ChurnPrediction_johnson.ipynb

{
 "cells": [
  {
   "attachments": {
    "COUR_IPO.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![COUR_IPO.png](attachment:COUR_IPO.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Welcome to the Data Science Coding Challange!\n",
    "\n",
    "Test your skills in a real-world coding challenge. Coding Challenges provide CS & DS Coding Competitions with Prizes and achievement badges!\n",
    "\n",
    "CS & DS learners want to be challenged as a way to evaluate if they’re job ready. So, why not create fun challenges and give winners something truly valuable such as complimentary access to select Data Science courses, or the ability to receive an achievement badge on their Coursera Skills Profile - highlighting their performance to recruiters."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "\n",
    "In this challenge, you'll get the opportunity to tackle one of the most industry-relevant maching learning problems with a unique dataset that will put your modeling skills to the test. Subscription services are leveraged by companies across many industries, from fitness to video streaming to retail. One of the primary objectives of companies with subscription services is to decrease churn and ensure that users are retained as subscribers. In order to do this efficiently and systematically, many companies employ machine learning to predict which users are at the highest risk of churn, so that proper interventions can be effectively deployed to the right audience.\n",
    "\n",
    "In this challenge, we will be tackling the churn prediction problem on a very unique and interesting group of subscribers on a video streaming service! \n",
    "\n",
    "Imagine that you are a new data scientist at this video streaming company and you are tasked with building a model that can predict which existing subscribers will continue their subscriptions for another month. We have provided a dataset that is a sample of subscriptions that were initiated in 2021, all snapshotted at a particular date before the subscription was cancelled. Subscription cancellation can happen for a multitude of reasons, including:\n",
    "* the customer completes all content they were interested in, and no longer need the subscription\n",
    "* the customer finds themselves to be too busy and cancels their subscription until a later time\n",
    "* the customer determines that the streaming service is not the best fit for them, so they cancel and look for something better suited\n",
    "\n",
    "Regardless the reason, this video streaming company has a vested interest in understanding the likelihood of each individual customer to churn in their subscription so that resources can be allocated appropriately to support customers. In this challenge, you will use your machine learning toolkit to do just that!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Understanding the Datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train vs. Test\n",
    "In this competition, you’ll gain access to two datasets that are samples of past subscriptions of a video streaming platform that contain information about the customer, the customers streaming preferences, and their activity in the subscription thus far. One dataset is titled `train.csv` and the other is titled `test.csv`.\n",
    "\n",
    "`train.csv` contains 70% of the overall sample (243,787 subscriptions to be exact) and importantly, will reveal whether or not the subscription was continued into the next month (the “ground truth”).\n",
    "\n",
    "The `test.csv` dataset contains the exact same information about the remaining segment of the overall sample (104,480 subscriptions to be exact), but does not disclose the “ground truth” for each subscription. It’s your job to predict this outcome!\n",
    "\n",
    "Using the patterns you find in the `train.csv` data, predict whether the subscriptions in `test.csv` will be continued for another month, or not."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dataset descriptions\n",
    "Both `train.csv` and `test.csv` contain one row for each unique subscription. For each subscription, a single observation (`CustomerID`) is included during which the subscription was active. \n",
    "\n",
    "In addition to this identifier column, the `train.csv` dataset also contains the target label for the task, a binary column `Churn`.\n",
    "\n",
    "Besides that column, both datasets have an identical set of features that can be used to train your model to make predictions. Below you can see descriptions of each feature. Familiarize yourself with them so that you can harness them most effectively for this machine learning task!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Column_name</th>\n",
       "      <th>Column_type</th>\n",
       "      <th>Data_type</th>\n",
       "      <th>Description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AccountAge</td>\n",
       "      <td>Feature</td>\n",
       "      <td>integer</td>\n",
       "      <td>The age of the user's account in months.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>MonthlyCharges</td>\n",
       "      <td>Feature</td>\n",
       "      <td>float</td>\n",
       "      <td>The amount charged to the user on a monthly basis.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TotalCharges</td>\n",
       "      <td>Feature</td>\n",
       "      <td>float</td>\n",
       "      <td>The total charges incurred by the user over the account's lifetime.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SubscriptionType</td>\n",
       "      <td>Feature</td>\n",
       "      <td>object</td>\n",
       "      <td>The type of subscription chosen by the user (Basic, Standard, or Premium).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>PaymentMethod</td>\n",
       "      <td>Feature</td>\n",
       "      <td>string</td>\n",
       "      <td>The method of payment used by the user.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>PaperlessBilling</td>\n",
       "      <td>Feature</td>\n",
       "      <td>string</td>\n",
       "      <td>Indicates whether the user has opted for paperless billing (Yes or No).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ContentType</td>\n",
       "      <td>Feature</td>\n",
       "      <td>string</td>\n",
       "      <td>The type of content preferred by the user (Movies, TV Shows, or Both).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>MultiDeviceAccess</td>\n",
       "      <td>Feature</td>\n",
       "      <td>string</td>\n",
       "      <td>Indicates whether the user has access to the service on multiple devices (Yes or No).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>DeviceRegistered</td>\n",
       "      <td>Feature</td>\n",
       "      <td>string</td>\n",
       "      <td>The type of device registered by the user (TV, Mobile, Tablet, or Computer).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ViewingHoursPerWeek</td>\n",
       "      <td>Feature</td>\n",
       "      <td>float</td>\n",
       "      <td>The number of hours the user spends watching content per week.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>AverageViewingDuration</td>\n",
       "      <td>Feature</td>\n",
       "      <td>float</td>\n",
       "      <td>The average duration of each viewing session in minutes.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>ContentDownloadsPerMonth</td>\n",
       "      <td>Feature</td>\n",
       "      <td>integer</td>\n",
       "      <td>The number of content downloads by the user per month.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>GenrePreference</td>\n",
       "      <td>Feature</td>\n",
       "      <td>string</td>\n",
       "      <td>The preferred genre of content chosen by the user.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>UserRating</td>\n",
       "      <td>Feature</td>\n",
       "      <td>float</td>\n",
       "      <td>The user's rating for the service on a scale of 1 to 5.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>SupportTicketsPerMonth</td>\n",
       "      <td>Feature</td>\n",
       "      <td>integer</td>\n",
       "      <td>The number of support tickets raised by the user per month.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Gender</td>\n",
       "      <td>Feature</td>\n",
       "      <td>string</td>\n",
       "      <td>The gender of the user (Male or Female).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>WatchlistSize</td>\n",
       "      <td>Feature</td>\n",
       "      <td>float</td>\n",
       "      <td>The number of items in the user's watchlist.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>ParentalControl</td>\n",
       "      <td>Feature</td>\n",
       "      <td>string</td>\n",
       "      <td>Indicates whether parental control is enabled for the user (Yes or No).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>SubtitlesEnabled</td>\n",
       "      <td>Feature</td>\n",
       "      <td>string</td>\n",
       "      <td>Indicates whether subtitles are enabled for the user (Yes or No).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>CustomerID</td>\n",
       "      <td>Identifier</td>\n",
       "      <td>string</td>\n",
       "      <td>A unique identifier for each customer.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Churn</td>\n",
       "      <td>Target</td>\n",
       "      <td>integer</td>\n",
       "      <td>The target variable indicating whether a user has churned or not (1 for churned, 0 for not churned).</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Column_name Column_type Data_type   \n",
       "0                 AccountAge     Feature   integer  \\\n",
       "1             MonthlyCharges     Feature     float   \n",
       "2               TotalCharges     Feature     float   \n",
       "3           SubscriptionType     Feature    object   \n",
       "4              PaymentMethod     Feature    string   \n",
       "5           PaperlessBilling     Feature    string   \n",
       "6                ContentType     Feature    string   \n",
       "7          MultiDeviceAccess     Feature    string   \n",
       "8           DeviceRegistered     Feature    string   \n",
       "9        ViewingHoursPerWeek     Feature     float   \n",
       "10    AverageViewingDuration     Feature     float   \n",
       "11  ContentDownloadsPerMonth     Feature   integer   \n",
       "12           GenrePreference     Feature    string   \n",
       "13                UserRating     Feature     float   \n",
       "14    SupportTicketsPerMonth     Feature   integer   \n",
       "15                    Gender     Feature    string   \n",
       "16             WatchlistSize     Feature     float   \n",
       "17           ParentalControl     Feature    string   \n",
       "18          SubtitlesEnabled     Feature    string   \n",
       "19                CustomerID  Identifier    string   \n",
       "20                     Churn      Target   integer   \n",
       "\n",
       "                                                                                             Description  \n",
       "0                                                               The age of the user's account in months.  \n",
       "1                                                     The amount charged to the user on a monthly basis.  \n",
       "2                                    The total charges incurred by the user over the account's lifetime.  \n",
       "3                             The type of subscription chosen by the user (Basic, Standard, or Premium).  \n",
       "4                                                                The method of payment used by the user.  \n",
       "5                                Indicates whether the user has opted for paperless billing (Yes or No).  \n",
       "6                                 The type of content preferred by the user (Movies, TV Shows, or Both).  \n",
       "7                  Indicates whether the user has access to the service on multiple devices (Yes or No).  \n",
       "8                           The type of device registered by the user (TV, Mobile, Tablet, or Computer).  \n",
       "9                                         The number of hours the user spends watching content per week.  \n",
       "10                                              The average duration of each viewing session in minutes.  \n",
       "11                                                The number of content downloads by the user per month.  \n",
       "12                                                    The preferred genre of content chosen by the user.  \n",
       "13                                               The user's rating for the service on a scale of 1 to 5.  \n",
       "14                                           The number of support tickets raised by the user per month.  \n",
       "15                                                              The gender of the user (Male or Female).  \n",
       "16                                                          The number of items in the user's watchlist.  \n",
       "17                               Indicates whether parental control is enabled for the user (Yes or No).  \n",
       "18                                     Indicates whether subtitles are enabled for the user (Yes or No).  \n",
       "19                                                                A unique identifier for each customer.  \n",
       "20  The target variable indicating whether a user has churned or not (1 for churned, 0 for not churned).  "
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "data_descriptions = pd.read_csv('data_descriptions.csv')\n",
    "pd.set_option('display.max_colwidth', None)\n",
    "data_descriptions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## How to Submit your Predictions to Coursera\n",
    "Submission Format:\n",
    "\n",
    "In this notebook you should follow the steps below to explore the data, train a model using the data in `train.csv`, and then score your model using the data in `test.csv`. Your final submission should be a dataframe (call it `prediction_df` with two columns and exactly 104,480 rows (plus a header row). The first column should be `CustomerID` so that we know which prediction belongs to which observation. The second column should be called `predicted_probability` and should be a numeric column representing the __likellihood that the subscription will churn__.\n",
    "\n",
    "Your submission will show an error if you have extra columns (beyond `CustomerID` and `predicted_probability`) or extra rows. The order of the rows does not matter.\n",
    "\n",
    "The naming convention of the dataframe and columns are critical for our autograding, so please make sure to use the exact naming conventions of `prediction_df` with column names `CustomerID` and `predicted_probability`!\n",
    "\n",
    "To determine your final score, we will compare your `predicted_probability` predictions to the source of truth labels for the observations in `test.csv` and calculate the [ROC AUC](https://scikit-learn.org/stable/modules/generated/sklearn.metrics.roc_auc_score.html). We choose this metric because we not only want to be able to predict which subscriptions will be retained, but also want a well-calibrated likelihood score that can be used to target interventions and support most accurately."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import Python Modules\n",
    "\n",
    "First, import the primary modules that will be used in this project. Remember as this is an open-ended project please feel free to make use of any of your favorite libraries that you feel may be useful for this challenge. For example some of the following popular packages may be useful:\n",
    "\n",
    "- pandas\n",
    "- numpy\n",
    "- Scipy\n",
    "- Scikit-learn\n",
    "- keras\n",
    "- maplotlib\n",
    "- seaborn\n",
    "- etc, etc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import required packages\n",
    "\n",
    "# Data packages\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "# Machine Learning / Classification packages\n",
    "from sklearn.metrics import roc_auc_score, roc_curve, accuracy_score, classification_report\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.dummy import DummyClassifier\n",
    "\n",
    "# Visualization Packages\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import any other packages you may want to use\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load the Data\n",
    "\n",
    "Let's start by loading the dataset `train.csv` into a dataframe `train_df`, and `test.csv` into a dataframe `test_df` and display the shape of the dataframes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train_df Shape: (243787, 21)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>SubscriptionType</th>\n",
       "      <th>PaymentMethod</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>ContentType</th>\n",
       "      <th>MultiDeviceAccess</th>\n",
       "      <th>DeviceRegistered</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>...</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>GenrePreference</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>Gender</th>\n",
       "      <th>WatchlistSize</th>\n",
       "      <th>ParentalControl</th>\n",
       "      <th>SubtitlesEnabled</th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>20</td>\n",
       "      <td>11.055215</td>\n",
       "      <td>221.104302</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>No</td>\n",
       "      <td>Both</td>\n",
       "      <td>No</td>\n",
       "      <td>Mobile</td>\n",
       "      <td>36.758104</td>\n",
       "      <td>...</td>\n",
       "      <td>10</td>\n",
       "      <td>Sci-Fi</td>\n",
       "      <td>2.176498</td>\n",
       "      <td>4</td>\n",
       "      <td>Male</td>\n",
       "      <td>3</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>CB6SXPNVZA</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>57</td>\n",
       "      <td>5.175208</td>\n",
       "      <td>294.986882</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>32.450568</td>\n",
       "      <td>...</td>\n",
       "      <td>18</td>\n",
       "      <td>Action</td>\n",
       "      <td>3.478632</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>S7R2G87O09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>73</td>\n",
       "      <td>12.106657</td>\n",
       "      <td>883.785952</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Computer</td>\n",
       "      <td>7.395160</td>\n",
       "      <td>...</td>\n",
       "      <td>23</td>\n",
       "      <td>Fantasy</td>\n",
       "      <td>4.238824</td>\n",
       "      <td>6</td>\n",
       "      <td>Male</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>EASDC20BDT</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>32</td>\n",
       "      <td>7.263743</td>\n",
       "      <td>232.439774</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>27.960389</td>\n",
       "      <td>...</td>\n",
       "      <td>30</td>\n",
       "      <td>Drama</td>\n",
       "      <td>4.276013</td>\n",
       "      <td>2</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>NPF69NT69N</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57</td>\n",
       "      <td>16.953078</td>\n",
       "      <td>966.325422</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>TV</td>\n",
       "      <td>20.083397</td>\n",
       "      <td>...</td>\n",
       "      <td>20</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>3.616170</td>\n",
       "      <td>4</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>4LGYPK7VOL</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   AccountAge  MonthlyCharges  TotalCharges SubscriptionType   \n",
       "0          20       11.055215    221.104302          Premium  \\\n",
       "1          57        5.175208    294.986882            Basic   \n",
       "2          73       12.106657    883.785952            Basic   \n",
       "3          32        7.263743    232.439774            Basic   \n",
       "4          57       16.953078    966.325422          Premium   \n",
       "\n",
       "      PaymentMethod PaperlessBilling ContentType MultiDeviceAccess   \n",
       "0      Mailed check               No        Both                No  \\\n",
       "1       Credit card              Yes      Movies                No   \n",
       "2      Mailed check              Yes      Movies                No   \n",
       "3  Electronic check               No    TV Shows                No   \n",
       "4  Electronic check              Yes    TV Shows                No   \n",
       "\n",
       "  DeviceRegistered  ViewingHoursPerWeek  ...  ContentDownloadsPerMonth   \n",
       "0           Mobile            36.758104  ...                        10  \\\n",
       "1           Tablet            32.450568  ...                        18   \n",
       "2         Computer             7.395160  ...                        23   \n",
       "3           Tablet            27.960389  ...                        30   \n",
       "4               TV            20.083397  ...                        20   \n",
       "\n",
       "   GenrePreference UserRating  SupportTicketsPerMonth  Gender WatchlistSize   \n",
       "0           Sci-Fi   2.176498                       4    Male             3  \\\n",
       "1           Action   3.478632                       8    Male            23   \n",
       "2          Fantasy   4.238824                       6    Male             1   \n",
       "3            Drama   4.276013                       2    Male            24   \n",
       "4           Comedy   3.616170                       4  Female             0   \n",
       "\n",
       "   ParentalControl SubtitlesEnabled  CustomerID Churn  \n",
       "0               No               No  CB6SXPNVZA     0  \n",
       "1               No              Yes  S7R2G87O09     0  \n",
       "2              Yes              Yes  EASDC20BDT     0  \n",
       "3              Yes              Yes  NPF69NT69N     0  \n",
       "4               No               No  4LGYPK7VOL     0  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df = pd.read_csv(\"train.csv\")\n",
    "print('train_df Shape:', train_df.shape)\n",
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['AccountAge', 'MonthlyCharges', 'TotalCharges', 'SubscriptionType',\n",
       "       'PaymentMethod', 'PaperlessBilling', 'ContentType', 'MultiDeviceAccess',\n",
       "       'DeviceRegistered', 'ViewingHoursPerWeek', 'AverageViewingDuration',\n",
       "       'ContentDownloadsPerMonth', 'GenrePreference', 'UserRating',\n",
       "       'SupportTicketsPerMonth', 'Gender', 'WatchlistSize', 'ParentalControl',\n",
       "       'SubtitlesEnabled', 'CustomerID', 'Churn'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 243787 entries, 0 to 243786\n",
      "Data columns (total 21 columns):\n",
      " #   Column                    Non-Null Count   Dtype  \n",
      "---  ------                    --------------   -----  \n",
      " 0   AccountAge                243787 non-null  int64  \n",
      " 1   MonthlyCharges            243787 non-null  float64\n",
      " 2   TotalCharges              243787 non-null  float64\n",
      " 3   SubscriptionType          243787 non-null  object \n",
      " 4   PaymentMethod             243787 non-null  object \n",
      " 5   PaperlessBilling          243787 non-null  object \n",
      " 6   ContentType               243787 non-null  object \n",
      " 7   MultiDeviceAccess         243787 non-null  object \n",
      " 8   DeviceRegistered          243787 non-null  object \n",
      " 9   ViewingHoursPerWeek       243787 non-null  float64\n",
      " 10  AverageViewingDuration    243787 non-null  float64\n",
      " 11  ContentDownloadsPerMonth  243787 non-null  int64  \n",
      " 12  GenrePreference           243787 non-null  object \n",
      " 13  UserRating                243787 non-null  float64\n",
      " 14  SupportTicketsPerMonth    243787 non-null  int64  \n",
      " 15  Gender                    243787 non-null  object \n",
      " 16  WatchlistSize             243787 non-null  int64  \n",
      " 17  ParentalControl           243787 non-null  object \n",
      " 18  SubtitlesEnabled          243787 non-null  object \n",
      " 19  CustomerID                243787 non-null  object \n",
      " 20  Churn                     243787 non-null  int64  \n",
      "dtypes: float64(5), int64(5), object(11)\n",
      "memory usage: 39.1+ MB\n"
     ]
    }
   ],
   "source": [
    "train_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test_df Shape: (104480, 20)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>SubscriptionType</th>\n",
       "      <th>PaymentMethod</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>ContentType</th>\n",
       "      <th>MultiDeviceAccess</th>\n",
       "      <th>DeviceRegistered</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>AverageViewingDuration</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>GenrePreference</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>Gender</th>\n",
       "      <th>WatchlistSize</th>\n",
       "      <th>ParentalControl</th>\n",
       "      <th>SubtitlesEnabled</th>\n",
       "      <th>CustomerID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>38</td>\n",
       "      <td>17.869374</td>\n",
       "      <td>679.036195</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>TV</td>\n",
       "      <td>29.126308</td>\n",
       "      <td>122.274031</td>\n",
