customer_churn_binary_classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "customer_churn_binary_classification.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "IpBp1K2NA-uy",
        "VhBRCmRjBP8t",
        "ZlxHV7GxC4iC",
        "CkDCmhWGHMs1",
        "8rjMqTaGSYy7",
        "GyQYP6wkSgNM",
        "EgxbXEYtUhqz",
        "S2HgUeOZVmEC",
        "Z3tbK6KXZ9am",
        "HUHPPqDi6_Vv",
        "NOKnMgZg7NvT",
        "HFlf3wYm-Brr"
      ],
      "authorship_tag": "ABX9TyPZ0ph6doLPCFf06Aa3YXBw",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/wenjunsun/personal-machine-learning-projects/blob/master/customer_churn_binary_classification.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HkeE23F2OjND",
        "colab_type": "text"
      },
      "source": [
        "[Data Source](https://www.kaggle.com/adammaus/predicting-churn-for-bank-customers): from Kaggle. Mainly toy data, no description, no objective metrics, no explanation of how data was collected or what that data means specifically. It is about the task of giving the features of customers, predict whether he will churn or not (whether he will leave the bank)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IpBp1K2NA-uy",
        "colab_type": "text"
      },
      "source": [
        "##Download data from google drive shared link url and load it into variable df"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DdNpc3T2Omyj",
        "colab_type": "text"
      },
      "source": [
        "Could also use PyDrive or Mount Google Drive but this method is simper"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yN86D7Gk9p3T",
        "colab_type": "code",
        "outputId": "ee553bf6-cdc2-4cba-f3ae-5b86e36ab880",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
            "  import pandas.util.testing as tm\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BJy087rU9373",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# one way to read data, directly from the csv download link from google drive.\n",
        "# I think this is the easiest way, but of course there are other ways like mounting the drive, etc.\n",
        "file_id = '1yDcwAMmyeznHD2mZC4CYKqdoRhVZySPO'\n",
        "link=f'https://drive.google.com/uc?export=download&id={file_id}'\n",
        "df = pd.read_csv(link)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VhBRCmRjBP8t",
        "colab_type": "text"
      },
      "source": [
        "## Look at Data (size of data set, what are the features, any null values?)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ypStGEb4QP_Z",
        "colab_type": "code",
        "outputId": "ae9b6da1-0a6a-4850-92a1-4b89c5661231",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 224
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>RowNumber</th>\n",
              "      <th>CustomerId</th>\n",
              "      <th>Surname</th>\n",
              "      <th>CreditScore</th>\n",
              "      <th>Geography</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>15634602</td>\n",
              "      <td>Hargrave</td>\n",
              "      <td>619</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>15647311</td>\n",
              "      <td>Hill</td>\n",
              "      <td>608</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>15619304</td>\n",
              "      <td>Onio</td>\n",
              "      <td>502</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>15701354</td>\n",
              "      <td>Boni</td>\n",
              "      <td>699</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>15737888</td>\n",
              "      <td>Mitchell</td>\n",
              "      <td>850</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   RowNumber  CustomerId   Surname  ...  IsActiveMember EstimatedSalary Exited\n",
              "0          1    15634602  Hargrave  ...               1       101348.88      1\n",
              "1          2    15647311      Hill  ...               1       112542.58      0\n",
              "2          3    15619304      Onio  ...               0       113931.57      1\n",
              "3          4    15701354      Boni  ...               0        93826.63      0\n",
              "4          5    15737888  Mitchell  ...               1        79084.10      0\n",
              "\n",
              "[5 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 173
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nHoEojk-ARO5",
        "colab_type": "code",
        "outputId": "b97476eb-6ebe-4c71-852c-d33786ab1b4d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "df.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10000, 14)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 174
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "STruj8syQPcI",
        "colab_type": "code",
        "outputId": "d9ce88a5-c2cc-425e-808c-8ca64887ec67",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "df['Exited'].mean() # 20% of churned customers"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.2037"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 175
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NsKt6ewV_8ZV",
        "colab_type": "code",
        "outputId": "d4ced29e-7531-4948-8fa2-0fb818200017",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 374
        }
      },
      "source": [
        "df.info()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 10000 entries, 0 to 9999\n",
            "Data columns (total 14 columns):\n",
            " #   Column           Non-Null Count  Dtype  \n",
            "---  ------           --------------  -----  \n",
            " 0   RowNumber        10000 non-null  int64  \n",
            " 1   CustomerId       10000 non-null  int64  \n",
            " 2   Surname          10000 non-null  object \n",
            " 3   CreditScore      10000 non-null  int64  \n",
            " 4   Geography        10000 non-null  object \n",
            " 5   Gender           10000 non-null  object \n",
            " 6   Age              10000 non-null  int64  \n",
            " 7   Tenure           10000 non-null  int64  \n",
            " 8   Balance          10000 non-null  float64\n",
            " 9   NumOfProducts    10000 non-null  int64  \n",
            " 10  HasCrCard        10000 non-null  int64  \n",
            " 11  IsActiveMember   10000 non-null  int64  \n",
            " 12  EstimatedSalary  10000 non-null  float64\n",
            " 13  Exited           10000 non-null  int64  \n",
            "dtypes: float64(2), int64(9), object(3)\n",
            "memory usage: 1.1+ MB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5FVDhJXeCTWg",
        "colab_type": "text"
      },
      "source": [
        "* With 10000 rows this is rather a small data set  \n",
        "* need to get rid of rowNumber and Id and Surname in analysis because they don't provide information\n",
        "*   hasCard, IsActiveMember, Gender, Geography are all categorical, need to deal with it later.\n",
        "*   This is a supervised binary classification problem\n",
        "  * we can try things like SVM, KNN, RF, logistic regression, etc as base model.\n",
        "* There is no missing/null value so we don't need to worry about it.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dcy7lgf-Bdfb",
        "colab_type": "text"
      },
      "source": [
        "##Data exploration/cleaning"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mdgk1-Y9BdiM",
        "colab_type": "code",
        "outputId": "6a5b74f9-522d-4f6d-8bc5-db86cc70a462",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 224
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>RowNumber</th>\n",
              "      <th>CustomerId</th>\n",
              "      <th>Surname</th>\n",
              "      <th>CreditScore</th>\n",
              "      <th>Geography</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>15634602</td>\n",
              "      <td>Hargrave</td>\n",
              "      <td>619</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>15647311</td>\n",
              "      <td>Hill</td>\n",
              "      <td>608</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>15619304</td>\n",
              "      <td>Onio</td>\n",
              "      <td>502</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>15701354</td>\n",
              "      <td>Boni</td>\n",
              "      <td>699</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>15737888</td>\n",
              "      <td>Mitchell</td>\n",
              "      <td>850</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   RowNumber  CustomerId   Surname  ...  IsActiveMember EstimatedSalary Exited\n",
              "0          1    15634602  Hargrave  ...               1       101348.88      1\n",
              "1          2    15647311      Hill  ...               1       112542.58      0\n",
              "2          3    15619304      Onio  ...               0       113931.57      1\n",
              "3          4    15701354      Boni  ...               0        93826.63      0\n",
              "4          5    15737888  Mitchell  ...               1        79084.10      0\n",
              "\n",
              "[5 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QetnIAH4BdnV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# let's get rid of columns we don't want\n",
        "df.drop(['RowNumber', 'CustomerId', 'Surname'],axis=1,inplace=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RH7IC0u7BdqX",
        "colab_type": "code",
        "outputId": "0fc0fbba-0e7d-4626-804b-6160d04b6dc8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>CreditScore</th>\n",
              "      <th>Geography</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>619</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>608</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>502</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>699</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>850</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   CreditScore Geography  Gender  ...  IsActiveMember  EstimatedSalary  Exited\n",
              "0          619    France  Female  ...               1        101348.88       1\n",
              "1          608     Spain  Female  ...               1        112542.58       0\n",
              "2          502    France  Female  ...               0        113931.57       1\n",
              "3          699    France  Female  ...               0         93826.63       0\n",
              "4          850     Spain  Female  ...               1         79084.10       0\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CwL9A2FQFLsp",
        "colab_type": "text"
      },
      "source": [
        "let's look at distribution of salary data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SNNgEz29Fwiw",
        "colab_type": "code",
        "outputId": "6fd09eed-df05-4ed1-ac88-4d3000906c77",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df['EstimatedSalary'].describe()['mean']"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "100090.2398809998"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aOfLc1PDF5Og",
        "colab_type": "code",
        "outputId": "4779dbe0-34c3-4b92-f488-85155fd79743",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "plt.style.use('bmh')\n",
        "plt.figure(figsize=(5, 5))\n",
        "sns.distplot(df['EstimatedSalary'], color='g', bins = 100, hist_kws={'alpha': 0.4})"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fd2e8f3d400>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U2_joxzWGHTj",
        "colab_type": "text"
      },
      "source": [
        "we can see from the plot that the distirbution of salary is pretty uniform and there is no major outliers."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e4LBDNFRHFau",
        "colab_type": "text"
      },
      "source": [
        "now let's look at all numerical variable's distribution"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mdKl-kwHHyvU",
        "colab_type": "code",
        "outputId": "cb587fb1-677a-4a54-c803-0ba63a3ffbbb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "list(set(df.dtypes.to_list())) # to see what data types are there."
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[dtype('O'), dtype('int64'), dtype('float64')]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FRMr0oJuIGcy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df_numerical = df.select_dtypes(include=['float64','int64'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IfeLAKEmIjvE",
        "colab_type": "code",
        "outputId": "1aa0dd43-e701-47d5-b9a6-0404ccd4383f",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df_numerical.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>CreditScore</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>619</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>608</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>502</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>699</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>850</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   CreditScore  Age  Tenure  ...  IsActiveMember  EstimatedSalary  Exited\n",
              "0          619   42       2  ...               1        101348.88       1\n",
              "1          608   41       1  ...               1        112542.58       0\n",
              "2          502   42       8  ...               0        113931.57       1\n",
              "3          699   39       1  ...               0         93826.63       0\n",
              "4          850   43       2  ...               1         79084.10       0\n",
              "\n",
              "[5 rows x 9 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T9uEdUOqIlEt",
        "colab_type": "code",
        "outputId": "a2089dff-35c9-47bb-a896-6ea39bb5ee61",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df_numerical.hist(figsize=(15,15));"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x1080 with 9 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "guk71J0CIw6s",
        "colab_type": "text"
      },
      "source": [
        "*   credit score/age is about a normal distribution, which is what we expect.\n",
        "*   tenure distribution is skewed a little. 1400+ people have stayed about 10 years with our bank! That is a long time. Are they more likely to churn or less likely? we will find out. (There are 400 people who just joined our bank with 0-1 year of tenure) (assuming here tenure means the time the customer has stayed in the bank, and the unit is in years. The data source really doesn't tell us anything about it, so I am assuming here)\n",
        "* **There are many people with just a little balance in the account. That is interesting - maybe they are the ones who will leave the bank? - business insight**\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vISHywoNJv9u",
        "colab_type": "code",
        "outputId": "9ca8752a-dc31-458a-a551-64b1e04c4b74",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# we will explore linear correlations between numerical features.\n",
        "# why? - because some models don't work well with correlated features,\n",
        "# such as linear models.\n",
        "# note that 0 correlatoin != independence.\n",
        "sns.heatmap(df_numerical.corr());"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T6YXOH80KPXJ",
        "colab_type": "code",
        "outputId": "1baea61e-6a85-4eee-de89-836fe44e7800",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df_numerical.corr()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>CreditScore</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>CreditScore</th>\n",
              "      <td>1.000000</td>\n",
              "      <td>-0.003965</td>\n",
              "      <td>0.000842</td>\n",
              "      <td>0.006268</td>\n",
              "      <td>0.012238</td>\n",
              "      <td>-0.005458</td>\n",
              "      <td>0.025651</td>\n",
              "      <td>-0.001384</td>\n",
              "      <td>-0.027094</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Age</th>\n",
              "      <td>-0.003965</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>-0.009997</td>\n",
              "      <td>0.028308</td>\n",
              "      <td>-0.030680</td>\n",
              "      <td>-0.011721</td>\n",
              "      <td>0.085472</td>\n",
              "      <td>-0.007201</td>\n",
              "      <td>0.285323</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Tenure</th>\n",
              "      <td>0.000842</td>\n",
              "      <td>-0.009997</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>-0.012254</td>\n",
              "      <td>0.013444</td>\n",
              "      <td>0.022583</td>\n",
              "      <td>-0.028362</td>\n",
              "      <td>0.007784</td>\n",
              "      <td>-0.014001</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Balance</th>\n",
              "      <td>0.006268</td>\n",
              "      <td>0.028308</td>\n",
              "      <td>-0.012254</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>-0.304180</td>\n",
              "      <td>-0.014858</td>\n",
              "      <td>-0.010084</td>\n",
              "      <td>0.012797</td>\n",
              "      <td>0.118533</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>NumOfProducts</th>\n",
              "      <td>0.012238</td>\n",
              "      <td>-0.030680</td>\n",
              "      <td>0.013444</td>\n",
              "      <td>-0.304180</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.003183</td>\n",
              "      <td>0.009612</td>\n",
              "      <td>0.014204</td>\n",
              "      <td>-0.047820</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>HasCrCard</th>\n",
              "      <td>-0.005458</td>\n",
              "      <td>-0.011721</td>\n",
              "      <td>0.022583</td>\n",
              "      <td>-0.014858</td>\n",
              "      <td>0.003183</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>-0.011866</td>\n",
              "      <td>-0.009933</td>\n",
              "      <td>-0.007138</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>IsActiveMember</th>\n",
              "      <td>0.025651</td>\n",
              "      <td>0.085472</td>\n",
              "      <td>-0.028362</td>\n",
              "      <td>-0.010084</td>\n",
              "      <td>0.009612</td>\n",
              "      <td>-0.011866</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>-0.011421</td>\n",
              "      <td>-0.156128</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <td>-0.001384</td>\n",
              "      <td>-0.007201</td>\n",
              "      <td>0.007784</td>\n",
              "      <td>0.012797</td>\n",
              "      <td>0.014204</td>\n",
              "      <td>-0.009933</td>\n",
              "      <td>-0.011421</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.012097</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Exited</th>\n",
              "      <td>-0.027094</td>\n",
              "      <td>0.285323</td>\n",
              "      <td>-0.014001</td>\n",
              "      <td>0.118533</td>\n",
              "      <td>-0.047820</td>\n",
              "      <td>-0.007138</td>\n",
              "      <td>-0.156128</td>\n",
              "      <td>0.012097</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                 CreditScore       Age  ...  EstimatedSalary    Exited\n",
              "CreditScore         1.000000 -0.003965  ...        -0.001384 -0.027094\n",
              "Age                -0.003965  1.000000  ...        -0.007201  0.285323\n",
              "Tenure              0.000842 -0.009997  ...         0.007784 -0.014001\n",
              "Balance             0.006268  0.028308  ...         0.012797  0.118533\n",
              "NumOfProducts       0.012238 -0.030680  ...         0.014204 -0.047820\n",
              "HasCrCard          -0.005458 -0.011721  ...        -0.009933 -0.007138\n",
              "IsActiveMember      0.025651  0.085472  ...        -0.011421 -0.156128\n",
              "EstimatedSalary    -0.001384 -0.007201  ...         1.000000  0.012097\n",
              "Exited             -0.027094  0.285323  ...         0.012097  1.000000\n",
              "\n",
              "[9 rows x 9 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yj8ULorIKpR1",
        "colab_type": "text"
      },
      "source": [
        "we can see that age and IsActiveMember might have a strong impact on the churn probability of a customer. **Note: correlation only measures linear relationship and we cannot say for sure other variabels don't have strong impact!** "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rR_NsyfpLwc6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "corr = df.drop(columns=['Exited']).corr()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JTa-iUH2K6IX",
        "colab_type": "code",
        "outputId": "3f91c4e4-4dda-4eed-876a-30906fc20468",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        }
      },
      "source": [
        "sns.heatmap(corr[(corr >= 0.2) | (corr <= -0.2)], \n",
        "            cmap='viridis', vmax=1.0, vmin=-1.0, linewidths=0.1,\n",
        "            annot=True, annot_kws={\"size\": 8}, square=True);"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAE9CAYAAAAbNJn3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tXZwfB4-Man9",
        "colab_type": "text"
      },
      "source": [
        "\n",
        "Some ML models require non-correlated features, but since this data set's features doesn't have strong correlations (all correlatoins <= 0.3) so we don't need to worry about that."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "siAPB_nTfM-a",
        "colab_type": "text"
      },
      "source": [
        "Now let's visualize the relationship between target variable and some numerical variables in another way with box-whisker plot (using seaborn)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w2e3GxMtLhGW",
        "colab_type": "code",
        "outputId": "80154636-a8e6-48f3-8dbb-0d05480bdfb1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df.columns"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['RowNumber', 'CustomerId', 'Surname', 'CreditScore', 'Geography',\n",
              "       'Gender', 'Age', 'Tenure', 'Balance', 'NumOfProducts', 'HasCrCard',\n",
              "       'IsActiveMember', 'EstimatedSalary', 'Exited'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4Tqagyw1MbUk",
        "colab_type": "code",
        "outputId": "77d09c9e-2fe2-4130-9987-addc7c7da902",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "_ ,axes = plt.subplots(2, 3, figsize = (20,10))\n",
        "sns.boxplot(x='Exited', y ='CreditScore', data=df,ax=axes[0][0])\n",
        "sns.boxplot(x='Exited', y ='Age', data=df,ax=axes[0][1])\n",
        "sns.boxplot(x='Exited', y ='Tenure', data=df,ax=axes[0][2])\n",
        "sns.boxplot(x='Exited', y ='Balance', data=df,ax=axes[1][0])\n",
        "sns.boxplot(x='Exited', y ='NumOfProducts', data=df,ax=axes[1][1])\n",
        "sns.boxplot(x='Exited', y ='EstimatedSalary', data=df,ax=axes[1][2])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f79a4822b38>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABKUAAAJNCAYAAADgesaeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSA2z/AAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1440x720 with 6 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fz_hZ57xfyc2",
        "colab_type": "text"
      },
      "source": [
        "Notice that age distribution is a lot different for churned/non-churned customers! That will be worthwhile to investigate in the future"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZlxHV7GxC4iC",
        "colab_type": "text"
      },
      "source": [
        "#Now let's look at categorical data!"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0_f4GaIAG123",
        "colab_type": "code",
        "outputId": "bc4e70a3-c331-4ce2-ca81-ac6afdb8a0c1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>CreditScore</th>\n",
