code.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Aqb9XVKiB8PP",
        "outputId": "00da81da-8ad5-48bd-d32e-91dc1a874b84"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ]
        }
      ],
      "source": [
        "# Import librabry\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import numpy as np\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n",
        "pd.set_option('display.max_column', None)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R2xCSbGHR32d"
      },
      "source": [
        "# Data Cleansing"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Read dataset and solve datatype dataset"
      ],
      "metadata": {
        "id": "bZk99mIr-ZRz"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        },
        "id": "FbQuKkh-CUZB",
        "outputId": "92a7c3e3-ce16-4325-b650-17d1b3160036"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   Unnamed: 0  accountNumber  customerId  creditLimit  availableMoney  \\\n",
              "0           0      733493772   733493772         5000         5000.00   \n",
              "1           1      733493772   733493772         5000         4888.67   \n",
              "2           2      733493772   733493772         5000         4863.92   \n",
              "3           3      733493772   733493772         5000         4676.52   \n",
              "4           4      733493772   733493772         5000         4449.18   \n",
              "\n",
              "   transactionDateTime  transactionAmount merchantName acqCountry  \\\n",
              "0  2016-01-08T19:04:50             111.33         Lyft         US   \n",
              "1  2016-01-09T22:32:39              24.75         Uber         US   \n",
              "2  2016-01-11T13:36:55             187.40         Lyft         US   \n",
              "3  2016-01-11T22:47:46             227.34         Lyft         US   \n",
              "4  2016-01-16T01:41:11               0.00         Lyft         US   \n",
              "\n",
              "  merchantCountryCode  posEntryMode  posConditionCode merchantCategoryCode  \\\n",
              "0                  US           5.0               1.0            rideshare   \n",
              "1                  US           9.0               1.0            rideshare   \n",
              "2                  US           5.0               1.0            rideshare   \n",
              "3                  US           2.0               1.0            rideshare   \n",
              "4                  US           2.0               1.0            rideshare   \n",
              "\n",
              "  currentExpDate accountOpenDate dateOfLastAddressChange  cardCVV  enteredCVV  \\\n",
              "0        04/2020      2014-08-03              2014-08-03      492         492   \n",
              "1        06/2023      2014-08-03              2014-08-03      492         492   \n",
              "2        12/2027      2014-08-03              2014-08-03      492         492   \n",
              "3        09/2029      2014-08-03              2014-08-03      492         492   \n",
              "4        10/2024      2014-08-03              2014-08-03      492         492   \n",
              "\n",
              "   cardLast4Digits       transactionType  isFraud  echoBuffer  currentBalance  \\\n",
              "0             9184              PURCHASE     True         NaN            0.00   \n",
              "1             9184              PURCHASE    False         NaN          111.33   \n",
              "2             9184              PURCHASE    False         NaN          136.08   \n",
              "3             9184              PURCHASE     True         NaN          323.48   \n",
              "4             9184  ADDRESS_VERIFICATION    False         NaN          550.82   \n",
              "\n",
              "   merchantCity  merchantState  merchantZip  cardPresent  posOnPremises  \\\n",
              "0           NaN            NaN          NaN        False            NaN   \n",
              "1           NaN            NaN          NaN        False            NaN   \n",
              "2           NaN            NaN          NaN        False            NaN   \n",
              "3           NaN            NaN          NaN        False            NaN   \n",
              "4           NaN            NaN          NaN        False            NaN   \n",
              "\n",
              "   recurringAuthInd  expirationDateKeyInMatch  \n",
              "0               NaN                     False  \n",
              "1               NaN                     False  \n",
              "2               NaN                     False  \n",
              "3               NaN                     False  \n",
              "4               NaN                     False  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-5134725a-0db3-44bc-9bef-1b55a89d13d5\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>accountNumber</th>\n",
              "      <th>customerId</th>\n",
              "      <th>creditLimit</th>\n",
              "      <th>availableMoney</th>\n",
              "      <th>transactionDateTime</th>\n",
              "      <th>transactionAmount</th>\n",
              "      <th>merchantName</th>\n",
              "      <th>acqCountry</th>\n",
              "      <th>merchantCountryCode</th>\n",
              "      <th>posEntryMode</th>\n",
              "      <th>posConditionCode</th>\n",
              "      <th>merchantCategoryCode</th>\n",
              "      <th>currentExpDate</th>\n",
              "      <th>accountOpenDate</th>\n",
              "      <th>dateOfLastAddressChange</th>\n",
              "      <th>cardCVV</th>\n",
              "      <th>enteredCVV</th>\n",
              "      <th>cardLast4Digits</th>\n",
              "      <th>transactionType</th>\n",
              "      <th>isFraud</th>\n",
              "      <th>echoBuffer</th>\n",
              "      <th>currentBalance</th>\n",
              "      <th>merchantCity</th>\n",
              "      <th>merchantState</th>\n",
              "      <th>merchantZip</th>\n",