       "      <td>42</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>3.522724</td>\n",
       "      <td>2</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>O1W6BHP6RM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>77</td>\n",
       "      <td>9.912854</td>\n",
       "      <td>763.289768</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>TV</td>\n",
       "      <td>36.873729</td>\n",
       "      <td>57.093319</td>\n",
       "      <td>43</td>\n",
       "      <td>Action</td>\n",
       "      <td>2.021545</td>\n",
       "      <td>2</td>\n",
       "      <td>Female</td>\n",
       "      <td>22</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>LFR4X92X8H</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5</td>\n",
       "      <td>15.019011</td>\n",
       "      <td>75.095057</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Bank transfer</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Computer</td>\n",
       "      <td>7.601729</td>\n",
       "      <td>140.414001</td>\n",
       "      <td>14</td>\n",
       "      <td>Sci-Fi</td>\n",
       "      <td>4.806126</td>\n",
       "      <td>2</td>\n",
       "      <td>Female</td>\n",
       "      <td>22</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>QM5GBIYODA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>88</td>\n",
       "      <td>15.357406</td>\n",
       "      <td>1351.451692</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>No</td>\n",
       "      <td>Both</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>35.586430</td>\n",
       "      <td>177.002419</td>\n",
       "      <td>14</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>4.943900</td>\n",
       "      <td>0</td>\n",
       "      <td>Female</td>\n",
       "      <td>23</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>D9RXTK2K9F</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>91</td>\n",
       "      <td>12.406033</td>\n",
       "      <td>1128.949004</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>23.503651</td>\n",
       "      <td>70.308376</td>\n",
       "      <td>6</td>\n",
       "      <td>Drama</td>\n",
       "      <td>2.846880</td>\n",
       "      <td>6</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>ENTCCHR1LR</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AccountAge  MonthlyCharges  TotalCharges SubscriptionType   \n",
       "0          38       17.869374    679.036195          Premium  \\\n",
       "1          77        9.912854    763.289768            Basic   \n",
       "2           5       15.019011     75.095057         Standard   \n",
       "3          88       15.357406   1351.451692         Standard   \n",
       "4          91       12.406033   1128.949004         Standard   \n",
       "\n",
       "      PaymentMethod PaperlessBilling ContentType MultiDeviceAccess   \n",
       "0      Mailed check               No    TV Shows                No  \\\n",
       "1  Electronic check              Yes    TV Shows                No   \n",
       "2     Bank transfer               No    TV Shows               Yes   \n",
       "3  Electronic check               No        Both               Yes   \n",
       "4       Credit card              Yes    TV Shows               Yes   \n",
       "\n",
       "  DeviceRegistered  ViewingHoursPerWeek  AverageViewingDuration   \n",
       "0               TV            29.126308              122.274031  \\\n",
       "1               TV            36.873729               57.093319   \n",
       "2         Computer             7.601729              140.414001   \n",
       "3           Tablet            35.586430              177.002419   \n",
       "4           Tablet            23.503651               70.308376   \n",
       "\n",
       "   ContentDownloadsPerMonth GenrePreference  UserRating   \n",
       "0                        42          Comedy    3.522724  \\\n",
       "1                        43          Action    2.021545   \n",
       "2                        14          Sci-Fi    4.806126   \n",
       "3                        14          Comedy    4.943900   \n",
       "4                         6           Drama    2.846880   \n",
       "\n",
       "   SupportTicketsPerMonth  Gender  WatchlistSize ParentalControl   \n",
       "0                       2    Male             23              No  \\\n",
       "1                       2  Female             22             Yes   \n",
       "2                       2  Female             22              No   \n",
       "3                       0  Female             23             Yes   \n",
       "4                       6  Female              0              No   \n",
       "\n",
       "  SubtitlesEnabled  CustomerID  \n",
       "0               No  O1W6BHP6RM  \n",
       "1               No  LFR4X92X8H  \n",
       "2              Yes  QM5GBIYODA  \n",
       "3              Yes  D9RXTK2K9F  \n",
       "4               No  ENTCCHR1LR  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df = pd.read_csv(\"test.csv\")\n",
    "print('test_df Shape:', test_df.shape)\n",
    "test_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explore, Clean, Validate, and Visualize the Data (optional)\n",
    "\n",
    "Feel free to explore, clean, validate, and visualize the data however you see fit for this competition to help determine or optimize your predictive model. Please note - the final autograding will only be on the accuracy of the `prediction_df` predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>SubscriptionType</th>\n",
       "      <th>PaymentMethod</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>ContentType</th>\n",
       "      <th>MultiDeviceAccess</th>\n",
       "      <th>DeviceRegistered</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>AverageViewingDuration</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>GenrePreference</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>Gender</th>\n",
       "      <th>WatchlistSize</th>\n",
       "      <th>ParentalControl</th>\n",
       "      <th>SubtitlesEnabled</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>20</td>\n",
       "      <td>11.055215</td>\n",
       "      <td>221.104302</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>No</td>\n",
       "      <td>Both</td>\n",
       "      <td>No</td>\n",
       "      <td>Mobile</td>\n",
       "      <td>36.758104</td>\n",
       "      <td>63.531377</td>\n",
       "      <td>10</td>\n",
       "      <td>Sci-Fi</td>\n",
       "      <td>2.176498</td>\n",
       "      <td>4</td>\n",
       "      <td>Male</td>\n",
       "      <td>3</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>57</td>\n",
       "      <td>5.175208</td>\n",
       "      <td>294.986882</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>32.450568</td>\n",
       "      <td>25.725595</td>\n",
       "      <td>18</td>\n",
       "      <td>Action</td>\n",
       "      <td>3.478632</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>73</td>\n",
       "      <td>12.106657</td>\n",
       "      <td>883.785952</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Computer</td>\n",
       "      <td>7.395160</td>\n",
       "      <td>57.364061</td>\n",
       "      <td>23</td>\n",
       "      <td>Fantasy</td>\n",
       "      <td>4.238824</td>\n",
       "      <td>6</td>\n",
       "      <td>Male</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>32</td>\n",
       "      <td>7.263743</td>\n",
       "      <td>232.439774</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>27.960389</td>\n",
       "      <td>131.537507</td>\n",
       "      <td>30</td>\n",
       "      <td>Drama</td>\n",
       "      <td>4.276013</td>\n",
       "      <td>2</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57</td>\n",
       "      <td>16.953078</td>\n",
       "      <td>966.325422</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>TV</td>\n",
       "      <td>20.083397</td>\n",
       "      <td>45.356653</td>\n",
       "      <td>20</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>3.616170</td>\n",
       "      <td>4</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AccountAge  MonthlyCharges  TotalCharges SubscriptionType   \n",
       "0          20       11.055215    221.104302          Premium  \\\n",
       "1          57        5.175208    294.986882            Basic   \n",
       "2          73       12.106657    883.785952            Basic   \n",
       "3          32        7.263743    232.439774            Basic   \n",
       "4          57       16.953078    966.325422          Premium   \n",
       "\n",
       "      PaymentMethod PaperlessBilling ContentType MultiDeviceAccess   \n",
       "0      Mailed check               No        Both                No  \\\n",
       "1       Credit card              Yes      Movies                No   \n",
       "2      Mailed check              Yes      Movies                No   \n",
       "3  Electronic check               No    TV Shows                No   \n",
       "4  Electronic check              Yes    TV Shows                No   \n",
       "\n",
       "  DeviceRegistered  ViewingHoursPerWeek  AverageViewingDuration   \n",
       "0           Mobile            36.758104               63.531377  \\\n",
       "1           Tablet            32.450568               25.725595   \n",
       "2         Computer             7.395160               57.364061   \n",
       "3           Tablet            27.960389              131.537507   \n",
       "4               TV            20.083397               45.356653   \n",
       "\n",
       "   ContentDownloadsPerMonth GenrePreference  UserRating   \n",
       "0                        10          Sci-Fi    2.176498  \\\n",
       "1                        18          Action    3.478632   \n",
       "2                        23         Fantasy    4.238824   \n",
       "3                        30           Drama    4.276013   \n",
       "4                        20          Comedy    3.616170   \n",
       "\n",
       "   SupportTicketsPerMonth  Gender  WatchlistSize ParentalControl   \n",
       "0                       4    Male              3              No  \\\n",
       "1                       8    Male             23              No   \n",
       "2                       6    Male              1             Yes   \n",
       "3                       2    Male             24             Yes   \n",
       "4                       4  Female              0              No   \n",
       "\n",
       "  SubtitlesEnabled  Churn  \n",
       "0               No      0  \n",
       "1              Yes      0  \n",
       "2              Yes      0  \n",
       "3              Yes      0  \n",
       "4               No      0  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.drop(\"CustomerID\", axis=1, inplace=True)\n",
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Churn\n",
       "0    199605\n",
       "1     44182\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df[\"Churn\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AccountAge                  0\n",
       "MonthlyCharges              0\n",
       "TotalCharges                0\n",
       "SubscriptionType            0\n",
       "PaymentMethod               0\n",
       "PaperlessBilling            0\n",
       "ContentType                 0\n",
       "MultiDeviceAccess           0\n",
       "DeviceRegistered            0\n",
       "ViewingHoursPerWeek         0\n",
       "AverageViewingDuration      0\n",
       "ContentDownloadsPerMonth    0\n",
       "GenrePreference             0\n",
       "UserRating                  0\n",
       "SupportTicketsPerMonth      0\n",
       "Gender                      0\n",
       "WatchlistSize               0\n",
       "ParentalControl             0\n",
       "SubtitlesEnabled            0\n",
       "Churn                       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train_df.isnull().sum() # find missing values\n",
    "train_df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Show Columns unique Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AccountAge: [ 20  57  73  32 113  38  25  26  14 114   3  64  43  98  47  11  50   5\n",
      "  63  55  84  16  80 116  33  86  52  77   4  91 119  72  94  99  40  60\n",
      " 105  54  34  67  49 100   6  97  24  53  82  18   8  29   9 101  30  74\n",
      "  17  69  61  65  96  62  35  51 109  76   2  42 111  68  15  27  58 103\n",
      "  45  36  85 107 115  81 106  70  46  83  92   1  44 118  23  88   7  31\n",
      "  71  48  93 117  95  37 112  21 108  19  12  78 104  87  90 110  56  79\n",
      "  41  13  28  10  59  39  75 102  89  66  22]\n",
      "MonthlyCharges: [11.0552151   5.17520846 12.10665687 ... 14.51456873 18.14055532\n",
      " 11.59377449]\n",
      "TotalCharges: [ 221.10430197  294.98688225  883.78595161 ... 1640.14626689  126.98388725\n",
      " 1043.4397042 ]\n",
      "SubscriptionType: ['Premium' 'Basic' 'Standard']\n",
      "PaymentMethod: ['Mailed check' 'Credit card' 'Electronic check' 'Bank transfer']\n",
      "PaperlessBilling: ['No' 'Yes']\n",
      "ContentType: ['Both' 'Movies' 'TV Shows']\n",
      "MultiDeviceAccess: ['No' 'Yes']\n",
      "DeviceRegistered: ['Mobile' 'Tablet' 'Computer' 'TV']\n",
      "ViewingHoursPerWeek: [36.75810391 32.45056783  7.39516011 ... 10.62872846 30.46678245\n",
      " 24.97253673]\n",
      "AverageViewingDuration: [ 63.53137733  25.72559464  57.36406085 ... 176.18609505 153.38631494\n",
      "  84.824498  ]\n",
      "ContentDownloadsPerMonth: [10 18 23 30 20 35 28  0 49 26 15  6 41  3  1  7 37 25 22 43 13 17 40 12\n",
      "  8  2  4 16 34 29 19 31 11  9 42  5 47 36 39 48 27 46 45 33 21 24 44 38\n",
      " 14 32]\n",
      "GenrePreference: ['Sci-Fi' 'Action' 'Fantasy' 'Drama' 'Comedy']\n",
      "UserRating: [2.17649751 3.47863229 4.23882362 ... 4.01221731 2.13578882 1.42889572]\n",
      "SupportTicketsPerMonth: [4 8 6 2 9 0 1 7 3 5]\n",
      "Gender: ['Male' 'Female']\n",
      "WatchlistSize: [ 3 23  1 24  0  2 20 22  5 18  8 14 21 15 13 17  9  7 11 10 19  4 12 16\n",
      "  6]\n",
      "ParentalControl: ['No' 'Yes']\n",
      "SubtitlesEnabled: ['No' 'Yes']\n",
      "Churn: [0 1]\n"
     ]
    }
   ],
   "source": [
    "def printAllColumnsUniqueValues(df):\n",
    "    for col in df:\n",
    "        print(f'{col}: {df[col].unique()}') # show unique values in each column\n",
    "printAllColumnsUniqueValues(train_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AccountAge: [ 38  77   5  88  91 103  22  21  86 104  20  94 106  30 108  27  64  83\n",
      "  44  16  26  61   7  23  81   9  75  17  62  69  11  31  67  45 113  79\n",
      "  82  10  95  84  50  58 101 110  28 102  90  37  76  73  53   6  92  12\n",
      "  57  36  18 111  97 119  39 114  72  93  65  89  66 100  87 116 109  29\n",
      "  49  56  68  19  78   2  70  41  33  60  13  99 105  54  47   4  98  43\n",
      "  35  42 107  80  96  52   1  40  55  63  48  85  15  24 117 118   8  51\n",
      "  74  34 112  25  71  46 115  59   3  14  32]\n",
      "MonthlyCharges: [17.86937354  9.91285413 15.01901132 ... 18.13434307 19.77400987\n",
      "  7.43838545]\n",
      "TotalCharges: [ 679.03619453  763.28976817   75.09505659 ... 1922.2403651   909.60445412\n",
      "  818.22239896]\n",
      "SubscriptionType: ['Premium' 'Basic' 'Standard']\n",
      "PaymentMethod: ['Mailed check' 'Electronic check' 'Bank transfer' 'Credit card']\n",
      "PaperlessBilling: ['No' 'Yes']\n",
      "ContentType: ['TV Shows' 'Both' 'Movies']\n",
      "MultiDeviceAccess: ['No' 'Yes']\n",
      "DeviceRegistered: ['TV' 'Computer' 'Tablet' 'Mobile']\n",
      "ViewingHoursPerWeek: [29.12630786 36.87372943  7.6017294  ...  7.23630314 25.80928493\n",
      " 34.14617433]\n",
      "AverageViewingDuration: [122.27403067  57.09331949 140.41400107 ... 109.58315254 115.15356954\n",
      " 132.60680148]\n",
      "ContentDownloadsPerMonth: [42 43 14  6  1  9 35 44 19  3  7  5 39  0 20 18 36 47 40 24 32  4 38 26\n",
      " 34 46 10  2 21 11 23 31 16 15 25 28 37 27 41  8 17 49 22 29 48 12 13 33\n",
      " 45 30]\n",
      "GenrePreference: ['Comedy' 'Action' 'Sci-Fi' 'Drama' 'Fantasy']\n",
      "UserRating: [3.52272397 2.02154532 4.80612634 ... 2.99152743 4.99801873 4.67990992]\n",
      "SupportTicketsPerMonth: [2 0 6 1 4 9 5 7 3 8]\n",
      "Gender: ['Male' 'Female']\n",
      "WatchlistSize: [23 22  0  6  7  8  3  4 15 13  9 19 14 18 20  2 21 24 11  1 10 12 16 17\n",
      "  5]\n",
      "ParentalControl: ['No' 'Yes']\n",
      "SubtitlesEnabled: ['No' 'Yes']\n",
      "CustomerID: ['O1W6BHP6RM' 'LFR4X92X8H' 'QM5GBIYODA' ... 'IPDIA02ZE1' 'ITLFTPRJGV'\n",
      " 'Y204GZY6NE']\n"
     ]
    }
   ],
   "source": [
    "printAllColumnsUniqueValues(test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SubscriptionType: ['Premium' 'Basic' 'Standard']\n",
      "PaymentMethod: ['Mailed check' 'Credit card' 'Electronic check' 'Bank transfer']\n",
      "PaperlessBilling: ['No' 'Yes']\n",
      "ContentType: ['Both' 'Movies' 'TV Shows']\n",
      "MultiDeviceAccess: ['No' 'Yes']\n",
      "DeviceRegistered: ['Mobile' 'Tablet' 'Computer' 'TV']\n",
      "GenrePreference: ['Sci-Fi' 'Action' 'Fantasy' 'Drama' 'Comedy']\n",
      "Gender: ['Male' 'Female']\n",
      "ParentalControl: ['No' 'Yes']\n",
      "SubtitlesEnabled: ['No' 'Yes']\n"
     ]
    }
   ],
   "source": [
    "def printUniqueValues(df):\n",
    "    for col in df:\n",
    "        if df[col].dtype == 'object':\n",
    "            print(f'{col}: {df[col].unique()}')  # show unique values in each column train_df\n",
    "        \n",
    "printUniqueValues(train_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# your code here (optional)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## EDA Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_df[\"Churn\"].value_counts().plot(kind='bar', title='Churn Counts');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "noChurn = train_df[train_df[\"Churn\"] == 0]\n",
    "yesChurn = train_df[train_df[\"Churn\"] == 1]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>AverageViewingDuration</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>WatchlistSize</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>199605.000000</td>\n",
       "      <td>199605.000000</td>\n",
       "      <td>199605.000000</td>\n",
       "      <td>199605.000000</td>\n",
       "      <td>199605.000000</td>\n",
       "      <td>199605.000000</td>\n",
       "      <td>199605.000000</td>\n",
       "      <td>199605.000000</td>\n",
       "      <td>199605.000000</td>\n",
       "      <td>199605.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>63.273290</td>\n",
       "      <td>12.286129</td>\n",
       "      <td>780.402245</td>\n",
       "      <td>21.182697</td>\n",
       "      <td>95.754533</td>\n",
       "      <td>25.383853</td>\n",
       "      <td>2.990688</td>\n",
       "      <td>4.390576</td>\n",
       "      <td>11.944941</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>33.953685</td>\n",
       "      <td>4.317675</td>\n",
       "      <td>524.200292</td>\n",
       "      <td>11.198072</td>\n",
       "      <td>50.281449</td>\n",
       "      <td>14.358435</td>\n",
       "      <td>1.155398</td>\n",
       "      <td>2.866237</td>\n",
       "      <td>7.197050</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.990062</td>\n",
       "      <td>4.999861</td>\n",
       "      <td>1.000133</td>\n",
       "      <td>5.000547</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000039</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>35.000000</td>\n",
       "      <td>8.532456</td>\n",
       "      <td>362.799813</td>\n",
       "      <td>11.601957</td>\n",
       "      <td>52.664521</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>1.986598</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>65.000000</td>\n",
       "      <td>12.189954</td>\n",
       "      <td>684.445761</td>\n",
       "      <td>21.545178</td>\n",
       "      <td>97.505261</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>2.983423</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>93.000000</td>\n",
       "      <td>15.993218</td>\n",
       "      <td>1124.699587</td>\n",
       "      <td>30.931229</td>\n",
       "      <td>139.403329</td>\n",
       "      <td>38.000000</td>\n",
       "      <td>3.989193</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>119.000000</td>\n",
       "      <td>19.989957</td>\n",
       "      <td>2378.723844</td>\n",
       "      <td>39.999723</td>\n",
       "      <td>179.999275</td>\n",
       "      <td>49.000000</td>\n",
       "      <td>4.999982</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          AccountAge  MonthlyCharges   TotalCharges  ViewingHoursPerWeek   \n",
       "count  199605.000000   199605.000000  199605.000000        199605.000000  \\\n",
       "mean       63.273290       12.286129     780.402245            21.182697   \n",
       "std        33.953685        4.317675     524.200292            11.198072   \n",
       "min         1.000000        4.990062       4.999861             1.000133   \n",
       "25%        35.000000        8.532456     362.799813            11.601957   \n",
       "50%        65.000000       12.189954     684.445761            21.545178   \n",
       "75%        93.000000       15.993218    1124.699587            30.931229   \n",
       "max       119.000000       19.989957    2378.723844            39.999723   \n",
       "\n",
       "       AverageViewingDuration  ContentDownloadsPerMonth     UserRating   \n",
       "count           199605.000000             199605.000000  199605.000000  \\\n",
       "mean                95.754533                 25.383853       2.990688   \n",
       "std                 50.281449                 14.358435       1.155398   \n",
       "min                  5.000547                  0.000000       1.000039   \n",
       "25%                 52.664521                 13.000000       1.986598   \n",
       "50%                 97.505261                 26.000000       2.983423   \n",
       "75%                139.403329                 38.000000       3.989193   \n",
       "max                179.999275                 49.000000       4.999982   \n",
       "\n",
       "       SupportTicketsPerMonth  WatchlistSize     Churn  \n",
       "count           199605.000000  199605.000000  199605.0  \n",
       "mean                 4.390576      11.944941       0.0  \n",
       "std                  2.866237       7.197050       0.0  \n",
       "min                  0.000000       0.000000       0.0  \n",
       "25%                  2.000000       6.000000       0.0  \n",
       "50%                  4.000000      12.000000       0.0  \n",
       "75%                  7.000000      18.000000       0.0  \n",
       "max                  9.000000      24.000000       0.0  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "noChurn.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>AverageViewingDuration</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>WatchlistSize</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>44182.000000</td>\n",
       "      <td>44182.000000</td>\n",
       "      <td>44182.000000</td>\n",
       "      <td>44182.000000</td>\n",
       "      <td>44182.000000</td>\n",
       "      <td>44182.000000</td>\n",
       "      <td>44182.000000</td>\n",
       "      <td>44182.000000</td>\n",
       "      <td>44182.000000</td>\n",
       "      <td>44182.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>45.674121</td>\n",
       "      <td>13.414880</td>\n",
       "      <td>616.737816</td>\n",
       "      <td>17.427741</td>\n",
       "      <td>76.494840</td>\n",
       "      <td>20.526323</td>\n",
       "      <td>3.057037</td>\n",
       "      <td>5.017451</td>\n",
       "      <td>12.350867</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>32.001262</td>\n",
       "      <td>4.251322</td>\n",
       "      <td>496.333924</td>\n",
       "      <td>10.932147</td>\n",
       "      <td>48.466716</td>\n",
       "      <td>14.028993</td>\n",
       "      <td>1.153083</td>\n",
       "      <td>2.844926</td>\n",
       "      <td>7.165535</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.990112</td>\n",
       "      <td>4.991154</td>\n",
       "      <td>1.000065</td>\n",
       "      <td>5.002643</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000007</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>18.000000</td>\n",
       "      <td>9.963733</td>\n",
       "      <td>218.502198</td>\n",
       "      <td>7.898715</td>\n",
       "      <td>34.398096</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>2.065083</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>40.000000</td>\n",
       "      <td>13.855745</td>\n",
       "      <td>488.927265</td>\n",
       "      <td>16.079108</td>\n",
       "      <td>69.772525</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>3.084849</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>69.000000</td>\n",
       "      <td>17.132808</td>\n",
       "      <td>903.552235</td>\n",
       "      <td>26.218169</td>\n",
       "      <td>114.371954</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>4.057920</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>119.000000</td>\n",
       "      <td>19.989741</td>\n",
       "      <td>2356.614662</td>\n",
       "      <td>39.997750</td>\n",
       "      <td>179.991594</td>\n",
       "      <td>49.000000</td>\n",
