"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1              320\n",
       "More than 1    180\n",
       "Name: No_of_Credits_at_this_Bank, dtype: int64"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['No_of_Credits_at_this_Bank'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Duration_of_Credit_Month</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>No_of_Credits_at_this_Bank</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>21.615625</td>\n",
       "      <td>3314.081250</td>\n",
       "      <td>2.962500</td>\n",
       "      <td>2.340625</td>\n",
       "      <td>34.515625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>More than 1</th>\n",
       "      <td>21.111111</td>\n",
       "      <td>2997.133333</td>\n",
       "      <td>3.094444</td>\n",
       "      <td>2.394444</td>\n",
       "      <td>37.455556</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            Duration_of_Credit_Month  Credit_Amount  \\\n",
       "No_of_Credits_at_this_Bank                                            \n",
       "1                                          21.615625    3314.081250   \n",
       "More than 1                                21.111111    2997.133333   \n",
       "\n",
       "                            Instalment_per_cent  \\\n",
       "No_of_Credits_at_this_Bank                        \n",
       "1                                      2.962500   \n",
       "More than 1                            3.094444   \n",
       "\n",
       "                            Most_valuable_available_asset  Age_years  \n",
       "No_of_Credits_at_this_Bank                                            \n",
       "1                                                2.340625  34.515625  \n",
       "More than 1                                      2.394444  37.455556  "
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('No_of_Credits_at_this_Bank').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1692edb5788>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "table=pd.crosstab(data['No_of_Credits_at_this_Bank'],data['Credit_Application_Result'])\n",
    "table.div(table.sum(1).astype(float), axis=0).plot(kind='bar', stacked=True)\n",
    "plt.title('Stacked Bar Chart of Number of Credits (at this Bank) vs Credit Approval')\n",
    "plt.xlabel('Number of Credits at this Bank')\n",
    "plt.ylabel('Proportion of Credit Application Result')\n",
    "plt.xticks(rotation =360)\n",
    "plt.legend( bbox_to_anchor = (1, 0.5), loc = 6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> Finally I compare the customers who has only one credit with the customers with more than one credits at this bank. The customers with one credit only take more credit amount on average with lower instalment percentage. On the other hand, customers with more than one credits at this bank are more likely to be determined as creditworthy customers. Their available assets are a little bit higher than the other. This feature is regarded as a good predictor of the target variable."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> So far, we've conducted Data Cleaning and Data Exploration. Based on the cleaned dataset, I will continue the analysis using classification models in Part2. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Credit_Application_Result</th>\n",
       "      <th>Account_Balance</th>\n",
       "      <th>Duration_of_Credit_Month</th>\n",
       "      <th>Payment_Status_of_Previous_Credit</th>\n",
       "      <th>Purpose</th>\n",
       "      <th>Credit_Amount</th>\n",
       "      <th>Value_Savings_Stocks</th>\n",
       "      <th>Length_of_current_employment</th>\n",
       "      <th>Instalment_per_cent</th>\n",
       "      <th>Most_valuable_available_asset</th>\n",
       "      <th>Age_years</th>\n",
       "      <th>Type_of_apartment</th>\n",
       "      <th>No_of_Credits_at_this_Bank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>No Problems (in this bank) or Paid Up</td>\n",
       "      <td>Other</td>\n",
       "      <td>1494</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>33.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>No Problems (in this bank) or Paid Up</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>1494</td>\n",
       "      <td>£100-£1000</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>29.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>No Problems (in this bank) or Paid Up</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>1544</td>\n",
       "      <td>None</td>\n",
       "      <td>longer than 1 yrs employment</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>42.0</td>\n",
       "      <td>2</td>\n",
       "      <td>More than 1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>Some Balance</td>\n",
       "      <td>4</td>\n",
       "      <td>No Problems (in this bank) or Paid Up</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>3380</td>\n",
       "      <td>None</td>\n",
       "      <td>longer than 1 yrs employment</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>37.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Creditworthy</td>\n",
       "      <td>No Account</td>\n",
       "      <td>6</td>\n",
       "      <td>No Problems (in this bank) or Paid Up</td>\n",
       "      <td>Home Related</td>\n",
       "      <td>343</td>\n",
       "      <td>None</td>\n",
       "      <td>&lt; 1yr</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>27.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Credit_Application_Result Account_Balance  Duration_of_Credit_Month  \\\n",
       "0              Creditworthy    Some Balance                         4   \n",
       "1              Creditworthy    Some Balance                         4   \n",
       "2              Creditworthy    Some Balance                         4   \n",
       "3              Creditworthy    Some Balance                         4   \n",
       "4              Creditworthy      No Account                         6   \n",
       "\n",
       "       Payment_Status_of_Previous_Credit       Purpose  Credit_Amount  \\\n",
       "0  No Problems (in this bank) or Paid Up         Other           1494   \n",
       "1  No Problems (in this bank) or Paid Up  Home Related           1494   \n",
       "2  No Problems (in this bank) or Paid Up  Home Related           1544   \n",
       "3  No Problems (in this bank) or Paid Up  Home Related           3380   \n",
       "4  No Problems (in this bank) or Paid Up  Home Related            343   \n",
       "\n",
       "  Value_Savings_Stocks  Length_of_current_employment  Instalment_per_cent  \\\n",
       "0           £100-£1000                         < 1yr                    1   \n",
       "1           £100-£1000                         < 1yr                    1   \n",
       "2                 None  longer than 1 yrs employment                    2   \n",
       "3                 None  longer than 1 yrs employment                    1   \n",
       "4                 None                         < 1yr                    4   \n",
       "\n",
       "   Most_valuable_available_asset  Age_years Type_of_apartment  \\\n",
       "0                              1       33.0                 2   \n",
       "1                              1       29.0                 2   \n",
       "2                              1       42.0                 2   \n",
       "3                              1       37.0                 2   \n",
       "4                              1       27.0                 2   \n",
       "\n",
       "  No_of_Credits_at_this_Bank  \n",
       "0                          1  \n",
       "1                          1  \n",
       "2                More than 1  \n",
       "3                          1  \n",
       "4                          1  "
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.to_excel('credit-data-training-cleaned.xlsx', index = False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}