              "      <th>Geography</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>619</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>608</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>502</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>699</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>850</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   CreditScore Geography  Gender  ...  IsActiveMember  EstimatedSalary  Exited\n",
              "0          619    France  Female  ...               1        101348.88       1\n",
              "1          608     Spain  Female  ...               1        112542.58       0\n",
              "2          502    France  Female  ...               0        113931.57       1\n",
              "3          699    France  Female  ...               0         93826.63       0\n",
              "4          850     Spain  Female  ...               1         79084.10       0\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r_oXOTL-ENwA",
        "colab_type": "code",
        "outputId": "6c218bf9-0b70-4c73-e74d-91da86a65ab2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "sns.countplot(data=df, hue='Exited', x = 'Geography')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fd2e302bc18>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rvyk8IFOGxCh",
        "colab_type": "code",
        "outputId": "522d7653-ddcd-48c9-e0d7-7a3183b7c761",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "sns.countplot(data=df, hue='Exited', x= 'Gender')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fd2e2f66dd8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Gv0tmv5bG6sV",
        "colab_type": "code",
        "outputId": "c56c0e9b-b607-4c36-a7e2-506cea259562",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "sns.countplot(data=df, hue='Exited', x= 'IsActiveMember')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fd2e84c56a0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r0PnEVl7G7wP",
        "colab_type": "code",
        "outputId": "6545bf64-6863-4a93-f3d6-06a42c29a043",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "sns.countplot(data=df, hue='Exited', x= 'HasCrCard')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fd2e855f358>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "daETAt6oG9bg",
        "colab_type": "text"
      },
      "source": [
        "the relative frequencies of 0(didn't exit) and 1(exit) seems to be about the same across many different categorical vairbales. This means that maybe\n",
        "the categorical variabels aren't gonna help us very much in terms of predicting\n",
        "the target variable."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CkDCmhWGHMs1",
        "colab_type": "text"
      },
      "source": [
        "# Data preprocessing for models\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MJr-kVeNYVgF",
        "colab_type": "text"
      },
      "source": [
        "Deal with categorical variables - one hot encoding(gender, geography, hasCrCard, IsActiveMember). I think one hot encoding does the job here since\n",
        "we don't have many categories for each variable. (3 categories for geography, 2 categories for gender). If there are a lot of categoreis I will use other encoding methods."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fDB8DXpGIMdP",
        "colab_type": "code",
        "outputId": "bb949da0-06e0-4aeb-a78c-529d3ebcde49",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>CreditScore</th>\n",
              "      <th>Geography</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>619</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>608</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>502</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>699</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>850</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   CreditScore Geography  Gender  ...  IsActiveMember  EstimatedSalary  Exited\n",
              "0          619    France  Female  ...               1        101348.88       1\n",
              "1          608     Spain  Female  ...               1        112542.58       0\n",
              "2          502    France  Female  ...               0        113931.57       1\n",
              "3          699    France  Female  ...               0         93826.63       0\n",
              "4          850     Spain  Female  ...               1         79084.10       0\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NmJBGgv7QtmK",
        "colab_type": "code",
        "outputId": "5e452d7d-ab91-4673-d67c-7a45d3b88161",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>CreditScore</th>\n",
              "      <th>Geography</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>619</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>608</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>502</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>699</td>\n",
              "      <td>France</td>\n",
              "      <td>Female</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>850</td>\n",
              "      <td>Spain</td>\n",
              "      <td>Female</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   CreditScore Geography  Gender  ...  IsActiveMember  EstimatedSalary  Exited\n",
              "0          619    France  Female  ...               1        101348.88       1\n",
              "1          608     Spain  Female  ...               1        112542.58       0\n",
              "2          502    France  Female  ...               0        113931.57       1\n",
              "3          699    France  Female  ...               0         93826.63       0\n",
              "4          850     Spain  Female  ...               1         79084.10       0\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "crekVqzyQ0IL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y = df['Exited'] # target variable for prediction"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "evr6sI0PQ_c1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df['Gender'] = df['Gender'] == 'Female' # replace Female by 1 and male by 0"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TxBuvS7FRMHS",
        "colab_type": "code",
        "outputId": "56199e7f-4e12-4140-9e4f-74b6ef4c0b43",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>CreditScore</th>\n",
              "      <th>Geography</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>619</td>\n",
              "      <td>France</td>\n",
              "      <td>True</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>608</td>\n",
              "      <td>Spain</td>\n",
              "      <td>True</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>502</td>\n",
              "      <td>France</td>\n",
              "      <td>True</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>699</td>\n",
              "      <td>France</td>\n",
              "      <td>True</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>850</td>\n",
              "      <td>Spain</td>\n",
              "      <td>True</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   CreditScore Geography  Gender  ...  IsActiveMember  EstimatedSalary  Exited\n",
              "0          619    France    True  ...               1        101348.88       1\n",
              "1          608     Spain    True  ...               1        112542.58       0\n",
              "2          502    France    True  ...               0        113931.57       1\n",
              "3          699    France    True  ...               0         93826.63       0\n",
              "4          850     Spain    True  ...               1         79084.10       0\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 32
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Rch-poD1RNhQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df = pd.get_dummies(df, columns=['Geography'], drop_first = True)\n",
        "# create two more columns for Geography column. - one hot encoding"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V5rJmpHkRzu3",
        "colab_type": "code",
        "outputId": "a5c84a69-e7b7-4c2a-d70a-7171ca4d36e8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>CreditScore</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Exited</th>\n",
              "      <th>Geography_Germany</th>\n",
              "      <th>Geography_Spain</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>619</td>\n",
              "      <td>True</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>608</td>\n",
              "      <td>True</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>502</td>\n",
              "      <td>True</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>699</td>\n",
              "      <td>True</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>850</td>\n",
              "      <td>True</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   CreditScore  Gender  Age  ...  Exited  Geography_Germany  Geography_Spain\n",
              "0          619    True   42  ...       1                  0                0\n",
              "1          608    True   41  ...       0                  0                1\n",
              "2          502    True   42  ...       1                  0                0\n",
              "3          699    True   39  ...       0                  0                0\n",
              "4          850    True   43  ...       0                  0                1\n",
              "\n",
              "[5 rows x 12 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sTWNKupDSLRA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X = df.drop(['Exited'], axis=1) # drop the target variable column to create feature matrix"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fHi_uCvjSTDL",
        "colab_type": "code",
        "outputId": "ea44651f-17ad-4d67-d66a-b84621581d2b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "X.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>CreditScore</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Tenure</th>\n",
              "      <th>Balance</th>\n",
              "      <th>NumOfProducts</th>\n",
              "      <th>HasCrCard</th>\n",
              "      <th>IsActiveMember</th>\n",
              "      <th>EstimatedSalary</th>\n",
              "      <th>Geography_Germany</th>\n",
              "      <th>Geography_Spain</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>619</td>\n",
              "      <td>True</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>101348.88</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>608</td>\n",
              "      <td>True</td>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>83807.86</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>112542.58</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>502</td>\n",
              "      <td>True</td>\n",
              "      <td>42</td>\n",
              "      <td>8</td>\n",
              "      <td>159660.80</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113931.57</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>699</td>\n",
              "      <td>True</td>\n",
              "      <td>39</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>93826.63</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>850</td>\n",
              "      <td>True</td>\n",
              "      <td>43</td>\n",
              "      <td>2</td>\n",
              "      <td>125510.82</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79084.10</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   CreditScore  Gender  ...  Geography_Germany  Geography_Spain\n",
              "0          619    True  ...                  0                0\n",
              "1          608    True  ...                  0                1\n",
              "2          502    True  ...                  0                0\n",
              "3          699    True  ...                  0                0\n",
              "4          850    True  ...                  0                1\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1olpZGTuSUBh",
        "colab_type": "code",
        "outputId": "0052e113-d36f-48de-8292-81b417fdd068",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "y.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    1\n",
              "1    0\n",
              "2    1\n",
              "3    0\n",
              "4    0\n",
              "Name: Exited, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hv3zJiqeMn7J",
        "colab_type": "code",
        "outputId": "704ce748-eeff-4f73-dec9-4bec8fc266e6",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "len(X), len(y) # check the length matches."
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10000, 10000)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8rjMqTaGSYy7",
        "colab_type": "text"
      },
      "source": [
        "# Model Training and Result Evaluation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GyQYP6wkSgNM",
        "colab_type": "text"
      },
      "source": [
        "## Part1: split dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zw4MJcYHTEZi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn import model_selection"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MMJ7dCLgTgZl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_train, X_test, y_train, y_test = model_selection.train_test_split(X, y, test_size = 0.25, stratify = y)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_1QWixsRUQNN",
        "colab_type": "code",
        "outputId": "22c667ac-a60b-4103-da63-cf9005207872",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "X_train.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(7500, 11)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MoPhDYMmUbVX",
        "colab_type": "code",
        "outputId": "91b73223-c986-4c63-c74d-dcc1d9a92e73",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "X_test.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(2500, 11)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "asha92vzUdxa",
        "colab_type": "code",
        "outputId": "a4920e43-7d42-4057-be81-005fe947c852",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "y_train.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(7500,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hdZ97jy3UgWO",
        "colab_type": "code",
        "outputId": "90998c0a-63d2-4816-89f2-9e7c076b611b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "y_test.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(2500,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EgxbXEYtUhqz",
        "colab_type": "text"
      },
      "source": [
        "##Part 2. standardization"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "m28UxodeUtDJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.preprocessing import StandardScaler\n",
        "scalar = StandardScaler()\n",
        "scalar.fit(X_train)\n",
        "X_train = scalar.transform(X_train)\n",
        "X_test = scalar.transform(X_test)\n",
        "# same standardization must apply both to training and testing set"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "S2HgUeOZVmEC",
        "colab_type": "text"
      },
      "source": [
        "##Part 3. Base Model Training and Selection"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pOLpbcJEY52c",
        "colab_type": "text"
      },
      "source": [
        "\n",
        "1.   This is a binary classification problem, so SVM is probably the best off-the-shelf model. It utilizes the notion of distance. I will definitely try this model first. \n",
        "2.   logistic regression is used because we might want probabilistic interpretation of results. For example, for a customer with 0.9 probability of churning we might use more agressive strategy like giving him/her more credits or something, compared to a customer with 0.6 probability of churning.\n",
        "\n",
        "  I know SVM also can also output probability, I don't know how it does that, but that is not honest probability in a mathematical sense.\n",
        "  \n",
        "  Draw back is logistic regression prones to overfit, so we need to tune regularization constant.\n",
        "3.   Picked random forest because it is very powerful, not easy to overfit, and it usually has good results. Also has the benefit of gaining feature importance from this model.\n",
        "4. K Nearest Neighbors is a simple model also, which I suspect is going to underperform with respect to random forest.\n",
        "\n",
        "**Note**: This is a relatively simple data set with 10000 data points, and 10+ features, so I will not bother with complex models like XgBoost or neural net. Neural net is not desirable here because we want to not only know our customer will exit, but why is the case, so we can prevent it. And neural net is terrible at providing explainability"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Pz69Bco_V4cr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.svm import SVC\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "\n",
        "logisticClassifier = LogisticRegression()\n",
        "KNNClassifier = KNeighborsClassifier()\n",
        "RFClassifier = RandomForestClassifier()\n",
        "svc = SVC()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fiOoDNGuW1NL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# get 5-fold cross validation score of different models\n",
        "model_names = ['SVC', 'Logistic Regression', 'Random Forest', 'KNN']\n",
        "model_list =  [ svc, logisticClassifier, RFClassifier, KNNClassifier]\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FFV7PphWiuXN",
        "colab_type": "code",
        "outputId": "2be3538a-9fc8-446d-b62e-cf6b5f700b5f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        }
      },
      "source": [
        "for index, model in enumerate(model_list):\n",
        "  model_name = model_names[index]\n",
        "  cv_score = model_selection.cross_val_score(model, X_train, y_train, cv = 5)\n",
        "  print(\"scores of \" + model_name + \" are:\", end = \"\")\n",
        "  print(cv_score)\n",
        "  print(\"mean cross validation accuracy of \" + model_name + \" is \" + str(cv_score.mean()))\n",
        "  print()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "scores of SVC are:[0.85866667 0.85866667 0.85133333 0.84       0.86133333]\n",
            "mean cross validation accuracy of SVC is 0.8539999999999999\n",
            "\n",
            "scores of Logistic Regression are:[0.81       0.80733333 0.80666667 0.80333333 0.816     ]\n",
            "mean cross validation accuracy of Logistic Regression is 0.8086666666666666\n",
            "\n",
            "scores of Random Forest are:[0.864      0.868      0.858      0.83733333 0.86466667]\n",
            "mean cross validation accuracy of Random Forest is 0.8583999999999999\n",
            "\n",
            "scores of KNN are:[0.84266667 0.82733333 0.82266667 0.81666667 0.832     ]\n",
            "mean cross validation accuracy of KNN is 0.8282666666666666\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qfWDAsOnkFsf",
        "colab_type": "text"
      },
      "source": [
        "As we can see, SVM and random forest unsurprising did better than simple KNN or logistic regression. Next, we are going to fine tune hyperparameters for each model. Our model seems to not work so well - even if I guess all customers to not churn, without any information, I will get 0.8 accuracy, since there are 20% who churned."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z3tbK6KXZ9am",
        "colab_type": "text"
      },
      "source": [
        "## Part 4: finding optimal hyperparameters of each model using Grid Search"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XpIi0rNOaPSo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.model_selection import GridSearchCV"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eEVrDLbRbDL_",
        "colab_type": "text"
      },
      "source": [
        "possible parameters for logistic regression: \n",
        "1. $l_1$ or $l_2$ penalty - LASSO or ridge\n",
        "2. C is the inverse of regularization strength $\\frac{1}{\\lambda}$ "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Av84XxasbbHN",
        "colab_type": "code",
        "outputId": "b63073bb-931f-44ef-f82b-f8b134b40128",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 751
        }
      },
      "source": [
        "# linear regression Grid Search\n",
        "parameters = {\n",
        "    'penalty':('l1','l2', 'elasticnet'),\n",
        "    'C':(0.01, 0.05, 0.1, 0.2 ,1.5,10,25)\n",
        "}\n",
        "Grid_LR = GridSearchCV(LogisticRegression(solver='liblinear'),parameters, cv = 5)\n",
        "Grid_LR.fit(X_train, y_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/model_selection/_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
            "ValueError: Only 'saga' solver supports elasticnet penalty, got solver=liblinear.\n",
            "\n",
            "  FitFailedWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/sklearn/model_selection/_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
            "ValueError: Only 'saga' solver supports elasticnet penalty, got solver=liblinear.\n",
            "\n",
            "  FitFailedWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/sklearn/model_selection/_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
            "ValueError: Only 'saga' solver supports elasticnet penalty, got solver=liblinear.\n",
            "\n",
            "  FitFailedWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/sklearn/model_selection/_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
            "ValueError: Only 'saga' solver supports elasticnet penalty, got solver=liblinear.\n",
            "\n",
            "  FitFailedWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/sklearn/model_selection/_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
            "ValueError: Only 'saga' solver supports elasticnet penalty, got solver=liblinear.\n",
            "\n",
            "  FitFailedWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/sklearn/model_selection/_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
            "ValueError: Only 'saga' solver supports elasticnet penalty, got solver=liblinear.\n",
            "\n",
            "  FitFailedWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/sklearn/model_selection/_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
            "ValueError: Only 'saga' solver supports elasticnet penalty, got solver=liblinear.\n",
            "\n",
            "  FitFailedWarning)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GridSearchCV(cv=5, error_score=nan,\n",
              "             estimator=LogisticRegression(C=1.0, class_weight=None, dual=False,\n",
              "                                          fit_intercept=True,\n",
              "                                          intercept_scaling=1, l1_ratio=None,\n",
              "                                          max_iter=100, multi_class='auto',\n",
              "                                          n_jobs=None, penalty='l2',\n",
              "                                          random_state=None, solver='liblinear',\n",
              "                                          tol=0.0001, verbose=0,\n",
              "                                          warm_start=False),\n",
              "             iid='deprecated', n_jobs=None,\n",
              "             param_grid={'C': (0.01, 0.05, 0.1, 0.2, 1.5, 10, 25),\n",
              "                         'penalty': ('l1', 'l2', 'elasticnet')},\n",
              "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
              "             scoring=None, verbose=0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 99
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oYJkqr16dsHC",
        "colab_type": "code",
        "outputId": "ad9cf08d-030f-4bcc-c43b-a5208de6cdca",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "Grid_LR.best_score_"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8097333333333333"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 100
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D9AtrTPWd2Tq",
        "colab_type": "code",
        "outputId": "918a778e-3582-4985-f8ef-45397b1b3191",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "Grid_LR.best_params_ \n",
        "# seems like our best model requires pretty strong regularization - C is inverse of regularization strength."