              "      <th>cardPresent</th>\n",
              "      <th>posOnPremises</th>\n",
              "      <th>recurringAuthInd</th>\n",
              "      <th>expirationDateKeyInMatch</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>5000.00</td>\n",
              "      <td>2016-01-08T19:04:50</td>\n",
              "      <td>111.33</td>\n",
              "      <td>Lyft</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>04/2020</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>True</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.00</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>4888.67</td>\n",
              "      <td>2016-01-09T22:32:39</td>\n",
              "      <td>24.75</td>\n",
              "      <td>Uber</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>9.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>06/2023</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>111.33</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>4863.92</td>\n",
              "      <td>2016-01-11T13:36:55</td>\n",
              "      <td>187.40</td>\n",
              "      <td>Lyft</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>12/2027</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>136.08</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>4676.52</td>\n",
              "      <td>2016-01-11T22:47:46</td>\n",
              "      <td>227.34</td>\n",
              "      <td>Lyft</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>09/2029</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>True</td>\n",
              "      <td>NaN</td>\n",
              "      <td>323.48</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>4449.18</td>\n",
              "      <td>2016-01-16T01:41:11</td>\n",
              "      <td>0.00</td>\n",
              "      <td>Lyft</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>10/2024</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>ADDRESS_VERIFICATION</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>550.82</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5134725a-0db3-44bc-9bef-1b55a89d13d5')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-5134725a-0db3-44bc-9bef-1b55a89d13d5 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-5134725a-0db3-44bc-9bef-1b55a89d13d5');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ],
      "source": [
        "# Import and read dataset\n",
        "df = pd.read_csv('/content/drive/MyDrive/transactions.csv')\n",
        "df.head()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Check missing value\n",
        "df.isnull().sum()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "z6XY_f9Vt-Jc",
        "outputId": "c1f98e21-96b9-476b-b88e-497bcd572115"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Unnamed: 0                       0\n",
              "accountNumber                    0\n",
              "customerId                       0\n",
              "creditLimit                      0\n",
              "availableMoney                   0\n",
              "transactionDateTime              0\n",
              "transactionAmount                0\n",
              "merchantName                     0\n",
              "acqCountry                    3913\n",
              "merchantCountryCode            624\n",
              "posEntryMode                  3345\n",
              "posConditionCode               287\n",
              "merchantCategoryCode             0\n",
              "currentExpDate                   0\n",
              "accountOpenDate                  0\n",
              "dateOfLastAddressChange          0\n",
              "cardCVV                          0\n",
              "enteredCVV                       0\n",
              "cardLast4Digits                  0\n",
              "transactionType                589\n",
              "isFraud                          0\n",
              "echoBuffer                  641914\n",
              "currentBalance                   0\n",
              "merchantCity                641914\n",
              "merchantState               641914\n",
              "merchantZip                 641914\n",
              "cardPresent                      0\n",
              "posOnPremises               641914\n",
              "recurringAuthInd            641914\n",
              "expirationDateKeyInMatch         0\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 270
        },
        "id": "VgrCfz4oygyZ",
        "outputId": "8d9a463c-1eb9-4a93-a7fe-d7af4009b48c"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   accountNumber  customerId  creditLimit  availableMoney  \\\n",
              "0      733493772   733493772         5000         5000.00   \n",
              "1      733493772   733493772         5000         4888.67   \n",
              "2      733493772   733493772         5000         4863.92   \n",
              "3      733493772   733493772         5000         4676.52   \n",
              "4      733493772   733493772         5000         4449.18   \n",
              "\n",
              "   transactionDateTime  transactionAmount merchantName acqCountry  \\\n",
              "0  2016-01-08T19:04:50             111.33         Lyft         US   \n",
              "1  2016-01-09T22:32:39              24.75         Uber         US   \n",
              "2  2016-01-11T13:36:55             187.40         Lyft         US   \n",
              "3  2016-01-11T22:47:46             227.34         Lyft         US   \n",
              "4  2016-01-16T01:41:11               0.00         Lyft         US   \n",
              "\n",
              "  merchantCountryCode  posEntryMode  posConditionCode merchantCategoryCode  \\\n",
              "0                  US           5.0               1.0            rideshare   \n",
              "1                  US           9.0               1.0            rideshare   \n",
              "2                  US           5.0               1.0            rideshare   \n",
              "3                  US           2.0               1.0            rideshare   \n",
