Day-3-Practice.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "view-in-github"
   },
   "source": [
    "<a href=\"https://colab.research.google.com/github/stefanlessmann/ASE-ML/blob/main/Day-3-Practice.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "grvYBSqLwztN"
   },
   "source": [
    "# Wecome to our Section 12 Deep Learning Practice Session\n",
    "Today, we put everything together studied thus far and develop a fully-fledged pipeline for training, testing, and model selecting a neural network classifier to predict the risk of credit default. \n",
    "\n",
    "The outline of today's session is as follows:\n",
    "- Load real-world credit data from GitHub\n",
    "- Eyeballing the data using the `Pandas` library\n",
    "- Perform basic data preprocessing\n",
    "- Partition the data into training a testing data\n",
    "- Train and assess a neural network classifier. For assessment, compute the classification accuracy of your trained network.\n",
    "- Optional task for expert: Tuning the architecture of the neural network\n",
    "\n",
    "Much work to do, so let's go!\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "executionInfo": {
     "elapsed": 1829,
     "status": "ok",
     "timestamp": 1680268950925,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "h6vhKogUqLQY"
   },
   "outputs": [],
   "source": [
    "# Standard libraries for data data handling and plotting\n",
    "import numpy as np \n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "sbbga7IizBfT"
   },
   "source": [
    "## Load real-world credit data from GitHub\n",
    "The data set we use for this session is available at the following URL: <br>\n",
    "https://raw.githubusercontent.com/stefanlessmann/ASE-ML/master/hmeq.csv\n",
    "\n",
    "Create a variable named `data_url` in which you store this URL. Next, use the method `read_csv()` from the `Pandas` library to download the data right from the web and store it in a `DataFrame`. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "executionInfo": {
     "elapsed": 15,
     "status": "ok",
     "timestamp": 1680268950927,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "d16WAzKPuCFf",
    "outputId": "c97fd64b-d646-4981-e99d-4c3790acfe62"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-99ea0857-4fec-49a9-aa80-20151838b0f3\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>BAD</th>\n",
       "      <th>LOAN</th>\n",
       "      <th>MORTDUE</th>\n",
       "      <th>VALUE</th>\n",
       "      <th>REASON</th>\n",
       "      <th>JOB</th>\n",
       "      <th>YOJ</th>\n",
       "      <th>DEROG</th>\n",
       "      <th>DELINQ</th>\n",
       "      <th>CLAGE</th>\n",
       "      <th>NINQ</th>\n",
       "      <th>CLNO</th>\n",
       "      <th>DEBTINC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1100</td>\n",
       "      <td>25860.0</td>\n",
       "      <td>39025.0</td>\n",
       "      <td>HomeImp</td>\n",
       "      <td>Other</td>\n",
       "      <td>10.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>94.366667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1300</td>\n",
       "      <td>70053.0</td>\n",
       "      <td>68400.0</td>\n",
       "      <td>HomeImp</td>\n",
       "      <td>Other</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>121.833333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1500</td>\n",
       "      <td>13500.0</td>\n",
       "      <td>16700.0</td>\n",
       "      <td>HomeImp</td>\n",
       "      <td>Other</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>149.466667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1700</td>\n",
       "      <td>97800.0</td>\n",
       "      <td>112000.0</td>\n",
       "      <td>HomeImp</td>\n",
       "      <td>Office</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>93.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5955</th>\n",
       "      <td>0</td>\n",
       "      <td>88900</td>\n",
       "      <td>57264.0</td>\n",
       "      <td>90185.0</td>\n",
       "      <td>DebtCon</td>\n",
       "      <td>Other</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>221.808718</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>36.112347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5956</th>\n",
       "      <td>0</td>\n",
       "      <td>89000</td>\n",
       "      <td>54576.0</td>\n",
       "      <td>92937.0</td>\n",
       "      <td>DebtCon</td>\n",
       "      <td>Other</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>208.692070</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>35.859971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5957</th>\n",
       "      <td>0</td>\n",
       "      <td>89200</td>\n",
       "      <td>54045.0</td>\n",
       "      <td>92924.0</td>\n",
       "      <td>DebtCon</td>\n",
       "      <td>Other</td>\n",
       "      <td>15.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>212.279697</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>35.556590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5958</th>\n",
       "      <td>0</td>\n",
       "      <td>89800</td>\n",
       "      <td>50370.0</td>\n",
       "      <td>91861.0</td>\n",
       "      <td>DebtCon</td>\n",
       "      <td>Other</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>213.892709</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>34.340882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5959</th>\n",
       "      <td>0</td>\n",
       "      <td>89900</td>\n",
       "      <td>48811.0</td>\n",
       "      <td>88934.0</td>\n",
       "      <td>DebtCon</td>\n",
       "      <td>Other</td>\n",
       "      <td>15.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>219.601002</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>34.571519</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5960 rows × 13 columns</p>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-99ea0857-4fec-49a9-aa80-20151838b0f3')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-99ea0857-4fec-49a9-aa80-20151838b0f3 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-99ea0857-4fec-49a9-aa80-20151838b0f3');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "      BAD   LOAN  MORTDUE     VALUE   REASON     JOB   YOJ  DEROG  DELINQ  \\\n",
       "0       1   1100  25860.0   39025.0  HomeImp   Other  10.5    0.0     0.0   \n",
       "1       1   1300  70053.0   68400.0  HomeImp   Other   7.0    0.0     2.0   \n",
       "2       1   1500  13500.0   16700.0  HomeImp   Other   4.0    0.0     0.0   \n",
       "3       1   1500      NaN       NaN      NaN     NaN   NaN    NaN     NaN   \n",
       "4       0   1700  97800.0  112000.0  HomeImp  Office   3.0    0.0     0.0   \n",
       "...   ...    ...      ...       ...      ...     ...   ...    ...     ...   \n",
       "5955    0  88900  57264.0   90185.0  DebtCon   Other  16.0    0.0     0.0   \n",
       "5956    0  89000  54576.0   92937.0  DebtCon   Other  16.0    0.0     0.0   \n",
       "5957    0  89200  54045.0   92924.0  DebtCon   Other  15.0    0.0     0.0   \n",
       "5958    0  89800  50370.0   91861.0  DebtCon   Other  14.0    0.0     0.0   \n",
       "5959    0  89900  48811.0   88934.0  DebtCon   Other  15.0    0.0     0.0   \n",
       "\n",
       "           CLAGE  NINQ  CLNO    DEBTINC  \n",
       "0      94.366667   1.0   9.0        NaN  \n",
       "1     121.833333   0.0  14.0        NaN  \n",
       "2     149.466667   1.0  10.0        NaN  \n",
       "3            NaN   NaN   NaN        NaN  \n",
       "4      93.333333   0.0  14.0        NaN  \n",
       "...          ...   ...   ...        ...  \n",
       "5955  221.808718   0.0  16.0  36.112347  \n",
       "5956  208.692070   0.0  15.0  35.859971  \n",
       "5957  212.279697   0.0  15.0  35.556590  \n",
       "5958  213.892709   0.0  16.0  34.340882  \n",
       "5959  219.601002   0.0  16.0  34.571519  \n",
       "\n",
       "[5960 rows x 13 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Enter code to download the demo data set from the web \n",
    "data_url = \"https://raw.githubusercontent.com/stefanlessmann/ASE-ML/master/hmeq.csv\"\n",
    "df = pd.read_csv(data_url)\n",
    "\n",
    "# Taking a first glance at the data\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "rqvOwIik6WIC"
   },
   "source": [
    "#### Introducing the HMEQ data set (you can skip this section)\n",
    "Our data set, called the  \"Home Equity\" or, in brief, HMEQ data set, is provided by www.creditriskanalytics.net. It comprises  information about a set of borrowers, which are categorized along demographic variables and variables concerning their business relationship with the lender. A binary target variable called 'BAD' is  provided and indicates whether a borrower has repaid her/his debt. You can think of the data as a standard use case of binary classification.\n",
    "\n",