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'C': 0.05, 'penalty': 'l1'}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 101
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "s-X6Z_gNeObk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "best_LR_model = Grid_LR.best_estimator_"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OxuTleslOTkx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# KNN grid search\n",
        "parameters = {\n",
        "    'n_neighbors': [3,5,7,9,11],\n",
        "    'p': [1,2,3], # p is the l_p norm metric choice, 2 is euclidean\n",
        "    'weights':['uniform','distance']\n",
        "}"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZSx9iCzKOrbM",
        "colab_type": "code",
        "outputId": "6b760ff1-10d5-47cb-a2a2-d1f2a4175297",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "Grid_KNN = GridSearchCV(KNeighborsClassifier(),parameters, cv = 5)\n",
        "Grid_KNN.fit(X_train, y_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GridSearchCV(cv=5, error_score=nan,\n",
              "             estimator=KNeighborsClassifier(algorithm='auto', leaf_size=30,\n",
              "                                            metric='minkowski',\n",
              "                                            metric_params=None, n_jobs=None,\n",
              "                                            n_neighbors=5, p=2,\n",
              "                                            weights='uniform'),\n",
              "             iid='deprecated', n_jobs=None,\n",
              "             param_grid={'n_neighbors': [3, 5, 7, 9, 11], 'p': [1, 2, 3],\n",
              "                         'weights': ['uniform', 'distance']},\n",
              "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
              "             scoring=None, verbose=0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 104
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ds3i9rOTQ31s",
        "colab_type": "code",
        "outputId": "1508b8ae-473f-4402-f31f-d03aeed879f2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "Grid_KNN.best_params_"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'n_neighbors': 9, 'p': 1, 'weights': 'uniform'}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 105
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CHnKIqDFQ_rQ",
        "colab_type": "code",
        "outputId": "ee6e2971-2e1f-4844-c0b3-671750029b33",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "Grid_KNN.best_score_"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8361333333333333"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 106
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iQRmdFwTRUXv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "best_KNN_model = Grid_KNN.best_estimator_"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "__KtdtgaQOXb",
        "colab_type": "code",
        "outputId": "275cb829-7cf5-418f-835d-82eec3e3dc0e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 374
        }
      },
      "source": [
        "# RF Grid Search\n",
        "parameters = {\n",
        "    'n_estimators' : [40,60,80,100,200],\n",
        "    'criterion': ['gini', 'entropy']\n",
        "}\n",
        "Grid_RF = GridSearchCV(RandomForestClassifier(), parameters, cv = 5)\n",
        "Grid_RF.fit(X_train, y_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GridSearchCV(cv=5, error_score=nan,\n",
              "             estimator=RandomForestClassifier(bootstrap=True, ccp_alpha=0.0,\n",
              "                                              class_weight=None,\n",
              "                                              criterion='gini', max_depth=None,\n",
              "                                              max_features='auto',\n",
              "                                              max_leaf_nodes=None,\n",
              "                                              max_samples=None,\n",
              "                                              min_impurity_decrease=0.0,\n",
              "                                              min_impurity_split=None,\n",
              "                                              min_samples_leaf=1,\n",
              "                                              min_samples_split=2,\n",
              "                                              min_weight_fraction_leaf=0.0,\n",
              "                                              n_estimators=100, n_jobs=None,\n",
              "                                              oob_score=False,\n",
              "                                              random_state=None, verbose=0,\n",
              "                                              warm_start=False),\n",
              "             iid='deprecated', n_jobs=None,\n",
              "             param_grid={'criterion': ['gini', 'entropy'],\n",
              "                         'n_estimators': [40, 60, 80, 100, 200]},\n",
              "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
              "             scoring=None, verbose=0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 108
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gzDZf1apRrAP",
        "colab_type": "code",
        "outputId": "0126b9cb-d1e7-49c9-eea9-4b96a60c02f2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "Grid_RF.best_params_"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'criterion': 'entropy', 'n_estimators': 40}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 109
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Wpc9kNcxSDOA",
        "colab_type": "code",
        "outputId": "c854f5e8-f269-47c1-fc60-8e690044b777",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "Grid_RF.best_score_"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8620000000000001"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 110
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JTO6PeE7SF6m",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "best_RF_model = Grid_RF.best_estimator_"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c3xVk7W_SPlK",
        "colab_type": "code",
        "outputId": "00b3c850-4331-43d4-8e44-8cf79625bf79",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        }
      },
      "source": [
        "# SVM Grid Search\n",
        "parameters = {\n",
        "    'kernel':('rbf', 'poly'),\n",
        "    'degree':(3, 5, 10), # degree of polynomial\n",
        "    'gamma':('scale', 0.01, 0.1), # Kernel coefficient\n",
        "    'C':(0.01, 0.1, 1, 5),\n",
        "    'probability':[True] # want to plot it on ROC later, so turn this on.\n",
        "}\n",
        "Grid_SVM = GridSearchCV(SVC(),parameters, cv = 5) # this takes a long time - 25 minutes\n",
        "Grid_SVM.fit(X_train, y_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GridSearchCV(cv=5, error_score=nan,\n",
              "             estimator=SVC(C=1.0, break_ties=False, cache_size=200,\n",
              "                           class_weight=None, coef0=0.0,\n",
              "                           decision_function_shape='ovr', degree=3,\n",
              "                           gamma='scale', kernel='rbf', max_iter=-1,\n",
              "                           probability=False, random_state=None, shrinking=True,\n",
              "                           tol=0.001, verbose=False),\n",
              "             iid='deprecated', n_jobs=None,\n",
              "             param_grid={'C': (0.01, 0.1, 1, 5), 'degree': (3, 5, 10),\n",
              "                         'gamma': ('scale', 0.01, 0.1),\n",
              "                         'kernel': ('rbf', 'poly'), 'probability': [True]},\n",
              "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
              "             scoring=None, verbose=0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 138
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cJg8zXTknW1M",
        "colab_type": "code",
        "outputId": "d88ff734-92c4-4533-f7eb-a959d16a803a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "Grid_SVM.best_params_"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'C': 5, 'degree': 3, 'gamma': 'scale', 'kernel': 'rbf', 'probability': True}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 139
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wBCClwZFncfn",
        "colab_type": "code",
        "outputId": "7128f040-f0aa-4293-bc72-ee36152a973c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "Grid_SVM.best_score_"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8554666666666666"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 140
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mNTDjxQGndU4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "best_SVM_model = Grid_SVM.best_estimator_"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VQ61NpHjSVGd",
        "colab_type": "text"
      },
      "source": [
        "# Model Evaluation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GzhDTb9HuXlF",
        "colab_type": "text"
      },
      "source": [
        "Our exact evaluation metric will depend on our boss's opinion, but I think in this scenario, we need to maximize recall, since we really care about finding all the people who are gonna churn, so we can keep them here by some strategies. Yes we are going to predict people are actually not churning to churn, that is just gonna cost us some money to make them happier, but the other case of missing a person who will churn will definitely cost more profit in the long run.\n",
        "\n",
        "other possible metrics include: accuracy, F1, precision, AUC"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HUHPPqDi6_Vv",
        "colab_type": "text"
      },
      "source": [
        "## best Model with AUC - Random Forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uIqYSAf8Sb0x",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn import metrics"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OHDLq5rYbxqN",