              "4                  US           2.0               1.0            rideshare   \n",
              "\n",
              "  currentExpDate accountOpenDate dateOfLastAddressChange  cardCVV  enteredCVV  \\\n",
              "0        04/2020      2014-08-03              2014-08-03      492         492   \n",
              "1        06/2023      2014-08-03              2014-08-03      492         492   \n",
              "2        12/2027      2014-08-03              2014-08-03      492         492   \n",
              "3        09/2029      2014-08-03              2014-08-03      492         492   \n",
              "4        10/2024      2014-08-03              2014-08-03      492         492   \n",
              "\n",
              "   cardLast4Digits       transactionType  isFraud  currentBalance  \\\n",
              "0             9184              PURCHASE     True            0.00   \n",
              "1             9184              PURCHASE    False          111.33   \n",
              "2             9184              PURCHASE    False          136.08   \n",
              "3             9184              PURCHASE     True          323.48   \n",
              "4             9184  ADDRESS_VERIFICATION    False          550.82   \n",
              "\n",
              "   cardPresent  expirationDateKeyInMatch  \n",
              "0        False                     False  \n",
              "1        False                     False  \n",
              "2        False                     False  \n",
              "3        False                     False  \n",
              "4        False                     False  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-5923c010-1102-4799-83c0-eb250e4025a4\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>accountNumber</th>\n",
              "      <th>customerId</th>\n",
              "      <th>creditLimit</th>\n",
              "      <th>availableMoney</th>\n",
              "      <th>transactionDateTime</th>\n",
              "      <th>transactionAmount</th>\n",
              "      <th>merchantName</th>\n",
              "      <th>acqCountry</th>\n",
              "      <th>merchantCountryCode</th>\n",
              "      <th>posEntryMode</th>\n",
              "      <th>posConditionCode</th>\n",
              "      <th>merchantCategoryCode</th>\n",
              "      <th>currentExpDate</th>\n",
              "      <th>accountOpenDate</th>\n",
              "      <th>dateOfLastAddressChange</th>\n",
              "      <th>cardCVV</th>\n",
              "      <th>enteredCVV</th>\n",
              "      <th>cardLast4Digits</th>\n",
              "      <th>transactionType</th>\n",
              "      <th>isFraud</th>\n",
              "      <th>currentBalance</th>\n",
              "      <th>cardPresent</th>\n",
              "      <th>expirationDateKeyInMatch</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>5000.00</td>\n",
              "      <td>2016-01-08T19:04:50</td>\n",
              "      <td>111.33</td>\n",
              "      <td>Lyft</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>04/2020</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>True</td>\n",
              "      <td>0.00</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>4888.67</td>\n",
              "      <td>2016-01-09T22:32:39</td>\n",
              "      <td>24.75</td>\n",
              "      <td>Uber</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>9.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>06/2023</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>False</td>\n",
              "      <td>111.33</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>4863.92</td>\n",
              "      <td>2016-01-11T13:36:55</td>\n",
              "      <td>187.40</td>\n",
              "      <td>Lyft</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>12/2027</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>False</td>\n",
              "      <td>136.08</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>4676.52</td>\n",
              "      <td>2016-01-11T22:47:46</td>\n",
              "      <td>227.34</td>\n",
              "      <td>Lyft</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>09/2029</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>True</td>\n",
              "      <td>323.48</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>733493772</td>\n",
              "      <td>733493772</td>\n",
              "      <td>5000</td>\n",
              "      <td>4449.18</td>\n",
              "      <td>2016-01-16T01:41:11</td>\n",
              "      <td>0.00</td>\n",
              "      <td>Lyft</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>10/2024</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>2014-08-03</td>\n",
              "      <td>492</td>\n",
              "      <td>492</td>\n",
              "      <td>9184</td>\n",
              "      <td>ADDRESS_VERIFICATION</td>\n",
              "      <td>False</td>\n",
              "      <td>550.82</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5923c010-1102-4799-83c0-eb250e4025a4')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-5923c010-1102-4799-83c0-eb250e4025a4 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-5923c010-1102-4799-83c0-eb250e4025a4');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ],
      "source": [
        "# Drop column where missing value in column 100%\n",
        "df.drop(columns=['Unnamed: 0','echoBuffer','merchantCity','merchantState','merchantZip','posOnPremises','recurringAuthInd'],inplace=True)\n",
        "df.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yGCPOQPoyC9W",
        "outputId": "fad3ab3f-b6b6-45ef-edb2-bb86ff252712"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "accountNumber                 int64\n",
              "customerId                    int64\n",
              "creditLimit                   int64\n",
              "availableMoney              float64\n",
              "transactionDateTime          object\n",
              "transactionAmount           float64\n",
              "merchantName                 object\n",