       "      <td>4.999989</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         AccountAge  MonthlyCharges  TotalCharges  ViewingHoursPerWeek   \n",
       "count  44182.000000    44182.000000  44182.000000         44182.000000  \\\n",
       "mean      45.674121       13.414880    616.737816            17.427741   \n",
       "std       32.001262        4.251322    496.333924            10.932147   \n",
       "min        1.000000        4.990112      4.991154             1.000065   \n",
       "25%       18.000000        9.963733    218.502198             7.898715   \n",
       "50%       40.000000       13.855745    488.927265            16.079108   \n",
       "75%       69.000000       17.132808    903.552235            26.218169   \n",
       "max      119.000000       19.989741   2356.614662            39.997750   \n",
       "\n",
       "       AverageViewingDuration  ContentDownloadsPerMonth    UserRating   \n",
       "count            44182.000000              44182.000000  44182.000000  \\\n",
       "mean                76.494840                 20.526323      3.057037   \n",
       "std                 48.466716                 14.028993      1.153083   \n",
       "min                  5.002643                  0.000000      1.000007   \n",
       "25%                 34.398096                  8.000000      2.065083   \n",
       "50%                 69.772525                 19.000000      3.084849   \n",
       "75%                114.371954                 32.000000      4.057920   \n",
       "max                179.991594                 49.000000      4.999989   \n",
       "\n",
       "       SupportTicketsPerMonth  WatchlistSize    Churn  \n",
       "count            44182.000000   44182.000000  44182.0  \n",
       "mean                 5.017451      12.350867      1.0  \n",
       "std                  2.844926       7.165535      0.0  \n",
       "min                  0.000000       0.000000      1.0  \n",
       "25%                  3.000000       6.000000      1.0  \n",
       "50%                  5.000000      13.000000      1.0  \n",
       "75%                  8.000000      19.000000      1.0  \n",
       "max                  9.000000      24.000000      1.0  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yesChurn.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xlabel(\"Monthly Charges\")\n",
    "plt.ylabel(\"Number Of Customers\")\n",
    "plt.title(\"Customer Churn Prediction Visualiztion\")\n",
    "plt.hist([noChurn.MonthlyCharges, yesChurn.MonthlyCharges], color = ['red', 'green'], label = ['Churn=No', 'Churn=Yes'])\n",
    "plt.legend();\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['AccountAge', 'MonthlyCharges', 'TotalCharges', 'SubscriptionType',\n",
       "       'PaymentMethod', 'PaperlessBilling', 'ContentType', 'MultiDeviceAccess',\n",
       "       'DeviceRegistered', 'ViewingHoursPerWeek', 'AverageViewingDuration',\n",
       "       'ContentDownloadsPerMonth', 'GenrePreference', 'UserRating',\n",
       "       'SupportTicketsPerMonth', 'Gender', 'WatchlistSize', 'ParentalControl',\n",
       "       'SubtitlesEnabled', 'Churn'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "noChurn.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xlabel(\"Total Charges\")\n",
    "plt.ylabel(\"Number Of Customers\")\n",
    "plt.title(\"Customer Churn Prediction Visualiztion\")\n",
    "plt.hist([noChurn.TotalCharges, yesChurn.TotalCharges],color= ['red', 'green'], label=['Churn=No', 'Churn=Yes'])\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xlabel(\"Viewing Hours Per Week\")\n",
    "plt.ylabel(\"Number Of Customers\")\n",
    "plt.title(\"Customer Churn Prediction Visualiztion\")\n",
    "plt.hist([noChurn.ViewingHoursPerWeek, yesChurn.ViewingHoursPerWeek], color=['red', 'green'], label=['Churn=No', 'Churn=Yes'])\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xlabel(\"Average Viewing Duration\")\n",
    "plt.ylabel(\"Number Of Customers\")\n",
    "plt.title(\"Customer Churn Prediction Visualiztion\")\n",
    "plt.hist([noChurn.AverageViewingDuration, yesChurn.AverageViewingDuration], color=['red', 'green'], label=['Churn=No', 'Churn=Yes'])\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Cleaning\n",
    "\n",
    "## Convert yes or no unique values to 1 or 0 , One Hot Enoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define yes or no column\n",
    "yesNoColumns = [\"PaperlessBilling\", \"MultiDeviceAccess\", \"ParentalControl\", \"SubtitlesEnabled\"]\n",
    "# define several columns for one hot encoding\n",
    "columnsForOneHotEncoding = [\"SubscriptionType\", \"PaymentMethod\", \"ContentType\", \"DeviceRegistered\", \"GenrePreference\"]\n",
    "# define the columbns to be scaled\n",
    "columnsToScale = [\"AccountAge\",\"MonthlyCharges\", \"TotalCharges\", \"ViewingHoursPerWeek\", \"AverageViewingDuration\", \n",
    "                  \"ContentDownloadsPerMonth\", \"UserRating\", \"SupportTicketsPerMonth\" ]\n",
    "\n",
    "# for col in yesNoColumns:\n",
    "#     train_df[col].replace({'Yes': 1, 'No': 0}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>SubscriptionType</th>\n",
       "      <th>PaymentMethod</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>ContentType</th>\n",
       "      <th>MultiDeviceAccess</th>\n",
       "      <th>DeviceRegistered</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>AverageViewingDuration</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>GenrePreference</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>Gender</th>\n",
       "      <th>WatchlistSize</th>\n",
       "      <th>ParentalControl</th>\n",
       "      <th>SubtitlesEnabled</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>20</td>\n",
       "      <td>11.055215</td>\n",
       "      <td>221.104302</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>No</td>\n",
       "      <td>Both</td>\n",
       "      <td>No</td>\n",
       "      <td>Mobile</td>\n",
       "      <td>36.758104</td>\n",
       "      <td>63.531377</td>\n",
       "      <td>10</td>\n",
       "      <td>Sci-Fi</td>\n",
       "      <td>2.176498</td>\n",
       "      <td>4</td>\n",
       "      <td>Male</td>\n",
       "      <td>3</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>57</td>\n",
       "      <td>5.175208</td>\n",
       "      <td>294.986882</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>32.450568</td>\n",
       "      <td>25.725595</td>\n",
       "      <td>18</td>\n",
       "      <td>Action</td>\n",
       "      <td>3.478632</td>\n",
       "      <td>8</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>73</td>\n",
       "      <td>12.106657</td>\n",
       "      <td>883.785952</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Computer</td>\n",
       "      <td>7.395160</td>\n",
       "      <td>57.364061</td>\n",
       "      <td>23</td>\n",
       "      <td>Fantasy</td>\n",
       "      <td>4.238824</td>\n",
       "      <td>6</td>\n",
       "      <td>Male</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>32</td>\n",
       "      <td>7.263743</td>\n",
       "      <td>232.439774</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>27.960389</td>\n",
       "      <td>131.537507</td>\n",
       "      <td>30</td>\n",
       "      <td>Drama</td>\n",
       "      <td>4.276013</td>\n",
       "      <td>2</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57</td>\n",
       "      <td>16.953078</td>\n",
       "      <td>966.325422</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>TV</td>\n",
       "      <td>20.083397</td>\n",
       "      <td>45.356653</td>\n",
       "      <td>20</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>3.616170</td>\n",
       "      <td>4</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AccountAge  MonthlyCharges  TotalCharges SubscriptionType   \n",
       "0          20       11.055215    221.104302          Premium  \\\n",
       "1          57        5.175208    294.986882            Basic   \n",
       "2          73       12.106657    883.785952            Basic   \n",
       "3          32        7.263743    232.439774            Basic   \n",
       "4          57       16.953078    966.325422          Premium   \n",
       "\n",
       "      PaymentMethod PaperlessBilling ContentType MultiDeviceAccess   \n",
       "0      Mailed check               No        Both                No  \\\n",
       "1       Credit card              Yes      Movies                No   \n",
       "2      Mailed check              Yes      Movies                No   \n",
       "3  Electronic check               No    TV Shows                No   \n",
       "4  Electronic check              Yes    TV Shows                No   \n",
       "\n",
       "  DeviceRegistered  ViewingHoursPerWeek  AverageViewingDuration   \n",
       "0           Mobile            36.758104               63.531377  \\\n",
       "1           Tablet            32.450568               25.725595   \n",
       "2         Computer             7.395160               57.364061   \n",
       "3           Tablet            27.960389              131.537507   \n",
       "4               TV            20.083397               45.356653   \n",
       "\n",
       "   ContentDownloadsPerMonth GenrePreference  UserRating   \n",
       "0                        10          Sci-Fi    2.176498  \\\n",
       "1                        18          Action    3.478632   \n",
       "2                        23         Fantasy    4.238824   \n",
       "3                        30           Drama    4.276013   \n",
       "4                        20          Comedy    3.616170   \n",
       "\n",
       "   SupportTicketsPerMonth  Gender  WatchlistSize ParentalControl   \n",
       "0                       4    Male              3              No  \\\n",
       "1                       8    Male             23              No   \n",
       "2                       6    Male              1             Yes   \n",
       "3                       2    Male             24             Yes   \n",
       "4                       4  Female              0              No   \n",
       "\n",
       "  SubtitlesEnabled  Churn  \n",
       "0               No      0  \n",
       "1              Yes      0  \n",
       "2              Yes      0  \n",
       "3              Yes      0  \n",
       "4               No      0  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SubscriptionType: ['Premium' 'Basic' 'Standard']\n",
      "PaymentMethod: ['Mailed check' 'Credit card' 'Electronic check' 'Bank transfer']\n",
      "PaperlessBilling: ['No' 'Yes']\n",
      "ContentType: ['Both' 'Movies' 'TV Shows']\n",
      "MultiDeviceAccess: ['No' 'Yes']\n",
      "DeviceRegistered: ['Mobile' 'Tablet' 'Computer' 'TV']\n",
      "GenrePreference: ['Sci-Fi' 'Action' 'Fantasy' 'Drama' 'Comedy']\n",
      "Gender: ['Male' 'Female']\n",
      "ParentalControl: ['No' 'Yes']\n",
      "SubtitlesEnabled: ['No' 'Yes']\n"
     ]
    }
   ],
   "source": [
    "printUniqueValues(train_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Balance the training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(44182, 199605)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(yesChurn) , len(noChurn) # seem to be imbalanced , yes churn is less than no churn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(44182, 20)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sampledNoChurn = noChurn.sample(n=len(yesChurn), random_state=42)\n",
    "sampledNoChurn.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### combine undersample no churn with yes churn become balanced training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(88364, 20)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "balanceTrainDF = pd.concat([sampledNoChurn, yesChurn], axis=0)\n",
    "balanceTrainDF.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "balanceTrainDF[\"Churn\"].value_counts().plot(kind='bar', title='Churn Counts');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# random order the row data\n",
    "balanceTrainDF = balanceTrainDF.sample(frac=1).reset_index(drop=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>SubscriptionType</th>\n",
       "      <th>PaymentMethod</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>ContentType</th>\n",
       "      <th>MultiDeviceAccess</th>\n",
       "      <th>DeviceRegistered</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>AverageViewingDuration</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>GenrePreference</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>Gender</th>\n",
       "      <th>WatchlistSize</th>\n",
       "      <th>ParentalControl</th>\n",
       "      <th>SubtitlesEnabled</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>58</td>\n",
       "      <td>19.098906</td>\n",
       "      <td>1107.736549</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>No</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>30.069979</td>\n",
       "      <td>134.730400</td>\n",
       "      <td>18</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>3.451714</td>\n",
       "      <td>9</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>102</td>\n",
       "      <td>15.949807</td>\n",
       "      <td>1626.880317</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Mobile</td>\n",
       "      <td>36.894715</td>\n",
       "      <td>11.358375</td>\n",
       "      <td>26</td>\n",
       "      <td>Action</td>\n",
       "      <td>4.699298</td>\n",
       "      <td>1</td>\n",
       "      <td>Male</td>\n",
       "      <td>22</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18</td>\n",
       "      <td>8.189600</td>\n",
       "      <td>147.412808</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>No</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Mobile</td>\n",
       "      <td>6.646635</td>\n",
       "      <td>17.284568</td>\n",
       "      <td>47</td>\n",
       "      <td>Action</td>\n",
       "      <td>2.964508</td>\n",
       "      <td>9</td>\n",
       "      <td>Female</td>\n",
       "      <td>2</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8</td>\n",
       "      <td>15.407258</td>\n",
       "      <td>123.258063</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Bank transfer</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>Mobile</td>\n",
       "      <td>17.450793</td>\n",
       "      <td>40.764098</td>\n",
       "      <td>28</td>\n",
       "      <td>Fantasy</td>\n",
       "      <td>3.371891</td>\n",
       "      <td>9</td>\n",
       "      <td>Female</td>\n",
       "      <td>17</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57</td>\n",
       "      <td>8.172963</td>\n",
       "      <td>465.858896</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>Computer</td>\n",
       "      <td>28.409069</td>\n",
       "      <td>158.098937</td>\n",
       "      <td>20</td>\n",
       "      <td>Fantasy</td>\n",
       "      <td>2.936275</td>\n",
       "      <td>9</td>\n",
       "      <td>Male</td>\n",
       "      <td>13</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>82</td>\n",
       "      <td>15.831980</td>\n",
       "      <td>1298.222382</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Both</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV</td>\n",
       "      <td>12.454230</td>\n",
       "      <td>155.668823</td>\n",
       "      <td>21</td>\n",
       "      <td>Sci-Fi</td>\n",
       "      <td>4.148611</td>\n",
       "      <td>4</td>\n",
       "      <td>Female</td>\n",
       "      <td>23</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>110</td>\n",
       "      <td>13.700209</td>\n",
       "      <td>1507.022965</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Bank transfer</td>\n",
       "      <td>No</td>\n",
       "      <td>Both</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV</td>\n",
       "      <td>5.542316</td>\n",
       "      <td>29.776039</td>\n",
       "      <td>5</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>3.804662</td>\n",
       "      <td>0</td>\n",
       "      <td>Male</td>\n",
       "      <td>15</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>64</td>\n",
       "      <td>18.505298</td>\n",
       "      <td>1184.339046</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Bank transfer</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Movies</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV</td>\n",
       "      <td>20.884378</td>\n",
       "      <td>163.245189</td>\n",
       "      <td>13</td>\n",
       "      <td>Drama</td>\n",
       "      <td>4.515523</td>\n",
       "      <td>9</td>\n",
       "      <td>Female</td>\n",
       "      <td>12</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>70</td>\n",
       "      <td>12.823614</td>\n",
       "      <td>897.653000</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Bank transfer</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>TV</td>\n",
       "      <td>29.145658</td>\n",
       "      <td>129.543308</td>\n",
       "      <td>12</td>\n",
       "      <td>Sci-Fi</td>\n",
       "      <td>4.880622</td>\n",
       "      <td>5</td>\n",
       "      <td>Male</td>\n",
       "      <td>13</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>68</td>\n",
       "      <td>9.913542</td>\n",
       "      <td>674.120828</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Both</td>\n",
       "      <td>No</td>\n",
       "      <td>Mobile</td>\n",
       "      <td>12.740308</td>\n",
       "      <td>20.186340</td>\n",
       "      <td>45</td>\n",
       "      <td>Action</td>\n",
       "      <td>1.861949</td>\n",
       "      <td>8</td>\n",
       "      <td>Female</td>\n",
       "      <td>5</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10</td>\n",
       "      <td>7.015992</td>\n",
       "      <td>70.159917</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>No</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>TV</td>\n",
       "      <td>5.224344</td>\n",
       "      <td>124.820080</td>\n",
       "      <td>7</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>4.399641</td>\n",
       "      <td>5</td>\n",
       "      <td>Female</td>\n",
       "      <td>6</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>60</td>\n",
       "      <td>14.779844</td>\n",
       "      <td>886.790665</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Bank transfer</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>33.229756</td>\n",
       "      <td>170.459844</td>\n",
       "      <td>0</td>\n",
       "      <td>Fantasy</td>\n",
       "      <td>4.126783</td>\n",
       "      <td>2</td>\n",
       "      <td>Female</td>\n",
       "      <td>4</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>48</td>\n",
       "      <td>10.017704</td>\n",
       "      <td>480.849792</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Both</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV</td>\n",
       "      <td>34.046250</td>\n",
       "      <td>57.809372</td>\n",
       "      <td>5</td>\n",
       "      <td>Drama</td>\n",
       "      <td>3.716187</td>\n",
       "      <td>9</td>\n",
       "      <td>Male</td>\n",
       "      <td>22</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>21</td>\n",
       "      <td>15.740959</td>\n",
       "      <td>330.560147</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>Computer</td>\n",
       "      <td>30.258227</td>\n",
       "      <td>162.546118</td>\n",
       "      <td>0</td>\n",
       "      <td>Action</td>\n",
       "      <td>4.489807</td>\n",
       "      <td>9</td>\n",
       "      <td>Male</td>\n",
       "      <td>14</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>107</td>\n",
       "      <td>15.428435</td>\n",
       "      <td>1650.842553</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Bank transfer</td>\n",
       "      <td>No</td>\n",
       "      <td>Movies</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Computer</td>\n",
       "      <td>14.134814</td>\n",
       "      <td>82.741481</td>\n",
       "      <td>1</td>\n",
       "      <td>Drama</td>\n",
       "      <td>3.810450</td>\n",
       "      <td>6</td>\n",
       "      <td>Female</td>\n",
       "      <td>24</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>5</td>\n",
       "      <td>14.842484</td>\n",
       "      <td>74.212418</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>No</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>6.129204</td>\n",
       "      <td>105.528936</td>\n",
       "      <td>6</td>\n",
       "      <td>Sci-Fi</td>\n",
       "      <td>4.805941</td>\n",
       "      <td>5</td>\n",
       "      <td>Male</td>\n",
       "      <td>2</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>34</td>\n",
       "      <td>15.837546</td>\n",
       "      <td>538.476573</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Credit card</td>\n",
       "      <td>No</td>\n",
       "      <td>Movies</td>\n",
       "      <td>Yes</td>\n",
       "      <td>TV</td>\n",
       "      <td>27.152253</td>\n",
       "      <td>139.206792</td>\n",
       "      <td>28</td>\n",
       "      <td>Action</td>\n",
       "      <td>3.094346</td>\n",
       "      <td>3</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>17</td>\n",
       "      <td>11.906368</td>\n",
       "      <td>202.408263</td>\n",
       "      <td>Premium</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>TV</td>\n",
       "      <td>6.744785</td>\n",
       "      <td>52.381826</td>\n",
       "      <td>24</td>\n",
       "      <td>Drama</td>\n",
       "      <td>3.575485</td>\n",
       "      <td>4</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>46</td>\n",
       "      <td>11.799564</td>\n",
       "      <td>542.779939</td>\n",
       "      <td>Basic</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>No</td>\n",
       "      <td>TV Shows</td>\n",
       "      <td>No</td>\n",
       "      <td>Tablet</td>\n",
       "      <td>19.076220</td>\n",
       "      <td>140.128335</td>\n",
       "      <td>28</td>\n",
       "      <td>Action</td>\n",
       "      <td>1.979869</td>\n",
       "      <td>5</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>21</td>\n",
       "      <td>17.281776</td>\n",
       "      <td>362.917286</td>\n",
       "      <td>Standard</td>\n",
       "      <td>Bank transfer</td>\n",
       "      <td>No</td>\n",
       "      <td>Movies</td>\n",
       "      <td>No</td>\n",
       "      <td>Computer</td>\n",
       "      <td>27.762390</td>\n",
       "      <td>120.694662</td>\n",
       "      <td>22</td>\n",
       "      <td>Sci-Fi</td>\n",
       "      <td>1.789349</td>\n",
       "      <td>7</td>\n",
       "      <td>Female</td>\n",
       "      <td>10</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    AccountAge  MonthlyCharges  TotalCharges SubscriptionType   \n",
       "0           58       19.098906   1107.736549            Basic  \\\n",
       "1          102       15.949807   1626.880317         Standard   \n",
       "2           18        8.189600    147.412808         Standard   \n",
       "3            8       15.407258    123.258063            Basic   \n",
       "4           57        8.172963    465.858896            Basic   \n",
       "5           82       15.831980   1298.222382          Premium   \n",
       "6          110       13.700209   1507.022965            Basic   \n",
       "7           64       18.505298   1184.339046         Standard   \n",
       "8           70       12.823614    897.653000         Standard   \n",
       "9           68        9.913542    674.120828            Basic   \n",
       "10          10        7.015992     70.159917            Basic   \n",
       "11          60       14.779844    886.790665            Basic   \n",
       "12          48       10.017704    480.849792            Basic   \n",
       "13          21       15.740959    330.560147          Premium   \n",
       "14         107       15.428435   1650.842553         Standard   \n",
       "15           5       14.842484     74.212418          Premium   \n",
       "16          34       15.837546    538.476573         Standard   \n",
       "17          17       11.906368    202.408263          Premium   \n",
       "18          46       11.799564    542.779939            Basic   \n",
       "19          21       17.281776    362.917286         Standard   \n",
       "\n",
       "       PaymentMethod PaperlessBilling ContentType MultiDeviceAccess   \n",
       "0        Credit card               No      Movies                No  \\\n",
       "1        Credit card               No    TV Shows               Yes   \n",
       "2       Mailed check               No      Movies                No   \n",
       "3      Bank transfer              Yes    TV Shows                No   \n",
       "4        Credit card               No    TV Shows                No   \n",
       "5   Electronic check              Yes        Both               Yes   \n",
       "6      Bank transfer               No        Both               Yes   \n",
       "7      Bank transfer              Yes      Movies               Yes   \n",
       "8      Bank transfer              Yes      Movies                No   \n",
       "9        Credit card              Yes        Both                No   \n",
       "10       Credit card               No      Movies                No   \n",
       "11     Bank transfer              Yes    TV Shows                No   \n",
       "12      Mailed check              Yes        Both               Yes   \n",