    "You obtain the data, together with other interesting finance data sets, directly from www.creditriskanalytics.net. The website also provides a brief description of the data set. Specifically, the data set consists of 5,960 observations and 13 features including the target variable. The variables are defined as follows:\n",
    "\n",
    "- BAD: the target variable, 1=default; 0=non-default \n",
    "- LOAN: amount of the loan request\n",
    "- MORTDUE: amount due on an existing mortgage\n",
    "- VALUE: value of current property\n",
    "- REASON: DebtCon=debt consolidation; HomeImp=home improvement\n",
    "- JOB: occupational categories\n",
    "- YOJ: years at present job\n",
    "- DEROG: number of major derogatory reports\n",
    "- DELINQ: number of delinquent credit lines\n",
    "- CLAGE: age of oldest credit line in months\n",
    "- NINQ: number of recent credit inquiries\n",
    "- CLNO: number of credit lines\n",
    "- DEBTINC: debt-to-income ratio\n",
    "\n",
    "As you can see, the features aim at describing the financial situation of a borrower. We will keep using the data set for many modeling tasks in this demo notebook and future demo notebook. So it makes sense to familiarize yourself with the above features. Make sure you understand what type of information they provide and what this information might reveal about the risk of defaulting.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9kBNd8886D_o"
   },
   "source": [
    "## Eyeballing data using the Pandas library\n",
    "In this part, we briefly demonstrate some practices for obtaining a first intuition about a data set. Let's first identify some standard methods that the `Pandas` library provides for this purpose. This is not meant to give you a fully-comprehensive list. With that disclaimer, however, it is fair to say that you will apply all of the following methods to a new data frame.  \n",
    "- `info()` \n",
    "- `describe()`\n",
    "- `head()` / `tail()`\n",
    "\n",
    "Try them out and inspect the output. Briefly summarize your findings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 455,
     "status": "ok",
     "timestamp": 1680268951372,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "HXKMAQNV_y1Z",
    "outputId": "e2ab7fe3-a8d4-4dbf-83ba-485861c4a923"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5960 entries, 0 to 5959\n",
      "Data columns (total 13 columns):\n",
      " #   Column   Non-Null Count  Dtype  \n",
      "---  ------   --------------  -----  \n",
      " 0   BAD      5960 non-null   int64  \n",
      " 1   LOAN     5960 non-null   int64  \n",
      " 2   MORTDUE  5442 non-null   float64\n",
      " 3   VALUE    5848 non-null   float64\n",
      " 4   REASON   5708 non-null   object \n",
      " 5   JOB      5681 non-null   object \n",
      " 6   YOJ      5445 non-null   float64\n",
      " 7   DEROG    5252 non-null   float64\n",
      " 8   DELINQ   5380 non-null   float64\n",
      " 9   CLAGE    5652 non-null   float64\n",
      " 10  NINQ     5450 non-null   float64\n",
      " 11  CLNO     5738 non-null   float64\n",
      " 12  DEBTINC  4693 non-null   float64\n",
      "dtypes: float64(9), int64(2), object(2)\n",
      "memory usage: 605.4+ KB\n"
     ]
    }
   ],
   "source": [
    "# Standard methods to exploring a DataFrame \n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 300
    },
    "executionInfo": {
     "elapsed": 19,
     "status": "ok",
     "timestamp": 1680268951373,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "AvJd8Ol0_tlS",
    "outputId": "b9e8efc8-bff8-43bf-d0e8-f500e47dcdf0"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-a1df16c2-a68a-432a-b7a4-448b60c6e19b\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>BAD</th>\n",
       "      <th>LOAN</th>\n",
       "      <th>MORTDUE</th>\n",
       "      <th>VALUE</th>\n",
       "      <th>YOJ</th>\n",
       "      <th>DEROG</th>\n",
       "      <th>DELINQ</th>\n",
       "      <th>CLAGE</th>\n",
       "      <th>NINQ</th>\n",
       "      <th>CLNO</th>\n",
       "      <th>DEBTINC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5442.000000</td>\n",
       "      <td>5848.000000</td>\n",
       "      <td>5445.000000</td>\n",
       "      <td>5252.000000</td>\n",
       "      <td>5380.000000</td>\n",
       "      <td>5652.000000</td>\n",
       "      <td>5450.000000</td>\n",
       "      <td>5738.000000</td>\n",
       "      <td>4693.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.199497</td>\n",
       "      <td>18607.969799</td>\n",
       "      <td>73760.817200</td>\n",
       "      <td>101776.048741</td>\n",
       "      <td>8.922268</td>\n",
       "      <td>0.254570</td>\n",
       "      <td>0.449442</td>\n",
       "      <td>179.766275</td>\n",
       "      <td>1.186055</td>\n",
       "      <td>21.296096</td>\n",
       "      <td>33.779915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.399656</td>\n",
       "      <td>11207.480417</td>\n",
       "      <td>44457.609458</td>\n",
       "      <td>57385.775334</td>\n",
       "      <td>7.573982</td>\n",
       "      <td>0.846047</td>\n",
       "      <td>1.127266</td>\n",
       "      <td>85.810092</td>\n",
       "      <td>1.728675</td>\n",
       "      <td>10.138933</td>\n",
       "      <td>8.601746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1100.000000</td>\n",
       "      <td>2063.000000</td>\n",
       "      <td>8000.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.524499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>11100.000000</td>\n",
       "      <td>46276.000000</td>\n",
       "      <td>66075.500000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>115.116702</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>29.140031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>16300.000000</td>\n",
       "      <td>65019.000000</td>\n",
       "      <td>89235.500000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>173.466667</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>34.818262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>23300.000000</td>\n",
       "      <td>91488.000000</td>\n",
       "      <td>119824.250000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>231.562278</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>39.003141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>89900.000000</td>\n",
       "      <td>399550.000000</td>\n",
       "      <td>855909.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>1168.233561</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>71.000000</td>\n",
       "      <td>203.312149</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a1df16c2-a68a-432a-b7a4-448b60c6e19b')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-a1df16c2-a68a-432a-b7a4-448b60c6e19b button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-a1df16c2-a68a-432a-b7a4-448b60c6e19b');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "               BAD          LOAN        MORTDUE          VALUE          YOJ  \\\n",
       "count  5960.000000   5960.000000    5442.000000    5848.000000  5445.000000   \n",
       "mean      0.199497  18607.969799   73760.817200  101776.048741     8.922268   \n",
       "std       0.399656  11207.480417   44457.609458   57385.775334     7.573982   \n",
       "min       0.000000   1100.000000    2063.000000    8000.000000     0.000000   \n",
       "25%       0.000000  11100.000000   46276.000000   66075.500000     3.000000   \n",
       "50%       0.000000  16300.000000   65019.000000   89235.500000     7.000000   \n",
       "75%       0.000000  23300.000000   91488.000000  119824.250000    13.000000   \n",
       "max       1.000000  89900.000000  399550.000000  855909.000000    41.000000   \n",
       "\n",
       "             DEROG       DELINQ        CLAGE         NINQ         CLNO  \\\n",
       "count  5252.000000  5380.000000  5652.000000  5450.000000  5738.000000   \n",
       "mean      0.254570     0.449442   179.766275     1.186055    21.296096   \n",
       "std       0.846047     1.127266    85.810092     1.728675    10.138933   \n",
       "min       0.000000     0.000000     0.000000     0.000000     0.000000   \n",
       "25%       0.000000     0.000000   115.116702     0.000000    15.000000   \n",
       "50%       0.000000     0.000000   173.466667     1.000000    20.000000   \n",
       "75%       0.000000     0.000000   231.562278     2.000000    26.000000   \n",
       "max      10.000000    15.000000  1168.233561    17.000000    71.000000   \n",
       "\n",
       "           DEBTINC  \n",
       "count  4693.000000  \n",
       "mean     33.779915  \n",
       "std       8.601746  \n",
       "min       0.524499  \n",
       "25%      29.140031  \n",
       "50%      34.818262  \n",
       "75%      39.003141  \n",
       "max     203.312149  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "executionInfo": {
     "elapsed": 17,
     "status": "ok",
     "timestamp": 1680268951374,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "SppRdAqGuQtd",