              "acqCountry                   object\n",
              "merchantCountryCode          object\n",
              "posEntryMode                float64\n",
              "posConditionCode            float64\n",
              "merchantCategoryCode         object\n",
              "currentExpDate               object\n",
              "accountOpenDate              object\n",
              "dateOfLastAddressChange      object\n",
              "cardCVV                       int64\n",
              "enteredCVV                    int64\n",
              "cardLast4Digits               int64\n",
              "transactionType              object\n",
              "isFraud                        bool\n",
              "currentBalance              float64\n",
              "cardPresent                    bool\n",
              "expirationDateKeyInMatch       bool\n",
              "dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "# Check data types\n",
        "df.dtypes"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V_1j1eGCyVE9",
        "outputId": "17c4165c-107a-459f-baf4-ee476ddfa130"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "accountNumber                  0\n",
              "customerId                     0\n",
              "creditLimit                    0\n",
              "availableMoney                 0\n",
              "transactionDateTime            0\n",
              "transactionAmount              0\n",
              "merchantName                   0\n",
              "acqCountry                  3913\n",
              "merchantCountryCode          624\n",
              "posEntryMode                3345\n",
              "posConditionCode             287\n",
              "merchantCategoryCode           0\n",
              "currentExpDate                 0\n",
              "accountOpenDate                0\n",
              "dateOfLastAddressChange        0\n",
              "cardCVV                        0\n",
              "enteredCVV                     0\n",
              "cardLast4Digits                0\n",
              "transactionType              589\n",
              "isFraud                        0\n",
              "currentBalance                 0\n",
              "cardPresent                    0\n",
              "expirationDateKeyInMatch       0\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "source": [
        "# Check missing value again\n",
        "df.isnull().sum()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4WvuCtWIydmc",
        "outputId": "d1fc73f0-c23b-463d-cf9b-7c94f841ff1d"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(641914, 23)"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ],
      "source": [
        "#Check shape dataset\n",
        "df.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uUcaB18Qy4a-",
        "outputId": "6420fdb5-59ed-4948-cbfe-cf99f4cc7e4b"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "accountNumber                object\n",
              "customerId                   object\n",
              "creditLimit                   int64\n",
              "availableMoney              float64\n",
              "transactionDateTime          object\n",
              "transactionAmount           float64\n",
              "merchantName                 object\n",
              "acqCountry                   object\n",
              "merchantCountryCode          object\n",
              "posEntryMode                 object\n",
              "posConditionCode             object\n",
              "merchantCategoryCode         object\n",
              "currentExpDate               object\n",
              "accountOpenDate              object\n",
              "dateOfLastAddressChange      object\n",
              "cardCVV                      object\n",
              "enteredCVV                   object\n",
              "cardLast4Digits              object\n",
              "transactionType              object\n",
              "isFraud                        bool\n",
              "currentBalance              float64\n",
              "cardPresent                    bool\n",
              "expirationDateKeyInMatch       bool\n",
              "dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "#Convert Data to Object\n",
        "df = df.astype({\"accountNumber\":'object', \"customerId\":'object',\"posEntryMode\":'object',\"posConditionCode\":'object',\"cardCVV\":'object',\"enteredCVV\":'object',\"cardLast4Digits\":'object'})\n",
        "df.dtypes"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yS2D36uN9VOt",
        "outputId": "6f8106be-315b-4885-b419-d774bc929c7e"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "accountNumber                       object\n",
              "customerId                          object\n",
              "creditLimit                          int64\n",
              "availableMoney                     float64\n",
              "transactionDateTime         datetime64[ns]\n",
              "transactionAmount                  float64\n",
              "merchantName                        object\n",
              "acqCountry                          object\n",
              "merchantCountryCode                 object\n",
              "posEntryMode                        object\n",
              "posConditionCode                    object\n",
              "merchantCategoryCode                object\n",
              "currentExpDate              datetime64[ns]\n",
              "accountOpenDate             datetime64[ns]\n",
              "dateOfLastAddressChange     datetime64[ns]\n",
              "cardCVV                             object\n",
              "enteredCVV                          object\n",
              "cardLast4Digits                     object\n",
              "transactionType                     object\n",
              "isFraud                               bool\n",
              "currentBalance                     float64\n",
              "cardPresent                           bool\n",
              "expirationDateKeyInMatch              bool\n",
              "dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ],