       "13  Electronic check               No    TV Shows                No   \n",
       "14     Bank transfer               No      Movies               Yes   \n",
       "15      Mailed check               No      Movies                No   \n",
       "16       Credit card               No      Movies               Yes   \n",
       "17  Electronic check               No    TV Shows                No   \n",
       "18      Mailed check               No    TV Shows                No   \n",
       "19     Bank transfer               No      Movies                No   \n",
       "\n",
       "   DeviceRegistered  ViewingHoursPerWeek  AverageViewingDuration   \n",
       "0            Tablet            30.069979              134.730400  \\\n",
       "1            Mobile            36.894715               11.358375   \n",
       "2            Mobile             6.646635               17.284568   \n",
       "3            Mobile            17.450793               40.764098   \n",
       "4          Computer            28.409069              158.098937   \n",
       "5                TV            12.454230              155.668823   \n",
       "6                TV             5.542316               29.776039   \n",
       "7                TV            20.884378              163.245189   \n",
       "8                TV            29.145658              129.543308   \n",
       "9            Mobile            12.740308               20.186340   \n",
       "10               TV             5.224344              124.820080   \n",
       "11           Tablet            33.229756              170.459844   \n",
       "12               TV            34.046250               57.809372   \n",
       "13         Computer            30.258227              162.546118   \n",
       "14         Computer            14.134814               82.741481   \n",
       "15           Tablet             6.129204              105.528936   \n",
       "16               TV            27.152253              139.206792   \n",
       "17               TV             6.744785               52.381826   \n",
       "18           Tablet            19.076220              140.128335   \n",
       "19         Computer            27.762390              120.694662   \n",
       "\n",
       "    ContentDownloadsPerMonth GenrePreference  UserRating   \n",
       "0                         18          Comedy    3.451714  \\\n",
       "1                         26          Action    4.699298   \n",
       "2                         47          Action    2.964508   \n",
       "3                         28         Fantasy    3.371891   \n",
       "4                         20         Fantasy    2.936275   \n",
       "5                         21          Sci-Fi    4.148611   \n",
       "6                          5          Comedy    3.804662   \n",
       "7                         13           Drama    4.515523   \n",
       "8                         12          Sci-Fi    4.880622   \n",
       "9                         45          Action    1.861949   \n",
       "10                         7          Comedy    4.399641   \n",
       "11                         0         Fantasy    4.126783   \n",
       "12                         5           Drama    3.716187   \n",
       "13                         0          Action    4.489807   \n",
       "14                         1           Drama    3.810450   \n",
       "15                         6          Sci-Fi    4.805941   \n",
       "16                        28          Action    3.094346   \n",
       "17                        24           Drama    3.575485   \n",
       "18                        28          Action    1.979869   \n",
       "19                        22          Sci-Fi    1.789349   \n",
       "\n",
       "    SupportTicketsPerMonth  Gender  WatchlistSize ParentalControl   \n",
       "0                        9    Male              0              No  \\\n",
       "1                        1    Male             22             Yes   \n",
       "2                        9  Female              2              No   \n",
       "3                        9  Female             17              No   \n",
       "4                        9    Male             13              No   \n",
       "5                        4  Female             23             Yes   \n",
       "6                        0    Male             15             Yes   \n",
       "7                        9  Female             12             Yes   \n",
       "8                        5    Male             13             Yes   \n",
       "9                        8  Female              5              No   \n",
       "10                       5  Female              6              No   \n",
       "11                       2  Female              4              No   \n",
       "12                       9    Male             22              No   \n",
       "13                       9    Male             14             Yes   \n",
       "14                       6  Female             24              No   \n",
       "15                       5    Male              2             Yes   \n",
       "16                       3    Male             23              No   \n",
       "17                       4    Male              0             Yes   \n",
       "18                       5  Female              0             Yes   \n",
       "19                       7  Female             10              No   \n",
       "\n",
       "   SubtitlesEnabled  Churn  \n",
       "0               Yes      1  \n",
       "1               Yes      0  \n",
       "2                No      1  \n",
       "3               Yes      1  \n",
       "4               Yes      1  \n",
       "5               Yes      1  \n",
       "6               Yes      0  \n",
       "7                No      0  \n",
       "8                No      1  \n",
       "9               Yes      1  \n",
       "10               No      0  \n",
       "11              Yes      0  \n",
       "12               No      1  \n",
       "13              Yes      1  \n",
       "14               No      0  \n",
       "15               No      1  \n",
       "16              Yes      0  \n",
       "17               No      1  \n",
       "18               No      1  \n",
       "19              Yes      0  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "balanceTrainDF.head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convert Gender column to 1 or 0, Create One-hot Encoding for several columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler, StandardScaler\n",
    "def preprocessData(df, scale=True):\n",
    "\n",
    "    # Convert Yes/No columns to 1/0\n",
    "    for col in yesNoColumns:\n",
    "        df[col].replace({'Yes': 1, 'No': 0}, inplace=True)\n",
    "        \n",
    "    # Convert gender to 1/0\n",
    "    df.Gender.replace({\"Male\": 1, \"Female\": 0}, inplace=True)\n",
    "\n",
    "    # One hot encoding\n",
    "    df = pd.get_dummies(df, columns=columnsForOneHotEncoding, dtype=int)\n",
    "\n",
    "    if scale:\n",
    "        # Scale columns\n",
    "        scaler = MinMaxScaler()#StandardScaler()\n",
    "        df[columnsToScale] = scaler.fit_transform(df[columnsToScale])\n",
    "\n",
    "    return df\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train_df = preprocessData(train_df)\n",
    "balanceTrainDF = preprocessData(balanceTrainDF) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>MultiDeviceAccess</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>AverageViewingDuration</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>...</th>\n",
       "      <th>ContentType_TV Shows</th>\n",
       "      <th>DeviceRegistered_Computer</th>\n",
       "      <th>DeviceRegistered_Mobile</th>\n",
       "      <th>DeviceRegistered_TV</th>\n",
       "      <th>DeviceRegistered_Tablet</th>\n",
       "      <th>GenrePreference_Action</th>\n",
       "      <th>GenrePreference_Comedy</th>\n",
       "      <th>GenrePreference_Drama</th>\n",
       "      <th>GenrePreference_Fantasy</th>\n",
       "      <th>GenrePreference_Sci-Fi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>33154</th>\n",
       "      <td>0.686441</td>\n",
       "      <td>0.852045</td>\n",
       "      <td>0.613071</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.371704</td>\n",
       "      <td>0.293106</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>0.645176</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86974</th>\n",
       "      <td>0.338983</td>\n",
       "      <td>0.395824</td>\n",
       "      <td>0.187034</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.483538</td>\n",
       "      <td>0.794224</td>\n",
       "      <td>0.244898</td>\n",
       "      <td>0.027790</td>\n",
       "      <td>0.222222</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39634</th>\n",
       "      <td>0.415254</td>\n",
       "      <td>0.337317</td>\n",
       "      <td>0.210028</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.213354</td>\n",
       "      <td>0.092258</td>\n",
       "      <td>0.877551</td>\n",
       "      <td>0.246572</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10419</th>\n",
       "      <td>0.186441</td>\n",
       "      <td>0.714459</td>\n",
       "      <td>0.150404</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.076586</td>\n",
       "      <td>0.183578</td>\n",
       "      <td>0.755102</td>\n",
       "      <td>0.394621</td>\n",
       "      <td>0.222222</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2801</th>\n",
       "      <td>0.957627</td>\n",
       "      <td>0.711650</td>\n",
       "      <td>0.751791</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.929427</td>\n",
       "      <td>0.356005</td>\n",
       "      <td>0.551020</td>\n",
       "      <td>0.679868</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 34 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       AccountAge  MonthlyCharges  TotalCharges  PaperlessBilling   \n",
       "33154    0.686441        0.852045      0.613071                 0  \\\n",
       "86974    0.338983        0.395824      0.187034                 1   \n",
       "39634    0.415254        0.337317      0.210028                 0   \n",
       "10419    0.186441        0.714459      0.150404                 1   \n",
       "2801     0.957627        0.711650      0.751791                 1   \n",
       "\n",
       "       MultiDeviceAccess  ViewingHoursPerWeek  AverageViewingDuration   \n",
       "33154                  1             0.371704                0.293106  \\\n",
       "86974                  1             0.483538                0.794224   \n",
       "39634                  0             0.213354                0.092258   \n",
       "10419                  0             0.076586                0.183578   \n",
       "2801                   1             0.929427                0.356005   \n",
       "\n",
       "       ContentDownloadsPerMonth  UserRating  SupportTicketsPerMonth  ...   \n",
       "33154                  0.571429    0.645176                0.777778  ...  \\\n",
       "86974                  0.244898    0.027790                0.222222  ...   \n",
       "39634                  0.877551    0.246572                0.666667  ...   \n",
       "10419                  0.755102    0.394621                0.222222  ...   \n",
       "2801                   0.551020    0.679868                0.555556  ...   \n",
       "\n",
       "       ContentType_TV Shows  DeviceRegistered_Computer   \n",
       "33154                     0                          1  \\\n",
       "86974                     0                          1   \n",
       "39634                     0                          0   \n",
       "10419                     1                          0   \n",
       "2801                      1                          0   \n",
       "\n",
       "       DeviceRegistered_Mobile  DeviceRegistered_TV  DeviceRegistered_Tablet   \n",
       "33154                        0                    0                        0  \\\n",
       "86974                        0                    0                        0   \n",
       "39634                        0                    0                        1   \n",
       "10419                        1                    0                        0   \n",
       "2801                         1                    0                        0   \n",
       "\n",
       "       GenrePreference_Action  GenrePreference_Comedy  GenrePreference_Drama   \n",
       "33154                       0                       0                      0  \\\n",
       "86974                       0                       1                      0   \n",
       "39634                       0                       0                      0   \n",
       "10419                       0                       0                      0   \n",
       "2801                        0                       0                      0   \n",
       "\n",
       "       GenrePreference_Fantasy  GenrePreference_Sci-Fi  \n",
       "33154                        1                       0  \n",
       "86974                        0                       0  \n",
       "39634                        1                       0  \n",
       "10419                        0                       1  \n",
       "2801                         1                       0  \n",
       "\n",
       "[5 rows x 34 columns]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "\n",
    "# train_df = pd.get_dummies(train_df, columns=columnsForOneHotEncoding, dtype=int)\n",
    "# train_df.sample(5)\n",
    "balanceTrainDF.sample(5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>MultiDeviceAccess</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>AverageViewingDuration</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>...</th>\n",
       "      <th>ContentType_TV Shows</th>\n",
       "      <th>DeviceRegistered_Computer</th>\n",
       "      <th>DeviceRegistered_Mobile</th>\n",
       "      <th>DeviceRegistered_TV</th>\n",
       "      <th>DeviceRegistered_Tablet</th>\n",
       "      <th>GenrePreference_Action</th>\n",
       "      <th>GenrePreference_Comedy</th>\n",
       "      <th>GenrePreference_Drama</th>\n",
       "      <th>GenrePreference_Fantasy</th>\n",
       "      <th>GenrePreference_Sci-Fi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>66286</th>\n",
       "      <td>0.923729</td>\n",
       "      <td>0.454900</td>\n",
       "      <td>0.545905</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.361512</td>\n",
       "      <td>0.106066</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>0.030452</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40551</th>\n",
       "      <td>0.508475</td>\n",
       "      <td>0.897315</td>\n",
       "      <td>0.472502</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.524584</td>\n",
       "      <td>0.785168</td>\n",
       "      <td>0.918367</td>\n",
       "      <td>0.824719</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17224</th>\n",
       "      <td>0.677966</td>\n",
       "      <td>0.584253</td>\n",
       "      <td>0.467705</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.316024</td>\n",
       "      <td>0.545735</td>\n",
       "      <td>0.795918</td>\n",
       "      <td>0.430759</td>\n",
       "      <td>0.444444</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22313</th>\n",
       "      <td>0.118644</td>\n",
       "      <td>0.949968</td>\n",
       "      <td>0.119589</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.976702</td>\n",
       "      <td>0.801635</td>\n",
       "      <td>0.510204</td>\n",
       "      <td>0.886883</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20011</th>\n",
       "      <td>0.186441</td>\n",
       "      <td>0.231475</td>\n",
       "      <td>0.079963</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.482695</td>\n",
       "      <td>0.886358</td>\n",
       "      <td>0.836735</td>\n",
       "      <td>0.741351</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 34 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       AccountAge  MonthlyCharges  TotalCharges  PaperlessBilling   \n",
       "66286    0.923729        0.454900      0.545905                 0  \\\n",
       "40551    0.508475        0.897315      0.472502                 0   \n",
       "17224    0.677966        0.584253      0.467705                 0   \n",
       "22313    0.118644        0.949968      0.119589                 0   \n",
       "20011    0.186441        0.231475      0.079963                 1   \n",
       "\n",
       "       MultiDeviceAccess  ViewingHoursPerWeek  AverageViewingDuration   \n",
       "66286                  1             0.361512                0.106066  \\\n",
       "40551                  1             0.524584                0.785168   \n",
       "17224                  1             0.316024                0.545735   \n",
       "22313                  0             0.976702                0.801635   \n",
       "20011                  1             0.482695                0.886358   \n",
       "\n",
       "       ContentDownloadsPerMonth  UserRating  SupportTicketsPerMonth  ...   \n",
       "66286                  0.571429    0.030452                0.666667  ...  \\\n",
       "40551                  0.918367    0.824719                0.555556  ...   \n",
       "17224                  0.795918    0.430759                0.444444  ...   \n",
       "22313                  0.510204    0.886883                0.777778  ...   \n",
       "20011                  0.836735    0.741351                0.111111  ...   \n",
       "\n",
       "       ContentType_TV Shows  DeviceRegistered_Computer   \n",
       "66286                     0                          0  \\\n",
       "40551                     0                          0   \n",
       "17224                     0                          1   \n",
       "22313                     1                          1   \n",
       "20011                     0                          0   \n",
       "\n",
       "       DeviceRegistered_Mobile  DeviceRegistered_TV DeviceRegistered_Tablet   \n",
       "66286                        1                    0                       0  \\\n",
       "40551                        0                    0                       1   \n",
       "17224                        0                    0                       0   \n",
       "22313                        0                    0                       0   \n",
       "20011                        0                    0                       1   \n",
       "\n",
       "       GenrePreference_Action  GenrePreference_Comedy  GenrePreference_Drama   \n",
       "66286                       0                       0                      0  \\\n",
       "40551                       0                       0                      0   \n",
       "17224                       0                       0                      1   \n",
       "22313                       1                       0                      0   \n",
       "20011                       0                       0                      0   \n",
       "\n",
       "       GenrePreference_Fantasy  GenrePreference_Sci-Fi  \n",
       "66286                        0                       1  \n",
       "40551                        1                       0  \n",
       "17224                        0                       0  \n",
       "22313                        0                       0  \n",
       "20011                        0                       1  \n",
       "\n",
       "[5 rows x 34 columns]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df= preprocessData(test_df)\n",
    "test_df.sample(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AccountAge                        float64\n",
       "MonthlyCharges                    float64\n",
       "TotalCharges                      float64\n",
       "PaperlessBilling                    int64\n",
       "MultiDeviceAccess                   int64\n",
       "ViewingHoursPerWeek               float64\n",
       "AverageViewingDuration            float64\n",
       "ContentDownloadsPerMonth          float64\n",
       "UserRating                        float64\n",
       "SupportTicketsPerMonth            float64\n",
       "Gender                              int64\n",
       "WatchlistSize                       int64\n",
       "ParentalControl                     int64\n",
       "SubtitlesEnabled                    int64\n",
       "Churn                               int64\n",
       "SubscriptionType_Basic              int64\n",
       "SubscriptionType_Premium            int64\n",
       "SubscriptionType_Standard           int64\n",
       "PaymentMethod_Bank transfer         int64\n",
       "PaymentMethod_Credit card           int64\n",
       "PaymentMethod_Electronic check      int64\n",
       "PaymentMethod_Mailed check          int64\n",
       "ContentType_Both                    int64\n",
       "ContentType_Movies                  int64\n",
       "ContentType_TV Shows                int64\n",
       "DeviceRegistered_Computer           int64\n",
       "DeviceRegistered_Mobile             int64\n",
       "DeviceRegistered_TV                 int64\n",
       "DeviceRegistered_Tablet             int64\n",
       "GenrePreference_Action              int64\n",
       "GenrePreference_Comedy              int64\n",
       "GenrePreference_Drama               int64\n",
       "GenrePreference_Fantasy             int64\n",
       "GenrePreference_Sci-Fi              int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train_df.dtypes # check data types \n",
    "balanceTrainDF.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AccountAge                        float64\n",
       "MonthlyCharges                    float64\n",
       "TotalCharges                      float64\n",
       "PaperlessBilling                    int64\n",
       "MultiDeviceAccess                   int64\n",
       "ViewingHoursPerWeek               float64\n",
       "AverageViewingDuration            float64\n",
       "ContentDownloadsPerMonth          float64\n",
       "UserRating                        float64\n",
       "SupportTicketsPerMonth            float64\n",
       "Gender                              int64\n",
       "WatchlistSize                       int64\n",
       "ParentalControl                     int64\n",
       "SubtitlesEnabled                    int64\n",
       "CustomerID                         object\n",
       "SubscriptionType_Basic              int64\n",
       "SubscriptionType_Premium            int64\n",
       "SubscriptionType_Standard           int64\n",
       "PaymentMethod_Bank transfer         int64\n",
       "PaymentMethod_Credit card           int64\n",
       "PaymentMethod_Electronic check      int64\n",
       "PaymentMethod_Mailed check          int64\n",
       "ContentType_Both                    int64\n",
       "ContentType_Movies                  int64\n",
       "ContentType_TV Shows                int64\n",
       "DeviceRegistered_Computer           int64\n",
       "DeviceRegistered_Mobile             int64\n",
       "DeviceRegistered_TV                 int64\n",
       "DeviceRegistered_Tablet             int64\n",
       "GenrePreference_Action              int64\n",
       "GenrePreference_Comedy              int64\n",
       "GenrePreference_Drama               int64\n",
       "GenrePreference_Fantasy             int64\n",
       "GenrePreference_Sci-Fi              int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AccountAge: [ 20  57  73  32 113  38  25  26  14 114   3  64  43  98  47  11  50   5\n",
      "  63  55  84  16  80 116  33  86  52  77   4  91 119  72  94  99  40  60\n",
      " 105  54  34  67  49 100   6  97  24  53  82  18   8  29   9 101  30  74\n",
      "  17  69  61  65  96  62  35  51 109  76   2  42 111  68  15  27  58 103\n",
      "  45  36  85 107 115  81 106  70  46  83  92   1  44 118  23  88   7  31\n",
      "  71  48  93 117  95  37 112  21 108  19  12  78 104  87  90 110  56  79\n",
      "  41  13  28  10  59  39  75 102  89  66  22]\n",
      "MonthlyCharges: [11.0552151   5.17520846 12.10665687 ... 14.51456873 18.14055532\n",
      " 11.59377449]\n",
      "TotalCharges: [ 221.10430197  294.98688225  883.78595161 ... 1640.14626689  126.98388725\n",
      " 1043.4397042 ]\n",
      "SubscriptionType: ['Premium' 'Basic' 'Standard']\n",
      "PaymentMethod: ['Mailed check' 'Credit card' 'Electronic check' 'Bank transfer']\n",
      "PaperlessBilling: ['No' 'Yes']\n",
      "ContentType: ['Both' 'Movies' 'TV Shows']\n",
      "MultiDeviceAccess: ['No' 'Yes']\n",
      "DeviceRegistered: ['Mobile' 'Tablet' 'Computer' 'TV']\n",
      "ViewingHoursPerWeek: [36.75810391 32.45056783  7.39516011 ... 10.62872846 30.46678245\n",
      " 24.97253673]\n",
      "AverageViewingDuration: [ 63.53137733  25.72559464  57.36406085 ... 176.18609505 153.38631494\n",
      "  84.824498  ]\n",
      "ContentDownloadsPerMonth: [10 18 23 30 20 35 28  0 49 26 15  6 41  3  1  7 37 25 22 43 13 17 40 12\n",
      "  8  2  4 16 34 29 19 31 11  9 42  5 47 36 39 48 27 46 45 33 21 24 44 38\n",
      " 14 32]\n",
      "GenrePreference: ['Sci-Fi' 'Action' 'Fantasy' 'Drama' 'Comedy']\n",
      "UserRating: [2.17649751 3.47863229 4.23882362 ... 4.01221731 2.13578882 1.42889572]\n",
      "SupportTicketsPerMonth: [4 8 6 2 9 0 1 7 3 5]\n",
      "Gender: ['Male' 'Female']\n",
      "WatchlistSize: [ 3 23  1 24  0  2 20 22  5 18  8 14 21 15 13 17  9  7 11 10 19  4 12 16\n",