        "colab_type": "code",
        "outputId": "ab0d9173-c30d-409e-f501-2a1124b4e659",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "confusion_matrix(y_test, best_RF_model.predict(X_test))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[1924,   67],\n",
              "       [ 265,  244]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 121
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZSzRrtjAcSw1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_pred_rf = best_RF_model.predict_proba(X_test)[:, 1]\n",
        "fpr_rf, trp_rf, _ = roc_curve(y_test, y_pred_rf)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C83dPnZ3cbMZ",
        "colab_type": "code",
        "outputId": "8433802d-a836-4868-bb17-b7d94638dbdf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_pred_rf"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.2  , 0.15 , 0.8  , ..., 0.025, 0.2  , 0.05 ])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 123
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fU4EckwKzbFg",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_pred_lr = best_LR_model.predict_proba(X_test)[:, 1]\n",
        "fpr_lr, trp_lr, _ = roc_curve(y_test, y_pred_lr)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "haEBn6W-zdNx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_pred_knn = best_KNN_model.predict_proba(X_test)[:, 1]\n",
        "fpr_knn, trp_knn, _ = roc_curve(y_test, y_pred_knn)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lw8g7UiDdER0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_pred_svm = best_SVM_model.predict_proba(X_test)[:, 1]\n",
        "fpr_svm, trp_svm, _ = roc_curve(y_test, y_pred_svm)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4ZO08J1rdTLN",
        "colab_type": "code",
        "outputId": "2d41f7f2-6c30-45fa-9e4c-00667d312d86",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        }
      },
      "source": [
        "plt.figure(1)\n",
        "plt.plot([0,1],[0,1],'k--')\n",
        "plt.plot(fpr_rf, trp_rf, label=\"RF\")\n",
        "plt.plot(fpr_lr, trp_lr, label=\"LR\")\n",
        "plt.plot(fpr_svm, trp_svm, label=\"SVM\")\n",
        "plt.plot(fpr_knn, trp_knn, label=\"KNN\")\n",
        "plt.xlabel('False positive rate')\n",
        "plt.ylabel('True positive rate')\n",
        "plt.title('ROC curve of various models')\n",
        "plt.legend(loc = 'best')\n",
        "plt.savefig('roc.png')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAEXCAYAAABLZvh6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOydeXxTVdrHvydJl3RfgbKVrSwqorIJCsqmKA6OzujguM6r44yiuO8OoqLiqIi8rjPuu87r6DiKC6u4ALIJshXK2kK3dEmbNnvO+8dNQ1qSNim5bZrk+/nkk9x7z733/HKS85x7znOeI6SUxIgRI0aM6ETT2RmIESNGjBidR8wIxIgRI0YUEzMCMWLEiBHFxIxAjBgxYkQxMSMQI0aMGFFMzAjEiBEjRhQTMwIxYvhACHG2EGKbEMIuhFjVCfe/Rgjh6Oj7djbt0S2EmCeEKFIrT5FOzAh0MYQQbwohpPvlFEKUCCHeFkL08pF2oDv9YSGETQhxRAjxlhBioI+0SUKIB4UQW4UQjUKIaiHEOiHEzUKIpI5RF1a8BGwCBgAXd8L9PwKOKdMYMUJNzAh0Tb4H8oC+wB+BU4F/eScQQpwKbAB6u9MMAmYBPYENQohTvNKmAT8CNwMvAOOBkcDTwKXAOerKaY4QIr4j7+eHAmCplLJYSlndUTcVCnFSSrOUsryj7hsjipFSxl5d6AW8CSxrse9mQAJp7m0BbAG2AroWaXXAr8AvgHDv+1/ADPT3cT8BZLSSnxRgEVAMWIEDwP3uY/3c+TqzxTlFwDyvbQnMAd4HjCit4B+Bf/i4305gvtf2LLcWi/veC4HkNr7DIcCXgMn9+i8wyH3sbHd+vF/X+LjGNMAJ9G6x/w9Ao1dZPObOc6P7O3oZSPdKfw3gACYBmwEbcF7T/hbXPh/Y6P6eK4AXvbX6+W1cofzNPdu9gU8Ag/s72wfc1cp31fR9nA+scf9ONgInul8/uLX9DJwQZH41wKPuYyZ3ud/mQ/c09+/BDBwG3gCyvY7PA4raqzHaX52egdgryAJr8UdHadl/565Ikt37Rrj/uFf4ucaV7uMnu/+I1cCr7ciLAFa5/2S/Rek6mQj82X28H4EbgSrgJmAgSiv8eqAGSPBKN8addrB7+xp3miu97r0VeKeVPOuBg8BylKedkcBKd57i3a8e7vvMdn/W+7iOBigB7mmxfwnwvtf2g8AE93cxBdgFvOV1/BrA5a5EJ7l15NLCCLjLygE8CwxFMRSHvLW2/G2497U0Ap8Dy4BT3HmaBFzWyvd1tvu72AxMBk5AMQZbgdVuTcNQjMG6IPN7C9AAXA0MBu4GalvonoxiZG52/y5Gu8vrO442YubR3AgEpTHaX52egdgryAJT/ugOlJZTI0dbq097pbnUve9UP9c4zX38EqCb+/Pt7cjLFPe5o/wc70fgRuC1FmkyUFp+l3jtex5Y47V9APhri/Mmuq+X6SdP17q/txyvfd3d97qqRZ58GlGvNAuAbS2u4wDObeWci1Baxhr39jXue01oke6aFpXhO8DPLdJciGJA8r1+G20ZgS3e330AZXy2O3+/9dp3iXvf71rokkBKEPktAR5rkeb/WuheBSxokaav+16nuLfn0dwIBKUx2l+xMYGuyTqUVs4YlMfpNSgtzvYgjiMfI4EaKeWG47hGEz97b0gpa1FadFcCCCHiULp+3nZv5wL5wEIhhKnpBXzlvsQgP/c5EdghpTR43ascKHQfC4a3gBOFEKe5ty9H6dpY1pRACHGxEGK1e1DeBLzH0acNb9a3ca8TUVre3nyHUn4nBJHnRcD97kH/J4UQEwM8b4vX5zL3+1Yf+7oFkl/3OFQv4KcWaX5osT0auLVFGe9wHyvwk9f2aoxKYkaga2KWUhZJKbdJKecC+1H69ZvY7X4/yc/5TZVdIVCJ0qUSTEUSKC73e0tDE+cjbYOPfW8D090V/gyU8YcP3ceafru3oBjEptcIlMrh1/ZnOzCklDtRBt+vcu+6CnhXSukEEEKMRRmwX43SUj4N+Ks7rffgt1NKaQlBlly08V1LKd9AMZ4vozgXfCWEeDeAa9u9L9PKvlDXKRrgSZqX8SkoZfyVrxOOQ2NUEjMCkcE84E9CiFHu7S3ANuAuIYTOO6F7+y6UVtyvUkoXyoDs5UKI/i0v7PZWSfdz341Aptd9W1Lpfu/pdb1uBO76+A3KeMUslAr2CyllDXha78XAELdBbPnyV6luR2mJ5njlqTvKYPG2APPlzVvAZe6ngRG4n1TcnAkYpJQPSinXSSl3owxatoftKF1d3pyFUvlud29X4PVduzmtxTZSylIp5RtSyqtQuscud7fMQ0mr+ZVS1qEM8o5vkeaMFtsbgBP9lLHJ3807SGNk0Nn9UbFXcC989Pu6938KfOO1PRLF02Ypyp+xD8oA5bcog2+neqVNRzEK5SgDsiOA/iit19V49Qe3uKdwH9+L0t/bH+VPfJ1Xmh9QjMUId56+Rmn1z/NK09og9kKUx38r8JsWx65E8aZ5AOWpZwjKAPUrrXx/3gPDp9FiYDiQPLW4XrY7D5uBjS2OXYDSOr8WZcD3KpR+cAn0c6e5hhbeML72c+xA63SOHWidytEB7YHAn91lKr3SPI/itTMQ5YnwY/d1hB99Z7uv2dtr35neGtz7TnfvGxREfm9DGdu6EqVlfwfKU6m37kkoTxwLUZ4ABrqv9RruAXuOHRMISmO0vzo9A7FXkAXm3wiMd/8Jz/baV4DSUj3i/iOVurcH+jg/GZiL0o1idv8Z17krlGO8Y7zOS0Xpiip1V4b7gXu9jg9G6QtuAPagTLzyNTDszwg0eTpV0MLd1X38tyhjIo1AHYq76Nw2vsMhKF48TS6iXzRVXoHkycf1PnWnv8XHsUdRKuIG9z0vox1GwL3P2+WyEmVCW3KLNA+gtLBNwAfu8pNex19A6S40o3hkfYnS0van7WzaYQQCyS9KT8TjKK6cDSiDwr5cRCegjLPUu9PtROn317mPz6O5EQhKY7S/mlysYsSIESNGFBIbE4gRI0aMKCZmBGLEiBEjiokZgRgxYsSIYmJGIEaMGDGimJgRiBEjRowoRtd2kvBi1apVMiEhobOzESNGjBhdhsbGRsOUKVNyfR3rckYgISGBoUOHtuvcgwcPkp+fH+IchTcxzZFPtOmFmOZg2bRp00F/x6KqOyguzlfImsgmpjnyiTa9ENMcSqLKCKSn+wuBE7nENEc+0aYXYppDSVQZAYPB0HaiCCOmOfKJNr0Q0xxKosoIxFoP0UG0aY42vRDTHEo6xAgIIV4XQlQIIXyG6nWHK14shCgSQmz1WqQjpNhsNjUuG9bENEc+0aYXYppDSUc9CbyJEv7VH+ehRLxsWlv2JTUyYTab1bhsWBPTHPlEm16IaQ4lHeIiKqVcLYTo10qSC4G3pRLSdK0QIkMIkSelLA1lPnr0aLmiX+QT0xz5RJteiFzNUkpsTonV4cLicGF1uCgrq+bQkX0Y6ouZld2N9BR9SO8ZLvMEeqGsEtVEiXtfSI1AWVlZ1PkWxzRHPtGmFzpHs9PVvHJuerc6XFid3tuy+bGmtHYnNqMFKupxuKTn5XRJ7E4XNpcBm6sCFxUkOMpxUoFNY8AmjCCUkP/DBwxn/HB/q8a2j3AxAgFTUVHBtddei06nw+l0cvHFFzN79mzKyspITk5Gq9VSV1dHbm4u1dXVSCnJzc2lvLwcu91OVVUVJpOJ7t27U1lZiRCCrKwsKisrSUtLw+l00tDQQI8ePSgrKyMuLo709HQMBgPp6enYbDbMZrPneHx8PKmpqVRVVZGZmYnZbMZisXiOJyYmotfrqampITs7m/r6emw2m+e4Xq8nPj4eo9FITk4ORqMRu93uOd6WppSUFAC/mhwOBwaDIaI0tVVOTqeTioqKiNLUWjmZTCbq6uoiSlNb5WQymTCZTH41Gapr0aZkcNhgpKbRRmJKOhXVtaCNwy4FdY1mdAl6jA1mbA6JiIunrsGCU2iwOSRmuxOn0GC2OdE12tBZnbi81l7pWW9BSHnMis7dSw6SbKpvtk8DJAGOvgXoRD1WTRVWrQGLxuB+r8KqqUYKp+cci1fNLKSGeGcWyY1J2O1Wjhw50q5y8keHLSrj7g76Qkp5jBkTQrwCrJJSfuDeLkRZIeuYJ4E1a9bI9s4YNplMbX4hkUZMc+QTDXqllDTYnNRaHNSYHZTW1GOROmrNDmrNDmrMds+xWrOdRrsr6Hvo7U4mFBuwawRNtWK8K/j60SHMSgWvqcKidb9rDFi0BlzC/+BuoiWRZHMSSeZkkhuTSW5MIq+whJRaSUJ9HVlnjWHsR4uCzg/Apk2bNk6ZMsXnWuDh8iTwOXCTEOJDYCxgDPV4AEBVVVXE/1laEtMc+XRVvU6XpK6p4rbYqTErn41mu3ufUrkr+xzYg6iQdRpBRqKODL3ySo7XkqjTEG93QnENjtI6dAk6hBBoBOB00VjVCECcn/ukHtx1NO/CgSW+HlG9BVOKDVOqFVOqDVOKFWui0+f5AMkJ6fTMzicvqy95mfnkZfYhLyuf7hl9SIzXU1hYyN13381TTz3E4MGDm5178KDfyA/HRYcYASHEByhrleYIIUqAh4A4ACnlyyhrr56PsvZsI/AnNfKRmZmpxmXDmpjmyCec9NocrmMq9VqzXWmxe1XqtWYHdRYHwbSz9XEaMvU6MhLjSImD3NREMvRx7n06MvRxZOh1ZOp1WGvN/LKuGJfTAWYHmGHr+hKsXtdz1lt93if3l9Vkb1+HS0gaUl3UpTtpSLVSmikxD0ynoVciVXVlSD+5T4hLpEdmX/Iy+5KXle9575HZh1R9hs9z7HY7zzzzDE899RQ2m4358+fz9ttvN0ujVjl3lHfQZW0clygLYquK2WwmLS1N7duEFTHNkU9H6bU6XOyvNrOv2oyhwX7c3TACSE/UNWuxZ7ao1FN1Sos+Xa8jQXfUo72iooJu3bp5to3VZor3GSh3Sv6zci9Wi6PVe6ceKiR933Z0DUYsiU5MaTbqU+0wMJGK6/qzpTqD8toSnK6W16mBOtBqtHRL7+1u0R+t7Htk9SUzJReNCNz7/pdffuHmm29m+/btAFxxxRU8+uijx6RTq5zDpTuoQ7BYLJ2dhQ4npjnyUUNvg83J3iozRVWNFFWZKTI0cqjWQls9Mi27YTL1cWQkKq1z75Z6hnu/VqOMrDocLr79dBtOh2JEHMCPv5Ydt47MXRvRV5Vi1zkwJZuxikosWgP7MySN5yRRl+7CYm9sflLRIc/HrNTu7kre3X3jfs9Nz0OnPb6Abmazmb///e88//zzOJ1O8vPzWbRoEWeddZbP9Gr9rqPKCESqb3FrxDRHPsert9ZsVyr6qkaKDGaKqswcqTu2q0QjID8jkYHZevLSEnxW7inxWoQQPu5yFKfThbnBhtnkZMvPxWz44QB2m/9+dABdnLbZtstqRTqbP3U44wTxxT9ijjdic5VipYLS4QnU52qpa6zxcVUT2CFVn04Prwq+6b1HZm8S4kLrk+9NWVkZr7zyCi6Xi7/+9a888MADJCcn+02v1u86qoxAzJ86Oog2zYHqlVJS2WD3quyVVr6hwX5M2jiNoF9WIoOykxiUrWdQThL9s/Qk6vx3c1SU1vH5Z9tJ1PtvIUsJB/b4D4SWkZXEmdMK2Pf8O9Rt20N8fQ1JhiOe4y4hMaVBXaaLukyJbXh3nCP7UFpzkKq6cmRGy0cVKzQq/fRKxd63WRdOa/30amAymUhOTkYIQf/+/Vm4cCEDBw5k9OjRbZ6r1u86qoxAYmJiZ2ehw4lpjnx86XVJSWmdlT0GM3urGtnj7tKpsx7b4k7UaRiYrfdU9oOy9fTNSCROq6G0uJbCbWWUH6iivMV5v6wrRgjQaESbffC+0DUq/vSOpFQG/uefxNdVo7Vb2f+0xJIM5kxJeZ6k7gRJXYakLlNSnyFxNXsoOAIHFSPhr58+LyufzJTcNp9Q1Gbp0qXcfvvt3HPPPVxxxRUAzJo1K+Dz1fpdR5UR0OvVe7QLV2KaI5+4hET2efffVzWyr8rsc5A2NUGrVPbZSQzKUd57piWg1QicThf7dlXSuL+aXcDeXRUU7agIOj8jz8gnf1BOs31Ff38V45adnm29oRSnNFGXqVTuu4ZIjO6Kvi5T4oj3f/3s1O7kpvWiT+6AowOymX1D0k+vBtXV1TzwwAN89NFHAHzyySdcfvnlQRsltX7XUWUEampqosprBGKaIw2T1UGx0eoZtN1bpbT0HT6ccrKT4jyt+/zkOHKQZOp1zSufOgsvLv6hzZb8mLP6o086tmZOSoln4NCjXjoJiTo2XXEnu5evAcChU1rvxkxJXZ5SwTf2T8GUq6Wu0f/EqVR9hlfXzbH99F1heUkpJZ999hn33HMPBoOBxMRE7rvvPm644YZ2PZWo9buOKiOQnZ3d2VnocGKaux42h4sj9VZKjFYOG62UGC3udyu1firrvNR4T1fOoOwkZEkt65fvIakiAadLssHQEPD9E/VxDDpBqdgddhcTpw8mLcN/K3T95bexf+NapZ/e3Zo3/k55b/BZZ9X57qdv6r7J7EuKvvXY+eFexrW1tdx0000sWbIEgDPOOIPnnnuOAQMGtPuaammOKiNQX1/fJWdWHg8xzeGJ0yWpbLBRYmyq7C2ezxUmm99JVAlaQa/0BPpnKZV9v9Q49n+/h3gnUG6HciPbDSYMZSYAzI3NB33z+hxbuUoJ+qQ4LrpqJAIQmmNbqRsuv51Da9coLXqvlzFTYjpJ4hrhO79ajZbuGb3dlX2fkPXTh3sZJyUlsX//flJSUnjkkUe46qqr0GiOL3K/WpqjygjEFqKIDsJFs5SSWovD04r3VPR1Vo7UWbE7fVf1GgF5qQn0Tk+gV3oCvdMS6J2eSK/0BHKS49AIgd3m5J0XfmJlZest/IuvHklqujKgmJmTjK4V7x4Ak6WO0uqDlNYcYvPrr1FaU6xU+MMkDj8VPUCqLYH8ghHHDMjmpueh1YS+mgmXMvbmwIEDpKamkp2dTXx8PK+++iqpqan07t07JNdXS3NUGYFo8x+HmOaOoNHm5HBd84q+abuhFf/37KS4ZhV9r/REeqcn0CM1njitBumSFO+vpvDXMgorTez1qsAP7Klqdq3cvFTGnnW0q0Gr1ZA/KJv4hGP/4ja7hbLaYkqrD1Fac9D9fojigzsw67yeHLq7X270jjj65J9wzOSp7hnq+tP7Ipx+106nk1deeYXHH3+cGTNm8MorrwAwbNiwkN4nNk8gBESb/zjENKvFkTory4uqWbm3hhKj7xg0AMnxWnqnN7XqE92t+gR6piWQFK/4Orpckt2/lvHFK2vQxWlp6jWw+XDnbEmv/AzGTOnOwEH9m+13uhxUGkspLVFa9U2t+9Lqg1TVt3T2dKMDnQ3SagVpNcqrd88Cxt1/b0D99B1JuPyud+7cyZw5c9i4cSMALpcLu91OXFzovZRi8wRCQLS5DkJMcyiptzr4bl8ty4uq2V5+tBsmTivomebutslQWvNKyz6B9ESdz35vKSVNYdxfW7gaY7WydKDD7rvi71eQzSmn56PVHr1WTvcU7Jp6duzbwv5fNlNac5Ayd+teiXvj+1oaKUipgfSao5V9Wo0gvUbQ9/RxjHpvYbu/o46is3/XNpuNRYsW8cwzz2C328nLy+OZZ55h+vTWVtE9PmIuoiEgPr4V5+MIJab