      "source": [
        "#converting in datetime format\n",
        "df['transactionDateTime'] = pd.to_datetime(df['transactionDateTime'])\n",
        "df['currentExpDate'] = pd.to_datetime(df['currentExpDate'])\n",
        "df['accountOpenDate'] = pd.to_datetime(df['accountOpenDate'])\n",
        "df['dateOfLastAddressChange'] = pd.to_datetime(df['dateOfLastAddressChange'])\n",
        "df.dtypes"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Handling Outlier tidak dilakukan, karena akan berpengaruh dalam data yang berlabel Fraud\n",
        "\n",
        "---"
      ],
      "metadata": {
        "id": "Ogk72Bb7uaUv"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "s6mVk4bTBcL5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fc67d527-149c-48bf-941a-2631f739dec1"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "US     632303\n",
            "MEX      2626\n",
            "CAN      1870\n",
            "PR       1202\n",
            "Name: acqCountry, dtype: int64\n",
            "US     635577\n",
            "MEX      2636\n",
            "CAN      1874\n",
            "PR       1203\n",
            "Name: merchantCountryCode, dtype: int64\n",
            "5.0     255615\n",
            "9.0     193193\n",
            "2.0     160589\n",
            "90.0     16251\n",
            "80.0     12921\n",
            "Name: posEntryMode, dtype: int64\n",
            "1.0     514144\n",
            "8.0     121507\n",
            "99.0      5976\n",
            "Name: posConditionCode, dtype: int64\n",
            "PURCHASE                608685\n",
            "ADDRESS_VERIFICATION     16478\n",
            "REVERSAL                 16162\n",
            "Name: transactionType, dtype: int64\n",
            "False    630612\n",
            "True      11302\n",
            "Name: isFraud, dtype: int64\n",
            "False    340453\n",
            "True     301461\n",
            "Name: cardPresent, dtype: int64\n",
            "False    640945\n",
            "True        969\n",
            "Name: expirationDateKeyInMatch, dtype: int64\n"
          ]
        }
      ],
      "source": [
        "#Unique Value\n",
        "#When an attribute has less number of Unique Values it is always a good criteria to look at them keenly to know more about the structure as a whole and how the other attributes are behaving around these small buckets.\n",
        "for col in df.columns:\n",
        "    if(df[col].nunique() <= 5):\n",
        "        print(df[col].value_counts())"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Handling Missing Value"
      ],
      "metadata": {
        "id": "Qtwp7p57-PlG"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "R6kUw6FTNxYk"
      },
      "outputs": [],
      "source": [
        "#Handling Missing Value\n",
        "df['acqCountry'] = df['acqCountry'].fillna('US')\n",
        "df['merchantCountryCode'] = df['merchantCountryCode'].fillna('US')\n",
        "df.dropna(subset=[\"posEntryMode\"], axis=0, inplace=True)\n",
        "df.dropna(subset=[\"posConditionCode\"], axis=0, inplace=True)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check mean from transaction type by transaction amount\n",
        "df.groupby('transactionType').mean()['transactionAmount']"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SfjGXaoyumld",
        "outputId": "a30d6a8b-4ebe-43e2-f5a6-8b3f3db51ad7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "transactionType\n",
              "ADDRESS_VERIFICATION      0.000000\n",
              "PURCHASE                138.730479\n",
              "REVERSAL                138.751706\n",
              "Name: transactionAmount, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check median from transaction type by transaction amount\n",
        "df.groupby('transactionType').median()['transactionAmount']"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5Z4ZFReiuokb",
        "outputId": "6ec33c9a-661b-486c-b25b-44f5996b2845"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "transactionType\n",
              "ADDRESS_VERIFICATION     0.00\n",
              "PURCHASE                89.62\n",
              "REVERSAL                91.25\n",
              "Name: transactionAmount, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check row where missing value in transaction type and transaction equal 0\n",
        "df[(df['transactionAmount'] == 0)&(df['transactionType'].isnull() == True)]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 113
        },
        "id": "GFfjaO_8urIg",
        "outputId": "0aae812d-ffd6-47bb-9382-b3981ee476af"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Empty DataFrame\n",
              "Columns: [accountNumber, customerId, creditLimit, availableMoney, transactionDateTime, transactionAmount, merchantName, acqCountry, merchantCountryCode, posEntryMode, posConditionCode, merchantCategoryCode, currentExpDate, accountOpenDate, dateOfLastAddressChange, cardCVV, enteredCVV, cardLast4Digits, transactionType, isFraud, currentBalance, cardPresent, expirationDateKeyInMatch]\n",
              "Index: []"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-7915ad41-c01d-4c88-96a3-6b3941904e8c\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>accountNumber</th>\n",
              "      <th>customerId</th>\n",
              "      <th>creditLimit</th>\n",
              "      <th>availableMoney</th>\n",
              "      <th>transactionDateTime</th>\n",
              "      <th>transactionAmount</th>\n",
              "      <th>merchantName</th>\n",
              "      <th>acqCountry</th>\n",
              "      <th>merchantCountryCode</th>\n",
              "      <th>posEntryMode</th>\n",
              "      <th>posConditionCode</th>\n",
              "      <th>merchantCategoryCode</th>\n",
              "      <th>currentExpDate</th>\n",
              "      <th>accountOpenDate</th>\n",
              "      <th>dateOfLastAddressChange</th>\n",
              "      <th>cardCVV</th>\n",
              "      <th>enteredCVV</th>\n",
              "      <th>cardLast4Digits</th>\n",
              "      <th>transactionType</th>\n",