    "outputId": "a8a9d776-5015-45f2-f04b-a854b00a92c0"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-d609a2d1-9929-4809-9008-644301b4ec6d\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>BAD</th>\n",
       "      <th>LOAN</th>\n",
       "      <th>MORTDUE</th>\n",
       "      <th>VALUE</th>\n",
       "      <th>REASON</th>\n",
       "      <th>JOB</th>\n",
       "      <th>YOJ</th>\n",
       "      <th>DEROG</th>\n",
       "      <th>DELINQ</th>\n",
       "      <th>CLAGE</th>\n",
       "      <th>NINQ</th>\n",
       "      <th>CLNO</th>\n",
       "      <th>DEBTINC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1100</td>\n",
       "      <td>25860.0</td>\n",
       "      <td>39025.0</td>\n",
       "      <td>HomeImp</td>\n",
       "      <td>Other</td>\n",
       "      <td>10.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>94.366667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1300</td>\n",
       "      <td>70053.0</td>\n",
       "      <td>68400.0</td>\n",
       "      <td>HomeImp</td>\n",
       "      <td>Other</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>121.833333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1500</td>\n",
       "      <td>13500.0</td>\n",
       "      <td>16700.0</td>\n",
       "      <td>HomeImp</td>\n",
       "      <td>Other</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>149.466667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1700</td>\n",
       "      <td>97800.0</td>\n",
       "      <td>112000.0</td>\n",
       "      <td>HomeImp</td>\n",
       "      <td>Office</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>93.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d609a2d1-9929-4809-9008-644301b4ec6d')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-d609a2d1-9929-4809-9008-644301b4ec6d button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-d609a2d1-9929-4809-9008-644301b4ec6d');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "   BAD  LOAN  MORTDUE     VALUE   REASON     JOB   YOJ  DEROG  DELINQ  \\\n",
       "0    1  1100  25860.0   39025.0  HomeImp   Other  10.5    0.0     0.0   \n",
       "1    1  1300  70053.0   68400.0  HomeImp   Other   7.0    0.0     2.0   \n",
       "2    1  1500  13500.0   16700.0  HomeImp   Other   4.0    0.0     0.0   \n",
       "3    1  1500      NaN       NaN      NaN     NaN   NaN    NaN     NaN   \n",
       "4    0  1700  97800.0  112000.0  HomeImp  Office   3.0    0.0     0.0   \n",
       "\n",
       "        CLAGE  NINQ  CLNO  DEBTINC  \n",
       "0   94.366667   1.0   9.0      NaN  \n",
       "1  121.833333   0.0  14.0      NaN  \n",
       "2  149.466667   1.0  10.0      NaN  \n",
       "3         NaN   NaN   NaN      NaN  \n",
       "4   93.333333   0.0  14.0      NaN  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HXqdAZMZAKol"
   },
   "source": [
    "Creating outputs in the form of lists and tables is fine and can provide useful insights. However, you probably know the saying \"*a picture says more than a thousand words*\". Suitable graphics are often more insightful than (big) tables and can convey information in a more accessible way. Fortunately, `Pandas` offers many useful features to visualize the data within a DataFrame.\n",
    "\n",
    "Try  accomplish the following tasks. Web search using competendly selected search phrases will easily reveal the `Pandas` methods you need and how these are invoked. \n",
    "- Create a *histogram* for the the dependent (aka target) variable **BAD**\n",
    "- Create a density plot to depict the distribution of the variable **LOAN**\n",
    "- It is more common to visualize the distribution of numerical variables using *boxplots*. Considering again the variable, **LOAN** and depict its distribution using a *boxplot*\n",
    "- Create one last boxplot of the variable **LOAN**. This time, depict the distribution of loan amounts seperately for good and bad borrowers. To that end, use the target variable **BAD** and use it to add a grouping to your boxplot."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 435
    },
    "executionInfo": {
     "elapsed": 960,
     "status": "ok",
     "timestamp": 1680268952320,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "hbaaz89TBCv7",
    "outputId": "284cca85-b83d-4729-9b50-6215a4631d99"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a histogram for the the dependent (aka target) variable BAD\n",
    "df[\"BAD\"].plot.hist();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 445
    },
    "executionInfo": {
     "elapsed": 418,
     "status": "ok",
     "timestamp": 1680268952731,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "1T6u04_kBXS2",
    "outputId": "e8add1f5-00fb-4d3b-c0ca-2174e4ac724d"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a density plot to depict the distribution of the variable LOAN\n",
    "df['LOAN'].plot.density();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 430
    },
    "executionInfo": {
     "elapsed": 13,
     "status": "ok",
     "timestamp": 1680268952732,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "2YTuJy_wBXVu",
    "outputId": "e377608f-dfa5-4651-fab7-41ba4bad49ec"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a box plot to depict the distribution of the variable LOAN\n",
    "df['LOAN'].plot.box();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 452
    },
    "executionInfo": {
     "elapsed": 410,
     "status": "ok",
     "timestamp": 1680268953132,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "trsGDeANBXXf",
    "outputId": "2136095d-0678-4952-d1f4-87fd5fa91d18"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a box plot of the variable LOAN while grouping by the target\n",
    "df.plot.box(by=\"BAD\", column=\"LOAN\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9qRx_48h-Ydc"
   },
   "source": [
    "## Perform basic data preprocessing\n",
    "Data preparation is a topic of great importance. We could easily devote an entire lecture to this topic. Below, we sketch only a tiny little bit of the tasks typically performed in the scope of data preparation. \n",
    "\n",
    "In fact, we could not proceed with the data in its present form. For example, going back to the tabular preview of our data, you might notice that several variables show some missing values (denoted as NaN for *not a number* in the preview*). Leaving those missing values w/o treatment would break our neural network.\n",
    "\n",
    "Similarly, the data preview highlights that different variables show substantialyl different value ranges. Simply compare the values (or descriptive statistics also shown above) of the variables **LOAN** to those of **VALUE** to see this. For example, the maximal value of the variable **LOAN** is 89.900,00 whreas the largest value for the variable **VALUE** is 855.909,00 so almost ten times bigger. Differences in the value range of variables are normal but, unless suitably treated in data prepration, will also harm our neural network. \n",
    "\n",
    "Last, the data set includes some categorical variables. These would also need special treatment, which we do not have the time to detail. Thus, we simply delete them.\n",
    "\n",
    "In sum, we need to accomplish the following tasks:\n",
    "- Deleting the categorical variables **REASON** and **JOB**\n",
    "- Replacing all missing values in the data using *mean imputation*\n",
    "- Transforming the value range of all variables such that values are scaled between zero and one.\n",
    "\n",
    "Let's go..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 235,
     "status": "ok",
     "timestamp": 1680268953364,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "FbWoD2YsBXZf",
    "outputId": "ad1f9b2d-50c4-4b69-f78c-6b8662a2cc91"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5960 entries, 0 to 5959\n",
      "Data columns (total 11 columns):\n",
      " #   Column   Non-Null Count  Dtype  \n",
      "---  ------   --------------  -----  \n",
      " 0   BAD      5960 non-null   int64  \n",
      " 1   LOAN     5960 non-null   int64  \n",
      " 2   MORTDUE  5442 non-null   float64\n",
      " 3   VALUE    5848 non-null   float64\n",
      " 4   YOJ      5445 non-null   float64\n",
      " 5   DEROG    5252 non-null   float64\n",
      " 6   DELINQ   5380 non-null   float64\n",
      " 7   CLAGE    5652 non-null   float64\n",
      " 8   NINQ     5450 non-null   float64\n",
      " 9   CLNO     5738 non-null   float64\n",
      " 10  DEBTINC  4693 non-null   float64\n",
      "dtypes: float64(9), int64(2)\n",
      "memory usage: 512.3 KB\n"
     ]
    }
   ],
   "source": [
    "# Delete selected columns\n",
    "df = df.drop(columns=[\"REASON\", \"JOB\"], axis=1)\n",
    "df.info() # verify the columns are no longer part of the DataFrame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "executionInfo": {
     "elapsed": 11,
     "status": "ok",
     "timestamp": 1680268953365,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "ZWOTgAr7H2xK",