      "  6]\n",
      "ParentalControl: ['No' 'Yes']\n",
      "SubtitlesEnabled: ['No' 'Yes']\n",
      "Churn: [0 1]\n"
     ]
    }
   ],
   "source": [
    "printAllColumnsUniqueValues(train_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['AccountAge', 'MonthlyCharges', 'TotalCharges', 'PaperlessBilling',\n",
       "       'MultiDeviceAccess', 'ViewingHoursPerWeek', 'AverageViewingDuration',\n",
       "       'ContentDownloadsPerMonth', 'UserRating', 'SupportTicketsPerMonth',\n",
       "       'Gender', 'WatchlistSize', 'ParentalControl', 'SubtitlesEnabled',\n",
       "       'Churn', 'SubscriptionType_Basic', 'SubscriptionType_Premium',\n",
       "       'SubscriptionType_Standard', 'PaymentMethod_Bank transfer',\n",
       "       'PaymentMethod_Credit card', 'PaymentMethod_Electronic check',\n",
       "       'PaymentMethod_Mailed check', 'ContentType_Both', 'ContentType_Movies',\n",
       "       'ContentType_TV Shows', 'DeviceRegistered_Computer',\n",
       "       'DeviceRegistered_Mobile', 'DeviceRegistered_TV',\n",
       "       'DeviceRegistered_Tablet', 'GenrePreference_Action',\n",
       "       'GenrePreference_Comedy', 'GenrePreference_Drama',\n",
       "       'GenrePreference_Fantasy', 'GenrePreference_Sci-Fi'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train_df.columns\n",
    "balanceTrainDF.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AccountAge</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>MultiDeviceAccess</th>\n",
       "      <th>ViewingHoursPerWeek</th>\n",
       "      <th>AverageViewingDuration</th>\n",
       "      <th>ContentDownloadsPerMonth</th>\n",
       "      <th>UserRating</th>\n",
       "      <th>SupportTicketsPerMonth</th>\n",
       "      <th>...</th>\n",
       "      <th>ContentType_TV Shows</th>\n",
       "      <th>DeviceRegistered_Computer</th>\n",
       "      <th>DeviceRegistered_Mobile</th>\n",
       "      <th>DeviceRegistered_TV</th>\n",
       "      <th>DeviceRegistered_Tablet</th>\n",
       "      <th>GenrePreference_Action</th>\n",
       "      <th>GenrePreference_Comedy</th>\n",
       "      <th>GenrePreference_Drama</th>\n",
       "      <th>GenrePreference_Fantasy</th>\n",
       "      <th>GenrePreference_Sci-Fi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "      <td>88364.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.452716</td>\n",
       "      <td>0.524889</td>\n",
       "      <td>0.293093</td>\n",
       "      <td>0.500079</td>\n",
       "      <td>0.499525</td>\n",
       "      <td>0.469796</td>\n",
       "      <td>0.463487</td>\n",
       "      <td>0.468561</td>\n",
       "      <td>0.504589</td>\n",
       "      <td>0.522250</td>\n",
       "      <td>...</td>\n",
       "      <td>0.330146</td>\n",
       "      <td>0.251245</td>\n",
       "      <td>0.250928</td>\n",
       "      <td>0.245892</td>\n",
       "      <td>0.251935</td>\n",
       "      <td>0.193065</td>\n",
       "      <td>0.206000</td>\n",
       "      <td>0.198305</td>\n",
       "      <td>0.198327</td>\n",
       "      <td>0.204303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.289105</td>\n",
       "      <td>0.287808</td>\n",
       "      <td>0.218293</td>\n",
       "      <td>0.500003</td>\n",
       "      <td>0.500003</td>\n",
       "      <td>0.287309</td>\n",
       "      <td>0.287710</td>\n",
       "      <td>0.293825</td>\n",
       "      <td>0.289157</td>\n",
       "      <td>0.319507</td>\n",
       "      <td>...</td>\n",
       "      <td>0.470268</td>\n",
       "      <td>0.433731</td>\n",
       "      <td>0.433550</td>\n",
       "      <td>0.430617</td>\n",
       "      <td>0.434127</td>\n",
       "      <td>0.394706</td>\n",
       "      <td>0.404433</td>\n",
       "      <td>0.398725</td>\n",
       "      <td>0.398742</td>\n",
       "      <td>0.403193</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",
       "      <td>...</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",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.194915</td>\n",
       "      <td>0.280156</td>\n",
       "      <td>0.115459</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.218211</td>\n",
       "      <td>0.210575</td>\n",
       "      <td>0.204082</td>\n",
       "      <td>0.255012</td>\n",
       "      <td>0.222222</td>\n",
       "      <td>...</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",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.432203</td>\n",
       "      <td>0.537784</td>\n",
       "      <td>0.246315</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.454755</td>\n",
       "      <td>0.445425</td>\n",
       "      <td>0.448980</td>\n",
       "      <td>0.506689</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>...</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",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.694915</td>\n",
       "      <td>0.775546</td>\n",
       "      <td>0.429843</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.713754</td>\n",
       "      <td>0.706420</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.755581</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</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>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",
       "      <td>...</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",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 34 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         AccountAge  MonthlyCharges  TotalCharges  PaperlessBilling   \n",
       "count  88364.000000    88364.000000  88364.000000      88364.000000  \\\n",
       "mean       0.452716        0.524889      0.293093          0.500079   \n",
       "std        0.289105        0.287808      0.218293          0.500003   \n",
       "min        0.000000        0.000000      0.000000          0.000000   \n",
       "25%        0.194915        0.280156      0.115459          0.000000   \n",
       "50%        0.432203        0.537784      0.246315          1.000000   \n",
       "75%        0.694915        0.775546      0.429843          1.000000   \n",
       "max        1.000000        1.000000      1.000000          1.000000   \n",
       "\n",
       "       MultiDeviceAccess  ViewingHoursPerWeek  AverageViewingDuration   \n",
       "count       88364.000000         88364.000000            88364.000000  \\\n",
       "mean            0.499525             0.469796                0.463487   \n",
       "std             0.500003             0.287309                0.287710   \n",
       "min             0.000000             0.000000                0.000000   \n",
       "25%             0.000000             0.218211                0.210575   \n",
       "50%             0.000000             0.454755                0.445425   \n",
       "75%             1.000000             0.713754                0.706420   \n",
       "max             1.000000             1.000000                1.000000   \n",
       "\n",
       "       ContentDownloadsPerMonth    UserRating  SupportTicketsPerMonth  ...   \n",
       "count              88364.000000  88364.000000            88364.000000  ...  \\\n",
       "mean                   0.468561      0.504589                0.522250  ...   \n",
       "std                    0.293825      0.289157                0.319507  ...   \n",
       "min                    0.000000      0.000000                0.000000  ...   \n",
       "25%                    0.204082      0.255012                0.222222  ...   \n",
       "50%                    0.448980      0.506689                0.555556  ...   \n",
       "75%                    0.714286      0.755581                0.777778  ...   \n",
       "max                    1.000000      1.000000                1.000000  ...   \n",
       "\n",
       "       ContentType_TV Shows  DeviceRegistered_Computer   \n",
       "count          88364.000000               88364.000000  \\\n",
       "mean               0.330146                   0.251245   \n",
       "std                0.470268                   0.433731   \n",
       "min                0.000000                   0.000000   \n",
       "25%                0.000000                   0.000000   \n",
       "50%                0.000000                   0.000000   \n",
       "75%                1.000000                   1.000000   \n",
       "max                1.000000                   1.000000   \n",
       "\n",
       "       DeviceRegistered_Mobile  DeviceRegistered_TV  DeviceRegistered_Tablet   \n",
       "count             88364.000000         88364.000000             88364.000000  \\\n",
       "mean                  0.250928             0.245892                 0.251935   \n",
       "std                   0.433550             0.430617                 0.434127   \n",
       "min                   0.000000             0.000000                 0.000000   \n",
       "25%                   0.000000             0.000000                 0.000000   \n",
       "50%                   0.000000             0.000000                 0.000000   \n",
       "75%                   1.000000             0.000000                 1.000000   \n",
       "max                   1.000000             1.000000                 1.000000   \n",
       "\n",
       "       GenrePreference_Action  GenrePreference_Comedy  GenrePreference_Drama   \n",
       "count            88364.000000            88364.000000           88364.000000  \\\n",
       "mean                 0.193065                0.206000               0.198305   \n",
       "std                  0.394706                0.404433               0.398725   \n",
       "min                  0.000000                0.000000               0.000000   \n",
       "25%                  0.000000                0.000000               0.000000   \n",
       "50%                  0.000000                0.000000               0.000000   \n",
       "75%                  0.000000                0.000000               0.000000   \n",
       "max                  1.000000                1.000000               1.000000   \n",
       "\n",
       "       GenrePreference_Fantasy  GenrePreference_Sci-Fi  \n",
       "count             88364.000000            88364.000000  \n",
       "mean                  0.198327                0.204303  \n",
       "std                   0.398742                0.403193  \n",
       "min                   0.000000                0.000000  \n",
       "25%                   0.000000                0.000000  \n",
       "50%                   0.000000                0.000000  \n",
       "75%                   0.000000                0.000000  \n",
       "max                   1.000000                1.000000  \n",
       "\n",
       "[8 rows x 34 columns]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train_df.describe()\n",
    "balanceTrainDF.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ContentType_TV Shows\n",
       "0    59191\n",
       "1    29173\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train_df[\"ContentType_TV Shows\"].value_counts()\n",
    "balanceTrainDF[\"ContentType_TV Shows\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# printAllColumnsUniqueValues(train_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make predictions (required)\n",
    "\n",
    "Remember you should create a dataframe named `prediction_df` with exactly 104,480 entries plus a header row attempting to predict the likelihood of churn for subscriptions in `test_df`. Your submission will throw an error if you have extra columns (beyond `CustomerID` and `predicted_probaility`) or extra rows.\n",
    "\n",
    "The file should have exactly 2 columns:\n",
    "`CustomerID` (sorted in any order)\n",
    "`predicted_probability` (contains your numeric predicted probabilities between 0 and 1, e.g. from `estimator.predict_proba(X, y)[:, 1]`)\n",
    "\n",
    "The naming convention of the dataframe and columns are critical for our autograding, so please make sure to use the exact naming conventions of `prediction_df` with column names `CustomerID` and `predicted_probability`!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Defaulting to user installation because normal site-packages is not writeable\n",
      "Requirement already satisfied: xgboost in /home/johnsonhk88/.local/lib/python3.10/site-packages (1.7.6)\n",
      "Requirement already satisfied: numpy in /home/johnsonhk88/.local/lib/python3.10/site-packages (from xgboost) (1.24.3)\n",
      "Requirement already satisfied: scipy in /usr/lib/python3/dist-packages (from xgboost) (1.8.0)\n",
      "Defaulting to user installation because normal site-packages is not writeable\n",
      "Requirement already satisfied: lightgbm in /home/johnsonhk88/.local/lib/python3.10/site-packages (4.3.0)\n",
      "Requirement already satisfied: numpy in /home/johnsonhk88/.local/lib/python3.10/site-packages (from lightgbm) (1.24.3)\n",
      "Requirement already satisfied: scipy in /usr/lib/python3/dist-packages (from lightgbm) (1.8.0)\n"
     ]
    }
   ],
   "source": [
    "## Install  xgboost and lightgbm\n",
    "!pip install xgboost\n",
    "!pip install lightgbm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Use Different Machine Learning Models for Predicting, selecting best model for final prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "#import ML different model\n",
    "from xgboost import XGBClassifier\n",
    "from lightgbm import LGBMClassifier\n",
    "# from catboost import CatBoostClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### RandomForest Model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "nEstimate=  200, #850#1000#500\n",
    "rounds=10\n",
    "maxDepth = 7#10\n",
    "\n",
    "rf = RandomForestClassifier(n_estimators=200, \n",
    "                            class_weight='balanced', \n",
    "                            random_state=42, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XgBoost Initial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'objective': 'binary:logistic',\n",
       " 'use_label_encoder': None,\n",
       " 'base_score': None,\n",
       " 'booster': None,\n",
       " 'callbacks': None,\n",
       " 'colsample_bylevel': None,\n",
       " 'colsample_bynode': None,\n",
       " 'colsample_bytree': None,\n",
       " 'early_stopping_rounds': None,\n",
       " 'enable_categorical': False,\n",
       " 'eval_metric': None,\n",
       " 'feature_types': None,\n",
       " 'gamma': None,\n",
       " 'gpu_id': None,\n",
       " 'grow_policy': None,\n",
       " 'importance_type': None,\n",
       " 'interaction_constraints': None,\n",
       " 'learning_rate': None,\n",
       " 'max_bin': None,\n",
       " 'max_cat_threshold': None,\n",
       " 'max_cat_to_onehot': None,\n",
       " 'max_delta_step': None,\n",
       " 'max_depth': 6,\n",
       " 'max_leaves': None,\n",
       " 'min_child_weight': None,\n",
       " 'missing': nan,\n",
       " 'monotone_constraints': None,\n",
       " 'n_estimators': 300,\n",
       " 'n_jobs': -1,\n",
       " 'num_parallel_tree': None,\n",
       " 'predictor': None,\n",
       " 'random_state': 42,\n",
       " 'reg_alpha': None,\n",
       " 'reg_lambda': None,\n",
       " 'sampling_method': None,\n",
       " 'scale_pos_weight': None,\n",
       " 'subsample': None,\n",
       " 'tree_method': None,\n",
       " 'validate_parameters': None,\n",
       " 'verbosity': None}"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nEstimate=300 #1000\n",
    "rounds=10\n",
    "maxDepth = 6#10\n",
    "\n",
    "xgb = XGBClassifier(n_estimators=nEstimate,\n",
    "                     max_depth=maxDepth, \n",
    "                     objective='binary:logistic',\n",
    "                    #  tree_method='gpu_hist', # GPU accelerated training.\n",
    "                     n_jobs=-1, # for CPU parallelism\n",
    "                     random_state=42)\n",
    "xgb.get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LGBM inital"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "lgbmModel = LGBMClassifier(n_estimators=nEstimate, \n",
    "                           max_depth=maxDepth, \n",
    "                           n_jobs=-1,\n",
    "                           class_weight='balanced',\n",
    "                           random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Example prediction submission:\n",
    "\n",
    "The code below is a very naive prediction method that simply predicts churn using a Dummy Classifier. This is used as just an example showing the submission format required. Please change/alter/delete this code below and create your own improved prediction methods for generating `prediction_df`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define X and Y train data and Train Test data split into training and testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "# X = train_df.drop(\"Churn\", axis=1)\n",
    "# y = train_df[\"Churn\"]\n",
    "X = balanceTrainDF.drop(\"Churn\", axis=1)\n",
    "y = balanceTrainDF[\"Churn\"]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, stratify=y, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((70691, 33), (17673, 33))"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**PLEASE CHANGE CODE BELOW TO IMPLEMENT YOUR OWN PREDICTIONS**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "### PLEASE CHANGE THIS CODE TO IMPLEMENT YOUR OWN PREDICTIONS\n",
    "\n",
    "# Fit a dummy classifier on the feature columns in train_df:\n",
    "# dummy_clf = DummyClassifier(strategy=\"stratified\")\n",
    "# dummy_clf.fit(train_df.drop(['CustomerID', 'Churn'], axis=1), train_df.Churn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomForestClassifier(class_weight=&#x27;balanced&#x27;, n_estimators=200, n_jobs=-1,\n",
       "                       random_state=42)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestClassifier</label><div class=\"sk-toggleable__content\"><pre>RandomForestClassifier(class_weight=&#x27;balanced&#x27;, n_estimators=200, n_jobs=-1,\n",
       "                       random_state=42)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomForestClassifier(class_weight='balanced', n_estimators=200, n_jobs=-1,\n",
       "                       random_state=42)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Random forest train \n",
    "rf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "evalRF = rf.predict(X_test)\n",
    "evalRFProb = rf.predict_proba(X_test)[:,1] # get the probability of the positive class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "reportClassificationRF = classification_report(y_test, evalRF)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.68      0.68      0.68      8837\n",
      "           1       0.68      0.68      0.68      8836\n",
      "\n",
      "    accuracy                           0.68     17673\n",
      "   macro avg       0.68      0.68      0.68     17673\n",
      "weighted avg       0.68      0.68      0.68     17673\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(reportClassificationRF)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7454363541435239"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc_auc = roc_auc_score(y_test, evalRFProb)\n",
    "rfc_fpr, rfc_tpr, rfc_th = roc_curve(y_test, evalRFProb)\n",
    "rfc_auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(rfc_fpr, rfc_tpr, label='ROC curve (area = %0.2f)' % rfc_auc)\n",
    "plt.plot([0, 1], [0, 1], 'k--', label='No Skill')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve for Random Forest Classification')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XBoost training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {color: black;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>XGBClassifier(base_score=None, booster=None, callbacks=None,\n",
       "              colsample_bylevel=None, colsample_bynode=None,\n",
       "              colsample_bytree=None, early_stopping_rounds=None,\n",
       "              enable_categorical=False, eval_metric=None, feature_types=None,\n",
       "              gamma=None, gpu_id=None, grow_policy=None, importance_type=None,\n",
       "              interaction_constraints=None, learning_rate=None, max_bin=None,\n",
       "              max_cat_threshold=None, max_cat_to_onehot=None,\n",
       "              max_delta_step=None, max_depth=6, max_leaves=None,\n",
       "              min_child_weight=None, missing=nan, monotone_constraints=None,\n",
       "              n_estimators=300, n_jobs=-1, num_parallel_tree=None,\n",
       "              predictor=None, random_state=42, ...)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" checked><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">XGBClassifier</label><div class=\"sk-toggleable__content\"><pre>XGBClassifier(base_score=None, booster=None, callbacks=None,\n",
       "              colsample_bylevel=None, colsample_bynode=None,\n",
       "              colsample_bytree=None, early_stopping_rounds=None,\n",
       "              enable_categorical=False, eval_metric=None, feature_types=None,\n",
       "              gamma=None, gpu_id=None, grow_policy=None, importance_type=None,\n",
       "              interaction_constraints=None, learning_rate=None, max_bin=None,\n",
       "              max_cat_threshold=None, max_cat_to_onehot=None,\n",
       "              max_delta_step=None, max_depth=6, max_leaves=None,\n",
       "              min_child_weight=None, missing=nan, monotone_constraints=None,\n",
       "              n_estimators=300, n_jobs=-1, num_parallel_tree=None,\n",
       "              predictor=None, random_state=42, ...)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "XGBClassifier(base_score=None, booster=None, callbacks=None,\n",
       "              colsample_bylevel=None, colsample_bynode=None,\n",
       "              colsample_bytree=None, early_stopping_rounds=None,\n",
       "              enable_categorical=False, eval_metric=None, feature_types=None,\n",
       "              gamma=None, gpu_id=None, grow_policy=None, importance_type=None,\n",
       "              interaction_constraints=None, learning_rate=None, max_bin=None,\n",
       "              max_cat_threshold=None, max_cat_to_onehot=None,\n",
       "              max_delta_step=None, max_depth=6, max_leaves=None,\n",
       "              min_child_weight=None, missing=nan, monotone_constraints=None,\n",
       "              n_estimators=300, n_jobs=-1, num_parallel_tree=None,\n",
       "              predictor=None, random_state=42, ...)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Random forest train \n",
    "xgb.fit(X_train, y_train, verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "evalXgb = xgb.predict(X_test)\n",
    "evalXgbProb = xgb.predict_proba(X_test)[:,1] # get the probability of the positive class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.68      0.65      0.67      8837\n",
      "           1       0.67      0.69      0.68      8836\n",
      "\n",
      "    accuracy                           0.67     17673\n",
      "   macro avg       0.67      0.67      0.67     17673\n",
      "weighted avg       0.67      0.67      0.67     17673\n",
      "\n"
     ]
    }
   ],
   "source": [
    "reportClassificationXgb = classification_report(y_test, evalXgb)\n",
    "print(reportClassificationXgb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.731328626813073"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgb_auc = roc_auc_score(y_test, evalXgbProb)\n",
    "xgb_auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_fpr, xgb_tpr, xgb_th = roc_curve(y_test, evalXgbProb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(xgb_fpr, xgb_tpr, label='ROC curve (area = %0.2f)' % xgb_auc)\n",
    "plt.plot([0, 1], [0, 1], 'k--', label='No Skill')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve for XGBoost Classification')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## lightgbm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] Accuracy may be bad since you didn't explicitly set num_leaves OR 2^max_depth > num_leaves. (num_leaves=31).\n",
      "[LightGBM] [Warning] Found whitespace in feature_names, replace with underlines\n",
      "[LightGBM] [Warning] Accuracy may be bad since you didn't explicitly set num_leaves OR 2^max_depth > num_leaves. (num_leaves=31).\n",
      "[LightGBM] [Info] Number of positive: 35346, number of negative: 35345\n",
      "[LightGBM] [Info] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000480 seconds.\n",
      "You can set `force_row_wise=true` to remove the overhead.\n",
      "And if memory is not enough, you can set `force_col_wise=true`.\n",
      "[LightGBM] [Info] Total Bins 1527\n",
      "[LightGBM] [Info] Number of data points in the train set: 70691, number of used features: 33\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=-0.000000\n",
      "[LightGBM] [Info] Start training from score -0.000000\n",