              "      <th>isFraud</th>\n",
              "      <th>currentBalance</th>\n",
              "      <th>cardPresent</th>\n",
              "      <th>expirationDateKeyInMatch</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-7915ad41-c01d-4c88-96a3-6b3941904e8c')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-7915ad41-c01d-4c88-96a3-6b3941904e8c button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-7915ad41-c01d-4c88-96a3-6b3941904e8c');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Tidak ditemukan data dengan jumlah transaksi 0 ketika missing value pada kolom tipe transaksi\n",
        "\n",
        "---"
      ],
      "metadata": {
        "id": "OXLl_8PF9eoo"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Fill with modus\n",
        "df['transactionType'] = df['transactionType'].fillna('PURCHASE')"
      ],
      "metadata": {
        "id": "rnIEh9SUus83"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Nlv4oCbcPDbQ",
        "outputId": "2ea0a12a-c611-4c3e-c103-5a346217af58"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "accountNumber               0\n",
              "customerId                  0\n",
              "creditLimit                 0\n",
              "availableMoney              0\n",
              "transactionDateTime         0\n",
              "transactionAmount           0\n",
              "merchantName                0\n",
              "acqCountry                  0\n",
              "merchantCountryCode         0\n",
              "posEntryMode                0\n",
              "posConditionCode            0\n",
              "merchantCategoryCode        0\n",
              "currentExpDate              0\n",
              "accountOpenDate             0\n",
              "dateOfLastAddressChange     0\n",
              "cardCVV                     0\n",
              "enteredCVV                  0\n",
              "cardLast4Digits             0\n",
              "transactionType             0\n",
              "isFraud                     0\n",
              "currentBalance              0\n",
              "cardPresent                 0\n",
              "expirationDateKeyInMatch    0\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ],
      "source": [
        "# check the missing values again to make sure the missing values are resolved\n",
        "df.isnull().sum()"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Find Duplicate Data"
      ],
      "metadata": {
        "id": "BFKp6RmVv35V"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 113
        },
        "id": "dUK3IsH6AKyG",
        "outputId": "78f19dce-b7b6-4752-bc5d-f7dea0b62b52"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Empty DataFrame\n",
              "Columns: [accountNumber, customerId, creditLimit, availableMoney, transactionDateTime, transactionAmount, merchantName, acqCountry, merchantCountryCode, posEntryMode, posConditionCode, merchantCategoryCode, currentExpDate, accountOpenDate, dateOfLastAddressChange, cardCVV, enteredCVV, cardLast4Digits, transactionType, isFraud, currentBalance, cardPresent, expirationDateKeyInMatch]\n",
              "Index: []"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b3175084-881f-48fc-8197-8ee1a95bca35\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>accountNumber</th>\n",
              "      <th>customerId</th>\n",
              "      <th>creditLimit</th>\n",
              "      <th>availableMoney</th>\n",
              "      <th>transactionDateTime</th>\n",
              "      <th>transactionAmount</th>\n",
              "      <th>merchantName</th>\n",
              "      <th>acqCountry</th>\n",
              "      <th>merchantCountryCode</th>\n",
              "      <th>posEntryMode</th>\n",
              "      <th>posConditionCode</th>\n",
              "      <th>merchantCategoryCode</th>\n",
              "      <th>currentExpDate</th>\n",
              "      <th>accountOpenDate</th>\n",
              "      <th>dateOfLastAddressChange</th>\n",
              "      <th>cardCVV</th>\n",
              "      <th>enteredCVV</th>\n",
              "      <th>cardLast4Digits</th>\n",
              "      <th>transactionType</th>\n",
              "      <th>isFraud</th>\n",
              "      <th>currentBalance</th>\n",
              "      <th>cardPresent</th>\n",
              "      <th>expirationDateKeyInMatch</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b3175084-881f-48fc-8197-8ee1a95bca35')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-b3175084-881f-48fc-8197-8ee1a95bca35 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-b3175084-881f-48fc-8197-8ee1a95bca35');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ],
      "source": [
        "#find transactions of same amount within the same account from the same merchant by transactionType and transactionDateTime\n",
        "duplicated = df[df.duplicated(['accountNumber', 'customerId', 'transactionAmount', 'merchantName','transactionType','transactionDateTime'], keep=False)]\n",
        "duplicated"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_VGj7q8qGLdw"
      },
      "source": [
        "Duplicate data not found"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Summary of dataset by data types"
      ],
      "metadata": {
        "id": "nSunbjgV-vu6"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "hYUMfilZIH1H",
        "outputId": "6a71fb03-6c2e-4423-d2b8-9c32e1f0593d"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "         creditLimit  availableMoney  transactionAmount  currentBalance\n",
              "count  638285.000000   638285.000000      638285.000000   638285.000000\n",
              "mean    10697.229294     6651.964333         135.176189     4045.264961\n",
              "std     11460.791551     9226.588766         147.058492     5947.629920\n",
              "min       250.000000    -1244.930000           0.000000        0.000000\n",
              "25%      5000.000000     1115.330000          32.320000      502.760000\n",
              "50%      7500.000000     3577.570000          85.820000     2152.150000\n",