    "outputId": "45b84e2d-b70f-4fda-9764-8969e2a1b143"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-8e8beb7c-e255-43f5-9b26-58503e4ed62e\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>BAD</th>\n",
       "      <th>LOAN</th>\n",
       "      <th>MORTDUE</th>\n",
       "      <th>VALUE</th>\n",
       "      <th>YOJ</th>\n",
       "      <th>DEROG</th>\n",
       "      <th>DELINQ</th>\n",
       "      <th>CLAGE</th>\n",
       "      <th>NINQ</th>\n",
       "      <th>CLNO</th>\n",
       "      <th>DEBTINC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1100</td>\n",
       "      <td>25860.0000</td>\n",
       "      <td>39025.000000</td>\n",
       "      <td>10.500000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>94.366667</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>33.779915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1300</td>\n",
       "      <td>70053.0000</td>\n",
       "      <td>68400.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>121.833333</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>33.779915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1500</td>\n",
       "      <td>13500.0000</td>\n",
       "      <td>16700.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>149.466667</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>33.779915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1500</td>\n",
       "      <td>73760.8172</td>\n",
       "      <td>101776.048741</td>\n",
       "      <td>8.922268</td>\n",
       "      <td>0.25457</td>\n",
       "      <td>0.449442</td>\n",
       "      <td>179.766275</td>\n",
       "      <td>1.186055</td>\n",
       "      <td>21.296096</td>\n",
       "      <td>33.779915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1700</td>\n",
       "      <td>97800.0000</td>\n",
       "      <td>112000.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>93.333333</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>33.779915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5955</th>\n",
       "      <td>0</td>\n",
       "      <td>88900</td>\n",
       "      <td>57264.0000</td>\n",
       "      <td>90185.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>221.808718</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>36.112347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5956</th>\n",
       "      <td>0</td>\n",
       "      <td>89000</td>\n",
       "      <td>54576.0000</td>\n",
       "      <td>92937.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>208.692070</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>35.859971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5957</th>\n",
       "      <td>0</td>\n",
       "      <td>89200</td>\n",
       "      <td>54045.0000</td>\n",
       "      <td>92924.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>212.279697</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>35.556590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5958</th>\n",
       "      <td>0</td>\n",
       "      <td>89800</td>\n",
       "      <td>50370.0000</td>\n",
       "      <td>91861.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>213.892709</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>34.340882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5959</th>\n",
       "      <td>0</td>\n",
       "      <td>89900</td>\n",
       "      <td>48811.0000</td>\n",
       "      <td>88934.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>219.601002</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>34.571519</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5960 rows × 11 columns</p>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8e8beb7c-e255-43f5-9b26-58503e4ed62e')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-8e8beb7c-e255-43f5-9b26-58503e4ed62e button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-8e8beb7c-e255-43f5-9b26-58503e4ed62e');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "      BAD   LOAN     MORTDUE          VALUE        YOJ    DEROG    DELINQ  \\\n",
       "0       1   1100  25860.0000   39025.000000  10.500000  0.00000  0.000000   \n",
       "1       1   1300  70053.0000   68400.000000   7.000000  0.00000  2.000000   \n",
       "2       1   1500  13500.0000   16700.000000   4.000000  0.00000  0.000000   \n",
       "3       1   1500  73760.8172  101776.048741   8.922268  0.25457  0.449442   \n",
       "4       0   1700  97800.0000  112000.000000   3.000000  0.00000  0.000000   \n",
       "...   ...    ...         ...            ...        ...      ...       ...   \n",
       "5955    0  88900  57264.0000   90185.000000  16.000000  0.00000  0.000000   \n",
       "5956    0  89000  54576.0000   92937.000000  16.000000  0.00000  0.000000   \n",
       "5957    0  89200  54045.0000   92924.000000  15.000000  0.00000  0.000000   \n",
       "5958    0  89800  50370.0000   91861.000000  14.000000  0.00000  0.000000   \n",
       "5959    0  89900  48811.0000   88934.000000  15.000000  0.00000  0.000000   \n",
       "\n",
       "           CLAGE      NINQ       CLNO    DEBTINC  \n",
       "0      94.366667  1.000000   9.000000  33.779915  \n",
       "1     121.833333  0.000000  14.000000  33.779915  \n",
       "2     149.466667  1.000000  10.000000  33.779915  \n",
       "3     179.766275  1.186055  21.296096  33.779915  \n",
       "4      93.333333  0.000000  14.000000  33.779915  \n",
       "...          ...       ...        ...        ...  \n",
       "5955  221.808718  0.000000  16.000000  36.112347  \n",
       "5956  208.692070  0.000000  15.000000  35.859971  \n",
       "5957  212.279697  0.000000  15.000000  35.556590  \n",
       "5958  213.892709  0.000000  16.000000  34.340882  \n",
       "5959  219.601002  0.000000  16.000000  34.571519  \n",
       "\n",
       "[5960 rows x 11 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Replace missing values in a column with the mean of that column \n",
    "df = df.fillna(value=df.mean())\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 300
    },
    "executionInfo": {
     "elapsed": 817,
     "status": "ok",
     "timestamp": 1680268954173,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "emd4Y6srJdIE",
    "outputId": "14942a20-6eae-4548-a27e-30932b4c18cc"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-1edc8cde-683d-402f-801c-2393d39cfd44\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>LOAN</th>\n",
       "      <th>MORTDUE</th>\n",
       "      <th>VALUE</th>\n",
       "      <th>YOJ</th>\n",
       "      <th>DEROG</th>\n",
       "      <th>DELINQ</th>\n",
       "      <th>CLAGE</th>\n",
       "      <th>NINQ</th>\n",
       "      <th>CLNO</th>\n",
       "      <th>DEBTINC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "      <td>5960.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.197162</td>\n",
       "      <td>0.180378</td>\n",
       "      <td>0.110597</td>\n",
       "      <td>0.217616</td>\n",
       "      <td>0.025457</td>\n",
       "      <td>0.029963</td>\n",
       "      <td>0.153879</td>\n",
       "      <td>0.069768</td>\n",
       "      <td>0.299945</td>\n",
       "      <td>0.163991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.126210</td>\n",
       "      <td>0.106875</td>\n",
       "      <td>0.067040</td>\n",
       "      <td>0.176568</td>\n",
       "      <td>0.079420</td>\n",
       "      <td>0.071400</td>\n",
       "      <td>0.071529</td>\n",
       "      <td>0.097238</td>\n",
       "      <td>0.140117</td>\n",
       "      <td>0.037639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.112613</td>\n",
       "      <td>0.115918</td>\n",
       "      <td>0.068981</td>\n",
       "      <td>0.073171</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.100469</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.211268</td>\n",
       "      <td>0.149115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.171171</td>\n",
       "      <td>0.169731</td>\n",
       "      <td>0.096708</td>\n",
       "      <td>0.195122</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.152432</td>\n",
       "      <td>0.058824</td>\n",
       "      <td>0.295775</td>\n",
       "      <td>0.163991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.216705</td>\n",
       "      <td>0.130916</td>\n",
       "      <td>0.292683</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.029963</td>\n",
       "      <td>0.194433</td>\n",
       "      <td>0.117647</td>\n",
       "      <td>0.366197</td>\n",
       "      <td>0.184555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1edc8cde-683d-402f-801c-2393d39cfd44')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-1edc8cde-683d-402f-801c-2393d39cfd44 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-1edc8cde-683d-402f-801c-2393d39cfd44');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "              LOAN      MORTDUE        VALUE          YOJ        DEROG  \\\n",
       "count  5960.000000  5960.000000  5960.000000  5960.000000  5960.000000   \n",
       "mean      0.197162     0.180378     0.110597     0.217616     0.025457   \n",