5+LA7XWwoqWdZUTVrDxo94YwTdRrO7JfO1IIsRuSlemLgBMKH/1hHyQFfYQxg+u9OouDEo4/9Go3A6mqgtPogR2rWU2pQum/KapQuHKvd/7qzOWk9SD5sJXGvUankawVpNRpSjIoh8CZ3yjhGvvdMwBo6m878XZvNZs455xxPwLerr76ahx9+WHW3ZLU0R5URMBqNZGR03BTxcCCmOXiklOw2NLJsTzWr9tVidLtlCuDUnqlMK8jijH7p6FvEs2l5jZa+9+YGG2tW7PVpAOL0cOH1Ayir3cWSzd96KvnS6oPUm2v93kfviCO13HG0RV+rtOhTawU6R9N9ju2a6GqVfks683et1+sZM2YMjY2NLFq0iAkTJnTIfdXSHFVGICcnp+1EEUZMc+BUmGwsL6pm6Z7qZv38+ZmJTBuUxaRBmeQmt90aK/y1jP9+8Eub6f7y4Hg271vNut0r2LB/DWve8u39obNxtMvGq78+rVaQYBFA63nq6hW+Lzr6d/3DDz+g0+k4/fTTAXj44YfRaDQkJSV1WB7U0hxVRsBoNLYapS8SiWlunQabk+/3K/38W0pNnv0ZiTomD8pk6qAsBmbrA/Znr6lqaGYA4uK1aLy6ihqsdej6FGNO281fX3zIE69eIMjLzCfpgInEPbXN+un1Dcpxf6SeeRpn/N/zAeUvUuio33VdXR0PPfQQb731Fv379+f7778nKSmpU+YoqKU5qoyA3X5stMRIJ6b5WJwuycbDdSwvquHHA7XY3P768VrB+Hyln39kr7SA+vltNgerv96Noaz+mG6ea245k5zuKdQ2VLFhzyrWFS5n+6ENuOqcUAdCaDix72jGDpnCmIJJZKTk8HWP8bT8W7bVkldr2cFwpiN+19988w233347paWlxMXF8Yc//AGdrvOqTLU0R5URCCff4o4ipllBSsm+ajNL9yhunTXmo/31J/dIYWpBFhP6Z5Ac77+fv4ntmw/z1b9+RRen9evNM2JiNhsOf8m6FcvZVbIZKZXgPlqNlhH9xjF28BRGFZxNWpKyZODGy++g0h1vB2B62U/HpTfSUVOzwWDgvvvu45NPPgHgtNNOY/HixZxwwgmq3TMQYvMEQkC4+BZ3JNGuubrRzgp3P//+mqMrM/VOT2DqoCwmD8qkR2pCq9czN9r4+pNt1NeacdhdVLvj8HgbgNT0REZNy2FP9Vp2lP3Ia1u3eo7ptHGc3O8MxgyezKhBZzXzt29Z+YPS8m+v3mhBLc0ul4uZM2eya9cu9Ho9DzzwAH/5y1/QattuHKhNbJ5ACIi2vnGITs1xiUms3lfD0j3VrC+p8yyJmJagZdLATKYMymJIblLA/fwvzF/hc//FV48kLqOBjftWsX7PClYu2340D7oETumvtPhPGzSBpIRUzzFfFT+0fwA3GstYLc0ajYa7776bN954g0WLFtG/f/+2T+og1NIcVUYgHKx5RxMtmqWUFFY2snRPNSuKqmlwx9LXChifn860gizG9EkjTus/do7VYsduc7J7WxkOd2zmXzeUeI5nZCdx/iXDqTSVUFSzjn+seZsDFYWe4wlxiZw64EzGDpnCqQPOZPuf/kbl8vmsbiPvx+u9Ey1l7E2oNLtcLt5++23q6uqYM2cOABdeeCEXXnhhpy9C0xK1yjmqjEBdXR2ZmZmdnY0OJdI1VzXYWV5Uzbd7qjlUe7S7Z1C2nmkFWUwamElGK8sdAkiX5JkHv/F/HIlFU0H+WAvPLnueYsNezzF9fDKnDZzA2CFTGNF/HNuueZDK5fNY1Ua+Q+m2Gell7ItQaN63bx+33nqrx/1z5syZ9OvXL+wq/ybUKueoMgK5ubmdnYUOJxI1Wx0ufjpoZOmeKjYdrvd092Qk6pgyKJMJfZI4oVeWJ73N5qCmsoHCbWXH/MGN1Y3s2lrWbJ9GI8jITiItv5F9xp/ZV7eOWmsp291jtckJqfQrTaLHd1XkHbKjc66kmpWs9JHXjvDRj8Qybovj0exwOHjppZd44oknsFgs5OTksGDBgrAfV1GrnKPKCFRXV3fo5I5wIFI0SynZWdHI0j1VrNpX64nOqdMIT3fP6D5p6DSCkpISIItGk433XlqDscZ/aAVvMrOTOOuyLNbtXsG63cuoKDrsOZaqT2d0wSSyPtuL/vNCtC4b4PvxvKMnZ0VKGQdDezXv2LGDOXPmsGnTJgAuvfRSHnvssbBfpAbUK+eoMgJNAbuiia6uubLBxrI9x87iHZyT5OnukWYbVrOD6rJ6DhRVcbi4gg+3bzvmWvEJOvoOzKJHr6PeOVK6OGLaQ5VmK5tLvmfpu0efCtKTsxlTMInMT/aQ+PlONPIr9xHlaSJcZuJ29TJuD+3V/Oijj7Jp0yZ69uzJwoULOeecc0KcM/VQq5yjygjEHpu7BhaHix8P1LJ0TzWbD9d7VtnK0uuYPCiLib1S+OlfW6krqeIf79W1eb3ho3oz6YKhxMcrP3eXy8mukl9Yt3s5P+9eQY2p0pM2K6UbY4ZMZuzgKZjufZeqR//rPnK0GylcKv8mumIZHy/BaHY4HJ5JXn//+9/Jz8/n/vvv73LrUMe6g0JAeXl52Pf7hZquotnpkmwtNbG8qJofDtTS6PbuiRMwwWqlZ5KO7DgtFWv280WFyec1cvMUN8x6o5nxkwvo2TeDbnmpaLQanC4Hvx5Yx7rC5azfsxJjY7XnvJy0PMYOnszYIVOpu/tNDI98ykE+bX7tMKv4vekqZRxKAtHc2NjI448/ztatW/nss8/QaDT06dOHBQsWdFAuQ4ta5RxVRiCc1yRVi3DW3DSLd3lRDSv31lDVaAcpSbI7GS1cFGTpMWw5AkCF++XN0JPzGDWhHzqdluxuyZ5B36qqKrKzs3E47Ww5uIZ1hcvZWLSKerPRc273jN6MHTKFsYOnMqDHMIQQbLz8DgzL1za7RzhX/k2EcxmrRVuaV69eza233sqBAwfQaDRs2LCBMWPGdFDu1EGtco4qIxAjPKgw2Vi5t4blRdUc8JrFe5LFSs8jR8MmG4qbx+I57/fDAdDqNAwYmuvp3vHG5rCy9eAadq79mQ1F39FoPfrU0DOrH2OHTOH0IVOpvO1lDHM/ZA8fsqfFNbpCxR/DN0ajkblz5/LOO+8AcOKJJ7J48WJOPfXUTs5Z+BJVRsBkMnUJL4BQEi6aTVYH3x8wsqKomq2lJqWfX0pOqTbRzdiILl6Lo8V6vAOHdSMrJ5le/TIZNKyb32tb7Wa27F/D2sJlbCr6Hou90XOsb+4gxg6ewtghU+mdM4CNl9/BzuXv+r1WVzQA4VLGHYkvzd4B3+Lj47nzzju55ZZbVFnqsTNQq5yjygh079697UQRRmdqtjtdrC9RonWuPWTE7pQgJSdWm8gTEl2DFe7b8QEAACAASURBVJe779/bAFw95wxye6T6uywAZmsDm/f9yLrdy/hl349Y7UefKPJzhzBu2FTGDJ5Cz6x8Nl5+B9uWv0lLf6GuWOH7Iva7Vti1axelpaWMGjWKxYsXM3To0E7ImXqoVc4dZgSEENOB51Ccq1+VUi5ocbwv8BaQ4U5zr5RySSjzUFlZSZ8+fUJ5ybCnMzTvrzbz350GvttXQ717sXQBjMNJ6n6DJ53L65xrb59AQmIc+qQ4hJ8Qzo3WejYWfc+6wmVs2b8Gu/PoIiyD8k5i7OApjBkyGVu99Gj2F6AtEir/JqL1d927d2+Ki4vp27cvALNnz6Z79+5ccsklERlKQ61y7hAjIITQAi8A04ASYL0Q4nMp5Q6vZA8CH0spXxJCnAAsAfqFOB+hvFyXoKM0u6Rk3aE6Pt1ewZbD9WRY7Aw0NpLhdJKsj8Ne23zCVna3FKbMHIZGCHr0yUCn8x3Tx2Q2sqHoO9YVLmPrgXXNFmEZ0vsUpeIfPImctDzPOSUmJd6PtwGItIrfm2j8XZeXl3PnnXeyfv161qxZQ/fu3dHpdMyaNauzs6YaapVzRz0JjAGKpJT7AIQQHwIXAt5GQAJNjrvpwJFQZyIrK6vtRBGG2pobbU6+2V3Ff3ZUUlVt5qxDBqa1SGO3Nl9r99rbJ5CZ4z8iorGhmvV7VvHz7uVsP7Qep8v9NCE0nNBnJGOHTFVm76b69ps+cufTbFv1s2c7kg0ARNfv2uVy8eabbzJv3jxMJhPp6ekUFhZGRZeYWuXcUUagF1DstV0CjG2RZh7wrRDiZiAZmOrrQhUVFVx77bXodDqcTicXX3wxs2fPpqysjOTkZLRaLXV1deTm5lJdXY2UktzcXMrLyzGbzeTm5mIymejevTuVlZUIIcjKyqKyspK0tDScTicNDQ306NGDsrIy4uLiSE9Px2AwkJ6ejs1mw2w2e47Hx8eTmppKVVUVmZmZmM1mLBaL53hiYiJ6vZ6amhqys7Opr6/HZrN5juv1euLj4zEajeTk5GA0GrHb7Z7jbWlqchvzp8lsNpOTkxNyTXVOHZ9uK+enwzbSaxtJtTk4ydh4THmdcGoPeg1IRuKiR49uWOxGnJipqbE102RsrGJ/zTbW7PyWfZU7PIuwaISWgh4jGNH3DM48aTrWBodSTlYnB6sPUn7vIqpXrvP7w8uaNJacx+dQWloa1uV0PL+9ffv20a9fv4jS5Kuc1q1bx/z589mwYQMA06dP59577yU9PR2r1dolNQVTTocOHWLgwIHt1uQP0RFTzoUQvwemSymvc29fCYyVUt7kleZ2d36eEUKMA14DTpJNtYGbNWvWyPYO+NTU1ERdtMVQapZSsqXUxKfbKqjfUEyiw0mW5dgl74ae3IMZfxjR5uOroa6Mn3evYN3u5ewu2YJ0zw3WanSc3G8sY4dMZeSgiaTqM5qd5y8evzeR3vr3Jhp+1x988AF33HEHFouF3NxcHnroIS677LKo6go7nnLetGnTxilTpozydayjngQOA94jGr3d+7y5FpgOIKVcI4RIBHI4do5Qu3E6fS8FGMmEQnPhvmpWrt7PIaMVm8lKjwYrLX13NFrBqDP6MeyUnq169lTUHmbd7uWsK1xOUelRf504bTwj+o9jzJApjBw4keTEY68R6GIsBoOBnJyc4IV2UaLhdz1o0CBsNhuzZs1i/vz5uFyuqDIAoF45d5QRWA8UCCH6o1T+s4A/tkhzCJgCvCmEGAYkApWEkIaGhqiqHKD9mk1WB9/tqWLTt3vQG5QJV756JC+66jR69s1AnxTv91pHqg/ys7vi31++y7M/XpfAqQPPZOxgZREWfYL/cYKWBqC1ln60lXMk6rVarXzzzTfMnDkTgNGjR7NmzRoKCgoAOHjwYMRpbgu1yrlDjICU0iGEuAn4BsX983Up5XYhxCPABinl58AdwD+FELehDBJfI0PcVxVbkLt1TI02ftxWwfoDtdTsNZBXb0HvdbzbkFwK8jPQagSDhnUjK9d/X2OJYR/rCpexbvdyDlUWefYnxiVx2qAJjB08hRH9x5MYr/d7jfa6d0ZbOUea3p9//pk5c+awe/duPv74Y6ZOVYYHmwwARJ7mQOjyC827ff6XtNg31+vzDuAMNfMQW5D7WFxSsvVwPcs/3oJ0L6CeCOS1SPfnuyaSnuk/lrmUkkOVe1hbuIyfd6/gcNV+z7GkhBRGDjqLsYOncHL/04nX+V/YvbX+/kD7+aOtnCNFr8lk4rHHHuMf//gHUkoGDRrkN9JnpGgOhthC8yEgUqaPB4M/zYdqLSzfU82q3VUM3958eCY+U49eK+jZK52BQ7sx6MTufv34D1Xu4ccdX7OucDlltUcdwFIS0xlVcDZjh0xheP4YdNq2v/tQTeyKtnKOBL0rV67ktttu49ChQ2i1WubMmcNdd91FYmKiz/SRoDlY1NIcVUYgPT297UQRhrdms93Jd/tq+XqXgdp9VZxQWcdwr7S9h+Ry8WUjfAZma4nNYeWj1S+wZMP7Hq+etKRMxhRMZuyQKQzrc1pAFX8ToZzYFW3l3NX1fvjhh9x4440ADB8+nMWLFzNixIhWz+nqmtuDWpqjyggYDAaSk/0PPkYilZWVHGqArwur2Li9nH7ldfS22undIt2wU/KYcWnrf7wm9pbu4MUlczlctR+N0DLl5N8yfti5DO19ChpN+6brh3Jmb7SVc1fXe95559GvXz+uvPJKbrrppoBavF1dc3tQS3PARkAIMRS4BOghpZzt3o6XUm4Nea5UIppaD0aLg2V7qlmyo56Eg4foWW/hFKfrmHS/u2YkOd1TSUnz30/fhMNp59M1r/PpmtdwSSc9s/oxe8YjDMw78bjyuvHyOzyfQ+HbH03lDF1Pb0VFBc8++yxz585Fr9eTnp7O2rVriY/372HWkq6mORR06pOAEOISlNg//0Zx7ZwNpAAL8DOzNxyx2WxtJ+rCOF2SzUfq+bqwip/31ZBTb+YEQ/0x6SbNGErBid1Jy/DvmdOSEsM+XvxyLvvKdwJw3sg/ctnE2cTH+e6zDQRfbp+hINLLuSVdRa+Uko8++oj777+f2tpakpOTefDBBwGCMgDQdTSHErU0B/ok8AgwTUq5RQjxB/e+LUBg/QdhgtlsbjtRF2XT4Tqe/6mEEqMVISXT9jefY5eWkcjYswdy0sheaLW+B3l94XI5WbLhfT76/kXsThs5aXnccP48Tuzrc/JhQKgd2TOSy9kXXUFvcXExt912GytWrABg8uTJXH311e2+XlfQHGrU0hyoEegGNHX7SK939WNOhJBI9C2uabTz8rrDrNyrrMKVr4OhB6o8BZOcGs85F53EwKH+F2XxR3ltCS8tmceuks0ATBp+IVdOvp2khONb5i7QSV/tJRLLuTXCWa/L5eL111/nkUcewWQykZGRwWOPPcasWbOOa8ZvOGtWi86eJ7ARuBJ422vfLOBn38nDk0jyLXZJyVeFVbz28xFMNifJSCYcqcZlcTSzzOddNoB+/YIzAFJKlm/5lHdWLsRqN5OenM315z7IyEETjyvPLZ8Appf9dFzX80cklXMghLPelStXcvfddwMwc+ZMnnzyyZBE/AxnzWrR2fME5qBE+LwWSBZCfAMMBs4JeY5UJNh+x3Blf7WZ534opuxgDX1NFnrXK4+J3sO+I8b0YeL0IVTXBBd5o7q+kn988yi/7PsRgNOHTOPac+49JohbsKjV/++LSCnnQAk3vVJKTyt/8uTJXHXVVUyZMoXf/OY3IbtHuGnuCNTSHJARkFLucnsDXQB8gRIW+gsppan1M8OL1NTWlywMd6wOF+9uKmXtDwcYYqg/xs0ToF9BDhdfPRKNe3WuYDT/tPMbXlu6gAZLHcmJaVw77V7GDzv3uPIcTMyfUNHVyzlYwknv1q1bufPOO3nuuecYNmwYQggWLVoU8vuEk+aOQi3NgXoHLZZSzgE+brF/kZTyVlVypgJVVVVtxtYOV8x2Jw9/upOMLYcZ0iKk0tizBjBgaC49eqWjbTGzNxDN9eZaXvt2AWsLlwIwov94/jJ9rt9FW4Khow0AdO1ybg/hoNdisfD000/z3HPP4XQ6+fvf/84bb7yh2v3CQXNHo5bmQLuDrkHpEmrJlUCXMQJdLeZ6U/y8aqOFF17bSE5V8wev3/9pFH0HZnta/b5oS/Omvd/zj68fpbahioQ4PVdOup0pIy467jC9HdX/74uuVs7HS2frXbt2Lbfccgt79uxBCMH111/vcf1Ui87W3BmopblVIyCE+J+mdF6fmxgAGOhCmM1mvwGpwo0dm4+w5F9H5+F5h247bXw+Z00fckyr3xf+NDdaTbyzYiErf/0PAEN7n8oN58+je4avTqbAaC3ef0fSlco5FHSW3vr6eubPn8+rr76KlJKCggIWL17M2LEtFw0MPdFWxqCe5raeBK50v8d7fQbFNbQcaL+jbydgsVg6OwttIqXks3c3s3fnsWvpdOuTzsw/jCAjy380z5b40rz90AZeWjIPQ10pcdp4/jDhRs4f9cd2hXwIRdTPUNMVyjmUdJbe6upq3nvvPbRaLbfccgt33HGH34BvoSbayhjU09yqEZBSTgIQQsyXUqr7fNcBdAXf4t3bypsZgLW9skjvlsKT5w8iJzl47wBvzTa7hQ9Wv8BXG98HoH/3odw44xH65AwM6prhWPF70xXKOZR0pF6j0UhqaioajYb8/HwWL15MQUEBw4cPb/vkEBJtZQzqaQ5o6qi3ARAKmqaXKrlSibKyss7OQqtYLXb++8Evnu3l/XLJ7pHK0zMK2mUA4KjmvaXbufety/lq4/tohJbfjf8zj17xZtAGAPA523d62U9ML/up0w0AhH85h5qO0vv5558zduxY3n776HShiy++uMMNAERfGYN6mgP1DuqJEjtoItDSYbx9YSM7gY56VA2WkgM1rPxyJ+WH6zz7ijKTGdAtmSemDyItsf3BXuPidXz8/Ut8tvaNkAR98w721pGDvcEQruWsFmrrLSsr4+677+aLL74A4JtvvuHqq6/u1DV+o62MQT3NgdYurwCNKGsAf4diDObRYqWwcEevDzxgmto01Fv559OrkS4XTmdzl8/qxDgSB3dj/vSBpCS03wAUG/byv189wCHDHgSC80ddzqwJN7Y76FvLmP/hSjiVc0egll4pJe+//z4PPvggRqORlJQU5s2bxzXXXNPpi7xHWxmDepoDrWHGA32llA1CCOkOJHct8BPwT1VypgI1NTVh4VEgpeSlJ1Yes99S0I0NZif9eqfz+PSBJMW37yHL5XLy5Yb3+Pj7l7A7beSm9+SG8+ZxQt+R7c5zKBd9UZtwKeeOQg29VVVVXHfddXz33XcATJ06lYULF9K7d/u9x0JJtJUxqKc5UCPgBBzuz7VCiFygDugV8hypSHZ2dmdnAYBP39nk+dx/cA6/uewUVh2s5enVxSQlx3P/5H7tNgDltSW8uOQhCkuUsYUJw2bwP+fcgz4h+MUo1I72qRbhUs4dhRp609LSqKqqIjMzkyeeeIJLLrmk01v/3kRbGYN6mgM1AuuA84FPgW+AjwAzsEGVXKlEfX19p88y3FdYyb5dR+P5/O6aUVQ22Hhp7REAbhzXm24pwQ8CK0Hf/s07K5/FajeTkZzN9dP/Rl7SoHYZAPA9ABzuBgDCo5w7klDpLSwsJCsri9zcXOLi4njttdfIyMggN/f4Z46HmmgrY1BPc6BG4EqOehLdCtwBpAKhDwqiIp25EIXVYufz93/hYFGVZ991d0xESsnC1YdosDkZ2yeNaQVZQV+7ur6CV75+lC37lYHacUPP4X+m3UOqPoODBw8GfB1/rp/hOgDsj2hbcOR49drtdhYvXsxTTz3FjBkzeO211wAoKCgIRfZUIdrKGDpxURkhhBZ4DrgeQEppBuarkhuV6Szf4n2Flfz7rY3N9s34w8lkZCfx5S4DGw/Xk5qg5dYJfYN65JZS8uPOr3lj6ZM0WOt9Bn0LRHNbfv9djWjzIT8evb/88gs333wz27dvByAlJQWHw4FOF97Lj0dbGUMnricgpXQKIc6heaTiLklnxCBvqLc2MwADhuQy4ZzB5OalUlpv5ZW1hwG4aXwfspPaXmC7ibrGGl5b+gTrCpcDcMqAM7j+3L8dE/QtEM2dEeRNTaIt1nx79JrNZp588kmef/55XC4X+fn5LFq0iLPOOkulXIaWaCtj6Pz1BJ4FHhZCPCSltIc8Fx1ER7uVNdRbm3kBXXTVaZ4Vvg4bLTy24gAWh4uJ/TM4e0Dg8fo3Fq3mH9/Mx9hQRWJcEldOvo3JJ/sO+taW5q7g9x8s0eY+GKzexsZGJk2a5An4dsMNN3D//feTnNy+saPOINrKGDrfRfRmoAdwuxCiEq9lJaWUfdXImBp01EIU+3dXsuKLndQYGj37hgzvwcCh3XC6JJ9tr+SNDUewOSXdUuK4+Yw+AXUDNVpNvL1iIavcQd+G9T6NG86fR7cM/05abWnuCn7/wRJtC44EqzcpKYlJkyah0WhYvHgxo0ePViln6hFtZQydvKgMcIUqd+9gjEYjGRnHt0JWIHzyZvP+/1ET+nH2eUM5bLTyzOqDbCtvAGDqoEz+enrvgGYEbz+4npe+etgT9G3WxNmcN+qPaNqI3NGaZu+ngK7eBeRNR5VzuBCI3qVLl5KUlMQZZ5wBwNy5c9FqtSQkJHREFkNOtJUxqKc50JXFvgv5nTuBnJwcVa/vdLr4/tvdnu1TT+/LhHMHo4vX8um2Cl5ffwSrU5Kl13HLmX0Zl5/e5jWVoG/P89XGDwAY0H0YN854hN45AwLKkz/NXWX2b3tQu5zDjdb0VlVV8cADD/Dxxx/Tt29ffvjhB1JSUkhKCjwSbTgSbWUM6mkObxeAEGM0GlXr97SY7Tz/6PJm+yZfMIxCQyMvrilhV6XSNTR5YCY3jgus9W9zWHnkw79QVLoNjdBy8fjr+O3pf0KnDXwAuaXmzljusaNRs5zDEV96pZR89tln3HPPPRgMBhITE7nuuusiJuZOtJUxqKc5qoyA3a7emPae7eXNtv94x0Se/v4QS/dUA5Cl1zHnzD6Mzw/8ce7t5c9QVLqNnLQe3HbhUwzMOyHofDVp7qqzf9uDmuUcjrTUW1payl133cWSJUporzPPPJNFixYxYEBgT49dgWgrY1BPc4cZASHEdJT5BlrgVSnlAh9pLkUJTCeBLVLKP4YyD2r42Uop2bz2ECv+uxOAtEw96ecM5YYv9mC2u4jTCC4+KZfLTukRVCiI1du/ZNmWT4jTxnPHb5+mf49h7cpfk+ZIb/17E20+5N56XS4XF110Ebt37yY1NZWHH36Yq666Co2mS0V9b5NoK2PoxHkC3rjXD+gupSwN8jwtSijqaUAJsF4I8bmUcodXmgLgPuAMKWWNEKJbMPcIBDX8bL/4cAuFvx6N870rI5lt65UQEOP6pnP92F70Sg9u8K24sojXvn0cgGum3tVuAwCKZsP9iz3bkeIG2hrR5kPurVej0fC3v/2Nd999l6effppevbpUeK+AibYyhk6eJyCEyABeBH4P2IFkIcRMYEyAK46NAYqklPvc1/sQuBDY4ZXmz8ALUsoaACnlsesrHieh7k+zmO3NDEDxST3Z2eikV1oCs8f3ZlTv4CP+ma0NPPufu7HaLUw86QImn3zRceXxyJ3PYPxeCfEUaQPA/oimvmKn08m//vUvAO68804AZsyYwfnnnx9WAd9CTTSVcRNqaQ70SeBloAbI52jFvQZ4BgjECPQCir22S4CWq1EPBhBC/IjSZTRPSvl1gPkLCK02tOvf/Ou19Z7P5WP7s7OykV5pCSz8TQGZ+sAHb5uQUvLK149ypPogfXMHce20e9v9R46GAWB/hLqcw5WdO3cyZ84cNm7ciFar5dJLL6VvX2XaTiQbAIieMvZGLc2BGoEpQE8ppV0IIQGklJUh7rLRAQXA2UBvYLUQYriUstY7UUVFBddeey06nQ6n08nFF1/M7NmzKSsrIzk5Ga1WS11dHbm5uVRXVyOlJDc3l/LycsxmMy6XC5PJRPfu3amsrEQIQVZWFpWVlaSlpeF0OmloaKBHjx6UlZURFxdHeno6BoOB9PR0bDYbZrMZw2EX5UeUlcCcyfFsqWwkS6/ljtHp1FUcIcl9fmJiInq9npqaGrKzs6mvr8dms3mur9friY+Px2g0suXId6wtXEqCTs/s8+ZTdqSiTU1NUQVbavI2AOkTRpHz+BysVmurmpryFB8fT2pqqieUsNlsxmKxeI4HoyknJwej0Yjdbvccb6+mQMvJYrFgt9sjSpN3OWVlZfHEE0/w6quv4nA4yMnJYcGCBSQkJHDw4MEuqSnYciouLiYuLi6iNLVVTsXFxR5t7dHkDyGlbDUBgBCiCJggpSwVQlRLKbOEEH2Bb6WUQwM4fxxKy/5c9/Z9AFLKJ7zSvAysk1K+4d5eDtwrpVzvfa01a9bIoUPbvKVPGhsbQ+IfbW608cL8FZ7tVX1zSEiO5+kZBfTPat/U7j1HfmXe+9fhdDm49cInOX3I1HZdp+UTwMR9y7q8T3iwhKqcw5FNmzZx8803s3On4ohwzTXXcPfdd0fdQGkkl7E/jkfzpk2bNk6ZMmWUr2OBugy8CnwihJgEaNyV+lso3USBsB4oEEL0F0LEA7OAz1uk+QzlKQAhRA5K99C+AK8fENXV1SG5jveM4J96ZSESdDxyzoB2G4C6xhqe/c89OF0Ozhv5x5AZgNwp40KmuSsRyZqfeeYZdu7cSf/+/fnPf/7DwoULcTgcbZ8YYURyGftDLc2Bdgc9ibKIzAtAHPA6yrrDzwVyspTSIYS4CWVBGi3wupRyuxDiEWCDlPJz97FzhBA7UFYyu0tKWeX/qsETyFNPIBirlYlfdel6TAlx3DauNyd2b99iDy6Xk+e/eJDq+nIKep7M5WfPadd1/C3/WFxc3NppEUmoyjlcsFqtnvAOTz75JEOGDOHOO+/0tAojTW8gxDSHjkDDRkiUCj+gSt/PNZbQYmF6KeXcFve43f1ShVCskGS1ODA3KpM2tqfo6ZYSx7TB7V/27d9rXmPrgbWk6jO49cIFQc0GbqK19X/DcVUotYkUzXV1dcybN48dO3bw5ZdfotVq6d27N3Pnzm2WLlL0BkNMc+gIqDtICLFFCHGXECI8VpluJ+Xl5W0naoM9O45eoyFOx6Und0enaZ8nxpb9a/jkx38gENz8m8fITu3eruu0tgB8KDR3NSJB87fffsu4ceN488032bx5M5s3b/abNhL0BktMc+gIdExgHjAa2CWE+E4I8RchRPDrIHYyoVif02ZRngLq4nWkJ8dxbjufAgx1ZTz/xQNIJL8/43pO7nd6u67TViTQaFuHFbq2ZoPBwPXXX8+sWbMoLS1l5MiRrFq1ilGjfI7pAV1bb3uJaQ4dARkBKeWnUspLgTyU8YCLgGIhRMvB3YjG5XSx4otdANTH6/jdSd1I0AU/Hd/htPPc5/dRbzYyov84Lhp/XbvzFKmRQKORzz77jHHjxvF///d/6PV65s+fz9dff82wYe2fMR4jRlsEVYNJKeuB94GXgHXA+WpkSi1MJtNxnf/fD7d4PltSEpgxrH2hXd9b9Rx7jmwlO7U7s2c82uaaAP4IZD2A49XcFemqmktKSqiqqmLixIn8+OOP3HjjjQFNEOqqeo+HmObQEWjYCAFMBv6I8hRwEMUYXK1KrlSie/f29bkDVFeamkUKnTZ5IMlBBIRrYu2upXy18QO0Gh23XvgkaUmZQV/DlyuoP45Hc1elq2h2uVzs3buXgoICAG644Qb69OnDzJkzg5rx21X0hpKY5tARaBP0CMqcgCMoAd5OlVI+JaXsUv6HlZWV7T53xy9HY+YVDs3jN+14CjhSdYCXv3oEgCsn305Bz+FBXyPYcBDHo7mr0hU079u3j9/+9rdMmzaN0lLlt6XVarnwwguDDvnQFfSGmpjm0BHoPIELpZQ/q5KDDqS98VQaTFbWrtwLQGlKIlec2Zc4bXBdOBabmWf/czcWeyPjhp7DuadeGtT57Y0FFOkxZHwRzpodDgcvv/wyTzzxBGazmZycHPbv309eXl67rxnOetUipjl0+DUCQoh+UsoD7k2DEMLnihRNkUG7AllZwTs0SSl522uGcEJBLhP7B7fOp5SS15Y+QbFhLz2z+nH9uQ8G39prZzC49mju6oSr5h07djBnzhw2bdoEwKWXXsrjjz9+3PkNV71qEtMcOlprzv7q9bkI2ON+937tUSVXKhHs45RLSp767iCF9TYANDnJzLt4WNAV+Iqtn/L99i9JiEvktt/+HX1CcCFhvQeAp5f9FFQ00Nhjc3jw1ltvcfbZZ7Np0yZ69uzJRx99xMsvvxySP3Y46lWbmObQ4dcISClTvT5rpJRa97v3q0vFc01LCy6+/w8HallWVEPvegsA0yb2Jz7IbqD9ZTt5c9lTAPz5nAfokzMwqPOPd0H4YDVHAuGoefhwZfzn2muv5aeffmLatGkhu3Y46lWbmObQEeiM4cV+9i8KbXbUxel0BpxWSsmHv5STaD96TnpmcAHiTJY6nv3PPdidNqae8jvOPDE4j9rWwkEESjCaI4Vw0NzQ0MDHH3/s2T7ttNPYuHEjTz31VMj/zOGgt6OJaQ4dgTZrr/Gz/8oQ5aNDaGhoCDjtxsP1FFWZOaWqzrOv78DAZwe7pIuXvnyICuNhBvQ4gasn3xn4vS+/g697jD9uAwDBaY4UOlvz6tWrmTBhAn/961/55ptvPPv79Omjyv06W29nENMcOlr1DhJC/E9TOq/PTQwADKrkSiWCibn+0ZZyetabSWtUxgPy+qQHda8vfn6HjXtXk5yYxq0XPkmcLj6g80K9Ili0xZmHztNsNBqZO3cu77zzDgAnnnhih+QlVsbRgVqa23oSuNL9ivf6fCVwBTCQLjZZFCpy6AAAIABJREFUrKysrO1EQHGthS1H6jmp8uhTwEVXjQz4PjsObeSD1c8DMHvGI3RL7xnQeS27f4IdBPZFoJojic7Q/NVXXzF+/Hjeeecd4uPjeeCBB1ixYgUjRoxQ/d6xMo4O1NLc6pOAlHISgBBifoALyoc1cXGBhWleXlTNMEO9Z/uK2eNISg6sJV9jquS5/96HlC5+e/r/cNrACQHnLxTdPy0JVHMk0dGa3333XebMUdaBGD16NIsXL2bIkCEddv9YGUcHamn2+yQgmvtBzhVCaHy9VMmVSqSnt92l45KSFbur6VNv9uzr0SuwriCny8Hi/z6AsaGKE/uO4pIz/xJw3gKJA9QeAtEcaXS05pkzZzJ48GAWLFjAkiVLOtQAQKyMowW1NLdWiRu9PjsAe4tX074ug8HQ9hDGr6UmTt5W4tmeMy/wpR4/+v5FdhZvJDM5h5t/8zhaTWATso/XDbQ1AtEcaaituaSkhFtvvdUzUJeWlsYPP/zA9ddfH1DAt1ATK+PoQC3NrdVSJ3p97q/K3TuYQCzpsl1Hv+h+BTnExwdWkW/Y8x2fr3sLjdAyZ+YCMpID8yQKhRtoa8RaTKHD5XLx5ptvMm/ePEwmE1lZWZ5VvnS6QCOwhJ5YGUcHamn2+8v1Dg4npTzofUwIoQdcUkqrKrlSCZvN1upxh0uyZbeBk93bv/+T/4U8vCmvLeGlJQ8BcNlZNzGsz6kBnae2AYC2NUciamjeu3cvt9xyCz/99BMAF1xwAX/5S+DdfWoSK+PoQC3NgU4We1oIMcb9eQZQDdQIIX6jSq5Uwmw2t3p8V4WJkw8pa9trtIGHhnh5ycM0WOsZXXA2F4wObOpERxgAaFtzJBJKzQ6Hg8WLFzNhwgR++uknunXrxptvvsnbb78dNuGMY2UcHailOdCB3cuBbe7Pc1FcRGcCj6uRKbVoy892494az+f8QYGFit5z5Fd2lmwiOSGVv543L+C4Qh1hACDmT328rFq1innz5mGxWLjssstYs2YNM2fODNn1Q0GsjKODzpon0ESSlLJRCJENDJBSfiKlXAbkq5IrlWjLz7bwwFEj8LurA5sXsGTD+wBMHnExyYmpbaQ+FjUNAMT8qduDlNLzeerUqVx//fV8/PHHvPDCC2RmBr8IkNrEyjg6UEtzoEZgtxDicuAmYCmAECIH6FLPZPHx/n39LQ4XhvLglm8z1JWyrnA5GqHl3NMCXx/A2x1UbVrTHKkcj+aff/6ZiRMnsm3bNs++BQsWMHVq4F5iHU2sjKMDtTQHagRuBGajLDH5N/e+c4Fv1ciUWqSm+m+pbyszkdWgjHOPnhCYM9Q3mz7GJZ2cPmQqOWmBPaqp6Q7qi9Y0Ryrt0Wwymbjvvvs477zz2L59O88++6wKOVOHWBlHB2ppDsgISCnXSynHSynPklLude97T0rZpQLIVVVV+T22udhIjlkZfR92SturPFlsjSzf8m8Azh/9x4Du31GDwd60pjlSCVbzypUrOfPMM3nllVfQaDTcdtttvPDCCyrlLvTEyjg6UEtzwM7NQoizgauAXsBh4B0p5UpVcqUS/vpzXVLyS6HBMxmiW17boX5XbfsvjVYTg3uNYFDeSW2m7wwDAP41RzKBaq6treXBBx/k/feVcZ3hw4fzv//7v5x88sltnBlexMo4OlBLc6AuotcBHwNlwL+BUuADIcSfVcmVSvhzsVpfXIepRjmm0bTt3eOSLr7e8AEA549q+ymgswwAxFzpWqO+vp7PP/+chIQE/va3v7Fs2bIuZwAgVsbRglqaA30SuBuYJqXc0rRDCPER8AnwTzUypgYWi8Xn/k+2VZBicwDQrWfbTwGbir6nrLaYnLQ8Rhec3WrazjQA4F9zJNOaZoPBQFZWFhqNhj59+vDiiy8yePBgBg8e3IE5DC2xMo4O1NIc6MBwNrCjxb5CoEut9uzLz7bB5mTrkXr6GxsB0AYwSWzJhvcAOG/krDbjA3WmAYCYP3UTUko++OADxowZw+uvv+7Zf8EFF3RpAwCxMo4WOnuewA/AQiFEEoAQIhl4CvhJlVyphC8/262lJnrXNnq2Tx3X+tSHA+WF7CjeSGJcEpNOvjDge3eGAYCYPzVAcXExl1xyCbNnz6a2tpYffvihk3KmDrEyjg46e57AX4ERgFEIUQ7UurfDI3hKgCQmJh6zb/ORenp7hY0eenLrnkFLNiqDiJNOvpCkhNZdtjpyPoA/fGmOdJo0u1wu/vnPfzJ+/HhWrFhBRkYGL774Im+88UYn5zC0/H97Zx4fVXn9//fJTBKyEbIRZJGIihSUomURQVBRvohVKGKxlWotfHFBkKJY0LpU6lcQlZ9aKyoqrVSwFkTEhVargAJuuFSwKCARlEASQvZkMjPn98edDAmZJBMy+73v12tezNznufc5n7lhzn22c8x8j81EsDT7NSegqgeAESLSHegK/KCq+1s5rREiMgZ4BLABS1V1QTP1Lgf+AQxS1Y/b0kZrJCU1TRT/2fdl9PYkkx8++tQWzz9SUcTmr9YjCGN+cmWr7YVyP0Bz+NIc6yQlJVFYWMjVV1/NBx98ABgx/x944AE6d+4cZusCj1nvsdkIlma/k8KISCdgZP3L89nfc23A48DFQF/gFyLS10e9NOBm4AN/r90WSkpKGn0urXFSu++I93PfAS2ngfznpy/hdNUx8NTzyO3UvcW6wUoS01aO1WwGSkpKyMjIoLa2ltzcXP7617+ybNmymHQAYN57bDaCpdmvnoCIXICxNHQnkA+cCDwuIper6tt+XGIwsEtV93iutxIYR9PJ5vnAQmCOf+a3jaysxjH+/1NQQb8GeYQ7dmre0zrqavjXZ/8AYOzAq1psJ9S7glviWM2xzBdffEFubi5ZWVnY7XaeeeYZMjIy6NTJ7+eVqMRM97geS3Pg8HeJ6J+Aaar69/oDInIFxtN9Hz/O7wbsa/B5PzCkYQUROQvooaqviUizTuDQoUNMmTIFu92Oy+ViwoQJTJ8+nYKCAlJSUrDZbJSVlZGTk8Phw4dRVXJycjh48CB1dXVkZGRQUVFBbm4u7399gPo8UCf16URJSQkul4vKykq6dOlCQUEB8fHxpKen8/oHKyivPkKPrFNJchlPmQUFBSQkJJCWlkZxcTEZGRl8OeUOyjZ9AkDKOQPo/vBtlJWVUVJSQlZWFuXl5TgcDu/1k5KSSEhIoLS0lOzsbEpLS6mrq/OWt6YpNTUVwKupsLAQESEzM5PCwkKcTic1NTU+NRUVFZGeno7D4aC6utpbfqym6upqampqvOUdOnQgKSkpbJo6duzY6D7l5+fz9NNPs2zZMkaOHMkDDzxAVVUVcXFxJCUlkZ+fH3Wa2nKf8vPz6dGjR0xpau0+7du3j549e8aUptbu0/fff09eXt5xa2oOaRgxsdlKIkeALFV1NThmB4pUtdXHLBGZCIxR1amez78ChqjqTZ7PccC/gV+r6l4ReRe41decwJYtW7RPH3/8TlPy8/Pp2fPo6p+bn9tGj28OAXDDvPNJSUv0eZ6qMue5Sewv2s30S+Zzbr+xzbbxZpdzgPAtCT2WYzXHGlu3bmXmzJns2rULEWHatGlMnTqVk08+OdymhYxYv8e+sDS3jW3btn0yatQon1my/O0JPI8RQO7RBsduAP7q5/nfAz0afO7uOVZPGnA68K4nHn8XYK2IXBbIyeGG62wPV9WRuu/oGFtzDgDgi71b2V+0m4yUbIb2uajZepEyD9CQWF1PXV5ezvz581m6dCkAvXv35tFHH2Xw4MHU1kZVwrt2E6v3uCUszYHD34nhM4GHRGS/iHwgIvuBh4AzRWRj/auF8z8CThWRk0QkAbgSWFtfqKqlqpqtqnmqmgdsBQLqAKDxOtvPfigno6YOgCEje7V4Xn3OgNFnTcJui2+2XqTMAzQkFtdTV1VVMWLECJYuXYrdbufWW29lw4YNDB48GIhNzS1hNr1gaQ4k/vYEnqYd4SFU1SkiNwHrMZaIPquq20XkXuBjVV3b8hUCQ8MlVntLaqiy20h2ujjx5OYnXL4v/pbPv91Mgj2RCwdMaLZeJPYCIDaX0iUnJzN27Fi2bNnCo48+yumnNw7gF4uaW8JsesHSHEj83Sfwl/Y2pKqvA68fc+yuZuqe1972fNEwKUNhpYNkpzHF0bFT85sw6nsBI/r9lLSk5qc/IrEXALGRfENVWbt2Lenp6Zx33nkA3Hnnndjtduz2pn/CsaC5LZhNL1iaA4nf+wRigdLSUu/7koJy7/uEDr59YVlVCRu3vwbAxQN/4VcbkdQLgMaao5GCggKuueYarr32WmbMmEFFhZH9rUOHDj4dAES/5rZiNr1gaQ4kpnIC2dlHk8cnf1PofZ+S6ntS+O3PV1PnrGVAr2F0y/Iv21ik0VBzNKGq/O1vf2Po0KGsW7eO1NRUZs+eTXJycqvnRqvm48VsesHSHEhM5QTqPamqUudZGvujs3zvEnar25s57OKftNwLiIQYQc0RjU9M+fn5TJgwgRkzZlBaWspFF13E5s2bufbaa4mLa/1PNho1twez6QVLcyDxO7NYLFBXZ6wGOlLtJN2zMmjAoB4+6361bxtFZQVkd+zCGXlDfNapJ1LnA+Co5mjB7Xbz85//nG+++YbMzEzuv/9+Jk6ciGfpsF9Em+b2Yja9YGkOJP5mFksUkftEZI+IlHqOjfas+Ika6tfZHig/uo48IcG3H9z45ToAzu13CXHS/NcUqauC6om29dRxcXHMnz+f8ePHs2XLFq644oo2OQCIPs3txWx6wdIcSPwdDlqMsZnrKqB+i/F2jA1jUUP9Ottdu48mbM7KbbqlusZRzQc7jZBI5/a7pNnrRVKMoOaI9PXUdXV1PPTQQ9x3333eY6NHj+bZZ58lJyfnuK4Z6ZoDjdn0gqU5kPg7HPQz4BRVrRQRN4Cqfi8i3YJiVZBISUkB4NDBSuOAPc5nTuGPvnmHmroqTu16Bl0zfW/TDnfaSH+p1xyJfPbZZ8yYMYPt27djs9m4+uqr6dHD9/BcW4hkzcHAbHrB0hxI/O0JODjGYYhIDlDsu3pkYrPZqK2p48hH3wEQ38zS0I3bjaGgEf1+2uy1osEBgKE50qiuruaee+7hoosuYvv27eTl5bFq1aqAOACITM3BxGx6wdIcSPx1Ai8BfxGRkwBE5ASMyKIrg2JVkCgrK6OywuH9fOqwvCZ1issP8uXeD7Hb4puNExTp8wANKSsra71SCNm8eTMjRozg0UcfRVW58cYb2bRpEyNGjAhYG5GmOdiYTS9YmgOJv07gduBb4D9AJ+Ab4AfgD0GxKkg0HGOujLdx+o+bLg99b8cbKMpPThlBalK6z+tE+jxAQ453XD1YPPHEE+zevZvTTjuNN998kz/+8Y8B7+ZGmuZgYza9YGkOJP6GjXAAvwV+6xkGKlJ/YlBHGIcPH6ZDwtHQD51TG2/DVlXvqqDmhoKiqRcAhmZ/NlgFk6qqKq8NDzzwAP3792fmzJkkJjYfubU9RILmUGI2vWBpDiT+ZhY7NsxmWv2yvfpsYdGAqrLve6NLZVfoYG/cEdpT8BXfF39Lx+QMfnyS76f8aOoFgKE5XBQXF3PHHXewa9cu1q9fj81m44QTTmDOnKAkjvMShc8n7cJsesHSHEj8XR20C2NpaMOlNPUWRc0MTU5ODjvzjWVW8T6+0PoJ4WE/urjFkNEQHb0ACE+3WVV5+eWXmTt3LkVFRSQlJfHFF19w5plnhqR9sw0VmE0vWJoDiV9zAqoap6o2z79xQFfgKeBXQbEqSBw8eJDvvjIyiekxSWScrjo2f/UmACNO9703IJLDQzTHwYMHQ9regQMHmDx5MlOnTqWoqIjhw4ezadOmkDkACL3mcGM2vWBpDiTHFTtIVQuAWcD9gTUnuKSmplLl2S1sdzfuCXy65z3Kq0s5MecU8jqf1uTcaNgY5ovW8osGkpUrVzJ06FDeeOMN0tLSWLx4MWvWrKFXr5aT9gSaUGqOBMymFyzNgaQ9sYNOA6JuZqa2rAaA5BMb5wZoGCbCV5iCaNkXEE5KS0spKytjzJgxPPjgg3Tt6js4n4VFuFBVKioqonJOwe12t7pMVERITU1tU6gVfyeGN3F0DgCMH/9+wL1+txQBlJeXow4jkUx69lGvWlZVwrbd7yESx/C+F7d4jWhzABUVFWRlNZ85rT24XC7++9//0q9fPwCmTp1Kr169uPDCC9sc7yeQBFNzJGI2vXD8misqKkhMTIzKpDS1tbWtrqhzOBxUVFSQlpbm93X97QksPeZzJfC5qn7jd0sRQKf0o380XXpmeN9v/u8/cbmd/Pikc8hIbTr5Eo1zAfXk5uYG5bo7duxg5syZfP3112zevJnu3btjs9m46CLfG+xCSbA0Rypm0wvHr1lVo9IBAMTHt7xYBYzsYzU1NW26bqtzAiJiAy4AVqrqXzyvf0SbAwDY51kZpEBG2tE/hE1fGtnDRp7ue29ANM4F1FNYWNh6pTbgcDhYuHAh559/Ptu2bSMtLY0ffvghoG20l0BrjnTMphfMqdnpdAbluq32BFTVJSKjAXdQLAghjhpjKEiAJM8ege+Lv2V3wXaSElIYeMrIFs+PtqEgIKDDMtu2bWPGjBl89dVXAFx77bXcfffddOzYMWBtBIJwDkWFA7PpBXNqDhb+DgctBv4gIneratRmcyguMEyviLeRYDOcQP2E8NA+F5EQ33zC+WglMzMzINd5+umnmTdvHm63m169evHII48wbNiwgFw70ARKc7RgNr0Q3Zqzs7Pp27cvTqeTnj17smTJEtLT0/nuu+84++yzOeWUU7x133rrLe/wVVgCyIlIfV7FGcAcoFxE9onId/WvoFgVJBwOY6ysKt5Gol1wu11s2v46AOf6CBPxyVW38GaXc0JqY6AJVLd5yJAhxMfHM2PGDDZu3BixDgDMN1RgNr0Q3ZqTkpLYuHEjmzdvJiMjg6VLj0655uXlsXHjRu+r4fxFuIaDngRWAJOD0nqIqZ9ZL01MIN4Wx/bvPuZwxSE6p3ejT/cBTerXzwVAdM4HAMc9VFNWVsbLL7/MNddcA0D//v359NNPoyKjU6QNTwUbs+mF2NE8aNAgtm/f7lfdYPUEWnMCAqCqG4LSeohx69FpjUR7HBta2BvQcEXQmILNoTEwCLhcrjaf8+abb3LLLbdw4MABsrOzueQSYwd1NDgAOD7N0YzZ9EJgNI9e+mkALGnKP6f6tzve5XKxYcMGJk8++oy9d+9eb1j1IUOGsGjRIm9ZuGIH2UTkfBrHDGqEqv47sCYFD4ejwXSGVvPRN4bpx4aJiNbdwb6orKwkOzvbr7pFRUXMmzePVatWATBw4EBOPvnkYJoXFNqiORYwm16Ibs3V1dWMGDGCAwcO0Lt3b84//3xvWf1wkC/c7uCszWnNCSQCz9C8E1AgtDEB2kFScjJQjE3g013vUFtXQ5/uZ5LbqXujerG0O9ifp3dVZdWqVcydO9cbrvaOO+5g2rRpUZnBKVp6LIHCbHohMJr9fWIPNPVzAlVVVUycOJGlS5dy3XXXtXqeP/sEjofW9glUqmovVT2pmVfUOACAggNHALDbpEHegOYTyUe7AwD/klP/5S9/Ydq0aRw+fJiRI0fy3nvvccMNN0SlAwDzJSE3m16IDc3JycksWLCAxx9/3K9J37q64CzMPK4ActFK0fdGgvl4qWbHvk+ItyVwdp8LG9WJ5t3BvvDn6WHixImcccYZPPLII6xevZq8vLzgGxZEgvXEFKmYTS/Ejub+/fvTr18/7xBsSwRrb4RfE8OxQmJKAk5HDbUpleCALhk9SE5sHGMjVuYC6klPb5oic8+ePSxcuJAHH3yQtLQ0UlNTeeedd4iLi41nAl+aYxmz6YXo1rxv375Gn1esWOF9v3lz84tQwrJPQFX9j0IUBbhcxsSK2oxIfJlpjeOPRFvqSH8oKiryvnc6nTz22GMMHz6cl156iYceOqoxVhwANNZsBsymF8ypOVj7BEL2P19ExojIThHZJSJzfZTPFpEdIvKFiLwtIj2DYAMAqocByGrgBGJpRVBD6p+YduzYwZgxY7j77rupqalh0qRJzJw5M8zWBYdofko8HsymF8ypOVz7BAKCJwjd48BFwH7gIxFZq6o7GlT7FBioqlUicgPwADApkHZUlxoJZdRVAkBWWmegqQOIlV4AGKFzn3jiCRYvXozT6aRbt248/PDDERHtM1g4HI5wmxBSzKYXzKk5WPsEQtUTGAzsUtU9quoAVgLjGlZQ1XdUtcrzcSvQnSDhUo8T6GgsM4tVBwCwadMmFi1ahNPpZMqUKbz//vsx7QDAWIdtJsymF8ypOVz7BAJFN6DhbMh+YEgL9acAbwTSgCOHq7zvXZ6eQKanJ1BPrDgAl8vl7TpOmDCBnTt3Mnr0aM45J7rjIPmL2dbNm00vmFNzsFZEhcoJ+I2ITAYGAj7jOh86dIgpU6Zgt9txuVxMmDCB6dOnU1BQQEpKCjabjbKyMnJycjh8+DCqSk5ODh+9/1/vNRwOI/hUXaXy3uXTvcdLSkpwuVxUVlbSpUsXCgoKiI+PJz09naKiItLT03E4HFRXV3vLExISSEtLo7i4mIyMDKqrq6mpqfGWd+jQgaSkJEpKSsjKyqK8vByHw+EtT0pKIiEhgdLSUrKzsyktLaWurs5b3pKmgwcPevOOVlRUkJuby7p167jvvvtYsmQJXbp0obq6mptuuonKykpqa2ujUlNhYSEiQmZmJoWFhXTs2LHF+1RTU+O1O1Y0tXSf9uzZQ15eXkxpau0+7d27l169erVZU1JSEsnJybhcLu9vCIDdbqeurs778ORyuYiPj/dOxtpsNpxOJzabDVXF7XYTHx9PXV0dIuJ3ud1ux+12NyqPi4sjLi7OW+5yuVDVRuUigsPhIDExsdnyek1VVVWUlpY2uU/N/uaGItemiAwF7lHV//F8ngegqvcfU+9C4DFgpKoe8nWtLVu2aJ8+fdpsw8fv7eXd1/9LaYKN3al34nbXsWzWJt490dgnEO1DQaWlpdx5550sX74cgCuuuIInn3ySAwcOcMIJJ4TZutBiNs1m0wvHr7msrCzswed69OjRZJnoggULeP7558nKysLhcDBnzhwuv/zyRnUcDodfWdF8ady2bdsno0aNGuirfqjmBD4CThWRk0QkAbgSWNuwgoiciRG19LLmHEAgONyhDre7jpTENDokJHuPR7MDeP311xk6dCjLly8nISGB3//+9/zpT38CaFOu0VjBbJrNphdiU/P111/Pxo0b+dvf/sZvf/vbJjuEw7JPIFCoqhO4CVgPfAX8XVW3i8i9InKZp9oiIBV4SUQ+E5G1zVzuuHC6jR6PW3zPB0QjRUVF/OY3v2Hy5MkUFBQwePBgNmzYwOzZs73jh8XFxWG2MvSYTbPZ9EJsaz755JNJTk7myJEjjY6HLb1koFDV14HXjzl2V4P3FzY5KYAUVRpLyjrEV4Dz6MqgaKa2tpa33nqLlJQU7rzzTqZOndpk01dGRkaYrAsfZtNsNr0QGM3BShjV3tDzn3/+Ob169SInJ6fRcbs9OD/XETcxHCwqHcYEUJyt1HACablRGSfohx9+IDc3F5vNRrdu3XjmmWc47bTTOPHEE33Wr66uDvsYaKgxm2az6YXY1LxkyRJeeOEFdu/ezQsvvNCk3O12B2VIyDROoNYTMsKF0cXKTOtM4duvAdGxQ9jtdvPcc8/xhz/8gdtvv53rr78eoNU1/zU1NaEwL6Iwm2az6YXAaI60ZFHXX389M2bM4I033uDmm2/mk08+oUOHo3nPg7VPIHYCxrRCVYmxucRZv1GsQciISJ8U3rVrF5deeilz5syhoqKCL774wu9zzbie2myazaYXYlvzxRdfzIABA1i5cmWj4+HKJxAzODzDQQ5XUycQqTidTh555BHOPfdctmzZQufOnVm2bBl//vOf/b5GLMRdbytm02w2vRDdmquqqujXr5/39fjjjzepM2fOHP785z83evoPVj4B0wwHOT3B4+rEGA4quHdZGK1pnYMHD3LllVfy+eefA/CLX/yCP/7xj22eEGvYnTQLZtNsNr0Q3Zr9Wdk0YMAAPvzww0bHghXp1zxOoKwGQan19ARq/vkZ8UjEzgdkZ2djs9no0aMHixcv5oILLjiu6yQlJQXYssjHbJrNphfMqdlyAu1E61y4pBK3OkmogXinRNwu4Q8++IAePXrQtWtXbDYby5Yto1OnTq1u+26JkpKSmFtF0Rpm02w2vWBOzfVhKQKNaeYEbJUOHHGlAKSUG0NDkeIAKioqmDt3LmPHjuXWW2/1hozt3r17uxwAQFZWViBMjCrMptlsesGcmoO1T8A0TgDAIUedQKQMA/373/9m2LBhPPXUU8TFxdGvX7+ALgUrLy8P2LWiBbNpNpteMKfm+mB3gcYUw0GOWmO7dV2ckVYyuSL8vYAjR45wxx13ePOL9u/fn8cee4wzzjgjoO2YMfmG2TSbTS+YU3Owgn2awgnUf3cNewLhpLKykuHDh/PDDz+QmJjI7373O2666aagdPdieT11c5hNs9n0gjk1W/sEAkCdJ3hclx6nhNWOlJQUJk6cyNlnn83GjRuZNWtW0Mb7onk99fFiNs1m0wvRrfmhhx5i6NChDB8+nBEjRrBw4ULuvffeRnX+85//MGSIkXfrxz/+MWPHjm20T2DEiBEBSxJlip5AncMYDnJ4hoPOuvXmkLavqqxcuZLs7GxvmIfbb78du90etGVf9ZhxKZ3ZNJtNL0Sv5g8//JD169fz7rvvkpiYSHFxMTt37mT69OncdZc3niarV69ulE+goqKCAwcOkJeXx86dOwNqkyl6AuWlRpwRh60CCG0Y6e+++46JEycyffp0Zs2a5Z3QSkhICLoDqG/HbJhNs9n0QvRqPnjwIFlZWSQmJgLGKqdzzjnMWWeOAAASvElEQVSHTp068fHHH3vrrVmzppETGD9+PK+88goAq1atapJwpj2YoieACIpS55kTyEwNvhNwu90sXbqU+fPnU1lZSUZGBnfddVe7l3y2ldLSUjp16hTSNsON2TSbTS8ERvODt78ZIGsac+v/jWm27Pzzz2fRokUMGjSIkSNH8rOf/Yxhw4YxYcIEVq9ezcCBA/noo4/IyMjg5JNP9p536aWXMn36dG6++WbWr1/Pk08+yYsvvhgQe03REwBwSTXuOCfxNZCUmBLUtr7++msuueQS5s6dS2VlJePGjWPLli1MmjQJkdBOSmdnZ4e0vUjAbJrNpheiV3NqairvvPMOixcvJisriylTpvDCCy/ws5/9jLVr1+J2u1m9ejUTJkxodF5mZiadOnVi1apV9O7dm+Tk5GZaaDvm6AkADjHmA1Iqgvsj7Ha7mTx5Mrt27SI3N5dFixbx05/+NKhttkRpaSkpKcF1epGG2TSbTS8ERnNLT+zBxGazMXz4cIYPH07fvn1ZuXIlv/zlL+nZsyfvv/8+r776KuvXr29y3mWXXcZtt93mM+BcezCPE/DsFk6uCG47cXFx3H///axZs4b58+eHvZserMiDkYzZNJtNL0Sv5m+++Ya4uDjvUM+XX35Jjx49ALj88su54447yMvLo1u3bk3OvfjiiykuLuaCCy4I6Ooo0ziBuvqQEQHuCVRXV7No0SKcTqd3mdeoUaMYNWpUQNs5Xsy4ntpsms2mF6JXc2VlJb/73e8oLS3FbrfTq1cvFi9eDMC4ceOYO3cuCxcu9HluRkYGN98c+JWNpnEC9ctDkwO4UWzr1q3MnDmTXbt2YbfbmTZtGt27dw/Y9QNBQUEBPXv2DLcZIcVsms2mF6JX84ABA3wO9YCxUujQoUNNjteHk6+trfWuKjrxxBPZvDkwmdFMMTGsqkd3CwegJ1BeXs5tt93G2LFj2bVrF71792bdunUR5wAA040Vg/k0m00vmFOzFUq6HVQ5XEfjBrUz7tRbb73F7Nmz2b9/P3a7nVmzZnHLLbd4PXSkEYzQs5GO2TSbTS+YU3OwVhaaoidQUes6Gka6nT2B5cuXs3//fgYMGMA777zD7bffHrEOAKCsrCzcJoQcs2k2m14wp2Yrimg7qHW5cUoVAIlVbXMCqkp5ebk3gcXChQsZPHgw06ZNC1q8n0CSk5MTbhNCjtk0m00vmFOzlU+gnShGjH5bG0L1FxQUcPXVVzNu3DicTiP+UG5uLjfeeGNUOACAw4cPh9uEkGM2zWbTC+bUHKyegHmcgBhfYOdzB7deV5Xly5dz9tln89prr7F792527NgRbBODQrBikEcyZtNsNr1gTs3BwjxOAMMJDHxuUYv18vPzmTBhAjNnzqSsrIzRo0ezZcsW+vfvHwozA44Zu81m02w2vRDdmus3hwH861//YtCgQezbt48FCxbQrVs3CgsLfdbt0qULv//9772fH3vsMRYsWNBue0zhBFTV2xOw2Zofxlm2bBnDhg1jw4YNZGZm8tRTT7FixQqfu/eihYMHD4bbhJBjNs1m0wuxoXnDhg3MnTuXv//9794f+8zMzGbDQiQmJrJu3TqKi4sDaodJnIBnIkCFOGlessvloqqqigkTJrB161YmTpwY8oBvgSbUUUsjAbNpNpteiH7NmzdvZtasWaxYsYKTTjrJe/yqq67i5ZdfpqSkpMk5NpuNa665hieeeCKgtkTH7GY7casxqRunjR2Aw+Hgyy+/5KyzzgLg2muv5bTTTmP48OEht9HCwiK0XPnAT4Jy3ZW3fdJieW1tLZMnT+bVV1+ld+/ejcpSUlK46qqrWLJkCfPmzWty7pQpUzj33HOZOXNmwOw1RU/A5TacgDRwAp9++imjRo1i3Lhx7Nu3DzB25MWaA6ioCHLEvAjEbJrNpheiW3N8fDyDBw/m+eef91l+3XXXsXLlSm8CqoZ07NiRSZMm8eSTTwbMnpD1BERkDPAIYAOWquqCY8oTgb8CPwGKgUmqujcQbde5agGIUzvV1dUsWLCAxx9/HLfbTV5eHkVFRY0mYGKJ3NzccJsQcsym2Wx6ITCaW3tiDxZxcXE8++yzjB8/nocffpjZs2c3Kk9PT2fixIk888wzjY7XD03fcMMNnHfeefzyl78MjD0BuUoriIgNeBy4GOgL/EJE+h5TbQpQoqqnAIsB36H0jgOnxwnYNJ5zzz2Xxx57DIDp06fz3nvvceaZZwaqqYij4UoDs2A2zWbTC9GvOTk5mRdffJGXXnrJZ4/gxhtvZNmyZd79SXB0WWxGRgbjx49n+fLlAbElVMNBg4FdqrpHVR3ASmDcMXXGAX/xvP8HMEoCNCvrdHt6Am47e/bs4Uc/+hHr169n/vz5Ac3QE4lE+8T28WA2zWbTC7GhOSMjg5deeomHHnqIN954o1FZVlYWl1xyCbW1tT7PnT59esA2zEkoNl2IyERgjKpO9Xz+FTBEVW9qUOdLT539ns+7PXWKGl7rlVde0Xnz5mG323G5XEyYMIHp06dTUFBASkoKNpuNsrIycnJyOHz4MKrKB998zT8+vo3Ums50dw9h8uTJdO/encLCQkSEzMxMCgsL6dixIy6Xi8rKSrp06UJBQQHx8fGkp6dTVFREeno6DoeD6upqb3lCQgJpaWkUFxeTkZFBdXU1NTU13vIOHTqQlJRESUkJWVlZlJeX43A4vOVJSUkkJCRQWlpKdnY2paWl1NXVecub05STk8PBgwe9qyQqKirIzc1toikxMRG73R5Tmlq7Tx06dEBEYkpTS/epoKCAzp07x5Sm1u7ToUOH6NKlS5s1JSUlkZmZicvl8v6GgBGSoa6uzhuYzuVyER8f730St9lsOJ1ObDYbqorb7SY+Pp66ujpExO9yu92O2+1uVB4XF0dcXJy33OVyoaqNykUEp9NJfHx8s+X1mgoLC6mrq2t0n/Lz8z8ZNWrUQJ+/z9HmBLZs2aJ9+vRpU/v/3ZvPP95di7riuHPK9HaqiS7y8/OjMu56ezCbZrPphePXXFZW5o0DFm00zCfQEr40btu2rVknEKqJ4e+BhjOv3T3HfNXZLyJ2IB1jgrjd9Mnrye9/PcPn2ttYJ1r/4NuD2TSbTS+YU3OwwmeHak7gI+BUETlJRBKAK4G1x9RZC1zjeT8R+LcGuJsSrABMkYylOfYxm14wp+ZgjdqExAmoqhO4CVgPfAX8XVW3i8i9InKZp9ozQJaI7AJmA3MDbUdlZWWgLxnxWJpjH7PphePXLCI4HI4AWxMa3O7WQyA7HI42T5qHbJ+Aqr4OvH7MsbsavK8BrgimDdGanLo9WJpjH7PphePXnJqaSkVFBTU1NQG2KPi4XK5mVwvVIyJtDqlhirAR9URrcur2YGmOfcymF45fs4iQlpYWBIuCT7AWAJgibEQ9a9asCbcJIcfSHPuYTS9YmgOJqZzA6tWrw21CyLE0xz5m0wuW5kBiKifQcAu2WbA0xz5m0wuW5kASks1igeTtt98uBPKP59zDhw9nZ2ZmFrVeM3awNMc+ZtMLlubjoOeoUaN8pmOLOidgYWFhYRE4TDUcZGFhYWHRGMsJWFhYWJiYmHQCIjJGRHaKyC4RabLzWEQSReRFT/kHIpIXeisDhx96Z4vIDhH5QkTeFpGoX1TemuYG9S4XERURn8Gzogl/NIvIzz33eruIvBBqGwONH3/bJ4rIOyLyqefve2w47AwUIvKsiBzyBNT0VS4i8qjn+/hCRM5qd6OqGlMvjMxlu4FeQALwOdD3mDo3Aks8768EXgy33UHWez6Q7Hl/QzTr9Vezp14asBHYCgwMt90huM+nAp8CGZ7PncNtdwg0PwXc4HnfF9gbbrvbqXkEcBbwZTPlY4E3AAHOBj5ob5ux2BMIawKbMNCqXlV9R1WrPB+3YkRxjWb8uccA8zEy1EVfjICm+KP5f4HHVbUEQFUPhdjGQOOPZgXqQ4qmAz+E0L6Ao6obgZayxYwD/qoGW4FOInJCe9qMRSfQDdjX4PN+zzGfddQIblcKZIXEusDjj96GTMF4kohmWtXs6Sb3UNXXQmlYEPHnPvcGeovI+yKy1ZPXO5rxR/M9wGQR2Y8Rm2xGaEwLG239/94qpoodZHZEZDIwEBgZbluCiYjEAQ8Dvw6zKaHGjjEkdB5Gb2+jiJyhqkfCalVw+QWwTFUfEpGhwPMicrqqth5y0wKIzZ5AWxLYEOgENmHAH72IyIXAHcBlqtpyKMLIpzXNacDpwLsishdj7HRtlE8O+3Of9wNrVbVOVb8FvsZwCtGKP5qnAH8HUNUtQAcgOyTWhQe//r+3hVh0AhGRwCaEtKpXRM4EnsRwANE+TgytaFbVUlXNVtU8Vc3DmAe5TFU/Do+5AcGfv+s1GL0ARCQbY3hoTyiNDDD+aP4OGAUgIj/CcAKFIbUytKwFrvasEjobKFXVA+25YMwNB6mqU0TqE9jYgGfVk8AG+FhV12IksHnek8DmMMYfV1Tip95FQCrwkmf++ztVvazZi0Y4fmqOKfzUvB4YLSI7ABcwR1WjtYfrr+ZbgKdF5LcYk8S/juIHOkRkBYYjz/bMc9wNxAOo6hKMeY+xwC6gCri23W1G8fdlYWFhYdFOYnE4yMLCwsLCTywnYGFhYWFiLCdgYWFhYWIsJ2BhYWFhYiwnYGFhYWFiLCdgERWIyLsiMjXcdrSGiNwuIktbKL9KRP4ZSpssLFrCWiJqEXI8u3hzMday19NbVZsN/iUi7wLLVbXZH9hIwxOi/Fsg3hOjKmKIZNssQovVE7AIF5eqamqDV1RHf4w0RMQWbhssogPLCVhEBCKSISLrRKRQREo8732GvBaRU0Rkg4iUikiRiLzYoKyPiPxLRA57kpH8vIU23xWR+0XkQxEpE5FXRCSzQfllnuQsRzx1f9Sg7Hci8r2IlHvaqQ9dcI+ILPdU2+j594iIVIjIUBH5tYi856n7hIg8eIxNr4jIbM/7riKyyvOdfCsiM1vQssxzvddFpBI4X0QuESPZSpmI7BORexqc0sQ2z3V+IyJfee7BeomBBEQWLWM5AYtIIQ54DugJnAhUA39qpu584J9ABkYArccARCQF+BfwAtAZIxzIn0WkbwvtXg38BjgBcAKPeq7VG1gBzAJyMLbrvyoiCSJyGnATMEhV04D/Afb6uPYIz7+dPL2dLceUrwAmiSeWh4hkAKOBlWJEQn0VI5FKN4z4OLNE5H9a0PJL4D6MAHrvAZUefZ2AS4AbRGR8c7aJyDjgdmCCR/Mmj40WMYzlBCzCxRrPE/YREVmjqsWqukpVq1S1HOPHrLmQ13UYzqKrqtao6nue4z/FyCz1nKo6VfVTYBVwRQt2PK+qX6pqJXAn8HPPUMok4DVV/Zeq1gEPAknAORhzGYlAXxGJV9W9qrr7OL6DTRjxbs71fJ4IbPEMjQ0CclT1XlV1qOoe4GlajnP1iqq+r6puz/fyrqr+x/P5C4wf9JbCiF8P3K+qX3nmCf4PGGD1BmIbywlYhIvxqtrJ8xovIski8qSI5ItIGcZwRadmxrZvw0iv96FnuOY3nuM9gSENnMsR4CqgSwt2NEzQkY8RrCsb6Or5DIAnPv0+oJuq7sLoIdwDHBKRlSLSta1fgCfQ2UqMmPhgPMn/rYGWrsdouR1jQt0fLYjIEDHy7xaKSCnGj3xLYZZ7Ao80aO8wxvfcrqQlFpGN5QQsIoVbgNOAIarakaPDFU3Sfqpqgar+r6p2Ba7DGPI5BeNHcEMD51I/1HFDC+02jM1+IkYvowgjTaH3CdgzZNMDT+x2VX1BVYd76ihGGssmpvqhewUw0fO0PQSj54JHy7fHaElT1ZYSqR/b3gsYoYd7qGo6sISj36cv2/YB1x3TZpKqbvZDh0WUYjkBi0ghDWMe4Ihncvbu5iqKyBUNJo1LMH7Q3MA6jPSKvxKReM9rUMMJXR9MFpG+IpIM3Av8Q1VdGIlKLhGRUSISj+GkaoHNInKaiFwgIokY+YurPe0fS6HneK/mGvcMWRUBS4H1DbKAfQiUeyagk0TEJiKni8igFrQcSxpwWFVrRGQwRk+jJduWAPNEpB+AiKSLSEtDaRYxgOUELCKF/4cx5l6EkQTmzRbqDgI+EJEKjCfdmz3JyMsxJlavxHiSL8B4Qk9s4VrPA8s8dTsAMwFUdScwGWPSuQi4FGNZq8NzvQWe4wUYk9Dzjr2wqlZhzG287xliObsZG14ALvT8W3+uC2OOYwDGev56R5HegpZjuRG4V0TKgbvwZOBqzjZVfRnj+1rpGZL7Eri4De1ZRCHWZjEL0yJRuAHNwiLQWD0BCwsLCxNjOQELCwsLE2MNB1lYWFiYGKsnYGFhYWFiLCdgYWFhYWIsJ2BhYWFhYiwnYGFhYWFiLCdgYWFhYWIsJ2BhYWFhYv4/umcXMUGrV2AAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DSgN6zHldrgx",