      "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-3 {color: black;}#sk-container-id-3 pre{padding: 0;}#sk-container-id-3 div.sk-toggleable {background-color: white;}#sk-container-id-3 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-3 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-3 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-3 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-3 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-3 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-3 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-3 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-3 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-3 div.sk-item {position: relative;z-index: 1;}#sk-container-id-3 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-3 div.sk-item::before, #sk-container-id-3 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-3 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-3 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-3 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-3 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-3 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-3 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-3 div.sk-label-container {text-align: center;}#sk-container-id-3 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-3 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>LGBMClassifier(class_weight=&#x27;balanced&#x27;, max_depth=6, n_estimators=300,\n",
       "               n_jobs=-1, random_state=42)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" checked><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LGBMClassifier</label><div class=\"sk-toggleable__content\"><pre>LGBMClassifier(class_weight=&#x27;balanced&#x27;, max_depth=6, n_estimators=300,\n",
       "               n_jobs=-1, random_state=42)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "LGBMClassifier(class_weight='balanced', max_depth=6, n_estimators=300,\n",
       "               n_jobs=-1, random_state=42)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgbmModel.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] Accuracy may be bad since you didn't explicitly set num_leaves OR 2^max_depth > num_leaves. (num_leaves=31).\n",
      "[LightGBM] [Warning] Accuracy may be bad since you didn't explicitly set num_leaves OR 2^max_depth > num_leaves. (num_leaves=31).\n"
     ]
    }
   ],
   "source": [
    "evalLgbm = lgbmModel.predict(X_test)\n",
    "evalLgbmProb = lgbmModel.predict_proba(X_test)[:,1] # get the probability of the positive class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.69      0.66      0.68      8837\n",
      "           1       0.68      0.70      0.69      8836\n",
      "\n",
      "    accuracy                           0.68     17673\n",
      "   macro avg       0.68      0.68      0.68     17673\n",
      "weighted avg       0.68      0.68      0.68     17673\n",
      "\n"
     ]
    }
   ],
   "source": [
    "reportClassificationLgbm = classification_report(y_test, evalLgbm)\n",
    "print(reportClassificationLgbm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7463127402773218"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgbm_auc = roc_auc_score(y_test, evalLgbmProb)\n",
    "lgbm_auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "lgbm_fpr, lgbm_tpr, lgbm_th = roc_curve(y_test, evalLgbmProb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiMAAAGJCAYAAABYRTOkAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACKZElEQVR4nOzdd1gT2dcH8G8ChF5FqgiCXVAUEXtFwd7FjhW7rm3toqhgRSwo9u4PXVddV11dxYoNG1aKKIgNFJHek/v+4euskSLBwFDO53l4zNzcmTkZAzm5bQSMMQZCCCGEEJ4I+Q6AEEIIIRUbJSOEEEII4RUlI4QQQgjhFSUjhBBCCOEVJSOEEEII4RUlI4QQQgjhFSUjhBBCCOEVJSOEEEII4RUlI4QQQgjhFSUjhJRCa9asgaWlJRQUFGBra8t3ONi7dy8EAgGioqKKvO+9e/fkH1g5NmLECFhYWPB2/vz+z/N6b1pYWGDEiBElHuOSJUsgEAhK/LxE/igZIbl8+yP07UdRURGmpqYYMWIE3r17l+c+jDEcOHAArVu3ho6ODtTU1GBjYwMPDw+kpqbme64TJ06gc+fO0NfXh0gkgomJCQYMGIBLly4VKtaMjAysX78eDg4O0NbWhoqKCmrWrInJkycjPDy8SK+fb//++y9+//13tGjRAnv27IGnp2exnm/EiBHQ0NAo1nMU1pYtW7B37958n//48SPmzp0LGxsbaGhoQEVFBdWrV8fIkSMRGBgoVffH97FAIICBgQHatWuHf/75J9exv9UZM2ZMnudesGABVycuLq7IrzEpKQlLly5FgwYNoKGhAVVVVVhbW2POnDl4//59kY9bEkr6vQkAaWlpWLJkCa5cuVLs5yI8YoT8YM+ePQwA8/DwYAcOHGA7duxgo0ePZgoKCszKyoqlp6dL1c/JyWEDBgxgAFirVq3Y+vXr2bZt29jQoUOZUChk1tbWLCYmRmofiUTCRowYwQCwhg0bshUrVrBdu3ax5cuXMzs7OwaA3bhxo8A4P336xNXt1q0b8/HxYTt37mSzZ89mZmZmTElJSe7XpiTMmTOHCYVClpmZWSLnc3V1Zerq6gXWycnJYenp6Uwikch8/G/vp7t37/60br169VibNm3yfO7OnTtMX1+fKSsrM1dXV7Z582a2Y8cONn/+fFa3bl0GgF29ejXXeb+9j/fv38/WrFnD6tWrxwCwv//+W+r4AJiKigrT0dHJ89pXq1aNqaioMADs06dPsl2E//fy5UtWrVo1pqCgwAYOHMg2b97Mtm/fziZPnswqVarEatSowdV1dXVl5ubmRTqPPOT1f57fezMjI4NlZWUVSxyfPn1iAJi7u3uu57Kzs3P9PSJlEyUjJJf8PjzmzJnDALAjR45IlXt6ejIAbNasWbmOderUKSYUCpmzs7NU+Zo1axgA9ttvv+X5Abd//352586dAuPs2rUrEwqF7NixY7mey8jIYDNnzixw/8LKzs4uscSAMcZGjhz50+RAFhKJhKWlpeX7fGGSkV8hj2QkPj6eGRsbMyMjIxYSEpLreYlEwg4fPsyCgoJ+et74+HimpKTEBg8eLFUOgPXq1YsJhUJ28uRJqedu3LjBALC+ffsWORnJzs5mDRo0YGpqauz69eu5nk9MTGTz58/ntvlORvIi7/dmYRSUjJDyg5IRkkt+f8RPnz7NADBPT0+uLC0tjenq6rKaNWuy7OzsPI83cuRIBoDdunWL20dPT4/Vrl2b5eTkFCnG27dvMwBs7Nixharfpk2bPD/kfvyDHxkZyQCwNWvWsPXr1zNLS0smFArZ7du3mYKCAluyZEmuY4SGhjIAbNOmTVzZly9f2LRp01iVKlWYSCRiVlZWbOXKlUwsFhcYJ4BcP3v27GGMff0w8/DwYJaWlkwkEjFzc3M2b948lpGRIXUMc3Nz1rVrV3bu3DlmZ2fHlJWV2fr16/M9Z2GSkW/vicjISK5MLBYzd3d3ZmxszFRVVVnbtm3Zs2fPmLm5OXN1dc21b2BgIJs+fTrT19dnampqrFevXuzjx49Scf/42r/9n31LeP39/QuMM6+Yf3wfSyQSpqWlxYYPHy5VDoBNmjSJtW3blg0YMEDquYkTJzIbGxvm7u5e5GTE39+fAWArVqwoVP28kpE1a9awZs2aMT09PaaiosIaNWrE/vjjj1z7/vvvv6xFixZMW1ubqaurs5o1a7J58+ZJ1dm4cSOrW7cuU1VVZTo6OszOzo4dOnSIe/7H//OC3ps//p8z9vV34LfffmPm5uZMJBIxU1NTNmzYMO7aZWZmskWLFrFGjRoxLS0tpqamxlq2bMkuXbrEHePb7+OPP98Sk2//H9+T9ffk+vXrzN7enikrK7Nq1aqxffv2Ffj/QoqHovw6fEh5920gm66uLlcWGBiIL1++YNq0aVBUzPvtNHz4cOzZswenT59G06ZNERgYiPj4ePz2229QUFAoUiynTp0CAAwbNqxI+//Mnj17kJGRATc3NygrK8PY2Bht2rTB0aNH4e7uLlX3yJEjUFBQQP/+/QF87eNu06YN3r17h3HjxqFq1aq4efMm5s2bhw8fPsDHxyff8x44cADbt29HUFAQdu7cCQBo3rw5AGDMmDHYt28f+vXrh5kzZ+LOnTvw8vJCSEgITpw4IXWcsLAwDBo0COPGjcPYsWNRq1YtOV6dr+bNm4fVq1eje/fucHJywqNHj+Dk5ISMjIw860+ZMgW6urpwd3dHVFQUfHx8MHnyZBw5cgQA4OPjgylTpkBDQwMLFiwAABgaGgIA/v77b6iqqqJPnz4yx5mYmIi4uDgwxvDx40ds2rQJKSkpGDp0aJ71Bw8ejGnTpiElJQUaGhrIycnBH3/8gRkzZuT72gpDHu/ZDRs2oEePHhgyZAiysrLg7++P/v374/Tp0+jatSsA4NmzZ+jWrRvq168PDw8PKCsrIyIiAjdu3OCOs2PHDkydOhX9+vXDtGnTkJGRgcePH+POnTsYPHhwnucu6L35o5SUFLRq1QohISEYNWoUGjVqhLi4OJw6dQpv376Fvr4+kpKSsHPnTgwaNAhjx45FcnIydu3aBScnJwQFBcHW1haVK1fG1q1bMWHCBPTu3Zv7/69fv36+10iW35OIiAj069cPo0ePhqurK3bv3o0RI0bAzs4O9erVK/x/DPl1fGdDpPT59o3o4sWL7NOnT+zNmzfs2LFjrHLlykxZWZm9efOGq+vj48MAsBMnTuR7vPj4eAaA9enThzHG2IYNG366z8/07t2bAWBfvnwpVH1ZW0a0tLSkvrUzxti2bdsYAPbkyROp8rp167L27dtz28uWLWPq6uosPDxcqt7cuXOZgoICi46OLjDWvFoqgoODGQA2ZswYqfJZs2YxAFLfJr+1MJw7d67A8xR0vh/9+C05JiaGKSoqsl69eknVW7JkCQOQZ8uIo6OjVJfc9OnTmYKCAktISODK8uum0dXVZba2trnKk5KS2KdPn7iflJSUXOf98UdZWZnt3bs317Hw/y0j8fHxTCQSsQMHDjDGGDtz5gwTCAQsKirql1pGGjZsyLS1tQtdP6+WkR+727Kyspi1tbXU+2/9+vU/jbFnz56sXr16BZ4/r9aw/N4rP7aMLF68mAFgx48fz1X323sgJycnV/fnly9fmKGhIRs1ahRXVlA3zY8tI0X5Pbl27RpX9vHjR6asrCy3Ll5SeDSbhuTL0dERlStXhpmZGfr16wd1dXWcOnUKVapU4eokJycDADQ1NfM9zrfnkpKSpP4taJ+fkccxCtK3b19UrlxZqqxPnz5QVFTkvskDwNOnT/H8+XO4uLhwZX/88QdatWoFXV1dxMXFcT+Ojo4Qi8W4du2azPGcPXsWADBjxgyp8pkzZwIAzpw5I1VerVo1ODk5yXyewgoICEBOTg4mTpwoVT5lypR893Fzc5OahtmqVSuIxWK8fv36p+dLSkrKc8bPsGHDULlyZe5nzpw5uer4+vriwoULuHDhAg4ePIh27dphzJgxOH78eJ7n0tXVhbOzM/73v/8BAA4fPozmzZvD3Nz8p3H+7DX86vtVVVWVe/zlyxckJiaiVatWePDgAVeuo6MDAPjrr78gkUjyPI6Ojg7evn2Lu3fv/lI8+fnzzz/RoEED9O7dO9dz394DCgoKEIlEAACJRIL4+Hjk5OSgcePGUq9HFrL+ntStWxetWrXititXroxatWrh1atXRTo/KTpKRki+vv0RP3bsGLp06YK4uDgoKytL1fn2x/VbUpKXHxMWLS2tn+7zM/I4RkGqVauWq0xfXx8dOnTA0aNHubIjR45AUVFRqvvgxYsXOHfunNSHZOXKleHo6Ajg6/RUWb1+/RpCoRDVq1eXKjcyMoKOjk6uD/S84penb+f7MR49PT2pbrzvVa1aVWr7W70vX7789HyamppISUnJVe7h4cElGvlp0qQJHB0d4ejoiCFDhuDMmTOoW7cuJk+ejKysrDz3GTx4MC5cuIDo6GicPHky364LWWhpaf3y+/VbV6eKigr09PS4bozExESujouLC1q0aIExY8bA0NAQAwcOxNGjR6USkzlz5kBDQwNNmjRBjRo1MGnSJKlunF/18uVLWFtb/7Tevn37UL9+faioqKBSpUqoXLkyzpw5I/V6ZCHr78mP70ng6/uyMO9JIl+UjJB8ffsj3rdvX5w6dQrW1tYYPHiw1IdCnTp1AACPHz/O9zjfnqtbty4AoHbt2gCAJ0+eFDk2WY+R38JIYrE4z/Lvv4F+b+DAgQgPD0dwcDAA4OjRo+jQoQP09fW5OhKJBB07duQ+JH/86du3b6FiluV1FDZ+PuU3Pogx9tN9a9eujbCwMGRnZ0uV169fn0s0CksoFKJdu3b48OEDXrx4kWedHj16QFlZGa6ursjMzMSAAQMKffz81K5dG4mJiXjz5k2R9r9+/Tp69OgBFRUVbNmyBWfPnsWFCxcwePBgqWuoqqqKa9eu4eLFixg2bBgeP34MFxcXdOzYkXu/16lTB2FhYfD390fLli3x559/omXLlrnGQxWngwcPYsSIEbCyssKuXbtw7tw5XLhwAe3bt8+3RaewCvt78ivvSSJflIyQQlFQUICXlxfev3+PzZs3c+UtW7aEjo4ODh8+nO8H+/79+wEA3bp14/bR1dXF//73v3z3+Znu3bsD+PoHrTB0dXWRkJCQq7wwXQTf69WrF0QiEY4cOYLg4GCEh4dj4MCBUnWsrKyQkpLCfUj++JPXt7GfMTc3h0QiyfXhGRsbi4SEhF/uQihKPMDXAYDf+/z58y99q8zvQ6Rbt25IT0/PNQCxqHJycgAgz9YW4OsHeq9evXDlyhV07NhRKtksKlnfsz/6888/oaKigvPnz2PUqFHo3LlzvkmYUChEhw4d4O3tjefPn2PFihW4dOkSLl++zNVRV1eHi4sL9uzZg+joaHTt2hUrVqz4pUG631hZWeHp06cF1jl27BgsLS1x/PhxDBs2DE5OTnB0dMx1fllWWC1tvyek8CgZIYXWtm1bNGnSBD4+PtwfDDU1NcyaNQthYWHcDIjvnTlzBnv37oWTkxOaNm3K7TNnzhyEhIRgzpw5eX4LOXjwIIKCgvKNpVmzZnB2dsbOnTtx8uTJXM9nZWVh1qxZ3LaVlRVCQ0Px6dMnruzRo0cyN03r6OjAyckJR48ehb+/P0QiEXr16iVVZ8CAAbh16xbOnz+fa/+EhATug1AWXbp0AYBcM3G8vb0BgJtJUVI6dOgARUVFbN26Var8+0S1KNTV1fNMGidMmABDQ0NMnz49z5V1Zfkmm52djX///RcikYhr2cvLrFmz4O7ujkWLFhX62AXp168fbGxssGLFCty6dSvX88nJyXn+Dn2joKAAgUAglcBHRUXlev/Hx8fn2vfbsu2ZmZkAviaN3xOJRKhbty4YY7lan4qib9++ePToUZ7J47f/q2+tEt//3925cyfXtVFTUwOAPN8XPyptvyek8GhqL5HJ7Nmz0b9/f+zduxfjx48HAMydOxcPHz7EqlWrcOvWLfTt2xeqqqoIDAzEwYMHUadOHezbty/XcZ49e4Z169bh8uXL6NevH4yMjBATE4OTJ08iKCgIN2/eLDCW/fv3o1OnTujTpw+6d++ODh06QF1dHS9evIC/vz8+fPiAtWvXAgBGjRoFb29vODk5YfTo0fj48SP8/PxQr149bjBsYbm4uGDo0KHYsmULnJycuAGD37+2U6dOoVu3btw0wdTUVDx58gTHjh1DVFSUzN+0GzRoAFdXV2zfvh0JCQlo06YNgoKCsG/fPvTq1Qvt2rWT6Xg/ys7OxvLly3OV6+np5RqkCnydcjtt2jSsW7cOPXr0gLOzMx49eoR//vkH+vr6Rb5fiJ2dHbZu3Yrly5ejevXqMDAwQPv27aGnp4cTJ06ge/fuaNCgAQYOHAh7e3soKSnhzZs3+OOPPwDkPQbgn3/+QWhoKICv43UOHz6MFy9eYO7cudzYo7w0aNAADRo0KNLryIuSkhKOHz8OR0dHtG7dGgMGDECLFi2gpKSEZ8+e4fDhw9DV1cWKFSvy3L9r167w9vaGs7MzBg8ejI8fP8LX1xfVq1eX6ib18PDAtWvX0LVrV5ibm+Pjx4/YsmULqlSpgpYtWwIAOnXqBCMjI7Ro0QKGhoYICQnB5s2b0bVrV7kMCp89ezaOHTuG/v37Y9SoUbCzs0N8fDxOnToFPz8/NGjQAN26dcPx48fRu3dvdO3aFZGRkfDz80PdunWlWqxUVVVRt25dHDlyBDVr1oSenh6sra3zHJNS3L8npBjxN5GHlFYFrZgpFouZlZUVs7KyklqwTCwWsz179rAWLVowLS0tpqKiwurVq8eWLl0qNd3yR8eOHWOdOnVienp6TFFRkRkbGzMXFxd25cqVQsWalpbG1q5dy+zt7ZmGhgYTiUSsRo0abMqUKSwiIkKq7sGDB7mFkGxtbdn58+cLXPQsP0lJSUxVVZUBYAcPHsyzTnJyMps3bx6rXr06E4lETF9fnzVv3pytXbv2p8tm5zd9Mjs7my1dupRVq1aNKSkpMTMzswIXcyosV1fXPKfAAmBWVlaMsbyneebk5LBFixYxIyMjpqqqytq3b89CQkJYpUqV2Pjx47l6+b2fLl++zACwy5cvc2UxMTGsa9euTFNTU2rRs28+fPjAZs+ezS3WpayszCwtLdnw4cOlpmh+f97vf1RUVJitrS3bunVrrpV/8f9TewvyK1N7v/ny5QtbvHgxs7GxYWpqakxFRYVZW1uzefPmsQ8fPnD18prau2vXLlajRg2mrKzMateuzfbs2ZNremtAQADr2bMnMzExYSKRiJmYmLBBgwZJTTXftm0ba926NatUqRJTVlZmVlZWbPbs2SwxMZGr8ytTexlj7PPnz2zy5MnM1NSUiUQiVqVKFebq6sri4uIYY1+n+Hp6ejJzc3OmrKzMGjZsyE6fPp3n67558yazs7NjIpGoUIue/crvSX7LAJDiJWCMRuoQQuQjISEBurq6WL58eYFdDoQQ8j0aM0IIKZL09PRcZd/66tu2bVuywRBCyjQaM0IIKZIjR45g79696NKlCzQ0NBAYGIj//e9/6NSpE1q0aMF3eISQMoSSEUJIkdSvXx+KiopYvXo1kpKSuEGteQ2EJYSQgtCYEUIIIYTwisaMEEIIIYRXlIwQQgghhFcVbsyIRCLB+/fvoampWeSFmQghhJCKiDGG5ORkmJiYQCiUX3tGhUtG3r9/DzMzM77DIIQQQsqsN2/eoEqVKnI7XoVLRr4tdfzmzZsCl4ImhBBCiLSkpCSYmZnJ5bYB36twyci3rhktLS1KRgghhJAikPcwBxrASgghhBBeUTJCCCGEEF5RMkIIIYQQXlW4MSOFwRhDTk4OxGIx36EQUm4pKChAUVGRptgTQigZ+VFWVhY+fPiAtLQ0vkMhpNxTU1ODsbExRCIR36EQQnhEych3JBIJIiMjoaCgABMTE4hEIvrWRkgxYIwhKysLnz59QmRkJGrUqCHXBZQIIWULJSPfycrKgkQigZmZGdTU1PgOh5ByTVVVFUpKSnj9+jWysrKgoqLCd0iEEJ7w+lXk2rVr6N69O0xMTCAQCHDy5Mmf7nPlyhU0atQIysrKqF69Ovbu3Sv3uOgbGiElg37XCCEAz8lIamoqGjRoAF9f30LVj4yMRNeuXdGuXTsEBwfjt99+w5gxY3D+/PlijpQQQgghxYXXbprOnTujc+fOha7v5+eHatWqYd26dQCAOnXqIDAwEOvXr4eTk1NxhUkIIYSQYlSmxozcunULjo6OUmVOTk747bff8t0nMzMTmZmZ3HZSUlJxhUfKqLCwMLRp0wYvXryQ+/0WKrK4uDjUrVsXDx48kOsNtQghv0YsYUjPFiMrR4KUjBykZecgI1uCDwnpiPycCpGCEJk5EryITYaWqhIYA3IkEmSLGVKTi+cztEwlIzExMTA0NJQqMzQ0RFJSEtLT06GqqpprHy8vLyxdurSkQuTNiBEjsG/fPgCAoqIiqlSpgv79+8PDwyPXwMDTp09jzZo1ePDgAcRiMerVq4dJkyZhxIgRuY77559/YtOmTXj48CHEYjEsLS3Rr18/TJ48GXp6eiXx0ordvHnzMGXKlHKdiPj6+mLNmjWIiYlBgwYNsGnTJjRp0iTf+m3btsXVq1dzlXfp0gVnzpwBIP2e+8bJyQnnzp0DAOjr62P48OFwd3fHrl275PhqCCHfy8gWIy4lExnZEsSlZEIsYYhPzcKjNwl4+SkFUZ/TEBmXikrqInxOzSrSObITYvAlYDu0mw+Sc/RflalkpCjmzZuHGTNmcNvf7jhYHjk7O2PPnj3Izs7G/fv34erqCoFAgFWrVnF1Nm3ahN9++w1z5szB1q1bIRKJ8Ndff2H8+PF4+vQp1q5dy9VdsGABVq1ahenTp8PT0xMmJiZ48eIF/Pz8cODAAUybNq1EXldWVlaxrUMRHR2N06dPY9OmTb90nOKM8VcdOXIEM2bMgJ+fHxwcHODj4wMnJyeEhYXBwMAgz32OHz+OrKz//mh9/vwZDRo0QP/+/aXqfXvPfaOsrCz1/MiRI2FnZ4c1a9aUm+SVkJKSkS1GZrYEmTlifEjMQNTnVFwN+wRVkQJuv/qM+NQsfEnLLvTx8kpElBQEyBYzmOqoQkVJiI/JmTDSUkHDqjoQKQrxIT4V7wP/QMDhLcjOzIC5tiJi5Pki/1+ZSkaMjIwQGxsrVRYbGwstLa08W0WAr38cf/wDKQvGvjZn8UFVSUGmdU6UlZVhZGQEADAzM4OjoyMuXLjAJSNv3rzBzJkz8dtvv8HT05Pbb+bMmRCJRJg6dSr69+8PBwcHBAUFwdPTEz4+PlJJh4WFBTp27IiEhIR843j79i1mz56N8+fPIzMzE3Xq1IGvry8cHBwwYsQIJCQkSM2c+u233xAcHIwrV64A+Pqt3NraGoqKijh48CBsbGxgbGwMsViMI0eOcPtlZ2fD2NgY3t7eGD58OCQSCVatWoXt27cjJiYGNWvWxKJFi9CvX798Yz169CgaNGgAU1NTruzz58+YPHkyrl27hi9fvsDKygrz58/HoEH/fSPIK8bLly/j6dOnmD17Nq5fvw51dXV06tQJ69evh76+PgDg3LlzWL58OZ4+fQoFBQU0a9YMGzZsgJWVVQH/s7/G29sbY8eOxciRIwF8HXt15swZ7N69G3Pnzs1znx8TB39/f6ipqeVKRr5/z+WlXr16MDExwYkTJzB69OhffCWElB+MMbz9ko5n7xMRk5iBlMwcvEtIx8PoBKRliREdL/vCm9qqSmCMISkjBw2qaAMCATKzxWhmVQlWlTVgUUkdprpfkw4dVRFEikIoCPP/jAkKCsJY93F4/PgxAKB9+/ZYu3YtGjVqVOTXnZ8ylYw0a9YMZ8+elSq7cOECmjVrVmznTM8Wo+5ifmbrPPdwgpqoaP9FT58+xc2bN2Fubs6VHTt2DNnZ2Zg1a1au+uPGjcP8+fPxv//9Dw4ODjh06BA0NDQwceLEPI+vo6OTZ3lKSgratGkDU1NTnDp1CkZGRnjw4AEkEolM8e/btw8TJkzAjRs3AAARERHo378/UlJSoKGhAQA4f/480tLS0Lt3bwBfu+QOHjwIPz8/1KhRA9euXcPQoUNRuXJltGnTJs/zXL9+HY0bN5Yqy8jIgJ2dHebMmQMtLS2cOXMGw4YNg5WVlVTXxo8xJiQkoH379hgzZgzWr1+P9PR0zJkzBwMGDMClS5cAfJ1BNmPGDNSvXx8pKSlYvHgxevfujeDg4HynuXp6ekolj3l5/vw5qlatmqs8KysL9+/fx7x587gyoVAIR0dH3Lp1q8Bjfm/Xrl0YOHAg1NXVpcqvXLkCAwMD6Orqon379li+fDkqVaokVadJkya4fv06JSOkwvmYnIHHbxKRmpWD889ikJEtQVBkPFIyc2Q6jkAAqCgqID1bDBtTbQgFQLf6JsiRMNQy0kBNQ02Y6qjKbZHO5ORkLFiwAJs3bwZjDJUqVYK3tzeGDRuG5ORkuZzjR7wmIykpKYiIiOC2IyMjERwcDD09PVStWhXz5s3Du3fvsH//fgDA+PHjsXnzZvz+++8YNWoULl26hKNHj3J92BXd6dOnoaGhgZycHGRmZkIoFGLz5s3c8+Hh4dDW1oaxsXGufUUiESwtLREeHg4AePHiBSwtLaGkpCRTDIcPH8anT59w9+5d7tt19erVZX4tNWrUwOrVq7ltKysrqKur48SJExg2bBh3rh49ekBTUxOZmZnw9PTExYsXueTU0tISgYGB2LZtW77JyOvXr3MlI6amplIJ25QpU3D+/HkcPXpUKhn5Mcbly5ejYcOGUonD7t27YWZmhvDwcNSsWRN9+/aVOtfu3btRuXJlPH/+HNbW1nnGOH78eAwYMKDA62ViYpJneVxcHMRicZ5jrUJDQws85jdBQUF4+vRprnEfzs7O6NOnD6pVq4aXL19i/vz56Ny5M27dugUFBQWp2B4+fFiocxFSluSIJQiNSca9qHjcfPkZGsqKePo+EeGxKTIdR1dNCdUNNGCopQJbMx0YaqnAxlQb2qpK0FBRhKJQUKKrge/cuZPruh42bBjWrVuHypUrF+s5eU1G7t27h3bt2nHb38Z2uLq6Yu/evfjw4QOio6O556tVq4YzZ85g+vTp2LBhA6pUqYKdO3cW67ReVSUFPPfgZ9qwqpLCzyt9p127dti6dStSU1Oxfv16KCoq5vrwKyzGWJH2Cw4ORsOGDX95fICdnZ3UtqKiIgYMGIBDhw5h2LBhSE1NxV9//QV/f38AX1tO0tLS0LFjR6n9srKy0LBhw3zPk56enmuAr1gshqenJ44ePYp3794hKysLmZmZuVbl/THGR48e4fLly1zLzfdevnyJmjVr4sWLF1i8eDHu3LmDuLg4rsUoOjo632RET0+P1/EWu3btgo2NTa4BrwMHDuQe29jYoH79+rCyssKVK1fQoUMH7jlVVVW61xMpkxhjXJfJm/g03Hz5GZ9Ts/AlNQuBEXEyHau5VSUwBrStVRnGOqqoa6yJKrpqUJHx73xxYYxxCc+kSZNw9epVTJ48OdcM1uLCazLStm3bAj/08lpdtW3btiX6LUsgEBS5q6Skqaurc60Qu3fvRoMGDbBr1y6uebxmzZpITEzE+/fvc32TzsrKwsuXL7nksGbNmggMDER2drZMrSP5jd35RigU5vo/z87OPQDrx+4AABgyZAjatGmDjx8/4sKFC1BVVYWzszOAr61sAHDmzBmp8R9A7kGV39PX18eXL1+kytasWYMNGzbAx8cHNjY2UFdXx2+//SY1oDOvGFNSUtC9e3epAcPffGuN6t69O8zNzbFjxw6YmJhAIpHA2to617G/9yvdNPr6+lBQUMhzrFVBYz2+SU1Nhb+/Pzw8PH5a19LSEvr6+oiIiJBKRuLj44v9WxUhv4IxhrDYZNx//QXvvqQj5EMSAiPikC0u/Jcy80pq0NdQRq+GptBXF8G8kjqq6KlCS0W21uWSJhaL4efnh8OHD+Py5csQiUQQiUSFWhFdnsrGpyyRmVAoxPz58zFjxgwMHjwYqqqq6Nu3L+bMmYN169ZxC8d94+fnh9TUVG6Q5uDBg7Fx40Zs2bIlz1kzCQkJeY4bqV+/Pnbu3In4+Pg8v81XrlwZT58+lSoLDg4uVMLTvHlzmJmZ4ciRI/jnn3/Qv39/br+6detCWVkZ0dHR+XbJ5KVhw4Z4/vy5VNmNGzfQs2dPDB06FMDXGyiGh4ejbt26BR6rUaNG+PPPP2FhYQFFxdy/Wp8/f0ZYWBh27NiBVq1aAQACAwN/GuOvdNOIRCLY2dkhICAAvXr14l5PQEAAJk+e/NNz//HHH8jMzOSuRUHevn2Lz58/5+oGfPr0Kdq2bfvT/QkpbmIJQ2xSBp6/T8KlsI8I/ZAECQOC3yQUan/LyurQUxOhmVUlNDLXhY6qEqxNtaGkUDZva/DkyRO4ubnh9u3bAIADBw7wNraLkpFyrH///pg9ezZ8fX0xa9YsVK1aFatXr8bMmTOhoqKCYcOGQUlJCX/99Rfmz5+PmTNnwsHBAQDg4OCA33//HTNnzsS7d+/Qu3dvmJiYICIiAn5+fmjZsmWeScqgQYPg6emJXr16wcvLC8bGxnj48CFMTEzQrFkztG/fHmvWrMH+/fvRrFkzHDx4EE+fPi2wK+V7gwcPhp+fH8LDw3H58mWuXFNTE7NmzcL06dMhkUjQsmVLJCYm4saNG9DS0oKrq2uex3NycsKYMWMgFou5cQ41atTAsWPHcPPmTejq6sLb2xuxsbE/TUYmTZqEHTt2YNCgQfj999+hp6eHiIgI+Pv7Y+fOndDV1UWlSpWwfft2GBsbIzo6Ot/ZLN/71W6aGTNmwNXVFY0bN0aTJk3g4+OD1NRUbnYNAAwfPhympqbw8vKS2nfXrl3o1atXrkGpKSkpWLp0Kfr27QsjIyO8fPkSv//+O6pXry7VbZqWlob79+//tGWHEHnJyBYj5EMS3iWkIyouFccfvsMXGabAVlIXwc5cF3VNtNCqhj5qGGpCU1mxXN3BPT09HcuWLcOaNWuQk5MDTU1NrFy5Ms+1pkoKJSPlmKKiIiZPnozVq1djwoQJXHeDpaUl1q5diw0bNnCLnm3dulXqwwkAVq1aBTs7O/j6+sLPzw8SiQRWVlbo169fvh/uIpEI//77L2bOnIkuXbogJycHdevW5e4/5OTkhEWLFuH3339HRkYGRo0aheHDh+PJkyeFek1DhgzBihUrYG5ujhYtWkg9t2zZMlSuXBleXl549eoVdHR00KhRI8yfPz/f43Xu3BmKioq4ePEi9yG6cOFCvHr1Ck5OTlBTU4Obmxt69eqFxMTEAmMzMTHBjRs3MGfOHHTq1AmZmZkwNzeHs7MzhEIhBAIB/P39MXXqVFhbW6NWrVrYuHFjsbcauLi44NOnT1i8eDFiYmJga2uLc+fOSQ1qjY6OzjWbJywsDIGBgfj3339zHVNBQQGPHz/Gvn37kJCQABMTE3Tq1AnLli2T6hb766+/ULVqVa4liBB5EUsYHkZ/QfCbr1Nhg98k4HLYRxR2uJu6SAENzHTQwEwHza0qwdpEG7rqpXOtIHm6ePEixo8fj5cvXwIA+vTpg40bN+bq3i5pAlbUkYplVFJSErS1tZGYmAgtLS2p5zIyMhAZGYlq1arR7cwrEF9fX5w6dYpuuFgMmjZtiqlTp2Lw4MF5Pk+/c6QwcsQSJKRnIy4lEw9eJ2Ddv2E/XUlUV00Jza30UdtIE1X0VGGirQrzSurQ1xBBsYx2q/wqxhgcHR1x6dIlmJqawtfXFz179pTpGAV9hv4KahkhFd64ceOQkJCA5OTkcr0kfEmLi4tDnz59pBaLI+RnPiSmY2PAC5x8+B6KQgFEikLEp2Xl2+KhqqQAYx0VtLDSh1AA2FnooauNcYGLeVUkjDFkZWVBWVkZAoEAW7duxZYtW+Dh4SHXZOJXUcvId+hbGiEli37nKibGGF58TEFwdAIev0vAi9gU5EgYXn5KQcJPxnbUMNCAg6UeBjQ2Q11jrQrbylEYL168wPjx41GnTh2pNad+BbWMEEIIKXMYY4hNysTLTym4/eozIuNScfrxhwL3saysjhoGX5cv71DHEGZ6qtDXUC6zs1ZKWlZWFtasWYNly5YhMzMTd+7cweLFi/O9F1VpQMkIIYQQufmckomXn1Lx8lMK1p4PQ5ZYguSM/Jc/r6KrigZVdFDXRAtNLfVQvbImtFTL1+yVknTz5k24ubnh2bNnAICOHTvCz8+vVCciACUjhBBCZJSeJcbH5Ax8SMzAs/dJOPvkA+6//lLgPgaayjDVVUU1fXU4VNNDcyt9mOmpFbgPKbzExETMmzcPfn5+YIxBX18fPj4+GDx4cJlI7CgZIYQQkq/E9Gz8FfwO96K+4OGbL0jPkiAuJfOn++mpi2CqowqhUIDDYxygrkwfN8UpIyMD/v7+YIxh5MiRWLNmTa71gUozencQQgiREpuUgf23onAtPA5P3uW9vo5AABhrqcBERxUpmTlobKGLnramqGusRYlHCYmLi4O+vj6Arze//La44vf3fCsr6B1DCCEVFGMMqVli3I2Kx7knMXifmI57UV+Qni2WqidSEMJQWxkdahuiYVUdNLbQg6lOwfehIsVHLBZj8+bNWLhwIQ4ePMitFdKnTx+eIys6SkYIIaSCkEgYLod9xPlnMbgY8hHxBSwcVsNAA5U0RBjT0hKOdQ3zrUdKVnBwMNzc3HD37l0AwOHDh2VeuKw0omSEyEXbtm1ha2sLHx+fPJ8fMWIEEhISuDtB/ljfwsICv/32G3777bcSiZeQiuDtlzS8+pSKg7dfIzE9G3ci4/Osp62qhPRsMRqb66JJNT10q2+C6gYaJRwtKUhqaiqWLl0Kb29viMViaGtrY9WqVRg7dizfockFJSPlxIgRI7Bv3z54eXlJ3Xzt5MmT6N27N35lbTuxWIw1a9Zg7969eP36NVRVVVGjRg2MHTsWY8aMKdQxNmzY8EsxEEJ+jjGGD4kZ+Cv4PU4+fIew2OQ867WuWRkO1fRQx1gTDc10oaOmVCZmXFRUly9fxqhRoxAVFQXg601QN2zYkOsO2WUZJSPliIqKClatWoVx48ZBV1dXbsddunQptm3bhs2bN6Nx48ZISkrCvXv38OVLwVP5vqetrS23eAgh//l2w7gd11/h/LPYXM+b6qjCvJIaXn9Ow7QONdDZxgiaKko8REqKKj09HVFRUTAzM8OWLVvQrVs3vkOSO0pGCik1NTXf5xQUFKSWsi6orlAohKqq6k/rqquryxyjo6MjIiIi4OXlhdWrV+db788//8TixYsREREBY2NjTJkyBTNnzsy3/qlTpzBx4kT079+fK2vQoEGBsZw5cwaDBw/Gli1bMGTIkFzdNISQoknJzMHjNwk4/eQD/IOiIcmjwdFSXx09bE0w0L4qjLRpmf2yhjGG8PBw1KpVCwDQpUsX7Nu3D3369IGGRvnsPqNkpJAKegN06dIFZ86c4bYNDAyQlpaWZ902bdrgypUr3LaFhQXi4uJy1StKl4aCggI8PT0xePBgTJ06FVWqVMlV5/79+xgwYACWLFkCFxcX3Lx5ExMnTkSlSpUwYsSIPI9rZGSES5cuYeLEiahcufJP4zh8+DDGjx+Pw4cPl8sMnpCSlpCWhdH77v10YTGXxmYY27oaqhvQDR/LqrCwMIwfPx4PHjxASEgITExMAADDhw/nObLiRclIOdO7d2/Y2trC3d0du3btyvW8t7c3OnTogEWLFgEAatasiefPn2PNmjX5JiPe3t7o168fjIyMUK9ePTRv3hw9e/ZE586dc9X19fXFggUL8Pfff6NNmzZyfW2EVCTZYgkuPo/F4lPP8Ck59yJj2qpKcKxjiG4NjNGquj7dMK6My8zMxKpVq7BixQpkZWVBTU0N9+/f55KR8o6SkUJKSUnJ9zkFBQWp7Y8fP+ZbVyiU/oPxbUCSPK1atQrt27fHrFmzcj0XEhKSaxpYixYt4OPjA7FYnOu1AEDdunXx9OlT3L9/Hzdu3MC1a9fQvXt3jBgxAjt37uTqHTt2DB8/fsSNGzdgb28v99dFSHmXI5bg0dtEXH/xCT4XX0g9p6qkAJsq2ljcrS6sTWkMVnly/fp1uLm5ITQ0FADQuXNnbNmyBRYWFvwGVoIoGSkkWcZwFFfdwmrdujWcnJwwb968fFs7ZCUUCmFvbw97e3v89ttvOHjwIIYNG4YFCxagWrVqAICGDRviwYMH2L17Nxo3bkyj8wn5iYxsMW69/Iy/H7/Hv89ikZEtRs4Pg0Ba1dBH6xqVMaplNSgI6XeqPGGMYdKkSdi6dSuAr138GzduxIABAyrc309KRsqplStXwtbWlhsA9U2dOnVw48YNqbIbN26gZs2aebaK5Kdu3boApAfgWllZYd26dWjbti0UFBSwefPmX3gFhJQ/2WIJ7kbFY/2FcDyMTsiVeACAsqIQmipKGNzEDL0bVUE1ffl/YSGlg0Ag4CY/jBkzBqtWrYKenh7PUfGDkpFyysbGBkOGDMHGjRulymfOnAl7e3ssW7YMLi4uuHXrFjZv3owtW7bke6x+/fqhRYsWaN68OYyMjBAZGYl58+ahZs2aqF27tlTdmjVr4vLly2jbti0UFRXzXQSNkIqAMYYTD9/hr+D3uBr+Kd969atoo5llJTS1qoTWNSpTC0g59vr1a2RnZ6N69eoAAA8PD/Tu3RutWrXiOTJ+UTJSjnl4eODIkSNSZY0aNcLRo0exePFiLFu2DMbGxvDw8CiwO8fJyQn/+9//4OXlhcTERBgZGaF9+/ZYsmQJFBVzv4Vq1aqFS5cucS0k69atk/dLI6TUYozhY3Imjt1/izXnw/KsY2+hi0ZVddHZxhg2ptqUfFQAOTk52LhxIxYtWoRGjRrh6tWrEAqF0NDQqPCJCAAIWAVbFjMpKQna2tpITEyElpaW1HMZGRmIjIxEtWrVpNYNIYQUj/LwO5eZI8a5pzF4GJ2A/beiIBAIIP6h+8Wysjr625mhfW0D1DDQgJCSjwrlwYMHGDt2LB48eADg67i+EydOlMkumYI+Q38FtYwQQoiMssUS/P3oPfbciMKTd4nST/7/9ztlRSFMdVXh1dsGDpaVeIiS8C0lJQXu7u7w8fGBRCKBjo4O1q5di5EjR+aaWVnRUTJCCCGFIJYwBEbEYeU/oQj5kJTreTM9VXSxNoatmQ5a16wMdWX681qRRUREoEOHDoiOjgYADBo0COvXr4ehId0BOS/020IIIQW4+TIO+25G4UbEZ6Rk5kg9N9ihKtrUrIz2tQ2gRIuOke+Ym5tDV1cXQqEQW7duhbOzM98hlWqUjBBCyHdyxBKcexYD38sv82wBqWOshf52VdCroSn01EU8REhKI4lEgsOHD6N///5QVlaGkpISjh8/DkNDw2JZT6q8oWQkDxVsTC8hvCktv2spmTk4/zQGx+6/xa1Xn3M9b2+hC2tTbSzoUoeWXSe5PH/+HG5ubrhx4wZevnwJd3d3AIClpSXPkZUdlIx8R0np622109LSpO6sSwgpHt9uKPntd6+k5IglePIuEYfuRONmRBzeJ2bkWW9cG0sMdTCHmZ5aicZHyoaMjAx4enpi5cqVyM7Ohrq6OvT19fkOq0yiZOQ7CgoK0NHR4e4to6amVuGW5CWkJDDGkJaWho8fP0JHR0em1X9/RUa2GGeffIDn2VDEpeS++dygJmZobK6H7g1MIFKkFhCSvytXrmDcuHEIDw8HAHTr1g2+vr6oWrUqz5GVTZSM/MDIyAhAwTe7I4TIh46ODvc7V5x2B0Zi1blQiCVMagl2G1NtiBSFcGlshn52VWj9D1IoGzduxLRp0wB8/czYtGkT+vbtS19efwEtepYPsViM7OzsEoyMkIpFSUmp2FtE3n5Jw/iD9/H03X8DUQUCwEhLBQdGN0F1A81iPT8pn16+fIn69etj+PDh8PLygo6ODt8hlRha9KyEKSgolFjTMSFEvh6/TcDRe29w8HY0V6asKITfUDu0rVWZvsESmbx69Qrnzp3DxIkTAXy9KejLly9LpFWvoqBkhBBSLrz+nIpxB+4jLiVLajyIukgB3i62cKpHHxxENtnZ2Vi/fj2WLFmC9PR01K9fHy1btgQASkTkjJIRQkiZlZUjwf5bUbj58jMuhUqP87LUV0dfuypwa21JC5IRmQUFBWHs2LF4/PgxAKBdu3a0emoxomSEEFLm3IiIg//dN/j70XupckWhADUNNbG2fwPUNZFffzapOJKTk7Fw4UJs2rQJjDHo6elh3bp1cHV1pe69YkTJCCGkTLjz6jO2X3uFgNDcM90aVtXBwq510KiqLn1gkCJjjKF169YIDg4GAAwdOhTe3t6oXLkyv4FVAJSMEEJKreSMbJwMfo9FJ5/mek5ZUQgXezPM6FgTOmq0LDv5dQKBANOnT8fSpUvh5+eHjh078h1ShUFTewkhpU5mjhg7r0dizfkwqXINZUU0qaaH+V3qoLqBBk/RkfJCIpHAz88Ppqam6NmzJ4CvrSMZGRm0Cnc+aGovIaTck0gY3A7cx8WQWKlyO3NdzO9SB3bmujxFRsqbp0+fws3NDbdu3YKxsTHatm0LbW1tCAQCSkR4QMkIIYR3z94nYvTee4hJkr5HzNCmVTGrUy3qhiFyk56ejuXLl2P16tXIycmBpqYmFixYAA0NamnjEyUjhBDePHufiPEH7+NNfLpUuVtrS8x2qkVTcolcBQQEYPz48YiIiAAA9O7dGxs3bkSVKlV4joxQMkIIKVFiCcO+m1HYfu1VrpaQZpaV4DukEfTUqSWEyNezZ8/g6OgIADA1NcXmzZvRq1cvfoMiHEpGCCElgjEGj9PPsedGlFR5ZU1lLOxaBz1tTfkJjFQI9erVg6urKzQ1NbFixQqawFDKUDJCCClWaVk5OHDrNbz+CZUqb1OzMvo3roJu9U14ioyUZxEREZg1axY2bdoEMzMzAMDu3bshFFLXX2lEyQghRK4kEoYXH1Nw4uE7+F19mWed2/M6wEhbpYQjIxVBVlYW1q5di2XLliEjIwMikQhHjx4FAEpESjHe/2d8fX1hYWEBFRUVODg4ICgoqMD6Pj4+qFWrFlRVVWFmZobp06cjIyOjwH0IISVj/YVwWM4/Cyefa7kSEYdqepjSvjqiVnalRIQUi1u3bsHOzg4LFixARkYGOnbsCC8vL77DIoXAa8vIkSNHMGPGDPj5+cHBwQE+Pj5wcnJCWFgYDAwMctU/fPgw5s6di927d6N58+YIDw/HiBEjIBAI4O3tzcMrIIQAQMTHZDh6X5Mqq22kifpVtOFsbYR2tQxomXZSbBITEzF//nxs3boVjDHo6+tj/fr1GDJkCL3vygheV2B1cHCAvb09Nm/eDODranhmZmaYMmUK5s6dm6v+5MmTERISgoCAAK5s5syZuHPnDgIDAwt1TlqBlRD5+JCYjv8FvcGZx+/x8lMqV26srYKzU1tBl2bEkBLi4eEBd3d3AMCIESOwZs0a6Ovr8xxV+VRcn6G8ddNkZWXh/v373FQr4Gt/nqOjI27dupXnPs2bN8f9+/e5rpxXr17h7Nmz6NKlS77nyczMRFJSktQPIeTXnHn8Ac28LmFjwAupRGRcG0vcmteBEhFS7L7/Hj1r1iw4OTkhICAAe/bsoUSkDOKtmyYuLg5isRiGhoZS5YaGhggNDc1zn8GDByMuLg4tW7YEYww5OTkYP3485s+fn+95vLy8sHTpUrnGTkhF9CU1C8tOP8eLjyl48i6RK29ioYd6ploY28oSJjq0jDYpXmKxGL6+vjh9+jT++ecfKCgoQE1NDefOneM7NPILytRsmitXrsDT0xNbtmyBg4MDIiIiMG3aNCxbtgyLFi3Kc5958+ZhxowZ3HZSUhI3zYsQ8nPvE9Ix4eB9PHqbKFXuWMcQ6/o3gLaaEk+RkYrm0aNHGDt2LO7evQsAOHbsGFxcXHiOisgDb8mIvr4+FBQUEBsrfUOs2NhYGBkZ5bnPokWLMGzYMIwZMwYAYGNjg9TUVLi5uWHBggV5TttSVlaGsrKy/F8AIeUYYwxv4tOx5O9nuBz2Ed+PLLM108HSHvXQwEyHt/hIxZKWloYlS5bA29sbYrEYWlpaWLVqFfr37893aEROeEtGRCIR7OzsEBAQwC3JK5FIEBAQgMmTJ+e5T1paWq6EQ0FBAYB0/yEhpGguh32E76UI3Hv9Raq8hoEGpnaogW71jWl2AilR58+fx4QJExAZGQkA6NevHzZs2AATE1osrzzhtZtmxowZcHV1RePGjdGkSRP4+PggNTUVI0eOBAAMHz4cpqam3Dzx7t27w9vbGw0bNuS6aRYtWoTu3btzSQkhRHYRH1Mw+fADhMYkS5Vbm2qhi40xJrSxoiSElDiJRIIFCxYgMjISZmZm8PX1Rffu3fkOixQDXpMRFxcXfPr0CYsXL0ZMTAxsbW1x7tw5blBrdHS0VEvIwoULIRAIsHDhQrx79w6VK1dG9+7dsWLFCr5eAiFlVo5YgvPPYrH+YjgiPqZIPefZ2wZdbIygo0azYkjJ+jY5QUlJCUKhEDt27MD+/fuxbNkyaGho8B0eKSa8rjPCB1pnhFR018I/YdKhB0jPFiNHIv3rv7Z/A/SyNYGiAu+LM5MKKDw8HOPGjUOLFi2wfPlyvsMheSiuz1BKRgipIDKyxWiz5jJikzK5MpGiEFV0VTGtQw30aGBCXTGEF1lZWVi1ahVWrFiBzMxMaGtr4/Xr19DW1uY7NPKD4voMLVNTewkhRXPkbjTm/PlEqmxGx5qY1K46FISUgBD+BAYGws3NDSEhIQAAZ2dnbNmyhRKRCoaSEULKsTOPP+QaEzK8mTk8elrzGBUhQEJCAubMmYPt27cDAAwMDODj44OBAwdSC10FRMkIIeVQTGIGmnoF5CoPmt8BBlp0x1zCv4SEBBw8eBAAMGbMGKxatQp6eno8R0X4QskIIeXM2ScfMPHQA25bTaSAbcPs0LK6Pn3jJLz68uULdHV1AQAWFhbw8/ODubk5WrduzXNkhG+UjBBSTnxMykD/bbfw+nMaVzbbqRYmtavOY1SEADk5Odi0aRMWL16Mv//+G23btgUADBs2jN/ASKlByQghZRxjDIeDorHgxFOp8lvz2sNYm25cR/j18OFDjB07Fvfv3wcA7Nu3j0tGCPmGkhFCyrCH0V/Qe8tNqbI5zrUxvo0ldckQXqWmpsLd3R3r16+HRCKBjo4OVq9ejdGjR/MdGimFKBkhpIya++dj+N99w233aWiK351rw0ibBqgSfl24cAFjx47F69evAXxdbdvHxyffm6ASQskIIWUMYwy9ttzEozcJXNmekfZoV8uAv6AI+c6nT5/w+vVrmJubY+vWrejcuTPfIZFSjpIRQsqQB9Ff0Oe7bhkNZUXcW+gIFSW6USThj0QiQVRUFCwtLQEAgwYNQnJyMoYOHQp1dXWeoyNlASUjhJQB557GYPzB+1JlpjqqCJjZhhIRwquQkBC4ubkhPDwcoaGh0NXVhUAgwLhx4/gOjZQhlIwQUkp9Ss7ErsBI3H71GcHfdclU1VPD/C514GxN/e+EPxkZGfDy8oKXlxeys7Ohrq6Oe/fuoWPHjnyHRsqgX0pGMjIyoKJCg+UIkaeH0V+w9t8w3Ij4LFWuKBRgcfe6GN7Mgp/ACPl/V69exbhx4xAWFgYA6Nq1K3x9fWFubs5zZKSskjkZkUgkWLFiBfz8/BAbG4vw8HBYWlpi0aJFsLCwoGlbhBTRm/g0tFp9OVd5o6o6WNHbBnWM6S7ThF8SiQTjxo3Dzp07AQCGhobYtGkT+vXrR1PJyS8RyrrD8uXLsXfvXqxevRoikYgrt7a25t6ghJDCuxkRh+6bAnMlIhsG2uKVZxccn9iCEhFSKgiFQojFYgDAuHHjEBoaiv79+1MiQn6ZgDHGZNmhevXq2LZtGzp06ABNTU08evQIlpaWCA0NRbNmzfDly5fiilUukpKSoK2tjcTERGhp0R94wp/QmCSM2XcPb7+kS5WPaVkNC7vV5SkqQqRFRkZCSUkJVapUAQB8/vwZISEhaNmyJc+RET4U12eozN007969Q/Xque91IZFIkJ2dLZegCCnvroV/wvDdQVJl+0c1QeualXmKiBBpOTk5WL9+Pdzd3dG+fXv8/fffEAgEqFSpEiUiRO5kTkbq1q2L69ev5xqodOzYMTRs2FBugRFSHt2Likc/v1tSZfM618boltWgqCBzrykhxeLu3btwc3NDcHAwgK9Lu6ekpEBTU5PfwEi5JXMysnjxYri6uuLdu3eQSCQ4fvw4wsLCsH//fpw+fbo4YiSkXFh08ikO3H4tVXb+t9aoZUR/4EnpkJycjIULF2Lz5s2QSCTQ09PD2rVrMWLECBoXQoqVzGNGAOD69evw8PDAo0ePkJKSgkaNGmHx4sXo1KlTccQoVzRmhJS0rBwJVp8Lxc7ASK5sRW9rDGhsBiVqDSGlxLNnz+Ds7Iy3b98CAIYMGQJvb28YGNBtBsh/iusztEjJSFlGyQgpKamZOVhw4glOBr+XKn/k3gnaqko8RUVI3tLT01G/fn2IxWL4+fmViS+XpOQV12eozF/LLC0t8fnz51zlCQkJ3H0JCKnIssUS/Ob/EPXcz0slIjUMNBDi4UyJCCkVJBIJ/P39uam6qqqqOH36NJ4+fUqJCClxMo8ZiYqK4t6838vMzMS7d+/kEhQhZRVjDJ03XEfExxSurEk1PWwYaAtjbVUeIyPkP0+fPsW4ceNw8+ZNxMbGYtq0aQCAWrVq8RwZqagKnYycOnWKe3z+/Hloa2tz22KxGAEBAbCwsJBrcISUJUGR8Riw7b+ZMrZmOtg70h46aqIC9iKk5GRkZGD58uVYtWoVcnJyoKGhATU1Nb7DIqTwyUivXr0AAAKBAK6urlLPKSkpwcLCAuvWrZNrcISUdtliCfbfeo1lp59LlY9oboElPerxFBUhuV26dAnjxo1DREQEAKBnz57YvHkzt5gZIXwqdDIikUgAANWqVcPdu3ehr69fbEERUtpJJAzOG64hPDYl13O+gxuha31jHqIiJG+rVq3C3LlzAQAmJibYvHkzevfuzXNUhPxH5gGskZGRlIiQCu1LahZ6+t6QSkSq6qlhumNNRK3sSokIKXU6d+4MkUiESZMm4fnz55SIkFJH5gGswNfV+K5evYro6GhkZWVJPTd16lS5BEZIaXT07hv8/udjbtvGVBv+bk2hrlykXyVCisXLly9x7do1jBw5EgBQv359vHr1CqampjxHRkjeZP4L+vDhQ3Tp0gVpaWlITU2Fnp4e4uLioKamBgMDA0pGSLkUm5QBtwP38ehNAle2YaAtetrSH3dSemRnZ2Pt2rXw8PBAdnY2GjVqhAYNGgAAJSKkVJO5m2b69Ono3r07vnz5AlVVVdy+fRuvX7+GnZ0d1q5dWxwxEsKryLhUOHgGcImInroIZ6a2pESElCq3b9+GnZ0d5s+fj4yMDLRr147uJUPKDJmTkeDgYMycORNCoRAKCgrIzMyEmZkZVq9ejfnz5xdHjITw5kZEHNqtvcJtuzYzx4NFHVHPRDv/nQgpQUlJSZg8eTKaN2+OJ0+eQF9fHwcOHMC///5LC1GSMkPmbholJSUIhV9zGAMDA0RHR6NOnTrQ1tbGmzdv5B4gIXw5++QDJh56wG179KyH4c0s+AuIkB+IxWI4ODggNDQUAODq6oq1a9fSJANS5sicjDRs2BB3795FjRo10KZNGyxevBhxcXE4cOAArK2tiyNGQkpUbFIGxh+8j4fRCQAANZECjo5rBmtTag0hpYuCggImTJiATZs2wc/PDx06dOA7JEKKROYb5d27dw/Jyclo164dPn78iOHDh+PmzZuoUaMGdu3aBVtb22IKVT7oRnmkIAdvv8bCk0+57UZVdXBwjAPURDRbhvBPLBZjy5YtqFu3Lpd4iMViZGVlQVWVbjdAih/dtVdOKBkheYmMS5UaGwIA87vUhltrK34CIuQHjx8/xtixYxEUFARLS0s8ffqUEhBS4krNXXvz8+DBA3Tr1k1ehyOkxNx59VkqETHTU0XoMmdKREipkJaWhrlz56JRo0YICgqClpYWZs2aBWVlZb5DI0RuZEpGzp8/j1mzZmH+/Pl49eoVACA0NBS9evWCvb09t2Q8IWWF94VwuGy/zW3P6FgT139vDxUlBR6jIuSrf//9F9bW1li1ahXEYjH69u2LkJAQTJgwgZtIQEh5UOiO8F27dmHs2LHQ09PDly9fsHPnTnh7e2PKlClwcXHB06dPUadOneKMlRC58rv6EhsDXnDbO4c3hmNdQx4jIuQ/d+7cgZOTEwCgSpUq8PX1RY8ePXiOipDiUegxI/Xr18ewYcMwe/Zs/Pnnn+jfvz+aNm2Ko0ePlqm7PtKYEZIjlsB1TxBuRHzmyq7MagsLfXUeoyJEGmMMffr0gbm5OZYtW0YLmJFSgfcBrOrq6nj27BksLCzAGIOysjIuX76MFi1ayC2YkkDJSMWWLZbA0fsqXn9OAwDUM9HC35NbQigU8BwZqejCw8Mxb948bNu2jVsnRCwWQ0GBugxJ6VFcn6GF7qZJT0+HmpoaAEAgEEBZWRnGxnR3UlJ2iCUMfbbc5BIRt9aWmOtcmxIRwqusrCysWrUKK1asQGZmJvT09LBjxw4AoESEVBgyLZ6wc+dOaGhoAABycnKwd+/eXCv90Y3ySGkkkTBYzT/LbS/uVhejWlbjMSJCgMDAQLi5uSEkJAQA4OTkhHnz5vEcFSElr9DdNBYWFhAICv4GKRAIuFk2pRV101Q8jDG0Wn0Zb7+kAwB62ppgw8CGPEdFKrKEhATMnTsX27ZtAwBUrlwZGzZswMCBA3/6d5YQPvHeTRMVFSW3kxJSUp68TURP30BI/j/lrmeiRYkI4d3y5cu5RGT06NFYvXo19PT0eI6KEP7QGtekXMrIFqPDuqt4l5DOlZlXUsPpKS15jIpUZIwxrtVj4cKFuHfvHpYuXYo2bdrwHBkh/KNkhJQ7H5My0MQzQKrs0BgHtKhOdzIlJU8sFmPTpk0IDAzEH3/8AYFAAB0dHVy5coXv0AgpNSgZIeXK03eJ6LYpkNtuUk0PR9yaUj884cXDhw/h5uaGe/fuAQD++ecfdOnSheeoCCl9eF9P2NfXFxYWFlBRUYGDgwOCgoIKrJ+QkIBJkybB2NgYysrKqFmzJs6ePVvgPqRiiE3KkEpEhjatSokI4UVqaipmzZoFe3t73Lt3D9ra2ti+fTucnZ35Do2QUonXlpEjR45gxowZ8PPzg4ODA3x8fODk5ISwsDAYGBjkqp+VlYWOHTvCwMAAx44dg6mpKV6/fg0dHZ2SD56UKoEv4jB01x1ue0VvawxxMOcxIlJRnT17FhMnTsTr168BAC4uLvDx8YGRkRHPkRFSehV6au/3Xr58iT179uDly5fYsGEDDAwM8M8//6Bq1aqoV69eoY/j4OAAe3t7bN68GQAgkUhgZmaGKVOmYO7cubnq+/n5Yc2aNQgNDYWSkpKsYQOgqb3l0V/B7zDNP5jbPjC6CVrVqMxfQKTCysnJgbW1NcLCwlC1alVs2bIFXbt25TssQuSmuD5DZe6muXr1KmxsbHDnzh0cP34cKSkpAIBHjx7B3d290MfJysrC/fv34ejo+F8wQiEcHR1x69atPPc5deoUmjVrhkmTJsHQ0BDW1tbw9PSEWCzO9zyZmZlISkqS+iHlA2MM9Zecl0pEdrk2pkSElCiJRML9DVJUVMS2bdswY8YMPHv2jBIRQgpJ5mRk7ty5WL58OS5cuACRSMSVt2/fHrdv3y5gT2lxcXEQi8UwNJS+S6qhoSFiYmLy3OfVq1c4duwYxGIxzp49i0WLFmHdunVYvnx5vufx8vKCtrY292NmZlboGEnpdf91PKrNO4ukjBwAgL6GMm7ObY8Odeiuu6TkhIaGol27dli/fj1X1qZNG6xbt45brZoQ8nMyJyNPnjxB7969c5UbGBggLi5OLkHlRyKRwMDAANu3b4ednR1cXFywYMEC+Pn55bvPvHnzkJiYyP28efOmWGMkxe/+6y/ou/W/1rO6xlq4M78DTHRUeYyKVCSZmZlYsmQJGjRogGvXrmH16tVIS0vjOyxCyiyZB7Dq6Ojgw4cPqFZN+r4eDx8+hKmpaaGPo6+vDwUFBcTGxkqVx8bG5jvQy9jYGEpKSlI3j6pTpw5iYmKQlZUl1VLzjbKyMpSVlQsdFym9UjJzYO1+XqrszNSWqGeizVNEpCK6du0axo0bh9DQUABAly5d4Ovry91IlBAiO5lbRgYOHIg5c+YgJiYGAoEAEokEN27cwKxZszB8+PBCH0ckEsHOzg4BAf8tTiWRSBAQEIBmzZrluU+LFi0QEREBiUTClYWHh8PY2DjPRISUH55nQ3IlIntH2lMiQkrMly9fMHbsWLRp0wahoaEwNDTEkSNHcPr0aVhYWPAdHiFlmszJiKenJ2rXrg0zMzOkpKSgbt26aN26NZo3b46FCxfKdKwZM2Zgx44d2LdvH0JCQjBhwgSkpqZi5MiRAIDhw4dL3cFywoQJiI+Px7Rp0xAeHo4zZ87A09MTkyZNkvVlkDIiNTMH8088wfZr/92AsW+jKnjl2QVta+We/k1Icfnw4QP27dsHANyddgcMGEDr2BAiBzJ304hEIuzYsQOLFi3C06dPkZKSgoYNG6JGjRoyn9zFxQWfPn3C4sWLERMTA1tbW5w7d44b1BodHQ2h8L98yczMDOfPn8f06dNRv359mJqaYtq0aZgzZ47M5yalX2J6Nhos/Zfb1lFTwqExDtQaQkpMYmIitLW/vt/q1q2LDRs2wNraGq1ateI5MkLKF5nXGQkMDETLlmX3ZmO0zkjZsO9mFNxPPeO2hzhUxYreNjxGRCqSnJwc+Pj4wMPDA5cvX4adnR3fIRFSKpSadUbat2+PatWqYf78+Xj+/LncAiEE+Lp2SJ8tN6QSkcGUiJASdO/ePdjb22P27NlITk7Gnj17+A6JkHJP5mTk/fv3mDlzJq5evQpra2vY2tpizZo1ePv2bXHERyqQK2EfUW3eWTyITgAAmOmp4uGijvCkRISUgJSUFEyfPh0ODg4IDg6Grq4udu3ahU2bNvEdGiHlXpGWg/8mMjIShw8fxv/+9z+EhoaidevWuHTpkjzjkzvqpimdQj4kofOG69y2rZkOjk9oDqGQBgeS4vfPP/9g3Lhx3DpEgwcPxvr16/O8RxYhFVlxfYb+0o3yqlWrhrlz56JBgwZYtGgRrl69Kq+4SAVy+9VnDNz+3+q9u0c0RvvatJIqKTlRUVF48+YNqlWrhq1bt8LJyYnvkAipUIqcjNy4cQOHDh3CsWPHkJGRgZ49e8LLy0uesZFyjjEGn4svsCHgBVdGi5iRkiCRSPD27VtUrVoVADBu3Djk5ORg9OjRtHgZITyQuZtm3rx58Pf3x/v379GxY0cMGTIEPXv2LDO/wNRNU3pMPvwApx9/4Lb93ZqiqWUlHiMiFcGzZ8/g5uaGd+/e4dmzZ1BXV+c7JELKjFLTTXPt2jXMnj0bAwYMgL6+vtwCIRXLv89iuETEvJIaAma0gaKCzOOpCSm0jIwMrFixAqtWrUJ2djY0NDTw8OHDMr1UASHlhczJyI0bN4ojDlKB3Hn1GW4H7nPbl2a2hQINVCXF6PLlyxg3bhxevPjaJdijRw9s3ryZ7uJNSClRqGTk1KlT6Ny5M5SUlHDq1KkC6/bo0UMugZHy6fqLTxi2K4jbPju1FSUipNhkZ2dj3Lhx3FohxsbG2Lx5M3r37k3LuBNSihQqGenVqxdiYmJgYGCAXr165VtPIBBALBbLKzZSjjDG0HzlJXxIzODKNg1qiLomNG6HFB8lJSUkJCRAIBBgwoQJ8PT05JZ3J4SUHr+0zkhZRANYS17gizh4nH6G8NgUruzyrLaopk8DB4n8vXz5EpqamtwaIe/evUN0dHS+dwMnhBReqVkOfv/+/cjMzMxVnpWVhf3798slKFJ+hHxIwtBdd7hEpGFVHbxY0ZkSESJ32dnZWLVqFaytrTF9+nSu3NTUlBIRQko5mZORkSNHIjExMVd5cnIyRo4cKZegSPnAGMPc40+47Ysz2uDExBZQolkzRM7u3LmDxo0bY+7cucjIyEBsbCwyMjJ+viMhpFSQ+VOBMZbnwK+3b99SXyzhMMbQbu0VPHqTAABY2ccG1Q00+A2KlDtJSUmYMmUKmjVrhsePH6NSpUrYt28fLly4ABUVFb7DI4QUUqGn9jZs2BACgQACgQAdOnSAouJ/u4rFYkRGRsLZ2blYgiRlz9K/nyPqcxoAoIquKlzsaQolka+HDx+ie/fuePfuHQBg+PDhWLduHa1/REgZVOhk5NssmuDgYDg5OUFD479vuSKRCBYWFujbt6/cAyRlz+pzodh7MwoAoKwoROCc9vwGRMolS0tLMMZQvXp1+Pn5oUOHDnyHRAgpokInI+7u7gAACwsLuLi4UBMoydP2ay+x5cpLbvveQkceoyHliVgsxokTJ9C3b18IBAJoa2vj3LlzqF69OlRVVfkOjxDyC2QeM+Lq6kqJCMlT902B8DwbCuBr18xzDydoqijxHBUpDx4/fowWLVqgf//+2LdvH1duY2NDiQgh5UChWkb09PQQHh4OfX196OrqFrhyYXx8vNyCI2VHbFIGnrz7b5bV5VltadYM+WVpaWnw8PDAunXrkJOTQ2sDEVJOFSoZWb9+PTQ1NbnHtIwy+V7IhyR03nCd2w5f3pkSEfLLLly4gPHjx+PVq1cAgD59+mDjxo0wNTXlOTJCiLzRCqzkl1wO+4iRe+5y21uGNEIXG2MeIyLlgbu7Ozw8PAB8XbTM19cXPXv25DkqQkipWYH1wYMHePLkv4Ws/vrrL/Tq1Qvz589HVlaW3AIjpd/tV5+5RERDWRH7RzWhRITIhZOTExQUFDB16lSEhIRQIkJIOSdzMjJu3DiEh4cDAF69egUXFxeoqanhjz/+wO+//y73AEnpk5UjQd3F5zBw+22u7O8pLdG6ZmUeoyJlWXh4OI4cOcJtN2/eHK9evcKGDRu4LmJCSPklczISHh4OW1tbAMAff/yBNm3a4PDhw9i7dy/+/PNPecdHSpnw2GTUXPgP0rK+3p1ZQShA0PwOdK8ZUiRZWVlYvnw56tevjxEjRiAiIoJ7rmrVqjxGRggpSYVeZ+QbxhgkEgkA4OLFi+jWrRsAwMzMDHFxcfKNjpQqoTFJcPb5b6Bq+9oG2D3CnseISFl248YNuLm54fnz5wCATp06QUmJpoITUhHJnIw0btwYy5cvh6OjI65evYqtW7cCACIjI2FoaCj3AEnpwBhDny03ue09I+3RrpYBjxGRsiohIQHz5s2Dn58fAKBy5crw8fHBoEGDaKYeIRWUzMmIj48PhgwZgpMnT2LBggWoXr06AODYsWNo3ry53AMk/Lv/Oh59t97itn93rkWJCCmSzMxMNGzYEFFRUQCAUaNGYc2aNdDT0+M3MEIIr+Q2tTcjIwMKCgqlvpmVpvbK5kH0F6kWEed6RvAbZsdjRKSsW7ZsGQ4ePIht27ahbdu2fIdDCJFBcX2GFjkZuX//PkJCQgAAdevWRaNGjeQWVHGiZKTwvqRmoeGyC9z2LtfG6FCHuuJI4YnFYvj6+qJp06Zo0qQJgK+DViUSCd1WgpAyqLg+Q2Xupvn48SNcXFxw9epV6OjoAPjaB9yuXTv4+/ujcmWa3lkeBEXGY8C2/7pmNgy0pUSEyCQ4OBhubm64e/cu6tevj3v37kFJSQkikYjv0AghpYzMU3unTJmClJQUPHv2DPHx8YiPj8fTp0+RlJSEqVOnFkeMpIRdDv0olYis6VcfPW1pCW5SOKmpqZg9ezYaN26Mu3fvQltbGxMnToSCggLfoRFCSimZu2m0tbVx8eJF2NtLT+kMCgpCp06dkJCQIM/45I66aQoWGZeKdmuvcNsbBtpSIkIK7dy5c5gwYQI3QHXAgAHw8fGBsTGtzEtIeVBqumkkEkmeg1SVlJS49UdI2fT0XSK6bQrktq/ObgvzSrSYGSmcgIAAdO7cGcDXBcu2bNmCrl278hwVIaQskLmbpn379pg2bRrev3/Plb179w7Tp09Hhw4d5BocKTmfkjOlEpFtw+woESEyadeuHdq1a4cZM2bg2bNnlIgQQgpN5mRk8+bNSEpKgoWFBaysrGBlZYVq1aohKSkJmzZtKo4YSTFLTM+G/YqL3PbCrnXgVM+Ix4hIWRAWFoZBgwYhOTkZACAUCvHvv/9i3bp10NDQ4Dk6QkhZInM3jZmZGR48eICAgABuam+dOnXg6Ogo9+BI8ftx+u6I5hYY08qSx4hIaZeZmYmVK1fC09MTWVlZMDU1xdq1awEAiooy/0khhBDZkpEjR47g1KlTyMrKQocOHTBlypTiiouUkJF773KPl/WyxrCm5jxGQ0q769evw83NDaGhoQCAzp07Y/LkyTxHRQgp6wrdTbN161YMGjQI9+7dw4sXLzBp0iTMnj27OGMjxex/QdEIfpMAAJjWoQYlIiRfX758wdixY9G6dWuEhobC0NAQ/v7+OHPmDCwsLPgOjxBSxhU6Gdm8eTPc3d0RFhaG4OBg7Nu3D1u2bCnO2EgxEUsY5v75GPOOP+HKpnesyWNEpLSbO3cudu7cCQAYO3YsQkJC4OLiQje2I4TIRaHXGVFVVUVISAj3LUgikUBVVRVRUVFlag0BWmcE6LrxOp69TwIAKAgF2DvSHq1q0Mq5JH8xMTHo06cPVq1ahVatWvEdDiGEJ7yvM5KZmQl19f+megqFQohEIqSnp8stGFL8fC6Gc4mIqY4qbsxtz3NEpLTJycnBhg0b8OzZM+zevRsAYGRkhJs3b/5kT0IIKRqZBrAuWrQIampq3HZWVhZWrFgBbW1trszb21t+0RG5uv/6C3wuvuC2A2a24TEaUhrdv38fY8eOxcOHDwEAI0aMQOvWrXmOihBS3hU6GWndujXCwsKkypo3b45Xr15x29R/XHpJJAx9t/73zfbG3PZQUaJ7hZCvUlJSsHjxYmzYsAESiQS6urpYs2YNWrZsyXdohJAKoNDJyJUrV4oxDFLcBu+8zT0+/1trmOqo8hgNKU3OnDmDiRMnIjo6GgAwaNAgrF+/HoaGdJdmQkjJoBWKyrmsHAnarb2Cdwlfx/a0rK6PWkaaPEdFSouMjAxMmDABb968gYWFBbZu3QpnZ2e+wyKEVDCUjJRjyRnZsFnyL7dtqa+OA6Ob8BgRKQ2+3dBSKBRCRUUFW7ZswdWrV7FkyRKpQeqEEFJSZL43DSkbUjJzpBKRQU2q4tKstjSup4J79uwZWrVqxc2SAYBu3bphzZo1lIgQQnhTKpIRX19fWFhYQEVFBQ4ODggKCirUfv7+/hAIBOjVq1fxBlgG9dlyg3vconolePWx4TEawreMjAwsWrQIDRs2xM2bN+Hh4YHs7Gy+wyKEEAClIBk5cuQIZsyYAXd3dzx48AANGjSAk5MTPn78WOB+UVFRmDVrFi3AlIfIuFSEx6YAAHrZmuDQmKY8R0T4dPnyZdSvXx/Lly9HdnY2unfvjsDAQCgpKfEdGiGEAChiMnL9+nUMHToUzZo1w7t37wAABw4cQGBgoMzH8vb2xtixYzFy5EjUrVsXfn5+UFNTk2pG/pFYLMaQIUOwdOlSWFrSHWZ/NPnwA+7xehdb/gIhvPr8+TNGjRqF9u3b48WLFzA2NsaxY8fw119/oWrVqnyHRwghHJmTkT///BNOTk5QVVXFw4cPkZmZCQBITEyEp6enTMfKysrC/fv34ejo+F9AQiEcHR1x69atfPfz8PCAgYEBRo8e/dNzZGZmIikpSeqnPDv+4C23wur8LrVpjEgF9vLlS+zduxcCgQATJkxASEgI+vbtS+8JQkipI3Mysnz5cvj5+WHHjh1SzbwtWrTAgwcPCtgzt7i4OIjF4lzrGRgaGiImJibPfQIDA7Fr1y7s2LGjUOfw8vKCtrY292NmZiZTjGVJRrYYM44+AgA0s6yEsa2o1aiiSUlJ4R43adIEq1evRmBgILZs2SK1UjIhhJQmMicjYWFheS4Pra2tjYSEBHnElK/k5GQMGzYMO3bsgL6+fqH2mTdvHhITE7mfN2/eFGuMfHn7JQ21F53jtrcPt6NvwBVIdnY2Vq1aBTMzM6mVkmfNmoXmzZvzGBkhhPyczOuMGBkZISIigrt77zeBgYEyj9/Q19eHgoICYmNjpcpjY2NhZGSUq/7Lly8RFRWF7t27c2Xf1kxQVFREWFgYrKyspPZRVlaGsrKyTHGVRS1XXeYeT2hrBU0VGpxYUQQFBWHs2LF4/PgxAGD37t1YtWoVz1ERQkjhydwyMnbsWEybNg137tyBQCDA+/fvcejQIcyaNQsTJkyQ6VgikQh2dnYICAjgyiQSCQICAtCsWbNc9WvXro0nT54gODiY++nRowfatWuH4ODgct0FU5BHbxK4xz0amGCOc23+giElJjk5GVOnTkXTpk3x+PFjVKpUCfv27cPKlSv5Do0QQmQic8vI3LlzIZFI0KFDB6SlpaF169ZQVlbGrFmzMGXKFJkDmDFjBlxdXdG4cWM0adIEPj4+SE1NxciRIwEAw4cPh6mpKby8vKCiogJra2up/XV0dAAgV3lFkZKZg6E77wAAVJUUsGGgLb8BkRJx+vRpTJgwAW/fvgUADBs2DOvWrUPlypV5jowQQmQnczIiEAiwYMECzJ49GxEREUhJSUHdunWhoaFRpABcXFzw6dMnLF68GDExMbC1tcW5c+e4Qa3R0dEQCnlfDqVUev05Fb233ERyZg4A4K/JLWicSAXx5MkTvH37FlZWVvDz85OakUYIIWWNgDHG+A6iJCUlJUFbWxuJiYnQ0tLiO5xfYjH3DPfYe0AD9GlUhcdoSHESi8WIjY2FiYkJgK8DVjdt2oQJEyZAVZXuwEwIKRnF9Rkqc8tIu3btCvz2fenSpV8KiBSOo/dV7vHJSS1ga6bDXzCkWD158gRubm5ISkrCw4cPIRKJoKSkhBkzZvAdGiGEyIXMyYitra3UdnZ2NoKDg/H06VO4urrKKy5SgJefUhDx8et6EgIBKBEpp9LT07Fs2TKsWbMGOTk50NTUxKNHj2Bvb893aIQQIlcyJyPr16/Ps3zJkiVSCy6R4jPp0H+Ly71c0YXHSEhxuXjxIsaPH4+XL18CAHr37o1NmzbB1NSU58gIIUT+5DYydOjQoQXeT4bIx/3X8QiNSQYArOlXH0IhDVgtT9LT0zF8+HB07NgRL1++hKmpKU6cOIHjx49TIkIIKbfklozcunULKioq8jocyUffrf/ds6d/44q5rkp5pqKignfv3kEgEGDKlCl4/vw5evXqxXdYhBBSrGTupunTp4/UNmMMHz58wL1797Bo0SK5BUZy+83/Ifd4x/DGPEZC5OnFixcwMDCAtrY2BAIBtm/fjri4ODg4OPAdGiGElAiZW0a+v+mctrY29PT00LZtW5w9exbu7u7FESPB15vgnQx+DwAQCoCOdQ1/sgcp7bKysrBixQrY2Nhg/vz5XLmVlRUlIoSQCkWmlhGxWIyRI0fCxsYGurq6xRUT+UFCWhZsPS5w20ELaIGrsu7mzZtwc3PDs2fPAHy971JOTg4UFWVurCSEkDJPppYRBQUFdOrUqdjvzkukefz9nHv8m2MN6GuU/xv/lVeJiYmYOHEiWrZsiWfPnkFfXx8HDx7EP//8Q4kIIaTCkvmvn7W1NV69eoVq1aoVRzzkBxEfU3D84TsAgI2pNn5zrMlzRKSobt++jT59+uDDhw8AgJEjR2LNmjWoVKkSz5ERQgi/ZB4zsnz5csyaNQunT5/Ghw8fkJSUJPVD5CfkQ5LUSqu+gxvxGA35VZaWlsjIyECNGjVw6dIl7N69mxIRQgiBDPem8fDwwMyZM6Gpqfnfzt8tC88Yg0AggFgsln+UclSW7k3T3CsA7xMzAAC7XBujQx0atFqWiMVinDlzBj169ODK7t+/j3r16tE0eEJImVRcn6GFTkYUFBTw4cMHhISEFFivTZs2cgmsuJSVZOT+6y/ou/UmAGBZz3oY1syC34CITIKDg+Hm5oa7d+/ixIkTtFYIIaRc4P1Ged9yltKebJQHaVk5XCJirK1CiUgZkpqaiqVLl8Lb2xtisRja2tpIS0vjOyxCCCnVZBrAWtDdeon8HLj1mnvs1ceGx0iILM6fP4/x48cjKioKADBgwAD4+PjA2NiY38AIIaSUkykZqVmz5k8Tkvj4+F8KqKJ7E58Gr39CAQAtqldC21oGPEdECmPOnDlYvXo1AMDMzAxbtmxBt27deI6KEELKBpmSkaVLl0JbW7u4YqnwcsQStFp9mdue61yHx2iILNq1a4e1a9di2rRp8PDwgIaGBt8hEUJImSFTMjJw4EAYGNA39eKy4ux/g4P3j2oCmyqU+JVWYWFhCA8PR/fu3QEAzs7OePHiBSwtLXmOjBBCyp5CrzNC40WK354bUQAAMz1VtK5Zmd9gSJ4yMzPh4eGB+vXrY+jQoXj//j33HCUihBBSNDLPpiHFI/rzfzMu5jjX5jESkp/AwEC4ublx09s7dOgAiUTCc1SEEFL2FbplRCKRUBdNMdp46QX3uIs1zb4oTb58+QI3Nze0atUKISEhMDAwgL+/P86cOYMqVarwHR4hhJR5dGeuUiA+NQvH7r8FADhU04NQSF1ipUVqaipsbGzw7t3X+wONHTsWq1atortWE0KIHMl8bxoif1uvRHCPd7g25jES8iN1dXW4uLigdu3auHbtGrZv306JCCGEyBklIzxLzxJjx/VIAMCoFtWgpaLEc0QVW05ODry9vfH8+XOubNmyZQgODkarVq14jIwQQsovSkZ4dujOf6utTmlfncdIyIMHD+Dg4ICZM2fCzc2NG5yqpqYGZWVlnqMjhJDyi5IRHqVnibH8zNeZGdX01aGrLuI5ooopJSUFM2fOhL29PR48eAAdHR2MGDGC77AIIaTCoAGsPOqy8Tr3eHW/+jxGUnGdOXMGEydORHR0NICvC/utX78eRkZGPEdGCCEVByUjPNl+7SUi41IBAEMcqsLeQo/niCqeU6dOoWfPngAAc3NzbN26FZ07d+Y5KkIIqXgoGeEBYwyeZ0O5bY+e1jxGU3F16dIFjRs3Rps2bbB06VKoq6vzHRIhhFRINGaEB7OPPeYe/zu9NRRoXZES8fz5c7i6uiIzMxMAoKioiJs3b2Lt2rWUiBBCCI8oGSlh0Z/TuAXObEy1UdNQk+eIyr+MjAwsXrwYtra22L9/P1atWsU9p6REU6kJIYRv1E1Twsbuv8c9PjTWgcdIKoYrV65g3LhxCA8PBwB069aNZsoQQkgpQy0jJSgmMQNhsckAAOd6RrTAWTGKj4/H6NGj0a5dO4SHh8PIyAh//PEHTp06hapVq/IdHiGEkO9QMlKCpvo/5B6voqm8xWrSpEnYvXs3AGD8+PEICQlBv379IBDQ+BxCCCltqJumhFwKjUVQZDwAYFVfG2irUqtIcVqxYgUiIiLg4+ODFi1a8B0OIYSQAlAyUkJ+/24GTX87Mx4jKX+ys7Ph4+ODd+/ewcfHBwBgaWmJoKAgagkhhJAygJKRErDt6kvEpWQBAA6PcYCQpvLKzd27dzF27Fg8evQIAODq6oqGDRsCACUihBBSRtCYkWLGGMO6C19nctiZ66J5dX2eIyofkpOTMW3aNDg4OODRo0fQ09PDnj17YGtry3dohBBCZEQtI8XMdc9dZOV8vfvr9mF2PEdTPpw6dQqTJk3C27df12sZOnQovL29UblyZZ4jI4QQUhSUjBSjh9FfcC38EwCgpqEGKmnQbeh/VXJyMkaPHo24uDhYWlrCz88PHTt25DssQgghv4CSkWIiljD03nKT2z44hhY4KyqJRAKBQACBQABNTU1s2LABT548waJFi6CmpsZ3eIQQQn4RjRkpJntvRnGPd7k2hoGmCn/BlGFPnz5Fy5Yt8eeff3JlgwcPhpeXFyUihBBSTlAyUkyWnX4OABAKgA51DHmOpuxJT0/HggUL0LBhQ9y6dQsLFiyARCLhOyxCCCHFgJKRYnD9xSfusfcAW/4CKaMCAgJQv359eHp6IicnB71790ZAQACEQnq7EkJIeUR/3eUsLSsHI/bc5bZ7NTTlMZqyJS4uDq6urnB0dERERARMTU1x4sQJHD9+HFWqVOE7PEIIIcWEkhE5a7TsAsQSBgD4e3JLnqMpW4KDg7F//34IBAJMnjwZz58/R69evfgOixBCSDGj2TRy5Hs5AhnZX8c1TGhrBZsq2jxHVPqlpaVxA1EdHR2xZMkSODs7w8GBZh8RQkhFQS0jcsIYw5rzYQC+rikyx7k2zxGVbllZWfD09ISFhQXevHnDlbu7u1MiQgghFUypSEZ8fX1hYWEBFRUVODg4ICgoKN+6O3bsQKtWraCrqwtdXV04OjoWWL+kzD/xlHvs1ac+j5GUfrdu3YKdnR0WLFiAT58+Yffu3XyHRAghhEe8JyNHjhzBjBkz4O7ujgcPHqBBgwZwcnLCx48f86x/5coVDBo0CJcvX8atW7dgZmaGTp064d27dyUc+X+ev0/C/4KiuW07c13eYinNEhMTMWnSJLRo0QJPnz6Fvr4+Dhw4gMWLF/MdGiGEEB4JGGOMzwAcHBxgb2+PzZs3A/i62qaZmRmmTJmCuXPn/nR/sVgMXV1dbN68GcOHD/9p/aSkJGhrayMxMRFaWlq/HL9EwmA5/yy3HTinHaro0mJcPzp58iQmTZqE9+/fAwBGjBiBNWvWQF+fbhxICCFlhbw/Q7/htWUkKysL9+/fh6OjI1cmFArh6OiIW7duFeoYaWlpyM7Ohp6eXp7PZ2ZmIikpSepHnrZff8U9PjTGgRKRfNy+fRvv379H9erVERAQgD179lAiQgghBADPs2ni4uIgFothaCi9QqmhoSFCQ0MLdYw5c+bAxMREKqH5npeXF5YuXfrLseaFMYaV/3yN07GOIVpUpw/Xb8RiMT5//gwDAwMAwOLFi6Gnp4cpU6ZAVVWV5+gIIYSUJryPGfkVK1euhL+/P06cOAEVlbzv/TJv3jwkJiZyP9/P3PhVd6O+cI+X9aont+OWdY8ePUKzZs3Qo0cPiMViAICamhp+//13SkQIIYTkwmvLiL6+PhQUFBAbGytVHhsbCyMjowL3Xbt2LVauXImLFy+ifv38Z68oKytDWVlZLvH+aGPACwCAgaYyjLXpQzYtLQ1LliyBt7c3xGIxtLS0EBISAmtra75DI4QQUorx2jIiEolgZ2eHgIAArkwikSAgIADNmjXLd7/Vq1dj2bJlOHfuHBo3blwSoebCGENgRBwAWvIdAM6fPw9ra2usWbMGYrEY/fr1o0SEEEJIofC+AuuMGTPg6uqKxo0bo0mTJvDx8UFqaipGjhwJABg+fDhMTU3h5eUFAFi1ahUWL16Mw4cPw8LCAjExMQAADQ0NaGholFjc33fRjG5ZrcTOW9okJydj/PjxOHz4MADAzMwMvr6+6N69O8+REUIIKSt4T0ZcXFzw6dMnLF68GDExMbC1tcW5c+e4Qa3R0dFSd2vdunUrsrKy0K9fP6njuLu7Y8mSJSUSc1JGNgZs+2+2j6FW3uNVKgI1NTWEh4dDKBRi6tSpWLZsWYkmhYQQQso+3tcZKWnymCM959hjHLn3dSDsmn710b+xmTxDLPXCw8NhZmbGDUZ9+vQp0tPTYW9vz3NkhBBCilO5XGekLGKMcYmInbluhUpEMjMz4eHhARsbGyxfvpwrt7a2pkSEEEJIkfHeTVPWLP37Ofd4/QBb/gIpYYGBgXBzc0NISAgA4MmTJ5BIJFJdaIQQQkhR0CeJDCQShr03owAA+hrKqFqp/K+2mpCQgHHjxqFVq1YICQmBgYEBDh8+jL/++osSEUIIIXJBLSMyWPtvGPf4zwn5Tz0uL65duwYXFxduxtLo0aOxevXqfJfeJ4QQQoqCkhEZbLnyEgBgXkkN5pXUeY6m+JmbmyMpKQm1atXC9u3b0bp1a75DIoQQUg5RO3shvYhN5h5vH8bPQmvFLScnB+fPn+e2zc3NcfHiRTx69IgSEUIIIcWGkpFC2vD/S78DQC0jTR4jKR4PHjyAg4MDnJ2dcfnyZa68WbNmxbacPiGEEAJQMlIoSRnZOP34AwBgumNNnqORr5SUFMycORP29vZ48OABdHR08PnzZ77DIoQQUoHQmJFC2HHtFQBAQSjA5PbVeY5Gfs6ePYuJEyfi9evXAL6uhuvj4/PTmxQSQggh8kTJSCHc+P8b4jW11IOCUMBzNPIxZcoUbN68GcDXsSFbt25F586deY6KEEJIRUTdND+RkS3Gg+gEAOXrhngtWrSAUCjEzJkz8ezZM0pECCGE8IZaRn7iWxcNALSpacBjJL8mJCQE7969g6OjI4CvXTKNGjVCzZrlawwMIYSQsodaRn7i6P2v96FpYlE2u2gyMjLg7u6OBg0aYMiQIYiPjwcACAQCSkQIIYSUCtQyUoBHbxLwJj4dADCyhQW/wRTB1atXMW7cOISFfV051t7eHpmZmTxHRQghhEijlpECbLr039oinW2MeYxENvHx8RgzZgzatm2LsLAwGBoa4ujRo/j7779hbFx2XgchhJCKgVpGCnAx5CMAYHK7sjOdNz4+HnXr1kVsbCwAYNy4cVi5ciV0dHT4DYwQQgjJByUj+fiYlME9drE34zES2ejp6aFr1664desWtm/fjpYtW/IdEiGEEFIgSkbycfBONPfYTE+Nx0gKlpOTAx8fH/Tr1w8WFhYAAB8fH4hEIlrGnRBCSJlAY0bysfn/x4uY6qjyHEn+7t69C3t7e8yePRsTJ04EYwwAoKmpSYkIIYSQMoOSkTzkiCWQfP1cx2CHqvwGk4fk5GRMmzYNTZs2RXBwMPT09NC/f3++wyKEEEKKhLpp8rD3ZhT3uLStunrq1ClMmjQJb9++BQAMGTIE3t7eMDAouwuyEUIIqdgoGcnD4aCv40W0VZWgoqTAczT/OXz4MIYMGQIAqFatGvz8/NCpUyeeoyKEEEJ+DSUjeXj1KRVA6eui6d27N+rUqYMePXpg8eLFUFMrvQNrCSGEkMKiZOQHN1/GcY/72VXhMRLg6dOn2LRpE3x9faGoqAhVVVU8fPiQBqcSQggpV2gA6w/G7LsHAFBWFMKqsgYvMWRkZGDhwoVo2LAhtm/fDl9fX+45SkQIIYSUN9Qy8p0XsclIyxIDABZ2rcNLDJcuXcK4ceMQEREBAOjZsyf69u3LSyyEEEJISaCWke9svhzBPR7iYF6i546Li8OIESPQoUMHREREwMTEBMePH8fJkydRpQq/3UWEEEJIcaKWkf+XI5bgr+D3AIBZnWpCKBSU6PlHjhyJ06dPQyAQYOLEiVixYgW0tbVLNAZCCCGED5SM/L/1F8O5x2NbW5b4+T09PfH+/Xv4+vqiadOmJX5+QgghhC+UjADIFkvgd/UVAKB+FW0oKxbv2iLZ2dlYu3YtMjIysHTpUgCAjY0N7t27B4GgZFtkCCGEEL5RMgIgIOQjxP+//vuO4Y2L9Vy3b9+Gm5sbnjx5AgUFBQwePBi1atUCAEpECCGEVEg0gBXA+IP3AQAtq+vDUEulWM6RlJSEyZMno3nz5njy5AkqVaqEPXv2oGbNmsVyPkIIIaSsqPAtIzliCfe4ewPjYjnHiRMnMHnyZLx//3WArKurK9auXQt9ff1iOR8hhBBSllT4ZORS6EfucT87M7kfPy4uDq6urkhOTkb16tXh5+eHDh06yP08hBBCSFlV4ZORbzfFq2eiBQU5TedljHHjP/T19bF69Wq8efMGCxcuhKqqqlzOQQghhJQXFX7MyJWwTwCAhlV15HK8x48fo2nTprhw4QJXNn78eKxYsYISEUIIISQPFToZ+X68yK920aSlpWHu3Llo1KgRgoKCMGfOHDDGfjVEQgghpNyr0MnI/ddfuMc2pkVf7fTff/+FtbU1Vq1aBbFYjL59+3KrqRJCCCGkYBU6GfG7+hIAUMNAo0jjRT5+/IihQ4fCyckJkZGRqFKlCv766y8cO3YMJiYm8g6XEEIIKZcqdDISGZcKAKikISrS/oGBgTh06BAEAgGmTZuG58+fo0ePHvIMkRBCCCn3KuxsGsYYouPTAADd6he+FSMjIwMqKl8XRuvduzdmz56N/v37w97evljiJIQQQsq7Ctsy8ux9Iv5/BXh0q//zxc6ysrKwfPlyVK9eHXFxcQC+Lt++evVqSkQIIYSQX1Bhk5ENFyO4xzpqBXfT3LhxAw0bNsSiRYvw7t077N27t5ijI4QQQiqOCpuM3Hr1GQDQxEIv3zoJCQkYP348WrZsiefPn8PAwACHDx/GzJkzSypMQgghpNyrsGNGvpnaoUae5ceOHcOUKVMQExMDABg9ejRWr14NPb38kxdCCCGEyK7CJyMtqlfKs/zixYuIiYlBrVq1sG3bNrRp06aEIyOEEEIqhgqdjKiJFLiFycRiMRITE7mWj5UrV8Lc3BzTp0/nZs8QQgghRP4q7JgRAOjR4OuU3ocPH6Jp06YYOHAgt4S7jo4O5s2bR4kIIYQQUsxKRTLi6+sLCwsLqKiowMHBAUFBQQXW/+OPP1C7dm2oqKjAxsYGZ8+eLdJ5e1lXwqxZs2Bvb4979+4hKCgIr169KtKxCCGEEFI0vCcjR44cwYwZM+Du7o4HDx6gQYMGcHJywsePH/Osf/PmTQwaNAijR4/Gw4cP0atXL/Tq1QtPnz6V6bzpkQ8w0KkF1q1bB7FYDBcXF4SGhsLKykoeL4sQQgghhSRgPN9a1sHBAfb29ti8eTMAQCKRwMzMDFOmTMHcuXNz1XdxcUFqaipOnz7NlTVt2hS2trbw8/P76fmSkpKgrf3fTfHMzc2xZcsWdOnSRQ6vhhBCCCm/vn2GJiYmQktLS27H5XUAa1ZWFu7fv4958+ZxZUKhEI6Ojrh161ae+9y6dQszZsyQKnNycsLJkyfzrJ+ZmYnMzExuOzExkXs8adIkzJ8/HxoaGkhKSvqFV0IIIYSUf98+K+XdjsFrMhIXFwexWAxDQ0OpckNDQ4SGhua5T0xMTJ71v60H8iMvLy8sXbo0z+d8fX3h6+tbhMgJIYSQiuvz589SvQy/qtxP7Z03b55US0pCQgLMzc0RHR0t1wtJ8peUlAQzMzO8efNGrs16JH90zUseXfOSR9e85CUmJqJq1apyXwCU12REX18fCgoKiI2NlSqPjY2FkZFRnvsYGRnJVF9ZWRnKysq5yrW1tenNW8K0tLTompcwuuYlj655yaNrXvKEQvnOf+F1No1IJIKdnR0CAgK4MolEgoCAADRr1izPfZo1ayZVHwAuXLiQb31CCCGElG68d9PMmDEDrq6uaNy4MZo0aQIfHx+kpqZi5MiRAIDhw4fD1NQUXl5eAIBp06ahTZs2WLduHbp27Qp/f3/cu3cP27dv5/NlEEIIIaSIeE9GXFxc8OnTJyxevBgxMTGwtbXFuXPnuEGq0dHRUs1BzZs3x+HDh7Fw4ULMnz8fNWrUwMmTJ2FtbV2o8ykrK8Pd3T3PrhtSPOialzy65iWPrnnJo2te8orrmvO+zgghhBBCKjbeV2AlhBBCSMVGyQghhBBCeEXJCCGEEEJ4RckIIYQQQnhVLpMRX19fWFhYQEVFBQ4ODggKCiqw/h9//IHatWtDRUUFNjY2OHv2bAlFWn7Ics137NiBVq1aQVdXF7q6unB0dPzp/xHJTdb3+Tf+/v4QCATo1atX8QZYDsl6zRMSEjBp0iQYGxtDWVkZNWvWpL8vMpL1mvv4+KBWrVpQVVWFmZkZpk+fjoyMjBKKtuy7du0aunfvDhMTEwgEgnzv+/a9K1euoFGjRlBWVkb16tWxd+9e2U/Myhl/f38mEonY7t272bNnz9jYsWOZjo4Oi42NzbP+jRs3mIKCAlu9ejV7/vw5W7hwIVNSUmJPnjwp4cjLLlmv+eDBg5mvry97+PAhCwkJYSNGjGDa2trs7du3JRx52SXrNf8mMjKSmZqaslatWrGePXuWTLDlhKzXPDMzkzVu3Jh16dKFBQYGssjISHblyhUWHBxcwpGXXbJe80OHDjFlZWV26NAhFhkZyc6fP8+MjY3Z9OnTSzjysuvs2bNswYIF7Pjx4wwAO3HiRIH1X716xdTU1NiMGTPY8+fP2aZNm5iCggI7d+6cTOctd8lIkyZN2KRJk7htsVjMTExMmJeXV571BwwYwLp27SpV5uDgwMaNG1escZYnsl7zH+Xk5DBNTU22b9++4gqx3CnKNc/JyWHNmzdnO3fuZK6urpSMyEjWa75161ZmaWnJsrKySirEckfWaz5p0iTWvn17qbIZM2awFi1aFGuc5VVhkpHff/+d1atXT6rMxcWFOTk5yXSuctVNk5WVhfv378PR0ZErEwqFcHR0xK1bt/Lc59atW1L1AcDJySnf+kRaUa75j9LS0pCdnS33Gy+VV0W95h4eHjAwMMDo0aNLIsxypSjX/NSpU2jWrBkmTZoEQ0NDWFtbw9PTE2KxuKTCLtOKcs2bN2+O+/fvc105r169wtmzZ9GlS5cSibkiktdnKO8rsMpTXFwcxGIxt3rrN4aGhggNDc1zn5iYmDzrx8TEFFuc5UlRrvmP5syZAxMTk1xvaJK3olzzwMBA7Nq1C8HBwSUQYflTlGv+6tUrXLp0CUOGDMHZs2cRERGBiRMnIjs7G+7u7iURdplWlGs+ePBgxMXFoWXLlmCMIScnB+PHj8f8+fNLIuQKKb/P0KSkJKSnp0NVVbVQxylXLSOk7Fm5ciX8/f1x4sQJqKio8B1OuZScnIxhw4Zhx44d0NfX5zucCkMikcDAwADbt2+HnZ0dXFxcsGDBAvj5+fEdWrl15coVeHp6YsuWLXjw4AGOHz+OM2fOYNmyZXyHRn6iXLWM6OvrQ0FBAbGxsVLlsbGxMDIyynMfIyMjmeoTaUW55t+sXbsWK1euxMWLF1G/fv3iDLNckfWav3z5ElFRUejevTtXJpFIAACKiooICwuDlZVV8QZdxhXlfW5sbAwlJSUoKChwZXXq1EFMTAyysrIgEomKNeayrijXfNGiRRg2bBjGjBkDALCxsUFqairc3NywYMECud/2nuT/GaqlpVXoVhGgnLWMiEQi2NnZISAggCuTSCQICAhAs2bN8tynWbNmUvUB4MKFC/nWJ9KKcs0BYPXq1Vi2bBnOnTuHxo0bl0So5Yas17x27dp48uQJgoODuZ8ePXqgXbt2CA4OhpmZWUmGXyYV5X3eokULREREcIkfAISHh8PY2JgSkUIoyjVPS0vLlXB8SwYZ3YatWMjtM1S2sbWln7+/P1NWVmZ79+5lz58/Z25ubkxHR4fFxMQwxhgbNmwYmzt3Llf/xo0bTFFRka1du5aFhIQwd3d3mtorI1mv+cqVK5lIJGLHjh1jHz584H6Sk5P5eglljqzX/Ec0m0Z2sl7z6OhopqmpySZPnszCwsLY6dOnmYGBAVu+fDlfL6HMkfWau7u7M01NTfa///2PvXr1iv3777/MysqKDRgwgK+XUOYkJyezhw8fsocPHzIAzNvbmz18+JC9fv2aMcbY3Llz2bBhw7j636b2zp49m4WEhDBfX1+a2vvNpk2bWNWqVZlIJGJNmjRht2/f5p5r06YNc3V1lap/9OhRVrNmTSYSiVi9evXYmTNnSjjisk+Wa25ubs4A5Ppxd3cv+cDLMFnf59+jZKRoZL3mN2/eZA4ODkxZWZlZWlqyFStWsJycnBKOumyT5ZpnZ2ezJUuWMCsrK6aiosLMzMzYxIkT2ZcvX0o+8DLq8uXLef59/nadXV1dWZs2bXLtY2try0QiEbO0tGR79uyR+bwCxqjtihBCCCH8KVdjRgghhBBS9lAyQgghhBBeUTJCCCGEEF5RMkIIIYQQXlEyQgghhBBeUTJCCCGEEF5RMkIIIYQQXlEyQgghhBBeUTJCSDmzd+9e6Ojo8B1GkQkEApw8ebLAOiNGjECvXr1KJB5CSPGjZISQUmjEiBEQCAS5fiIiIvgODXv37uXiEQqFqFKlCkaOHImPHz/K5fgfPnxA586dAQBRUVEQCAQIDg6WqrNhwwbs3btXLufLz5IlS7jXqaCgADMzM7i5uSE+Pl6m41DiRMjPKfIdACEkb87OztizZ49UWeXKlXmKRpqWlhbCwsIgkUjw6NEjjBw5Eu/fv8f58+d/+dj53R7+e9ra2r98nsKoV68eLl68CLFYjJCQEIwaNQqJiYk4cuRIiZyfkIqCWkYIKaWUlZVhZGQk9aOgoABvb2/Y2NhAXV0dZmZmmDhxIlJSUvI9zqNHj9CuXTtoampCS0sLdnZ2uHfvHvd8YGAgWrVqBVVVVZiZmWHq1KlITU0tMDaBQAAjIyOYmJigc+fOmDp1Ki5evIj09HRIJBJ4eHigSpUqUFZWhq2tLc6dO8ftm5WVhcmTJ8PY2BgqKiowNzeHl5eX1LG/ddNUq1YNANCwYUMIBAK0bdsWgHRrw/bt22FiYgKJRCIVY8+ePTFq1Chu+6+//kKjRo2goqICS0tLLF26FDk5OQW+TkVFRRgZGcHU1BSOjo7o378/Lly4wD0vFosxevRoVKtWDaqqqqhVqxY2bNjAPb9kyRLs27cPf/31F9fKcuXKFQDAmzdvMGDAAOjo6EBPTw89e/ZEVFRUgfEQUl5RMkJIGSMUCrFx40Y8e/YM+/btw6VLl/D777/nW3/IkCGoUqUK7t69i/v372Pu3LlQUlICALx8+RLOzs7o27cvHj9+jCNHjiAwMBCTJ0+WKSZVVVVIJBLk5ORgw4YNWLduHdauXYvHjx/DyckJPXr0wIsXLwAAGzduxKlTp3D06FGEhYXh0KFDsLCwyPO4QUFBAICLFy/iw4cPOH78eK46/fv3x+fPn3H58mWuLD4+HufOncOQIUMAANevX8fw4cMxbdo0PH/+HNu2bcPevXuxYsWKQr/GqKgonD9/HiKRiCuTSCSoUqUK/vjjDzx//hyLFy/G/PnzcfToUQDArFmzMGDAADg7O+PDhw/48OEDmjdvjuzsbDg5OUFTUxPXr1/HjRs3oKGhAWdnZ2RlZRU6JkLKjV+93TAhRP5cXV2ZgoICU1dX53769euXZ90//viDVapUidves2cP09bW5rY1NTXZ3r1789x39OjRzM3NTars+vXrTCgUsvT09Dz3+fH44eHhrGbNmqxx48aMMcZMTEzYihUrpPaxt7dnEydOZIwxNmXKFNa+fXsmkUjyPD4AduLECcYYY5GRkQwAe/jwoVQdV1dX1rNnT267Z8+ebNSoUdz2tm3bmImJCROLxYwxxjp06MA8PT2ljnHgwAFmbGycZwyMMebu7s6EQiFTV1dnKioq3K3Uvb29892HMcYmTZrE+vbtm2+s385dq1YtqWuQmZnJVFVV2fnz5ws8PiHlEY0ZIaSUateuHbZu3cptq6urA/jaSuDl5YXQ0FAkJSUhJycHGRkZSEtLg5qaWq7jzJgxA2PGjMGBAwe4rgYrKysAX7twHj9+jEOHDnH1GWOQSCSIjIxEnTp18owtMTERGhoakEgkyMjIQMuWLbFz504kJSXh/fv3aNGihVT9Fi1a4NGjRwC+drF07NgRtWrVgrOzM7p164ZOnTr90rUaMmQIxo4diy1btkBZWRmHDh3CwIEDIRQKudd548YNqZYQsVhc4HUDgFq1auHUqVPIyMjAwYMHERwcjClTpkjV8fX1xe7duxEdHY309HRkZWXB1ta2wHgfPXqEiIgIaGpqSpVnZGTg5cuXRbgChJRtlIwQUkqpq6ujevXqUmVRUVHo1q0bJkyYgBUrVkBPTw+BgYEYPXo0srKy8vxQXbJkCQYPHowzZ87gn3/+gbu7O/z9/dG7d2+kpKRg3LhxmDp1aq79qlatmm9smpqaePDgAYRCIYyNjaGqqgoASEpK+unratSoESIjI/HPP//g4sWLGDBgABwdHXHs2LGf7puf7t27gzGGM2fOwN7eHtevX8f69eu551NSUrB06VL06dMn174qKir5HlckEnH/BytXrkTXrl2xdOlSLFu2DADg7++PWbNmYd26dWjWrBk0NTWxZs0a3Llzp8B4U1JSYGdnJ5UEflNaBikTUpIoGSGkDLl//z4kEgnWrVvHfev/Nj6hIDVr1kTNmjUxffp0DBo0CHv27EHv3r3RqFEjPH/+PFfS8zNCoTDPfbS0tGBiYoIbN26gTZs2XPmNGzfQpEkTqXouLi5wcXFBv3794OzsjPj4eOjp6Ukd79v4DLFYXGA8Kioq6NOnDw4dOoSIiAjUqlULjRo14p7/v3buHyS1Pg4D+HODFDGsQSQLIaJWDYcgl6CIxtZAiJYG41BT5GD/lqDFpSUMGmwwaWqIbEqIGjJCgqhjVBRBFEE4VYg8d7pSeR3iDoe39/mM5w/n+/tND+c8nGAwCNM0v73Or2KxGPr6+hCJRCrrDIVCGB8fr1zz9c2GzWarmj8YDCKdTsPj8cDlcv3TTCI/gQqsIv8hHR0dKJVKWF5exvX1NdbX17GyslLz+tfXVxiGgWw2i9vbWxwcHCCXy1U+v0xPT+Pw8BCGYSCfz+Py8hJbW1vfLrB+NDU1haWlJaTTaZimiWg0inw+j8nJSQBAPB5HKpXCxcUFCoUCNjc30dzc/NcftXk8HjgcDmQyGTw+PqJYLNZ8bjgcxvb2NtbW1irF1T9mZ2eRTCaxsLCAs7MznJ+fY2NjA7FY7Ftr6+npgd/vx+LiIgCgs7MTx8fH2N3dRaFQwMzMDHK53Kd72tracHp6CtM08fz8jFKphHA4DLfbjaGhIezv7+Pm5gbZbBYTExO4v7//1kwiP4LVpRURqfa30uMf8XicXq+XDoeDg4ODTCaTBMCXlxeSnwum7+/vHB4eps/no81mY0tLCw3D+FROPTo64sDAABsaGuh0Oun3+6sKqB99LbB+VS6XOT8/z9bWVtbX1zMQCHBnZ6dyPpFIsKuri06nky6Xi/39/Tw5Oamcx4cCK0murq7S5/Oxrq6Ovb29NfenXC7T6/USAK+urqrmymQyDIVCdDgcdLlc7O7uZiKRqLmOubk5BgKBquOpVIp2u513d3d8e3vj6OgoGxsb2dTUxEgkwmg0+um+p6enyv4C4N7eHkny4eGBIyMjdLvdtNvtbG9v59jYGIvFYs2ZRH6qXyRpbRwSERGR/zN9phERERFLKYyIiIiIpRRGRERExFIKIyIiImIphRERERGxlMKIiIiIWEphRERERCylMCIiIiKWUhgRERERSymMiIiIiKUURkRERMRSvwFyJtRJqlWqmQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(lgbm_fpr, lgbm_tpr, label='ROC curve (area = %0.2f)' % lgbm_auc)\n",
    "plt.plot([0, 1], [0, 1], 'k--', label='No Skill')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve for LightGBM  Classification')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Deep learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.utils.data \n",
    "import torch.nn.functional as F"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "device(type='cuda')"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "device"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MLPClassifier(nn.Module): # Multiple Layer Perceptron\n",
    "    def __init__(self, input_size, output_size= 1):\n",
    "        super(MLPClassifier, self).__init__()\n",
    "        self.fc1 = nn.Linear(input_size, 64)\n",
    "        \n",
    "        self.fc2 = nn.Linear(64, 64)\n",
    "        self.fc3 = nn.Linear(64, 32)\n",
    "        self.fc4 = nn.Linear(32, 32)\n",
    "        self.out = nn.Linear(32, output_size)\n",
    "        \n",
    "        # ReLU activation function\n",
    "        self.relu = nn.ReLU()\n",
    "\n",
    "        # Dropout with 0.2 probability\n",
    "        self.dropout = nn.Dropout(0.2)\n",
    "\n",
    "\n",
    "    def forward(self, x):\n",
    "        # layer 1 and layer 2\n",
    "        out = self.fc1(x)\n",
    "        out = self.relu(out)\n",
    "        out = self.fc2(out)\n",
    "        out = self.relu(out)\n",
    "        # add dropout layer\n",
    "        out = self.dropout(out)\n",
    "\n",
    "        # layer 3 and layer 4\n",
    "        out = self.fc3(out)\n",
    "        out = self.relu(out)\n",
    "        out = self.fc4(out)\n",
    "        out = self.relu(out)    \n",