              "75%     15000.000000     8163.550000         189.050000     5006.710000\n",
              "max     50000.000000    50000.000000        1825.250000    47496.500000"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-99c8b995-ee2b-4769-b23f-e13e9cec60cf\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>creditLimit</th>\n",
              "      <th>availableMoney</th>\n",
              "      <th>transactionAmount</th>\n",
              "      <th>currentBalance</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>638285.000000</td>\n",
              "      <td>638285.000000</td>\n",
              "      <td>638285.000000</td>\n",
              "      <td>638285.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>10697.229294</td>\n",
              "      <td>6651.964333</td>\n",
              "      <td>135.176189</td>\n",
              "      <td>4045.264961</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>11460.791551</td>\n",
              "      <td>9226.588766</td>\n",
              "      <td>147.058492</td>\n",
              "      <td>5947.629920</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>250.000000</td>\n",
              "      <td>-1244.930000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>5000.000000</td>\n",
              "      <td>1115.330000</td>\n",
              "      <td>32.320000</td>\n",
              "      <td>502.760000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>7500.000000</td>\n",
              "      <td>3577.570000</td>\n",
              "      <td>85.820000</td>\n",
              "      <td>2152.150000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>15000.000000</td>\n",
              "      <td>8163.550000</td>\n",
              "      <td>189.050000</td>\n",
              "      <td>5006.710000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>50000.000000</td>\n",
              "      <td>50000.000000</td>\n",
              "      <td>1825.250000</td>\n",
              "      <td>47496.500000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-99c8b995-ee2b-4769-b23f-e13e9cec60cf')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-99c8b995-ee2b-4769-b23f-e13e9cec60cf button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-99c8b995-ee2b-4769-b23f-e13e9cec60cf');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ],
      "source": [
        "#Summary by numerik type\n",
        "df.describe()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 239
        },
        "id": "zNayWPH-gOpg",
        "outputId": "f0adf9a0-8114-4b1b-f906-504a83360faf"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        accountNumber  customerId merchantName acqCountry merchantCountryCode  \\\n",
              "count          638285      638285       638285     638285              638285   \n",
              "unique           5000        5000         2492          4                   4   \n",
              "top         318001076   318001076         Lyft         US                  US   \n",
              "freq             9988        9988        25159     632623              632609   \n",
              "\n",
              "        posEntryMode  posConditionCode merchantCategoryCode  cardCVV  \\\n",
              "count       638285.0          638285.0               638285   638285   \n",
              "unique           5.0               3.0                   19      899   \n",
              "top              5.0               1.0        online_retail      633   \n",
              "freq        255509.0          511470.0               160637    11304   \n",
              "\n",
              "        enteredCVV  cardLast4Digits transactionType  \n",
              "count       638285           638285          638285  \n",
              "unique         980             5134               3  \n",
              "top            633             1789        PURCHASE  \n",
              "freq         11204             9988          605820  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-75784799-ce19-47f4-9fb5-56a9e81ad06d\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>accountNumber</th>\n",
              "      <th>customerId</th>\n",
              "      <th>merchantName</th>\n",
              "      <th>acqCountry</th>\n",
              "      <th>merchantCountryCode</th>\n",
              "      <th>posEntryMode</th>\n",
              "      <th>posConditionCode</th>\n",
              "      <th>merchantCategoryCode</th>\n",
              "      <th>cardCVV</th>\n",
              "      <th>enteredCVV</th>\n",
              "      <th>cardLast4Digits</th>\n",
              "      <th>transactionType</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285.0</td>\n",
              "      <td>638285.0</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>unique</th>\n",
              "      <td>5000</td>\n",
              "      <td>5000</td>\n",
              "      <td>2492</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>19</td>\n",
              "      <td>899</td>\n",
              "      <td>980</td>\n",
              "      <td>5134</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>top</th>\n",
              "      <td>318001076</td>\n",
              "      <td>318001076</td>\n",
              "      <td>Lyft</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>online_retail</td>\n",
              "      <td>633</td>\n",
              "      <td>633</td>\n",
              "      <td>1789</td>\n",
              "      <td>PURCHASE</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>freq</th>\n",
              "      <td>9988</td>\n",
              "      <td>9988</td>\n",
              "      <td>25159</td>\n",
              "      <td>632623</td>\n",
              "      <td>632609</td>\n",
              "      <td>255509.0</td>\n",
              "      <td>511470.0</td>\n",
              "      <td>160637</td>\n",
              "      <td>11304</td>\n",
              "      <td>11204</td>\n",
              "      <td>9988</td>\n",