        "colab_type": "code",
        "outputId": "cac9b952-b68e-4ee5-eed4-37ac9b85ba09",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print(\"RF AUC score: \" + str(metrics.auc(fpr_rf, trp_rf)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "RF AUC score: 0.854314947716591\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E3vwyhF8d6Ic",
        "colab_type": "code",
        "outputId": "b4b79f59-dfa6-42dd-db99-f226838cb9a6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print(\"SVM AUC score: \" + str(metrics.auc(fpr_svm, trp_svm)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "SVM AUC score: 0.8440156539397821\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R8bSo26YzLdA",
        "colab_type": "code",
        "outputId": "a3b0c352-d1e0-4355-ec08-596f6c1a1be4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print(\"logistic regression AUC score: \" + str(metrics.auc(fpr_lr, trp_lr)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "logistic regression AUC score: 0.7751453248853633\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8HhTvk0R6pgG",
        "colab_type": "code",
        "outputId": "21044db5-bf74-40a4-9c25-45d063074561",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print(\"KNN AUC score: \" + str(metrics.auc(fpr_knn, trp_knn)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "KNN AUC score: 0.8050258580113457\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_TZiAjK065DJ",
        "colab_type": "text"
      },
      "source": [
        "Looks like our best model is random forest according to AUC score."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NOKnMgZg7NvT",
        "colab_type": "text"
      },
      "source": [
        "## best Model with recall - random forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6pa1KJKU7Qlq",
        "colab_type": "code",
        "outputId": "f68ce006-0103-47b2-ed19-649a575589d2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_pred_rf = best_RF_model.predict(X_test)\n",
        "recall = metrics.recall_score(y_test, y_pred_rf)\n",
        "print(\"random forest recall: \" + str(recall))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "random forest recall: 0.4793713163064833\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2dhaUTrd7SGE",
        "colab_type": "code",
        "outputId": "31d6e3cc-176f-4274-921c-418c37bd2ebf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_pred_svm = best_SVM_model.predict(X_test)\n",
        "recall = metrics.recall_score(y_test, y_pred_svm)\n",
        "print(\"svm recall: \" + str(recall))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "svm recall: 0.46954813359528486\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cGIDZ-oT9Yeq",
        "colab_type": "code",
        "outputId": "511d7a08-edea-4069-adcd-9f39cbb61a3c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_pred_lr = best_LR_model.predict(X_test)\n",
        "recall = metrics.recall_score(y_test, y_pred_lr)\n",
        "print(\"logistic regression recall: \" + str(recall))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "logistic regression recall: 0.19646365422396855\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OVnmjaBq9auN",
        "colab_type": "code",
        "outputId": "40660c26-7618-46b8-d4ff-24e973764a4e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_pred_knn = best_KNN_model.predict(X_test)\n",
        "recall = metrics.recall_score(y_test, y_pred_knn)\n",
        "print(\"KNN recall: \" + str(recall))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "KNN recall: 0.35952848722986247\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZezRBWxK9sUB",
        "colab_type": "text"
      },
      "source": [
        "Looks like random forest is the best model with respect to recall as well, with a recall score of 0.48. At this point, should ask manager if this score is acceptable. If not, really need to look into more powerful models and add more features and collect more data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HFlf3wYm-Brr",
        "colab_type": "text"
      },
      "source": [
        "# feature importance detection"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "I8898PT_-OY-",
        "colab_type": "text"
      },
      "source": [
        "Suppose now manager asks why those predictions turn out to the way they did. Now I will use random forest's perk of being able to generate feature importance to answer that question. We could also use l1, l2 regularized coefficients to see feature importance, but I think using random forest is a more reliable way."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-WFfOl7R-h5W",
        "colab_type": "code",
        "outputId": "ebb5bada-dc38-46c2-cebc-9ba89008f5fc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "feature_importance_array = best_RF_model.feature_importances_\n",
        "print(feature_importance_array)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.14982928 0.02057751 0.23106551 0.08823552 0.14642457 0.11951589\n",
            " 0.0184491  0.03376532 0.15317792 0.02383227 0.01512713]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T19jEZ-J-l16",
        "colab_type": "code",
        "outputId": "a972877f-bcf0-4106-a363-cfa0c2e30c2f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "source": [
        "X.columns"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['CreditScore', 'Gender', 'Age', 'Tenure', 'Balance', 'NumOfProducts',\n",
              "       'HasCrCard', 'IsActiveMember', 'EstimatedSalary', 'Geography_Germany',\n",
              "       'Geography_Spain'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 165
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H69go16K-_Hl",
        "colab_type": "code",
        "outputId": "61d1b37c-e091-41db-efee-7411eeda61c0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "for index, feature in enumerate(X.columns):\n",
        "  print(feature + \"'s imprtance score: \" + str(feature_importance_array[index]))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "CreditScore's imprtance score: 0.1498292754019354\n",
            "Gender's imprtance score: 0.02057750935908157\n",
            "Age's imprtance score: 0.2310655061622185\n",
            "Tenure's imprtance score: 0.08823551846322243\n",
            "Balance's imprtance score: 0.14642456932407386\n",
            "NumOfProducts's imprtance score: 0.11951588519451901\n",
            "HasCrCard's imprtance score: 0.018449095510937933\n",
            "IsActiveMember's imprtance score: 0.03376531808985643\n",
            "EstimatedSalary's imprtance score: 0.1531779214440335\n",
            "Geography_Germany's imprtance score: 0.023832271399405622\n",
            "Geography_Spain's imprtance score: 0.015127129650715818\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LAsWIoWoAHPc",
        "colab_type": "text"
      },
      "source": [
        "According to our random forest, customer's age is the top influencing factors on our prediciton. The second and third important factors are salary and credit score. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nnIUf-OOAZsC",
        "colab_type": "text"
      },
      "source": [
        "# Conclusion/Analysis"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-ElDxTY9AckH",
        "colab_type": "text"
      },
      "source": [
        "\n",
        "\n",
        "* We first downloaded and uploaded dataset from Kaggle about customer churn. Generated google drive link for our dataset.\n",
        "* loaded csv into dataframe in google Colab using the download link, look at data and see its size is 10000 data points with 13 features and 1 target variable, no missing data\n",
        "* did EDA on data to see distributions of categorical and numerical data. Interestingly age seems like a important factor when we look at the distribution.\n",
        "* removed unnecessary columns like ID's. Did one-hot encoding on categorical columns.\n",
        "* Trained various optimal models, and evaluated them using AUC and recall.\n",
        "* the best model is random forest, both in AUC and recall, with AUC score of 0.854 and recall of 0.479.\n",
        "* Finally confirmed that age is the most important factor in predicting if customer will churn.\n",
        "* business advice to boss: older people (40 - 50 years old) are more likely to churn than (30 - 40 years old) younder people, so target on them: maybe give them some incentive credits/money to make themstay in our bank. Use random forest for prediction, as it turns out that is the best model \n",
        "\n"
      ]
    }
  ]
}