       "std       0.126210     0.106875     0.067040     0.176568     0.079420   \n",
       "min       0.000000     0.000000     0.000000     0.000000     0.000000   \n",
       "25%       0.112613     0.115918     0.068981     0.073171     0.000000   \n",
       "50%       0.171171     0.169731     0.096708     0.195122     0.000000   \n",
       "75%       0.250000     0.216705     0.130916     0.292683     0.000000   \n",
       "max       1.000000     1.000000     1.000000     1.000000     1.000000   \n",
       "\n",
       "            DELINQ        CLAGE         NINQ         CLNO      DEBTINC  \n",
       "count  5960.000000  5960.000000  5960.000000  5960.000000  5960.000000  \n",
       "mean      0.029963     0.153879     0.069768     0.299945     0.163991  \n",
       "std       0.071400     0.071529     0.097238     0.140117     0.037639  \n",
       "min       0.000000     0.000000     0.000000     0.000000     0.000000  \n",
       "25%       0.000000     0.100469     0.000000     0.211268     0.149115  \n",
       "50%       0.000000     0.152432     0.058824     0.295775     0.163991  \n",
       "75%       0.029963     0.194433     0.117647     0.366197     0.184555  \n",
       "max       1.000000     1.000000     1.000000     1.000000     1.000000  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Scale value ranges of variables to zero and one\n",
    "# REMARK: it is important we perform this operation only on the dependent \n",
    "#         variables and not on the dependent variable. Our target variable\n",
    "#         remains unchanged. \n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "scaler = MinMaxScaler(feature_range=(0,1))\n",
    "\n",
    "# Remove the target variable so that it remains unchanged\n",
    "y = df.pop(\"BAD\")\n",
    "\n",
    "# Scale the remaining columns\n",
    "# We also create a new DataFrame to store the transformed data. \n",
    "# This is because our scaler does not alter the original data but\n",
    "# first creates a copy of the data and then transforms this copy.  \n",
    "df_ready = pd.DataFrame(scaler.fit_transform(df), columns=df.columns)\n",
    "\n",
    "# Verify our transformed data shows the same value ranges for all columns\n",
    "df_ready.describe()  # We reuse the method .describe(), which computes min/max values for each column"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "hxYbZz2c-eIB"
   },
   "source": [
    "\n",
    "## Partition the data into training a testing data\n",
    "Data partitioning is standard practice to ensure we use disjoint data for training and evaluating a machine learning model. \n",
    "\n",
    "the `sklearn` library provides many methods to partition data for standard machine learning workflows. Arguably, the easiest approach is to use the method `train_test_split()`. Check out its documentation and partition our *prepared* data. Let's say we use 30% of the data for testing and the rest for training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 977,
     "status": "ok",
     "timestamp": 1680273999025,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "HuGa1nKXLUBx",
    "outputId": "83f3f7e3-5358-411e-ccf5-c6403e58f1d0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4172, 10) (1788, 10) (4172,) (1788,)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "test_fraction = 0.3  # Let's say we use 30% of the data for testing\n",
    "\n",
    "Xtrain, Xtest, ytrain, ytest = train_test_split(df_ready, y, test_size=test_fraction)\n",
    "print(Xtrain.shape, Xtest.shape, ytrain.shape, ytest.shape)  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Q8AyLv82Okw1"
   },
   "source": [
    "## Train and assess a neural network classifier\n",
    "Finally, we come to the point where neural networks come in. In fact, the following exercise of training and assessing a classifier will look pretty much the same with any learning algorithm. That is one of the nice features of the `sklearn` library. It offers a consistent interface to many different learning algorithms. This way, we can easily switch from one learning algorithm to another if we like. For now, however, we focus on neural networks. \n",
    "\n",
    "Here is the chain to tasks we need to perform\n",
    "- Import relevant libraries to train a neural network classifier\n",
    "- Train our NN classifier using the training data\n",
    "- Compute NN predictions for the test data\n",
    "- Compute a performance indicator over the test set predictions. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1717,
     "status": "ok",
     "timestamp": 1680274016012,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "idSazLynPXfd",
    "outputId": "82ba1fec-c892-4858-a5a3-64aca6aace78"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification accuracy:  0.8489932885906041\n"
     ]
    }
   ],
   "source": [
    "# Import relevant libraries to train a neural network classifier\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "\n",
    "# Train our NN classifier using the training data\n",
    "nn = MLPClassifier(max_iter=500, hidden_layer_sizes=(3,), random_state=123)\n",
    "nn.fit(Xtrain, ytrain)\n",
    "\n",
    "# Compute NN predictions for the test data\n",
    "yhat = nn.predict(Xtest)\n",
    "\n",
    "# Compute a performance indicator over the test set predictions. \n",
    "perf = nn.score(X=Xtest, y=ytest)  # The easiest way to compute performance using a default indicator\n",
    "print(\"Classification accuracy: \", perf)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 929,
     "status": "ok",
     "timestamp": 1680274018079,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "sY3D0b6IQlXi",
    "outputId": "0b68cd8e-f905-44e4-ec38-8a9e09b169f0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1406,   29],\n",
       "       [ 241,  112]])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# A more informative way to assess the NN involves creating a cross-table of\n",
    "# its predictions and the true classes of test set observations.\n",
    "# An easy - quick and dirty - way to do this is as follows\n",
    "from sklearn.metrics import confusion_matrix\n",
    "confusion_matrix(y_true=ytest, y_pred=yhat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 453
    },
    "executionInfo": {
     "elapsed": 656,
     "status": "ok",
     "timestamp": 1680274021743,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "M0oGw4P1Q6cC",
    "outputId": "5c9e14e3-7f7a-4598-b792-510179ad3682"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# A nice visualization of the so-called confusion matrix can be produced as follows\n",
    "from sklearn.metrics import ConfusionMatrixDisplay\n",
    "\n",
    "ConfusionMatrixDisplay.from_predictions(y_true=ytest, y_pred=yhat, labels=[0, 1], display_labels=['Good(0)', 'Bad(1)']);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "SODevpVI-hS-"
   },
   "source": [
    "\n",
    "## Optional task for expert: Tuning the architecture of the neural network\n",
    "Try to improve your NN classifier by systematically evaluating alternative network architectures. For example, you can consider networks with different numbers of layers and different numbers of nodes in those layers. \n",
    "\n",
    "In fact, there are many other configuration options (called meta-parameters) you could consider, as we discussed in the lecture. The point of this task is not to find the best NN classifier for our specific data set. Rather, the point is to familiarize you with relevant libraries and methods to carry out such tuning tasks using Python. In the lecture, we introduced the concept of *grid search*. Time to run a web search for, e.g., \"sklearn grid search\" and see what it produces ;) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "uM5EjufEST6r"
   },
   "outputs": [],
   "source": [
    "# Trying to improve our NN by grid searching meta-parameter options\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "print('Tuning neural network classifier')\n",
    "\n",
    "# Define meta-parameter grid of candidate settings\n",
    "# The following settings are just for illustration\n",
    "candidate_settings = {'hidden_layer_sizes': [(3,), (5,)]}\n",
    "\n",
    "# This is an example for a comprehensive grid search over many candidate neural\n",
    "# network architectures. Be careful with running this code as it will take some\n",
    "# time to execute (e.g., ~40 min) \n",
    "candidate_settings_big = {'hidden_layer_sizes': [(3,), (5,), (25,),(50,), (100,),\n",
    "                                             (5,2), (25,5), (50,7), (100,10), \n",
    "                                             (100,10, 3)],\n",
    "                      \"activation\": [\"logistic\", \"tanh\", \"relu\"],\n",
    "                      \"learning_rate_init\": [0.01, 0.001],\n",
    "                      \"solver\": [\"sgd\", \"adam\"],\n",
    "                      \"early_stopping\": [True, False]\n",
    "\n",
    "              }\n",