    "\n",
    "        # output layer\n",
    "        out  = self.out(out)\n",
    "        out =  nn.Sigmoid()(out) # for binary classification\n",
    "        # out = nn.Softmax(dim=1)# for multiply classes\n",
    "\n",
    "        return  out "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pandas.core.frame.DataFrame"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MLPClassifier(\n",
       "  (fc1): Linear(in_features=33, out_features=64, bias=True)\n",
       "  (fc2): Linear(in_features=64, out_features=64, bias=True)\n",
       "  (fc3): Linear(in_features=64, out_features=32, bias=True)\n",
       "  (fc4): Linear(in_features=32, out_features=32, bias=True)\n",
       "  (out): Linear(in_features=32, out_features=1, bias=True)\n",
       "  (relu): ReLU()\n",
       "  (dropout): Dropout(p=0.2, inplace=False)\n",
       ")"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mlp = MLPClassifier(input_size=X_train.shape[1], output_size=1).to(device)\n",
    "mlp.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create tensors for our train and test set. \n",
    "As you remember we need variable to accumulate gradients. \n",
    "Therefore first we create tensor, then we will create variable '''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "X_train_tensor = torch.tensor(X_train.values, dtype=torch.float32).to(device)\n",
    "X_test_tensor = torch.tensor(X_test.values, dtype=torch.float32).to(device)\n",
    "y_train_tensor = torch.tensor(y_train.values, dtype=torch.float32).view(-1,1).to(device)\n",
    "y_test_tensor = torch.tensor(y_test.values, dtype=torch.float32).view(-1,1).to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Optimizing the Model Parameters\n",
    "===============================\n",
    "\n",
    "To train a model, we need a [loss\n",
    "function](https://pytorch.org/docs/stable/nn.html#loss-functions) and an\n",
    "[optimizer](https://pytorch.org/docs/stable/optim.html).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "optimizer = torch.optim.Adam(mlp.parameters(), lr=0.001)\n",
    "# criterion = nn.CrossEntropyLoss()\n",
    "criterion = nn.BCELoss()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create Dataset and Data loader for pytorch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "from torch.utils.data import Dataset, DataLoader , TensorDataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "#   Make torch datasets from train and test sets\n",
    "# train = torch.utils.data.TensorDataset(X_train,y_train)\n",
    "# test = torch.utils.data.TensorDataset(X_test,y_test)\n",
    "trainDataSet = TensorDataset(X_train_tensor, y_train_tensor)\n",
    "testDataSet = TensorDataset(X_test_tensor, y_test_tensor)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create train and test data loaders\n",
    "\n",
    "batchSize = 32\n",
    "\n",
    "# train_loader = torch.utils.data.DataLoader(train, batch_size = 64, shuffle = True)\n",
    "# test_loader = torch.utils.data.DataLoader(test, batch_size = 64, shuffle = True)\n",
    "trainDataLoader = DataLoader(trainDataSet, batch_size=batchSize, shuffle=True)\n",
    "testDataLoader = DataLoader(testDataSet, batch_size=batchSize)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of X [N, C, H, W]: torch.Size([32, 33])\n",
      "Shape of y: torch.Size([32, 1]) torch.float32\n"
     ]
    }
   ],
   "source": [
    "for x, y in trainDataLoader:\n",
    "    print(f\"Shape of X [N, C, H, W]: {x.shape}\")\n",
    "    print(f\"Shape of y: {y.shape} {y.dtype}\")\n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of X [N, C, H, W]: torch.Size([32, 33])\n",
      "Shape of y: torch.Size([32, 1]) torch.float32\n"
     ]
    }
   ],
   "source": [
    "for x, y in testDataLoader:\n",
    "    print(f\"Shape of X [N, C, H, W]: {x.shape}\")\n",
    "    print(f\"Shape of y: {y.shape} {y.dtype}\")\n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "No. of batches in trainloader : 2210\n",
      "No. of batches in validloader : 553\n"
     ]
    }
   ],
   "source": [
    "print(f\"No. of batches in trainloader : {len(trainDataLoader)}\")\n",
    "print(f\"No. of batches in validloader : {len(testDataLoader)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([70691, 33])"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainDataLoader.dataset.tensors[0].shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create train/Valid loop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create some list to keep track of loss and accuracy during training\n",
    "train_losses = []\n",
    "test_acc = []\n",
    "y_predicted = []\n",
    "y_true = []\n",
    "# test_losses = []\n",
    "epochs_list = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(model, train_loader, criterion, optimizer):\n",
    "    model.train() # set the model to training mode calculate gradients and update weights\n",
    "    trainSize = len(train_loader)\n",
    "    # print(\"Length of Train Loader: \",trainSize)\n",
    "    running_loss = 0.0\n",
    "    avgLoss = 0.0\n",
    "    for i, data in enumerate(train_loader, 0):\n",
    "    # for i, data in tqdm(train_loader):\n",
    "        inputs, labels = data\n",
    "        inputs, labels = inputs.to(device), labels.to(device) # to GPU if available\n",
    "        optimizer.zero_grad() # reset gradients to zero\n",
    "        outputs = model(inputs)  # prediction \n",
    "        loss = criterion(outputs, labels) # calculate loss\n",
    "\n",
    "        #getting gradients\n",
    "        loss.backward()\n",
    "        optimizer.step() # update weights\n",
    "        running_loss += loss.item()   # add the loss\n",
    "        \n",
    "        if i % 500 == 499: # each 50 mini-batches print the loss\n",
    "            tempLoss = running_loss/ (i+1)\n",
    "            print(f\"Batch {i+1}, loss: {tempLoss}\")\n",
    "            # running_loss = 0.0\n",
    "    avgLoss = running_loss/ trainSize\n",
    "    print(f\"Training loss: {avgLoss}\")\n",
    "        \n",
    "    print('Finished Training')\n",
    "    return avgLoss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "def validate(model, test_loader):\n",
    "    model.eval() # set model to evaluation mode, no calcuate gradients\n",
    "    with torch.no_grad():\n",
    "        correct = 0\n",
    "        total = 0\n",
    "        # print(\"Length of test Loader: \", len(test_loader))\n",
    "        for i, data in enumerate(test_loader, 0):\n",
    "            inputs, labels = data\n",
    "            inputs, labels = inputs.to(device), labels.to(device) # to GPU if available\n",
    "            outputs = model(inputs) # prediction of the model\n",
    "            # _, predicted = torch.max(outputs.data, 1)  # get the predicted class\n",
    "            predicted = torch.round(outputs.data[0])\n",
    "            y_predicted.append(predicted.data.cpu().numpy()) # add list of predict for validation\n",
    "            y_true.append(labels[0].data.cpu().numpy()) # add list of true labels for validation\n",
    "            # print(\"Predicted  Result: \", predicted)\n",
    "            total += labels.size(0) # total number of labels\n",
    "            # total  += len(labels) # total number of labels\n",
    "            correct += (predicted == labels).sum() # add correct predictions \n",
    "            # print(\"labels size: \", labels.size(0))\n",
    "            # print(\"Total: \", total, \" Correct: \", correct)\n",
    "            if i % 500 == 499:\n",
    "                print(\"Total: \", total, \" Correct: \", correct)\n",
    "                print(f'Batch {i+1}, Accuracy: {100 * correct / total}')\n",
    "            # print(f'Batch {i+1}, Accuracy: {100 * correct / total}')\n",
    "        # print(f'Accuracy: {100 * correct / total}')\n",
    "        acc = 100 * (correct / total).item() # calulate accuracy\n",
    "    return acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 0/20 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.6444593051671982\n",
      "Batch 1000, loss: 0.6241445885002613\n",
      "Batch 1500, loss: 0.6177525495092074\n",
      "Batch 2000, loss: 0.613055008366704\n",
      "Training loss: 0.6121574758134816\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  5%|▌         | 1/20 [00:01<00:30,  1.61s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8053, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.33125305175781\n",
      "Epoch 2\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.6011042946577072\n",
      "Batch 1000, loss: 0.5987926816046238\n",
      "Batch 1500, loss: 0.5992376637061437\n",
      "Batch 2000, loss: 0.5988272910863162\n",
      "Training loss: 0.5985350738148884\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 10%|█         | 2/20 [00:03<00:28,  1.60s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(7985, device='cuda:0')\n",
      "Batch 500, Accuracy: 49.906253814697266\n",
      "Epoch 3\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5979283856749534\n",
      "Batch 1000, loss: 0.5975136299431324\n",
      "Batch 1500, loss: 0.5986038324435552\n",
      "Batch 2000, loss: 0.5971243099868297\n",
      "Training loss: 0.5970886873376315\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 15%|█▌        | 3/20 [00:04<00:26,  1.57s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8023, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.14375305175781\n",
      "Epoch 4\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.6002026990652084\n",
      "Batch 1000, loss: 0.5968617633581161\n",
      "Batch 1500, loss: 0.5968663161595662\n",
      "Batch 2000, loss: 0.5963173836767673\n",
      "Training loss: 0.5966462186558754\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20%|██        | 4/20 [00:06<00:24,  1.54s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8075, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.468753814697266\n",
      "Epoch 5\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5918212922811508\n",
      "Batch 1000, loss: 0.5931695275604725\n",
      "Batch 1500, loss: 0.5959028096000354\n",
      "Batch 2000, loss: 0.5954825653582811\n",
      "Training loss: 0.5953378912685144\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 25%|██▌       | 5/20 [00:07<00:23,  1.56s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8111, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.69375228881836\n",
      "Epoch 6\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5959014712572098\n",
      "Batch 1000, loss: 0.5964602429866791\n",
      "Batch 1500, loss: 0.5965948572357496\n",
      "Batch 2000, loss: 0.5949118440449238\n",
      "Training loss: 0.5951688109479878\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|███       | 6/20 [00:09<00:22,  1.64s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8079, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.493751525878906\n",
      "Epoch 7\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5851222065091133\n",
      "Batch 1000, loss: 0.591149443089962\n",
      "Batch 1500, loss: 0.5936235859791438\n",
      "Batch 2000, loss: 0.5933703399598599\n",
      "Training loss: 0.5942102384108764\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 35%|███▌      | 7/20 [00:11<00:21,  1.64s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8071, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.44375228881836\n",
      "Epoch 8\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5877835711240769\n",
      "Batch 1000, loss: 0.5917113311886787\n",
      "Batch 1500, loss: 0.5941420711080233\n",
      "Batch 2000, loss: 0.59378206756711\n",
      "Training loss: 0.5934020449268332\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 40%|████      | 8/20 [00:12<00:19,  1.63s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8079, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.493751525878906\n",
      "Epoch 9\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5919870836138725\n",
      "Batch 1000, loss: 0.5931383788585662\n",
      "Batch 1500, loss: 0.5925721935232481\n",
      "Batch 2000, loss: 0.5928987579941749\n",
      "Training loss: 0.5926884107055708\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 45%|████▌     | 9/20 [00:14<00:18,  1.64s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8067, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.41875076293945\n",
      "Epoch 10\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5920677190423012\n",
      "Batch 1000, loss: 0.592569535523653\n",
      "Batch 1500, loss: 0.5918537796139717\n",
      "Batch 2000, loss: 0.5924119225591421\n",
      "Training loss: 0.5922699391167627\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50%|█████     | 10/20 [00:16<00:16,  1.64s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8101, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.63125228881836\n",
      "Epoch 11\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5885468807220459\n",
      "Batch 1000, loss: 0.5904908609092235\n",
      "Batch 1500, loss: 0.5912300664782524\n",
      "Batch 2000, loss: 0.5920349511802196\n",
      "Training loss: 0.5920408444302114\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 55%|█████▌    | 11/20 [00:17<00:14,  1.62s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8133, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.83125305175781\n",
      "Epoch 12\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5886973899006843\n",
      "Batch 1000, loss: 0.5898656405806542\n",
      "Batch 1500, loss: 0.5910998265941938\n",
      "Batch 2000, loss: 0.5905878141373396\n",
      "Training loss: 0.5911139431986874\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 60%|██████    | 12/20 [00:19<00:13,  1.63s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8085, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.531253814697266\n",
      "Epoch 13\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5856991110444069\n",
      "Batch 1000, loss: 0.5859528475403786\n",
      "Batch 1500, loss: 0.5888871108690897\n",
      "Batch 2000, loss: 0.590116725474596\n",
      "Training loss: 0.5899766945731046\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 65%|██████▌   | 13/20 [00:21<00:11,  1.63s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8129, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.806251525878906\n",
      "Epoch 14\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5880598037838936\n",
      "Batch 1000, loss: 0.5875681533515453\n",
      "Batch 1500, loss: 0.5885109033385912\n",
      "Batch 2000, loss: 0.5890793721675873\n",
      "Training loss: 0.5900119516239987\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|███████   | 14/20 [00:22<00:09,  1.64s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8087, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.54375076293945\n",
      "Epoch 15\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5884471201300621\n",
      "Batch 1000, loss: 0.5872728008031846\n",
      "Batch 1500, loss: 0.5875125547647476\n",
      "Batch 2000, loss: 0.5887574945688248\n",
      "Training loss: 0.5894371016397735\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 75%|███████▌  | 15/20 [00:24<00:08,  1.63s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8091, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.56875228881836\n",
      "Epoch 16\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5824223284721375\n",
      "Batch 1000, loss: 0.586851443618536\n",
      "Batch 1500, loss: 0.5878012000719707\n",
      "Batch 2000, loss: 0.5880843205451965\n",
      "Training loss: 0.5881421436686322\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 80%|████████  | 16/20 [00:25<00:06,  1.61s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8127, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.79375076293945\n",
      "Epoch 17\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5868288985490799\n",
      "Batch 1000, loss: 0.5886521182060241\n",
      "Batch 1500, loss: 0.5898366784254709\n",
      "Batch 2000, loss: 0.5878173363208771\n",
      "Training loss: 0.5875421218338056\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 85%|████████▌ | 17/20 [00:27<00:04,  1.59s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8117, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.73125076293945\n",
      "Epoch 18\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5891309527158737\n",
      "Batch 1000, loss: 0.5904743082523346\n",
      "Batch 1500, loss: 0.5899602773388227\n",
      "Batch 2000, loss: 0.5875439270585775\n",
      "Training loss: 0.5870311323603893\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 90%|█████████ | 18/20 [00:28<00:03,  1.59s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8059, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.368751525878906\n",
      "Epoch 19\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5879736818671226\n",
      "Batch 1000, loss: 0.587192956238985\n",
      "Batch 1500, loss: 0.5857777720093728\n",
      "Batch 2000, loss: 0.58697541898489\n",
      "Training loss: 0.5869507630755998\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 95%|█████████▌| 19/20 [00:30<00:01,  1.62s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8079, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.493751525878906\n",
      "Epoch 20\n",
      "-------------------------------\n",
      "Batch 500, loss: 0.5835994929075241\n",
      "Batch 1000, loss: 0.5831057963371277\n",
      "Batch 1500, loss: 0.5844839237531027\n",
      "Batch 2000, loss: 0.5860444119721651\n",
      "Training loss: 0.586463978805693\n",
      "Finished Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 20/20 [00:32<00:00,  1.61s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total:  16000  Correct:  tensor(8051, device='cuda:0')\n",
      "Batch 500, Accuracy: 50.31875228881836\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# Define epochs (between 20-50)\n",
    "epochs = 20 #1000\n",
    "\n",
    "# for epoch in range(epochs):\n",
    "for epoch in tqdm(range(epochs)):\n",
    "    print(f\"Epoch {epoch+1}\\n-------------------------------\")\n",
    "    trainLoss = train(mlp, trainDataLoader, criterion, optimizer)\n",
    "    train_losses.append(trainLoss) # add the accuracy to the list\n",
    "    testAcc= validate(mlp, testDataLoader)\n",
    "    test_acc.append(testAcc)\n",
    "    epochs_list.append(epoch+1) # add the epoch to the list\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot training and validation Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(epochs_list, train_losses, label='Training loss')\n",
    "plt.title('Training loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(epochs_list, test_acc, label='Validation Accuracy')\n",
    "plt.title('Validation Accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  MLP classification report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.62      0.70      0.66      4400\n",
      "         1.0       0.79      0.71      0.75      6660\n",
      "\n",
      "    accuracy                           0.71     11060\n",
      "   macro avg       0.70      0.71      0.70     11060\n",
      "weighted avg       0.72      0.71      0.71     11060\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "reportClassificationMLP = classification_report(y_true, y_predicted)\n",
    "print(reportClassificationMLP)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7086175948675949"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mlp_auc = roc_auc_score(y_true, y_predicted)\n",
    "mlp_auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "mlp_fpr, mlp_tpr, mlp_th = roc_curve(y_true, y_predicted)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6, 4))\n",
    "plt.plot(mlp_fpr, mlp_tpr, label='ROC curve (area = %0.2f)' % mlp_auc)\n",
    "plt.plot([0, 1], [0, 1], 'k--', label='No Skill')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve for MLP  Classification')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Select best model for generate test results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] Accuracy may be bad since you didn't explicitly set num_leaves OR 2^max_depth > num_leaves. (num_leaves=31).\n"
     ]
    }
   ],
   "source": [
    "### PLEASE CHANGE THIS CODE TO IMPLEMENT YOUR OWN PREDICTIONS\n",
    "\n",
    "# Use our dummy classifier to make predictions on test_df using `predict_proba` method:\n",
    "# predicted_probability = dummy_clf.predict_proba(test_df.drop(['CustomerID'], axis=1))[:, 1]\n",
    "predicted_probability = lgbmModel.predict_proba(test_df.drop([\"CustomerID\"], axis=1))[:,1] # get the probability of the positive class\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "### PLEASE CHANGE THIS CODE TO IMPLEMENT YOUR OWN PREDICTIONS\n",
    "\n",
    "# Combine predictions with label column into a dataframe\n",
    "prediction_df = pd.DataFrame({'CustomerID': test_df[['CustomerID']].values[:, 0],\n",
    "                             'predicted_probability': predicted_probability})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(104480, 2)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>predicted_probability</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>O1W6BHP6RM</td>\n",
       "      <td>0.391903</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LFR4X92X8H</td>\n",
       "      <td>0.160941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>QM5GBIYODA</td>\n",
       "      <td>0.663871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>D9RXTK2K9F</td>\n",
       "      <td>0.058180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ENTCCHR1LR</td>\n",
       "      <td>0.460863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>7A88BB5IO6</td>\n",
       "      <td>0.680177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>70OMW9XEWR</td>\n",
       "      <td>0.306901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>EL1RMFMPYL</td>\n",
       "      <td>0.629730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>4IA2QPT6ZK</td>\n",
       "      <td>0.501073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>AEDCWHSJDN</td>\n",
       "      <td>0.545333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   CustomerID  predicted_probability\n",
       "0  O1W6BHP6RM               0.391903\n",
       "1  LFR4X92X8H               0.160941\n",
       "2  QM5GBIYODA               0.663871\n",
       "3  D9RXTK2K9F               0.058180\n",
       "4  ENTCCHR1LR               0.460863\n",
       "5  7A88BB5IO6               0.680177\n",
       "6  70OMW9XEWR               0.306901\n",
       "7  EL1RMFMPYL               0.629730\n",
       "8  4IA2QPT6ZK               0.501073\n",
       "9  AEDCWHSJDN               0.545333"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### PLEASE CHANGE THIS CODE TO IMPLEMENT YOUR OWN PREDICTIONS\n",
    "\n",
    "# View our 'prediction_df' dataframe as required for submission.\n",
    "# Ensure it should contain 104,480 rows and 2 columns 'CustomerID' and 'predicted_probaility'\n",
    "print(prediction_df.shape)\n",
    "prediction_df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**PLEASE CHANGE CODE ABOVE TO IMPLEMENT YOUR OWN PREDICTIONS**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Final Tests - **IMPORTANT** - the cells below must be run prior to submission\n",
    "\n",
    "Below are some tests to ensure your submission is in the correct format for autograding. The autograding process accepts a csv `prediction_submission.csv` which we will generate from our `prediction_df` below. Please run the tests below an ensure no assertion errors are thrown."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "# FINAL TEST CELLS - please make sure all of your code is above these test cells\n",
    "\n",
    "# Writing to csv for autograding purposes\n",
    "prediction_df.to_csv(\"prediction_submission.csv\", index=False)\n",
    "submission = pd.read_csv(\"prediction_submission.csv\")\n",
    "\n",
    "assert isinstance(submission, pd.DataFrame), 'You should have a dataframe named prediction_df.'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "# FINAL TEST CELLS - please make sure all of your code is above these test cells\n",
    "\n",
    "assert submission.columns[0] == 'CustomerID', 'The first column name should be CustomerID.'\n",
    "assert submission.columns[1] == 'predicted_probability', 'The second column name should be predicted_probability.'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "# FINAL TEST CELLS - please make sure all of your code is above these test cells\n",
    "\n",
    "assert submission.shape[0] == 104480, 'The dataframe prediction_df should have 104480 rows.'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "# FINAL TEST CELLS - please make sure all of your code is above these test cells\n",
    "\n",
    "assert submission.shape[1] == 2, 'The dataframe prediction_df should have 2 columns.'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "# FINAL TEST CELLS - please make sure all of your code is above these test cells\n",
    "\n",
    "## This cell calculates the auc score and is hidden. Submit Assignment to see AUC score."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SUBMIT YOUR WORK!\n",
    "\n",
    "Once we are happy with our `prediction_df` and `prediction_submission.csv` we can now submit for autograding! Submit by using the blue **Submit Assignment** at the top of your notebook. Don't worry if your initial submission isn't perfect as you have multiple submission attempts and will obtain some feedback after each submission!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}