    "# For your information, this is the best configuration coming out of the \n",
    "# comprehensive grid search \n",
    "# MLPClassifier(hidden_layer_sizes=(100, 10), learning_rate_init=0.01, max_iter=1000)\n",
    "\n",
    "# Train one NN for each candidate setting and assess the model\n",
    "gs_nn = GridSearchCV(estimator=MLPClassifier(max_iter=1000), param_grid=candidate_settings_big, verbose=3)\n",
    "gs_nn.fit(Xtrain, ytrain)\n",
    "\n",
    "# Print which configuration gave the best performance\n",
    "print(\"Best configuration: \", gs_nn.best_estimator_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 467
    },
    "executionInfo": {
     "elapsed": 730,
     "status": "ok",
     "timestamp": 1680274194378,
     "user": {
      "displayName": "Stefan Lessmann",
      "userId": "06342662613942148717"
     },
     "user_tz": -120
    },
    "id": "3ak61AZYT3_e",
    "outputId": "7fa628ae-398e-4929-9cfd-8e08b033ce19"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9256152125279642\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjMAAAGwCAYAAABcnuQpAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABLrElEQVR4nO3deVxU5f4H8M8MMICsgsJAEqK4gKKomCLmkigu1/RqmV0qNNJ+CeWSay7hSqm5b2nmUphZqSWZiUuKiqgk5YLkgoIL2A0BQWGAOb8/uJw6oSPDDHCGPu/X67xel+c8zznf4RJ8/T7Pc45CEAQBRERERCZKWdsBEBERERmCyQwRERGZNCYzREREZNKYzBAREZFJYzJDREREJo3JDBEREZk0JjNERERk0sxrO4B/Mq1Wi9u3b8POzg4KhaK2wyEiIj0IgoD79+/D3d0dSmX11QYKCwuh0WgMvo5KpYKVlZURIpIfJjO16Pbt2/Dw8KjtMIiIyAAZGRlo1KhRtVy7sLAQXp62yLxbavC11Go10tLS6mRCw2SmFtnZ2QEAbvzcGPa2nPGjuunfzf1qOwSialGCYhzDXvF3eXXQaDTIvFuKG0mNYW9X9b8Tefe18OxwHRqNhskMGVf51JK9rdKgH1IiOTNXWNR2CETV438vA6qJZQK2dgrY2lX9PlrU7aUMTGaIiIhkrlTQotSANymWClrjBSNDTGaIiIhkTgsBWlQ9mzFkrCng3AYRERGZNFZmiIiIZE4LLQyZKDJstPwxmSEiIpK5UkFAqVD1qSJDxpoCTjMRERGRSWNlhoiISOa4AFg3JjNEREQyp4WAUiYzj8VpJiIiIjJprMwQERHJHKeZdGMyQ0REJHPczaQbp5mIiIjIpLEyQ0REJHPa/x2GjK/LmMwQERHJXKmBu5kMGWsKmMwQERHJXKkAA9+abbxY5IhrZoiIiMiksTJDREQkc1wzoxuTGSIiIpnTQoFSKAwaX5dxmomIiIhMGiszREREMqcVyg5DxtdlTGaIiIhkrtTAaSZDxpoCTjMRERGRSWNlhoiISOZYmdGNyQwREZHMaQUFtIIBu5kMGGsKOM1EREREJo2VGSIiIpnjNJNuTGaIiIhkrhRKlBowmVJqxFjkiMkMERGRzAkGrpkRuGaGiIiISL5YmSEiIpI5rpnRjckMERGRzJUKSpQKBqyZqeOvM+A0ExEREZk0VmaIiIhkTgsFtAbUH7So26UZJjNEREQyxzUzunGaiYiIiEwaKzNEREQyZ/gCYE4zERERUS0qWzNjwIsmOc1EREREJF+szBAREcmc1sB3M3E3ExEREdUqrpnRjckMERGRzGmh5HNmdOCaGSIiIjJprMwQERHJXKmgQKlgwEPzDBhrCpjMEBERyVypgQuASznNRERERCRfrMwQERHJnFZQQmvAbiYtdzMRERFRbeI0k26cZiIiIqIKjh49ioEDB8Ld3R0KhQK7d+8WzxUXF2PKlCnw8/ODjY0N3N3d8dprr+H27duSa2RnZyM0NBT29vZwdHREeHg48vPzJX1+/fVXPPvss7CysoKHhwcWLlyod6xMZoiIiGROiz93NFXl0FbhngUFBWjbti1Wr15d4dyDBw/w888/Y+bMmfj555+xc+dOpKam4vnnn5f0Cw0NxYULFxAXF4fY2FgcPXoUo0ePFs/n5eWhT58+8PT0RFJSEhYtWoSoqCisX79er1g5zURERCRzhj80r2xsXl6epN3S0hKWlpaPHNOvXz/069fvkeccHBwQFxcnaVu1ahWeeeYZpKen4+mnn0ZKSgr27duH06dPIyAgAACwcuVK9O/fH4sXL4a7uztiYmKg0Wjw6aefQqVSoVWrVkhOTsaSJUskSc+TsDJDRET0D+Hh4QEHBwfxiI6ONtq1c3NzoVAo4OjoCABISEiAo6OjmMgAQHBwMJRKJRITE8U+3bp1g0qlEvuEhIQgNTUV9+7dq/S9WZkhIiKSOcPfzVQ2NiMjA/b29mL746oy+iosLMSUKVPw8ssvi9fPzMyEi4uLpJ+5uTmcnJyQmZkp9vHy8pL0cXV1Fc/Vr1+/UvdnMkNERCRzWiigRdWf4ls+1t7eXpLMGENxcTGGDRsGQRCwdu1ao167spjMEBERyZyxKjPGVp7I3LhxA4cOHZIkSmq1Gnfv3pX0LykpQXZ2NtRqtdgnKytL0qf86/I+lcE1M0RERKS38kTm8uXLOHDgAJydnSXnAwMDkZOTg6SkJLHt0KFD0Gq16NSpk9jn6NGjKC4uFvvExcWhRYsWlZ5iApjMEBERyV75Q/MMOfSVn5+P5ORkJCcnAwDS0tKQnJyM9PR0FBcX44UXXsCZM2cQExOD0tJSZGZmIjMzExqNBgDg4+ODvn37YtSoUTh16hSOHz+OyMhIDB8+HO7u7gCA//znP1CpVAgPD8eFCxfw5ZdfYvny5ZgwYYJesXKaiYiISOa0ggJaA958XZWxZ86cQc+ePcWvyxOMsLAwREVF4bvvvgMA+Pv7S8YdPnwYPXr0AADExMQgMjISvXr1glKpxNChQ7FixQqxr4ODA/bv34+IiAh06NABDRo0wKxZs/Talg0wmSEiIqJH6NGjBwQd73TSda6ck5MTtm3bprNPmzZtEB8fr3d8f8VkhoiISOa0Br6byZAH7pkCJjNEREQyZ/hbs+t2MlO3Px0RERHVeazMEBERyVwpFCg14KF5how1BUxmiIiIZI7TTLrV7U9HREREdR4rM0RERDJXCsOmikqNF4osMZkhIiKSOU4z6cZkhoiISObk+qJJuajbn46IiIjqPFZmiIiIZE6AAloD1swI3JpNREREtYnTTLrV7U9HREREdR4rM0RERDKnFRTQClWfKjJkrClgMkNERCRzpQa+NduQsaagbn86IiIiqvNYmSEiIpI5TjPpxmSGiIhI5rRQQmvAZIohY01B3f50REREVOexMkNERCRzpYICpQZMFRky1hQwmSEiIpI5rpnRjckMERGRzAkGvjVb4BOAiYiIiOSLlRkiIiKZK4UCpQa8LNKQsaaAyQwREZHMaQXD1r1oBSMGI0OcZiIiIiKTxsoMmZRzJ23w1RoXXD5XD9lZFnh/Yxq69MsVz3+2WI2fvnXE77ctYKES4O33ECOn3kHL9g/EPpd/tcbG+e747Zd6UJoJ6No/B29G3Ya1jVbsE+LuX+He09ZcR4/BOdX58Yj0NiwyC+HvZWLXhgZY9/5TAAALSy1Gv38bPZ7PgYWlgKSf7LBy2lPI+a9FLUdLVaU1cAGwIWNNQd3+dE+gUCiwe/duSdvGjRvRp08fva7TuXNnfPPNN0aMjB6n8IESTVo9ROSCm488/1STQkTMv4mPD6Xio91XoPbQYNrLTZHzhxkA4I9Mc0wd3hTuXkVYHvsb5sdcxY1UKywe93SFa727NB1fJJ8Xjy59cyv0IapNzds+wIBXsnHtgpWk/f+ibqNz7zzMe9MTE4c0hZNrMWZtvF47QZJRaKEw+KjLaj2ZyczMxNixY+Ht7Q0rKyu4uroiKCgIa9euxYMHD558ASMqLCzEzJkz8f7770vav/rqK7Rs2RJWVlbw8/PD3r17JednzJiBqVOnQqvVgqpXx+fuY8SUTAT1e3Ri8dyQHLTvlg83Tw0atyjE6KhbeHDfDGkXrQEAiQccYG4uIHLBTXh4F6GF/0O88+FNHPveEbfSVJJr2dqXwsmlRDxUVnV80plMilW9UkxZdQPLJjXC/Vwzsb2eXSlCXs7Gx1Hu+OW4Ha6cq4clEzzQquMDtGxfUIsRE1WfWk1mrl27hnbt2mH//v1YsGABzp49i4SEBEyePBmxsbE4cOBAjcbz9ddfw97eHkFBQWLbiRMn8PLLLyM8PBxnz57F4MGDMXjwYJw/f17s069fP9y/fx8//PBDjcZLuhVrFNj7uTNs7EvRxPdhWVuRAuYWApR/+clXWZUloRdO2UrGr5r+FF5s1Rpv92+GH79wgsBchmQkcsEtnDpoj7PxdpL2Zm0ewEIlSNozrlgh66YFfDrU7D8QyXjKnwBsyFGX1WoyM2bMGJibm+PMmTMYNmwYfHx80KRJEwwaNAjff/89Bg4cCABIT0/HoEGDYGtrC3t7ewwbNgxZWVmSa61duxZNmzaFSqVCixYt8Nlnn0nOX758Gd26dYOVlRV8fX0RFxdXIZ7t27eL9yy3fPly9O3bF5MmTYKPjw/mzp2L9u3bY9WqVWIfMzMz9O/fH9u3bzfWt4YMcDLOHoO8/TDQqw12bWiI6O1X4OBcCgBo2zUf9363wFdrGqJYo8D9HDN8usAdAJB9988lZK9NuoPp624gevtVdO2fi5XvNcK3GxvUyuch+rvug+7B2+8hPo12q3DOyaUEmiIFCvLMJO05v5vDyaW4pkIkIytfM2PIUZfV2qf7448/sH//fkRERMDGxuaRfRQKBbRaLQYNGoTs7GwcOXIEcXFxuHbtGl566SWx365duzB27Fi8++67OH/+PN58802MHDkShw8fBgBotVoMGTIEKpUKiYmJWLduHaZMmVLhfseOHUNAQICkLSEhAcHBwZK2kJAQJCQkSNqeeeYZxMfH6/zMRUVFyMvLkxxkfP5B+VgTl4ql311GQI/7mP9mY+T8tyxRadyiEBOX3cA3H7vg+aZt8LJ/K6g9NKjfsBiKv/zDJXR8Flo9UwBvv4d4KfIuXnzrLr5a61JLn4joTw3dNXhrzm18GPk0iovq9h8oosqqtd1MV65cgSAIaNGihaS9QYMGKCwsBABEREQgODgY586dQ1paGjw8PAAAW7duRatWrXD69Gl07NgRixcvxogRIzBmzBgAwIQJE3Dy5EksXrwYPXv2xIEDB3Dp0iX8+OOPcHcv+1f4ggUL0K9fP/G+OTk5yM3NFc+Xy8zMhKurq6TN1dUVmZmZkjZ3d3dkZGRAq9VCqXz0L5jo6GjMnj1b328V6cmqnhZPeWnwlJcGPh0eYGSQD/Z94YThb98FULau5rkhObj3uzms6mmhUAA71zeEm2fRY6/Zsv0DbFumhqZIAZUl55uo9ni3eYj6DUuw+sffxDYzc8CvcwGeH/lfvPefJlBZCrCxL5VUZxwbliD7LnczmSotDHw3ExcA16xTp04hOTkZrVq1QlFREVJSUuDh4SEmMgDg6+sLR0dHpKSkAABSUlIk61wAICgoSHLew8NDkqgEBgZK+j98WLamwspKuiugsqytraHValFU9Pg/iNOmTUNubq54ZGRkVOlepB9Bi0f+C7Z+wxJY22hx5FtHWFhq0b5b/mOvcfWCNWwdS5jIUK1LjrfF6J7N8VbvP4/UZGsc2lkfb/Vujt9+qYdijQLtut4XxzRqWgjXRsVISapXi5GTIQQDdzIJdTyZqbXKjLe3NxQKBVJTUyXtTZo0AVCWHNQkZ2dnKBQK3Lt3T9KuVqsrrM/JysqCWq2WtGVnZ8PGxkZn3JaWlrC0tDRe0P9ADwuUuJ325/cwM0OFq+etYedYAnunUmxb7orAPrlwci1GXrY5vtvUAP/NtMCzA3PEMd9+2gC+AQWwttHi56N2+GSuO15/7zZsHcrW1Zzcb497v5vDp8MDWFiW9dm+wgUv/N/vNf1xiSp4WGCGG6nS3zOFD5S4f+/P9h+/cMLoqNu4n2OOgvtKRMy/hYtn6uHSz4+e0if541uzdau1ZMbZ2Rm9e/fGqlWr8Pbbbz923YyPjw8yMjKQkZEhVmcuXryInJwc+Pr6in2OHz+OsLAwcdzx48cl5zMyMnDnzh24uZUtmDt58qTkPiqVCr6+vrh48aLkOTOBgYE4ePAgxo0bJ7bFxcVVqOycP38e7dq1q+J3gyrrt1/qYfIL3uLXH0eVPSSs97BsvPNBBm5escTcrxojL9scdvVL0bztA3y06zIatygUx6Qm18NnH6lRWKBEI+8ivLMwA8Ev/JnEmlkI2LO5AT6OsoQgAO6NNXgz6jb6hf5Rcx+UyADrotyhFYCZG67DwlLAmZ/ssGraU7UdFlG1qdUnAK9ZswZBQUEICAhAVFQU2rRpA6VSidOnT+PSpUvo0KEDgoOD4efnh9DQUCxbtgwlJSUYM2YMunfvLi7WnTRpEoYNG4Z27dohODgYe/bswc6dO8Wt3cHBwWjevDnCwsKwaNEi5OXlYfr06RXiCQkJwbFjxySJy9ixY9G9e3d89NFHGDBgALZv344zZ85g/fr1krHx8fF6P2yP9Ne2Sz5+vJ382POVeTDY5BXpOs937HkfHXve19mHSE7+muADZdOqq99rhNXvNaqliMjY+ARg3Wr10zVt2hRnz55FcHAwpk2bhrZt2yIgIAArV67ExIkTMXfuXCgUCnz77beoX78+unXrhuDgYDRp0gRffvmleJ3Bgwdj+fLlWLx4MVq1aoWPP/4YmzZtQo8ePQAASqUSu3btwsOHD/HMM8/gjTfewPz58yvEEx4ejr179yI3988HsnXp0gXbtm3D+vXr0bZtW3z99dfYvXs3WrduLfa5desWTpw4gZEjR1bfN4uIiP6xyqeZDDnqMoUg8FFgf/Xiiy+iffv2mDZtWqXHTJkyBffu3atQrXmSvLw8ODg44N5vTWBvV7ezZvrnetR7rojqghKhGD/hW+Tm5sLe3r5a7lH+d2LQ/tdhYaN68oDHKC7Q4Ns+n1ZrrLWJf0H/ZtGiRbC1tX1yx79wcXHB3LlzqykiIiL6p+O7mXTjW7P/pnHjxnj77bf1GvPuu+9WUzRERETczfQkrMwQERGRSWNlhoiISOZYmdGNlRkiIiKZq43dTEePHsXAgQPh7u4OhUKB3bt3S84LgoBZs2bBzc0N1tbWCA4OxuXLlyV9srOzERoaCnt7ezg6OiI8PBz5+dKnrf/666949tlnYWVlBQ8PDyxcuFDvWJnMEBERUQUFBQVo27YtVq9e/cjzCxcuxIoVK7Bu3TokJibCxsYGISEh4vsVASA0NBQXLlxAXFwcYmNjcfToUYwePVo8n5eXhz59+sDT0xNJSUlYtGgRoqKi9N4dzGkmIiIimauNaaZ+/fpJXsj8V4IgYNmyZZgxYwYGDRoEoOwl0K6urti9ezeGDx+OlJQU7Nu3D6dPnxYfcrty5Ur0798fixcvhru7O2JiYqDRaPDpp59CpVKhVatWSE5OxpIlSyRJz5OwMkNERCRzAgzbnl3+QLm8vDzJoevlyLqkpaUhMzMTwcHBYpuDgwM6deqEhIQEAEBCQgIcHR3FRAYoeyK/UqlEYmKi2Kdbt25Qqf58hk5ISAhSU1MrvCtRFyYzREREMmesNTMeHh5wcHAQj+jo6CrFk5mZCQBwdXWVtLu6uornMjMz4eLiIjlvbm4OJycnSZ9HXeOv96gMTjMRERH9Q2RkZEieAGxpaVmL0RgPkxkiIiKZM9aaGXt7e6O8zkCtVgMAsrKy4ObmJrZnZWXB399f7HP37l3JuJKSEmRnZ4vj1Wo1srKyJH3Kvy7vUxmcZiIiIpI5ub1o0svLC2q1GgcPHhTb8vLykJiYiMDAQABAYGAgcnJykJSUJPY5dOgQtFotOnXqJPY5evQoiouLxT5xcXFo0aIF6tevX+l4mMwQERFRBfn5+UhOTkZycjKAskW/ycnJSE9Ph0KhwLhx4zBv3jx89913OHfuHF577TW4u7tj8ODBAAAfHx/07dsXo0aNwqlTp3D8+HFERkZi+PDhcHd3BwD85z//gUqlQnh4OC5cuIAvv/wSy5cvx4QJE/SKldNMREREMlcbW7PPnDmDnj17il+XJxhhYWHYvHkzJk+ejIKCAowePRo5OTno2rUr9u3bBysrK3FMTEwMIiMj0atXLyiVSgwdOhQrVqwQzzs4OGD//v2IiIhAhw4d0KBBA8yaNUuvbdkAoBAEQXhyN6oO5a92v/dbE9jbsUhGdVOIu39th0BULUqEYvyEb5Gbm2uUdSiPUv53IujbSJjbVH2xbklBEY4PWlWtsdYm/gUlIiIik8ZpJiIiIpkrf/idIePrMiYzREREMse3ZuvGaSYiIiIyaazMEBERyZwgKCAYUF0xZKwpYDJDREQkc5xm0o3JDBERkcyxMqMb18wQERGRSWNlhoiISOYEA6eZ6nplhskMERGRzAkADHlef11/1D+nmYiIiMiksTJDREQkc1oooOATgB+LyQwREZHMcTeTbpxmIiIiIpPGygwREZHMaQUFFHxo3mMxmSEiIpI5QTBwN1Md387EaSYiIiIyaazMEBERyRwXAOvGZIaIiEjmmMzoxmSGiIhI5rgAWDeumSEiIiKTxsoMERGRzHE3k25MZoiIiGSuLJkxZM2MEYORIU4zERERkUljZYaIiEjmuJtJNyYzREREMif87zBkfF3GaSYiIiIyaazMEBERyRynmXRjMkNERCR3nGfSickMERGR3BlYmUEdr8xwzQwRERGZNFZmiIiIZI5PANaNyQwREZHMcQGwbpxmIiIiIpPGygwREZHcCQrDFvHW8coMkxkiIiKZ45oZ3TjNRERERCaNlRkiIiK540PzdGIyQ0REJHPczaRbpZKZ7777rtIXfP7556scDBEREZG+KpXMDB48uFIXUygUKC0tNSQeIiIiepQ6PlVkiEolM1qttrrjICIiosfgNJNuBu1mKiwsNFYcRERE9DiCEY46TO9kprS0FHPnzsVTTz0FW1tbXLt2DQAwc+ZMbNy40egBEhEREemidzIzf/58bN68GQsXLoRKpRLbW7dujU8++cSowREREREAKIxwVF5paSlmzpwJLy8vWFtbo2nTppg7dy6Evzx9TxAEzJo1C25ubrC2tkZwcDAuX74suU52djZCQ0Nhb28PR0dHhIeHIz8/v0rfAV30Tma2bt2K9evXIzQ0FGZmZmJ727ZtcenSJaMGR0RERKjxaaYPP/wQa9euxapVq5CSkoIPP/wQCxcuxMqVK8U+CxcuxIoVK7Bu3TokJibCxsYGISEhkiUooaGhuHDhAuLi4hAbG4ujR49i9OjRVf0uPJbez5m5desWvL29K7RrtVoUFxcbJSgiIiIyvry8PMnXlpaWsLS0rNDvxIkTGDRoEAYMGAAAaNy4Mb744gucOnUKQFlVZtmyZZgxYwYGDRoEoKzY4erqit27d2P48OFISUnBvn37cPr0aQQEBAAAVq5cif79+2Px4sVwd3c32ufSuzLj6+uL+Pj4Cu1ff/012rVrZ5SgiIiI6C+MVJnx8PCAg4ODeERHRz/ydl26dMHBgwfx22+/AQB++eUXHDt2DP369QMApKWlITMzE8HBweIYBwcHdOrUCQkJCQCAhIQEODo6iokMAAQHB0OpVCIxMdEY3xWR3pWZWbNmISwsDLdu3YJWq8XOnTuRmpqKrVu3IjY21qjBEREREYz21uyMjAzY29uLzY+qygDA1KlTkZeXh5YtW8LMzAylpaWYP38+QkNDAQCZmZkAAFdXV8k4V1dX8VxmZiZcXFwk583NzeHk5CT2MRa9KzODBg3Cnj17cODAAdjY2GDWrFlISUnBnj170Lt3b6MGR0RERMZjb28vOR6XzOzYsQMxMTHYtm0bfv75Z2zZsgWLFy/Gli1bajjiyqnSu5meffZZxMXFGTsWIiIiegRBKDsMGa+PSZMmYerUqRg+fDgAwM/PDzdu3EB0dDTCwsKgVqsBAFlZWXBzcxPHZWVlwd/fHwCgVqtx9+5dyXVLSkqQnZ0tjjeWKj8078yZM/jss8/w2WefISkpyZgxERER0V/V8G6mBw8eQKmUpghmZmbiGwG8vLygVqtx8OBB8XxeXh4SExMRGBgIAAgMDEROTo4kRzh06BC0Wi06deqkX0BPoHdl5ubNm3j55Zdx/PhxODo6AgBycnLQpUsXbN++HY0aNTJqgERERFSzBg4ciPnz5+Ppp59Gq1atcPbsWSxZsgSvv/46gLJ3MY4bNw7z5s1Ds2bN4OXlhZkzZ8Ld3V18n6OPjw/69u2LUaNGYd26dSguLkZkZCSGDx9u1J1MQBUqM2+88QaKi4uRkpKC7OxsZGdnIyUlBVqtFm+88YZRgyMiIiL8uQDYkEMPK1euxAsvvIAxY8bAx8cHEydOxJtvvom5c+eKfSZPnoy3334bo0ePRseOHZGfn499+/bByspK7BMTE4OWLVuiV69e6N+/P7p27Yr169cb7dtSTiEI+s2kWVtb48SJExW2YSclJeHZZ5/FgwcPjBpgXZaXlwcHBwfc+60J7O0Mek0WkWyFuPvXdghE1aJEKMZP+Ba5ubmSHULGVP53wmP5HCitrZ484DG0DwuRMXZWtcZam/SeZvLw8Hjkw/FKS0uNXjYiIiIiGP6ySL5oUmrRokV4++23cebMGbHtzJkzGDt2LBYvXmzU4IiIiIiepFKVmfr160Oh+HO+raCgAJ06dYK5ednwkpISmJub4/XXXxcX/hAREZGRGOmheXVVpZKZZcuWVXMYRERE9FicZtKpUslMWFhYdcdBREREVCVVegJwucLCQmg0GklbXVwlTUREVKtYmdFJ7wXABQUFiIyMhIuLC2xsbFC/fn3JQUREREZWw08ANjV6JzOTJ0/GoUOHsHbtWlhaWuKTTz7B7Nmz4e7ujq1bt1ZHjERERESPpfc00549e7B161b06NEDI0eOxLPPPgtvb294enoiJiZGfD04ERERGQl3M+mkd2UmOzsbTZo0AVC2PiY7OxsA0LVrVxw9etS40REREREUguFHXaZ3MtOkSROkpaUBAFq2bIkdO3YAKKvYlL94koiIiKim6J3MjBw5Er/88gsAYOrUqVi9ejWsrKwwfvx4TJo0yegBEhER/eNxAbBOeq+ZGT9+vPi/g4ODcenSJSQlJcHb2xtt2rQxanBERERET2LQc2YAwNPTE56ensaIhYiIiB5BAcPWvdTt5b+VTGZWrFhR6Qu+8847VQ6GiIiISF+VSmaWLl1aqYspFAomM1XwQkAXmCtUtR0GUfV4hpVbqqNKCoGkb2vmXtyarVOlkpny3UtERERUC/g6A5303s1EREREJCcGLwAmIiKiasbKjE5MZoiIiGTO0Kf48gnARERERDLGygwREZHccZpJpypVZuLj4/HKK68gMDAQt27dAgB89tlnOHbsmFGDIyIiIvB1Bk+gdzLzzTffICQkBNbW1jh79iyKiooAALm5uViwYIHRAyQiIiLSRe9kZt68eVi3bh02bNgACwsLsT0oKAg///yzUYMjIiKiPxcAG3LUZXqvmUlNTUW3bt0qtDs4OCAnJ8cYMREREdFf8QnAOuldmVGr1bhy5UqF9mPHjqFJkyZGCYqIiIj+gmtmdNI7mRk1ahTGjh2LxMREKBQK3L59GzExMZg4cSLeeuut6oiRiIiI6LH0nmaaOnUqtFotevXqhQcPHqBbt26wtLTExIkT8fbbb1dHjERERP9ofGiebnonMwqFAtOnT8ekSZNw5coV5Ofnw9fXF7a2ttURHxEREfE5MzpV+aF5KpUKvr6+xoyFiIiISG96JzM9e/aEQvH4VdGHDh0yKCAiIiL6G0O3V7MyI+Xv7y/5uri4GMnJyTh//jzCwsKMFRcRERGV4zSTTnonM0uXLn1ke1RUFPLz8w0OiIiIiEgfRntr9iuvvIJPP/3UWJcjIiKicnzOjE5Ge2t2QkICrKysjHU5IiIi+h9uzdZN72RmyJAhkq8FQcCdO3dw5swZzJw502iBEREREVWG3smMg4OD5GulUokWLVpgzpw56NOnj9ECIyIiIqoMvZKZ0tJSjBw5En5+fqhfv351xURERER/xd1MOum1ANjMzAx9+vTh27GJiIhqUPmaGUOOukzv3UytW7fGtWvXqiMWIiIiIr3pnczMmzcPEydORGxsLO7cuYO8vDzJQURERNWA27Ifq9JrZubMmYN3330X/fv3BwA8//zzktcaCIIAhUKB0tJS40dJRET0T8Y1MzpVujIze/ZsFBQU4PDhw+Jx6NAh8Sj/moiIiEzfrVu38Morr8DZ2RnW1tbw8/PDmTNnxPOCIGDWrFlwc3ODtbU1goODcfnyZck1srOzERoaCnt7ezg6OiI8PLxa3hZQ6cqMIJSldd27dzd6EERERPR4Nf3QvHv37iEoKAg9e/bEDz/8gIYNG+Ly5cuSncwLFy7EihUrsGXLFnh5eWHmzJkICQnBxYsXxYfohoaG4s6dO4iLi0NxcTFGjhyJ0aNHY9u2bVX/MI+g19ZsXW/LJiIiompSw9NMH374ITw8PLBp0yaxzcvL68/LCQKWLVuGGTNmYNCgQQCArVu3wtXVFbt378bw4cORkpKCffv24fTp0wgICAAArFy5Ev3798fixYvh7u5uwAeS0msBcPPmzeHk5KTzICIiInn6+6adoqKiR/b77rvvEBAQgBdffBEuLi5o164dNmzYIJ5PS0tDZmYmgoODxTYHBwd06tQJCQkJAMpec+To6CgmMgAQHBwMpVKJxMREo34uvSozs2fPrvAEYCIiIqpexppm8vDwkLS///77iIqKqtD/2rVrWLt2LSZMmID33nsPp0+fxjvvvAOVSoWwsDBkZmYCAFxdXSXjXF1dxXOZmZlwcXGRnDc3N4eTk5PYx1j0SmaGDx9eITAiIiKqZkaaZsrIyIC9vb3YbGlp+cjuWq0WAQEBWLBgAQCgXbt2OH/+PNatW4ewsDADAqkelZ5m4noZIiIi02Zvby85HpfMuLm5wdfXV9Lm4+OD9PR0AIBarQYAZGVlSfpkZWWJ59RqNe7evSs5X1JSguzsbLGPsVQ6mSnfzUREREQ1zJAH5lWhqhMUFITU1FRJ22+//QZPT08AZYuB1Wo1Dh48KJ7Py8tDYmIiAgMDAQCBgYHIyclBUlKS2OfQoUPQarXo1KmTfgE9QaWnmbRarVFvTERERJVT01uzx48fjy5dumDBggUYNmwYTp06hfXr12P9+vVl11MoMG7cOMybNw/NmjUTt2a7u7tj8ODBAMoqOX379sWoUaOwbt06FBcXIzIyEsOHDzfqTiZAzzUzREREVAtqeGt2x44dsWvXLkybNg1z5syBl5cXli1bhtDQULHP5MmTUVBQgNGjRyMnJwddu3bFvn37xGfMAEBMTAwiIyPRq1cvKJVKDB06FCtWrDDggzyaQuD8Ua3Jy8uDg4MDetm/AnOFqrbDIaoWpS09azsEompRUlKIn5KikZubK1lUa0zlfydajFsAM0urJw94jNKiQqQue69aY61NrMwQERHJHd/NpBOTGSIiIpmr6TUzpkavJwATERERyQ0rM0RERHLHaSadmMwQERHJHKeZdOM0ExEREZk0VmaIiIjkjtNMOjGZISIikjsmMzpxmomIiIhMGiszREREMqf432HI+LqMyQwREZHccZpJJyYzREREMset2bpxzQwRERGZNFZmiIiI5I7TTDoxmSEiIjIFdTwhMQSnmYiIiMiksTJDREQkc1wArBuTGSIiIrnjmhmdOM1EREREJo2VGSIiIpnjNJNuTGaIiIjkjtNMOnGaiYiIiEwaKzNEREQyx2km3ZjMEBERyR2nmXRiMkNERCR3TGZ04poZIiIiMmmszBAREckc18zoxmSGiIhI7jjNpBOnmYiIiMiksTJDREQkcwpBgEKoennFkLGmgMkMERGR3HGaSSdOMxEREZFJY2WGiIhI5ribSTcmM0RERHLHaSadOM1EREREJo2VGSIiIpnjNJNuTGaIiIjkjtNMOjGZISIikjlWZnTjmhkiIiIyaazMEBERyR2nmXRiMkNERGQC6vpUkSE4zUREREQmjZUZIiIiuROEssOQ8XUYkxkiIiKZ424m3TjNRERERDp98MEHUCgUGDdunNhWWFiIiIgIODs7w9bWFkOHDkVWVpZkXHp6OgYMGIB69erBxcUFkyZNQklJidHjYzJDREQkd4IRjio6ffo0Pv74Y7Rp00bSPn78eOzZswdfffUVjhw5gtu3b2PIkCHi+dLSUgwYMAAajQYnTpzAli1bsHnzZsyaNavqwTwGkxkiIiKZU2gNP6oiPz8foaGh2LBhA+rXry+25+bmYuPGjViyZAmee+45dOjQAZs2bcKJEydw8uRJAMD+/ftx8eJFfP755/D390e/fv0wd+5crF69GhqNxhjfFhGTGSIion+IvLw8yVFUVKSzf0REBAYMGIDg4GBJe1JSEoqLiyXtLVu2xNNPP42EhAQAQEJCAvz8/ODq6ir2CQkJQV5eHi5cuGDET8UFwFQHWduU4NV3bqBL8B9wcC7G1RQbfDy/KS6ftwMAhEbeQLf+v6OhugjFxUpcuWCLrcs8kfqrfS1HTlTRS0PPI6hzOjwa5UFTZIaLqQ2xcUs73LztIOnn0+J3jAhNRsvm/0WpVolrafXx3uznoNGU/Zq3sy3CmFGn0anjLQgCcCzhaaz9JACFhRa18bFIX0Z6aJ6Hh4ek+f3330dUVNQjh2zfvh0///wzTp8+XeFcZmYmVCoVHB0dJe2urq7IzMwU+/w1kSk/X37OmP6xlZmoqCj4+/tL2jQaDby9vXHixIlKX+fixYto1KgRCgoKjBwhVdXYuZfRrksOFk9pgTHPt8fZ4/WxYNM5OLuU/Qvk1nVrrJ3bFGOeb49JoW1w95Yl5m08D/v6xi17EhlDm1ZZ2PNDC4yb3BfTooJhZqbFgqhDsLT8cxGlT4vfMX/WISQlu+GdSf3wzsS++G5vCwhahdhnyvjj8Hw6F9Pe74VZ83rCz/cuxo1JrI2PRFVQvpvJkAMAMjIykJubKx7Tpk175P0yMjIwduxYxMTEwMrKqgY/adXIPpkZMWIEFAqFeDg7O6Nv37749ddfjX6vdevWwcvLC126dBHb5s+fjy5duqBevXoVMlAA8PX1RefOnbFkyRKjx0P6U1mWIqjPf/HpYi+cP+OAO+nWiFnlidvp1hjw8h0AwE+xLkhOqI/Mm9ZIv2KD9R80gY1dKbxaMCEl+Zk+pxfiDjXFjQxHXLteHx+t6AJXlwI0a/qH2OfN15Ow+/sW2LGzNW5kOOLmbQccPe6J4hIzAIBHo1x07HAbS1d1RurlBriQ4oI1Gzqie9frcKr/oLY+Gumj/DkzhhwA7O3tJYelpeUjb5eUlIS7d++iffv2MDc3h7m5OY4cOYIVK1bA3Nwcrq6u0Gg0yMnJkYzLysqCWq0GAKjV6gq7m8q/Lu9jLLJPZgCgb9++uHPnDu7cuYODBw/C3Nwc//rXv4x6D0EQsGrVKoSHh0vaNRoNXnzxRbz11luPHTty5EisXbu2WrabkX7MzAWYmQOaIoWkXVOohG+HvAr9zS206PdSJvLzzJB2ybamwiSqMpt6xQCA+/llf4QcHArh0+K/yMm1wtIP9mH75q+xaN5+tPK5K47xafE77uercPmqs9j28y9qCIICLZv/t2Y/AJmEXr164dy5c0hOThaPgIAAhIaGiv/bwsICBw8eFMekpqYiPT0dgYGBAIDAwECcO3cOd+/++bMYFxcHe3t7+Pr6GjVek0hmLC0toVaroVar4e/vj6lTpyIjIwO///47AGDKlClo3rw56tWrhyZNmmDmzJkoLi6WXOODDz6Aq6sr7OzsEB4ejsLCQsn5pKQkXL16FQMGDJC0z549G+PHj4efn99j4+vduzeys7Nx5MgRnZ+jqKiowuIrMq6HBea4eNYOL4/JgJNLEZRKAT0H3kVL/zw4NfxzGumZHn/gm6Tj2P3LcQwOu4Xpr/shL4drB0jeFAoB/xd+BucvNsSNdEcAgJvrfQDAqy/9ih/2N8P02c/hyjUnfDDnANzdyn7HONUvRE6u9F/gWq0S9++r4FRf+ruQ5MlY00yVZWdnh9atW0sOGxsbODs7o3Xr1nBwcEB4eDgmTJiAw4cPIykpCSNHjkRgYCA6d+4MAOjTpw98fX3x6quv4pdffsGPP/6IGTNmICIi4rEVoaoyiWTmr/Lz8/H555/D29sbzs5l/8qws7PD5s2bcfHiRSxfvhwbNmzA0qVLxTE7duxAVFQUFixYgDNnzsDNzQ1r1qyRXDc+Ph7NmzeHnZ2d3jGpVCr4+/sjPj5eZ7/o6Gg4ODiIx98XYpFxLJ7cAgqFgM+PnsK3vx7D86/ewpHvG0L7l62JvyQ6IvLf7fHuy22RFF8f05alwMGJa2ZI3iJHn4KnZw6iP+oqtin/V4Tcu78Z9h9qiqtpTvj40wDcvGWPkF5XaylSMrpafM7M4yxduhT/+te/MHToUHTr1g1qtRo7d+4Uz5uZmSE2NhZmZmYIDAzEK6+8gtdeew1z5swxeiwmsZspNjYWtrZlUwAFBQVwc3NDbGwslMqyXGzGjBli38aNG2PixInYvn07Jk+eDABYtmwZwsPDxSmkefPm4cCBA5LqzI0bN+Du7l7lGN3d3XHjxg2dfaZNm4YJEyaIX+fl5TGhqQaZGdaY8mpbWFqXop5tKe79rsLUJSnIzPhzEVvRQzPcSbfGnXRrpP5ijw37TiPkhSzsWM//P0ieIkadQqeOt/Due33w3z9sxPY/7lkDAG5kSHc3Zdx0gEvDsnVg2fes4Ogg3YKrVGphZ6dB9j35L+4kefjpp58kX1tZWWH16tVYvXr1Y8d4enpi79691RyZiVRmevbsKc7ZnTp1CiEhIejXr5+YPHz55ZcICgqCWq2Gra0tZsyYgfT0dHF8SkoKOnXqJLlm+ZxeuYcPHxq0Ytva2hoPHuheSGdpaVlh8RVVn6KHZrj3uwq29sVo3/UeTh5yfmxfpRKwUFXxqVJE1UpAxKhT6NI5A5NnBiPrrnRtV9ZdG/z3D2s0eko6bf2Uex7u/l6W9KSkNoSdrQbef1k07N8mEwqFgEu/Naj+j0AGq+lpJlNjEpUZGxsbeHt7i19/8skncHBwwIYNGzBgwACEhoZi9uzZCAkJgYODA7Zv346PPvpIr3s0aNAA586dq3KM2dnZaNq0aZXHk/G073oPCgi4mVYP7p4P8fqkNNy8Vg9xO11haV2K4f+XgZOHnHDvdxXs6xfjX/+5A2fXIsTv4y91kp/IN0+jZ7c0RC3ogYcPLVDf8SEAoOCBxf+eIaPA17t98erwX3EtrT6upTkh+Lmr8HgqD/MWdgNQVqU5neSOcWNOYuW6TjAz0yJi1GkcOdYY2ffq1eKno0rjW7N1Molk5u8UCgWUSiUePnyIEydOwNPTE9OnTxfP/326x8fHB4mJiXjttdfEtvLHLZdr164d1q5dC0EQoFBId8JUxvnz5/HCCy/oPY6Mz8a2BCMmXEcDdRHu55jjeFwDbFnaGKUlSiiVWjTyeoDpK7LgUL8YeTkW+O2cLSaFtkX6FZsnX5yohg3s9xsAYPH8OEn74hWBiDtU9g+oXXt8YGFRiv8LT4KdbRGuXa+PaVG9cCfzzzWAHy4NQsTo0/hgzgEIWgWOJTyNNZ8E1NwHIapGJpHMFBUViU8LvHfvHlatWoX8/HwMHDgQeXl5SE9Px/bt29GxY0d8//332LVrl2T82LFjMWLECAQEBCAoKAgxMTG4cOECmjRpIvbp2bMn8vPzceHCBbRu3VpsT09PR3Z2NtLT01FaWork5GQAgLe3t7iO5/r167h161aFxz1T7Yjf1xDx+xo+8lyxRon57xh3SyBRdQoZ/Eql+u3Y2Ro7drZ+7Pn7+Zb4YEnXx54neTN0qojTTDKwb98+uLm5ASjbudSyZUt89dVX6NGjB4CyN3dGRkaiqKgIAwYMwMyZMyWPZ37ppZdw9epVTJ48GYWFhRg6dCjeeust/Pjjj2IfZ2dn/Pvf/0ZMTAyio6PF9lmzZmHLli3i1+3atQMAHD58WLz/F198gT59+sDT07OavgNERPSPZqTXGdRVCkGo4xNpevj111/Ru3dvXL16Vay6PIlGo0GzZs2wbds2BAUF6XW/vLw8ODg4oJf9KzBXqKoSMpHslbZkkk91U0lJIX5KikZubm61bego/zsR2HcOzC2qvkmlpLgQCftmVWustckkdjPVlDZt2uDDDz9EWlpapcekp6fjvffe0zuRISIiqizuZtLNJKaZatKIESP06u/t7S3ZaUVERGR0WqHsMGR8HcZkhoiISO64ZkYnTjMRERGRSWNlhoiISOYUMHBrttEikScmM0RERHLHJwDrxGkmIiIiMmmszBAREckcnwCsG5MZIiIiueNuJp04zUREREQmjZUZIiIimVMIAhQGLOI1ZKwpYDJDREQkd9r/HYaMr8M4zUREREQmjZUZIiIimeM0k25MZoiIiOSOu5l0YjJDREQkd3wCsE5cM0NEREQmjZUZIiIimeMTgHVjMkNERCR3nGbSidNMREREZNJYmSEiIpI5hbbsMGR8XcZkhoiISO44zaQTp5mIiIjIpLEyQ0REJHd8aJ5OTGaIiIhkjq8z0I3TTERERGTSWJkhIiKSOy4A1onJDBERkdwJAAzZXl23cxkmM0RERHLHNTO6cc0MERERmTRWZoiIiOROgIFrZowWiSwxmSEiIpI7LgDWidNMREREZNJYmSEiIpI7LQCFgePrMCYzREREMsfdTLpxmomIiIhMGiszREREcscFwDqxMkNERCR35cmMIYceoqOj0bFjR9jZ2cHFxQWDBw9GamqqpE9hYSEiIiLg7OwMW1tbDB06FFlZWZI+6enpGDBgAOrVqwcXFxdMmjQJJSUlBn87/o7JDBEREUkcOXIEEREROHnyJOLi4lBcXIw+ffqgoKBA7DN+/Hjs2bMHX331FY4cOYLbt29jyJAh4vnS0lIMGDAAGo0GJ06cwJYtW7B582bMmjXL6PFymomIiEjuaniaad++fZKvN2/eDBcXFyQlJaFbt27Izc3Fxo0bsW3bNjz33HMAgE2bNsHHxwcnT55E586dsX//fly8eBEHDhyAq6sr/P39MXfuXEyZMgVRUVFQqVRV/zx/w8oMERGR3GmNcADIy8uTHEVFRZW6fW5uLgDAyckJAJCUlITi4mIEBweLfVq2bImnn34aCQkJAICEhAT4+fnB1dVV7BMSEoK8vDxcuHChKt+Fx2IyQ0REJHPlW7MNOQDAw8MDDg4O4hEdHf3Ee2u1WowbNw5BQUFo3bo1ACAzMxMqlQqOjo6Svq6ursjMzBT7/DWRKT9ffs6YOM1ERET0D5GRkQF7e3vxa0tLyyeOiYiIwPnz53Hs2LHqDM0gTGaIiIjkzkhrZuzt7SXJzJNERkYiNjYWR48eRaNGjcR2tVoNjUaDnJwcSXUmKysLarVa7HPq1CnJ9cp3O5X3MRZOMxEREcmdVjD80IMgCIiMjMSuXbtw6NAheHl5Sc536NABFhYWOHjwoNiWmpqK9PR0BAYGAgACAwNx7tw53L17V+wTFxcHe3t7+Pr6GvDNqIiVGSIiIpKIiIjAtm3b8O2338LOzk5c4+Lg4ABra2s4ODggPDwcEyZMgJOTE+zt7fH2228jMDAQnTt3BgD06dMHvr6+ePXVV7Fw4UJkZmZixowZiIiIqNT0lj6YzBAREcldDW/NXrt2LQCgR48ekvZNmzZhxIgRAIClS5dCqVRi6NChKCoqQkhICNasWSP2NTMzQ2xsLN566y0EBgbCxsYGYWFhmDNnTtU/x2MwmSEiIpI9A5MZ6D/N9CRWVlZYvXo1Vq9e/dg+np6e2Lt3r173rgqumSEiIiKTxsoMERGR3PFFkzoxmSEiIpI7rQB9p4oqjq+7OM1EREREJo2VGSIiIrkTtGWHIePrMCYzREREcsc1MzoxmSEiIpI7rpnRiWtmiIiIyKSxMkNERCR3nGbSickMERGR3AkwMJkxWiSyxGkmIiIiMmmszBAREckdp5l0YjJDREQkd1otAAOeFaOt28+Z4TQTERERmTRWZoiIiOSO00w6MZkhIiKSOyYzOnGaiYiIiEwaKzNERERyx9cZ6MRkhoiISOYEQQvBgDdfGzLWFDCZISIikjtBMKy6wjUzRERERPLFygwREZHcCQaumanjlRkmM0RERHKn1QIKA9a91PE1M5xmIiIiIpPGygwREZHccZpJJyYzREREMidotRAMmGaq61uzOc1EREREJo2VGSIiIrnjNJNOTGaIiIjkTisACiYzj8NpJiIiIjJprMwQERHJnSAAMOQ5M3W7MsNkhoiISOYErQDBgGkmgckMERER1SpBC8MqM9yaTURERCRbrMwQERHJHKeZdGMyQ0REJHecZtKJyUwtKs+USwRNLUdCVH1KSwprOwSialFSWgSgZqoeJSg26Jl5JSg2XjAyxGSmFt2/fx8AcOT+jlqOhKgaJdV2AETV6/79+3BwcKiWa6tUKqjVahzL3GvwtdRqNVQqlRGikh+FUNcn0mRMq9Xi9u3bsLOzg0KhqO1w6ry8vDx4eHggIyMD9vb2tR0OkdHxZ7xmCYKA+/fvw93dHUpl9e2nKSwshEZjeAVfpVLBysrKCBHJDysztUipVKJRo0a1HcY/jr29PX/RU53Gn/GaU10Vmb+ysrKqs0mIsXBrNhEREZk0JjNERERk0pjM0D+GpaUl3n//fVhaWtZ2KETVgj/j9E/FBcBERERk0liZISIiIpPGZIaIiIhMGpMZIiIiMmlMZoiIiMikMZkhAqBQKLB7925J28aNG9GnTx+9rtO5c2d88803RoyMSD9RUVHw9/eXtGk0Gnh7e+PEiROVvs7FixfRqFEjFBQUGDlCIuNjMkOykZmZibFjx8Lb2xtWVlZwdXVFUFAQ1q5diwcPHtRoLIWFhZg5cybef/99SftXX32Fli1bwsrKCn5+fti7V/q+lBkzZmDq1KnQauv2G2rJcCNGjIBCoRAPZ2dn9O3bF7/++qvR77Vu3Tp4eXmhS5cuYtv8+fPRpUsX1KtXD46OjhXG+Pr6onPnzliyZInR4yEyNiYzJAvXrl1Du3btsH//fixYsABnz55FQkICJk+ejNjYWBw4cKBG4/n6669hb2+PoKAgse3EiRN4+eWXER4ejrNnz2Lw4MEYPHgwzp8/L/bp168f7t+/jx9++KFG4yXT1LdvX9y5cwd37tzBwYMHYW5ujn/9619GvYcgCFi1ahXCw8Ml7RqNBi+++CLeeuutx44dOXIk1q5di5KSEqPGRGR0ApEMhISECI0aNRLy8/MfeV6r1QqCIAg3btwQnn/+ecHGxkaws7MTXnzxRSEzM1PSd82aNUKTJk0ECwsLoXnz5sLWrVsl53/77Tfh2WefFSwtLQUfHx9h//79AgBh165dYp8BAwYIEydOlIwbNmyYMGDAAElbp06dhDfffFPSNnLkSOGVV17R6/PTP09YWJgwaNAgSVt8fLwAQLh7964gCIIwefJkoVmzZoK1tbXg5eUlzJgxQ9BoNJIx0dHRgouLi2Brayu8/vrrwpQpU4S2bduK50+fPi0olUohLy/vkXFs2rRJcHBweOS5oqIiwdLSUjhw4ECVPydRTWBlhmrdH3/8gf379yMiIgI2NjaP7KNQKKDVajFo0CBkZ2fjyJEjiIuLw7Vr1/DSSy+J/Xbt2oWxY8fi3Xffxfnz5/Hmm29i5MiROHz4MICyN5UPGTIEKpUKiYmJWLduHaZMmVLhfseOHUNAQICkLSEhAcHBwZK2kJAQJCQkSNqeeeYZxMfHV+l7Qf9c+fn5+Pzzz+Ht7Q1nZ2cAgJ2dHTZv3oyLFy9i+fLl2LBhA5YuXSqO2bFjB6KiorBgwQKcOXMGbm5uWLNmjeS68fHxaN68Oezs7PSOSaVSwd/fnz/PJHt8azbVuitXrkAQBLRo0ULS3qBBAxQWFgIAIiIiEBwcjHPnziEtLQ0eHh4AgK1bt6JVq1Y4ffo0OnbsiMWLF2PEiBEYM2YMAGDChAk4efIkFi9ejJ49e+LAgQO4dOkSfvzxR7i7uwMAFixYgH79+on3zcnJQW5urni+XGZmJlxdXSVtrq6uyMzMlLS5u7sjIyMDWq0WSiX/vUCPFxsbC1tbWwBAQUEB3NzcEBsbK/7czJgxQ+zbuHFjTJw4Edu3b8fkyZMBAMuWLUN4eLg4hTRv3jwcOHBA/O8GAG7cuFHhZ1kf7u7uuHHjRpXHE9UE/qYl2Tp16hSSk5PRqlUrFBUVISUlBR4eHmIiA5QtUnR0dERKSgoAICUlRbLOBQCCgoIk5z08PCS/3AMDAyX9Hz58CACwsrKqUtzW1tbQarUoKiqq0nj65+jZsyeSk5ORnJyMU6dOISQkBP369ROThy+//BJBQUFQq9WwtbXFjBkzkJ6eLo5PSUlBp06dJNd81M9zVX+WgbKf55pegE+kLyYzVOu8vb2hUCiQmpoqaW/SpAm8vb1hbW1do/E4OztDoVDg3r17kna1Wo2srCxJW1ZWFtRqtaQtOzsbNjY2NR43mR4bGxt4e3vD29sbHTt2xCeffIKCggJs2LABCQkJCA0NRf/+/REbG4uzZ89i+vTp0Gg0et2jQYMGFX6W9ZGdnY2GDRtWeTxRTWAyQ7XO2dkZvXv3xqpVq3Q+08LHxwcZGRnIyMgQ2y5evIicnBz4+vqKfY4fPy4Zd/z4ccn5jIwM3LlzRzx/8uRJSX+VSgVfX19cvHhR0h4YGIiDBw9K2uLi4ir8S/j8+fNo167dkz42UQUKhQJKpRIPHz7EiRMn4OnpienTpyMgIADNmjWrMN3j4+ODxMRESdvff57btWuHS5cuQajiO4X580ymgGtmSBbWrFmDoKAgBAQEICoqCm3atIFSqcTp06dx6dIldOjQAcHBwfDz80NoaCiWLVuGkpISjBkzBt27dxcX606aNAnDhg1Du3btEBwcjD179mDnzp3i1u7g4GA0b94cYWFhWLRoEfLy8jB9+vQK8YSEhODYsWMYN26c2DZ27Fh0794dH330EQYMGIDt27fjzJkzWL9+vWRsfHy83g/bo3+moqIicc3VvXv3sGrVKuTn52PgwIHIy8tDeno6tm/fjo4dO+L777/Hrl27JOPHjh2LESNGICAgAEFBQYiJicGFCxfQpEkTsU/Pnj2Rn5+PCxcuoHXr1mJ7eno6srOzkZ6ejtLSUiQnJwMoq5SWr+O5fv06bt26VWHhO5Hs1PZ2KqJyt2/fFiIjIwUvLy/BwsJCsLW1FZ555hlh0aJFQkFBgSAIxtmanZqaKnTt2lVQqVRC8+bNhX379lXYmn3hwgXB2tpayMnJkYzdsWOH0Lx5c0GlUgmtWrUSvv/+e8n5mzdvChYWFkJGRoYRvzNUF4WFhQkAxMPOzk7o2LGj8PXXX4t9Jk2aJDg7Owu2trbCSy+9JCxdurTCNur58+cLDRo0EGxtbYWwsDBh8uTJkq3ZglD2WIGpU6fqvH/5cfjwYbHPggULhJCQEGN/dCKjUwhCFWuPRHXciy++iPbt22PatGmVHjNlyhTcu3evQrWGqDb9+uuv6N27N65evSpWXZ5Eo9GgWbNm2LZtW4VF9URywzUzRI+xaNGiSv/iL+fi4oK5c+dWU0REVdOmTRt8+OGHSEtLq/SY9PR0vPfee0xkyCSwMkNEREQmjZUZIiIiMmlMZoiIiMikMZkhIiIik8ZkhoiIiEwakxkiIiIyaUxmiP7hRowYgcGDB4tf9+jRQ/Lk45ry008/QaFQICcn57F9FAoFdu/eXelrRkVFwd/f36C4rl+/DoVCIT4hl4jkh8kMkQyNGDECCoUCCoUCKpUK3t7emDNnDkpKSqr93jt37qz0s3Iqk4AQEVU3vpuJSKb69u2LTZs2oaioCHv37kVERAQsLCwe+URijUYDlUpllPs6OTkZ5TpERDWFlRkimbK0tIRarYanpyfeeustBAcH47vvvgPw59TQ/Pnz4e7ujhYtWgAAMjIyMGzYMDg6OsLJyQmDBg3C9evXxWuWlpZiwoQJcHR0hLOzMyZPnlzhbcp/n2YqKirClClT4OHhAUtLS3h7e2Pjxo24fv06evbsCQCoX78+FAoFRowYAQDQarWIjo6Gl5cXrK2t0bZtW3z99deS++zduxfNmzeHtbU1evbsKYmzsqZMmYLmzZujXr16aNKkCWbOnIni4uIK/T7++GN4eHigXr16GDZsGHJzcyXnP/nkE/j4+MDKygotW7bEmjVr9I6FiGoPkxkiE2FtbQ2NRiN+ffDgQaSmpiIuLg6xsbEoLi5GSEgI7OzsEB8fj+PHj8PW1hZ9+/YVx3300UfYvHkzPv30Uxw7dgzZ2dkV3sT8d6+99hq++OILrFixAikpKfj4449ha2sLDw8PfPPNNwCA1NRU3LlzB8uXLwcAREdHY+vWrVi3bh0uXLiA8ePH45VXXsGRI0cAlCVdQ4YMwcCBA5GcnIw33ngDU6dO1ft7Ymdnh82bN+PixYtYvnw5NmzYgKVLl0r6XLlyBTt27MCePXuwb98+nD17FmPGjBHPx8TEYNasWZg/fz5SUlKwYMECzJw5E1u2bNE7HiKqJbX6mksieqSwsDBh0KBBgiAIglarFeLi4gRLS0th4sSJ4nlXV1ehqKhIHPPZZ58JLVq0ELRardhWVFQkWFtbCz/++KMgCILg5uYmLFy4UDxfXFwsNGrUSLyXIAhC9+7dhbFjxwqCUPaGcQBCXFzcI+M8fPiwAEC4d++e2FZYWCjUq1dPOHHihKRveHi48PLLLwuCIAjTpk0TfH19JeenTJlS4Vp/h7+93fzvFi1aJHTo0EH8+v333xfMzMyEmzdvim0//PCDoFQqhTt37giCIAhNmzYVtm3bJrnO3LlzhcDAQEEQBCEtLU0AIJw9e/ax9yWi2sU1M0QyFRsbC1tbWxQXF0Or1eI///kPoqKixPN+fn6SdTK//PILrly5Ajs7O8l1CgsLcfXqVeTm5uLOnTvo1KmTeM7c3BwBAQEVpprKJScnw8zMDN27d6903FeuXMGDBw/Qu3dvSbtGo0G7du0AACkpKZI4ACAwMLDS9yj35ZdfYsWKFbh69Sry8/NRUlICe3t7SZ+nn34aTz31lOQ+Wq0WqampsLOzw9WrVxEeHo5Ro0aJfUpKSuDg4KB3PERUO5jMEMlUz549sXbtWqhUKri7u8PcXPqfq42NjeTr/Px8dOjQATExMRWu1bBhwyrFYG1trfeY/Px8AMD3338vSSKAsnVAxpKQkIDQ0FDMnj0bISEhcHBwwPbt2/HRRx/pHeuGDRsqJFdmZmZGi5WIqheTGSKZsrGxgbe3d6X7t2/fHl9++SVcXFwqVCfKubm5ITExEd26dQNQVoFISkpC+/btH9nfz88PWq0WR44cQXBwcIXz5ZWh0tJSsc3X1xeWlpZIT09/bEXHx8dHXMxc7uTJk0/+kH9x4sQJeHp6Yvr06WLbjRs3KvRLT0/H7du34e7uLt5HqVSiRYsWcHV1hbu7O65du4bQ0FC97k9E8sEFwER1RGhoKBo0aIBBgwYhPj4eaWlp+Omnn/DOO+/g5s2bAICxY8figw8+wO7du3Hp0iWMGTNG5zNiGjdujLCwMLz++uvYvXu3eM0dO3YAADw9PaFQKBAbG4vff/8d+fn5sLOzw8SJEzF+/Hhs2bIFV69exc8//4yVK1eKi2r/7//+D5cvX8akSZOQmpqKbdu2YfPmzXp93mbNmiE9PR3bt2/H1atXsWLFikcuZrayskJYWBh++eUXxMfH45133sGwYcOgVqsBALNnz0Z0dDRWrFiB3377DefOncOmTZuwZMkSveIhotrDZIaojqhXrx6OHj2Kp59+GkOGDIGPjw/Cw8NRWFgoVmreffddvPrqqwgLC0NgYCDs7Ozw73//W+d1165dixdeeAFjxoxBy5YtMWrUKBQUFAAAnnrqKcyePRtTp06Fq6srIiMjAQBz587FzJkzER0dDR8fH/Tt2xfff/89vLy8AJStY/nmm2+we/dutG3bFuvWrcOCBQv0+rzPP/88xo8fj8jISPj7++PEiROYOXNmhX7e3t4YMmQI+vfvjz59+qBNmzaSrddvvPEGPvnkE2zatAl+fn7o3r07Nm/eLMZKRPKnEB638o+IiIjIBLAyQ0RERCaNyQwRERGZNCYzREREZNKYzBAREZFJYzJDREREJo3JDBEREZk0JjNERERk0pjMEBERkUljMkNEREQmjckMERERmTQmM0RERGTS/h9oOeFnsGkiswAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ConfusionMatrixDisplay.from_estimator(estimator=gs_nn, X=Xtest, y=ytest, labels=[0, 1], display_labels=['Good(0)', 'Bad(1)']);\n",
    "print(gs_nn.score(Xtest, ytest))"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}