              "      <td>605820</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-75784799-ce19-47f4-9fb5-56a9e81ad06d')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-75784799-ce19-47f4-9fb5-56a9e81ad06d button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-75784799-ce19-47f4-9fb5-56a9e81ad06d');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ],
      "source": [
        "#Summary by object type\n",
        "df.describe(include='object')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "id": "eZNtB4DygS5F",
        "outputId": "e3c25d94-caea-45fb-d3e6-743d9d31591c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:2: FutureWarning: Treating datetime data as categorical rather than numeric in `.describe` is deprecated and will be removed in a future version of pandas. Specify `datetime_is_numeric=True` to silence this warning and adopt the future behavior now.\n",
            "  \n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        transactionDateTime       currentExpDate      accountOpenDate  \\\n",
              "count                638285               638285               638285   \n",
              "unique               631920                  165                 1826   \n",
              "top     2016-10-16 12:53:14  2020-08-01 00:00:00  2015-12-11 00:00:00   \n",
              "freq                      3                 4171                10090   \n",
              "first   2016-01-01 00:00:37  2018-05-01 00:00:00  1985-12-25 00:00:00   \n",
              "last    2016-12-30 23:59:27  2032-01-01 00:00:00  2015-12-31 00:00:00   \n",
              "\n",
              "       dateOfLastAddressChange  \n",
              "count                   638285  \n",
              "unique                    2186  \n",
              "top        2016-07-20 00:00:00  \n",
              "freq                      3925  \n",
              "first      1985-12-25 00:00:00  \n",
              "last       2016-12-30 00:00:00  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ace61581-06d9-45ee-bdb3-5276c3adcb34\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>transactionDateTime</th>\n",
              "      <th>currentExpDate</th>\n",
              "      <th>accountOpenDate</th>\n",
              "      <th>dateOfLastAddressChange</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>unique</th>\n",
              "      <td>631920</td>\n",
              "      <td>165</td>\n",
              "      <td>1826</td>\n",
              "      <td>2186</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>top</th>\n",
              "      <td>2016-10-16 12:53:14</td>\n",
              "      <td>2020-08-01 00:00:00</td>\n",
              "      <td>2015-12-11 00:00:00</td>\n",
              "      <td>2016-07-20 00:00:00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>freq</th>\n",
              "      <td>3</td>\n",
              "      <td>4171</td>\n",
              "      <td>10090</td>\n",
              "      <td>3925</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>first</th>\n",
              "      <td>2016-01-01 00:00:37</td>\n",
              "      <td>2018-05-01 00:00:00</td>\n",
              "      <td>1985-12-25 00:00:00</td>\n",
              "      <td>1985-12-25 00:00:00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>last</th>\n",
              "      <td>2016-12-30 23:59:27</td>\n",
              "      <td>2032-01-01 00:00:00</td>\n",
              "      <td>2015-12-31 00:00:00</td>\n",
              "      <td>2016-12-30 00:00:00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ace61581-06d9-45ee-bdb3-5276c3adcb34')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ace61581-06d9-45ee-bdb3-5276c3adcb34 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ace61581-06d9-45ee-bdb3-5276c3adcb34');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ],
      "source": [
        "#Summary by datetime type\n",
        "df.describe(include='datetime')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "rQY_z7p7gasp",
        "outputId": "c77cf565-6f89-48f2-b717-6fe67c3434c7"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       isFraud cardPresent expirationDateKeyInMatch\n",
              "count   638285      638285                   638285\n",
              "unique       2           2                        2\n",
              "top      False       False                    False\n",
              "freq    627252      338522                   637322"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ef5b7920-1443-4550-b7bc-1c80477fbdda\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>isFraud</th>\n",
              "      <th>cardPresent</th>\n",
              "      <th>expirationDateKeyInMatch</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "      <td>638285</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>unique</th>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>top</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>freq</th>\n",
              "      <td>627252</td>\n",
              "      <td>338522</td>\n",
              "      <td>637322</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ef5b7920-1443-4550-b7bc-1c80477fbdda')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ef5b7920-1443-4550-b7bc-1c80477fbdda button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ef5b7920-1443-4550-b7bc-1c80477fbdda');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ],
      "source": [
        "#Summary by boolean type\n",
        "df.describe(include='boolean')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "alw_KF__Ulsb"
      },
      "source": [
        "# Exploratory Data Analysis"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 518
        },
        "id": "5NzQeA5QUlGJ",
        "outputId": "1a71967a-a255-4399-a7e6-17547b5ada14"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "False    627252\n",
            "True      11033\n",
            "Name: isFraud, dtype: int64\n",
            "False    0.982715\n",
            "True     0.017285\n",
            "Name: isFraud, dtype: float64\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variable as a keyword arg: x. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
            "  FutureWarning\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 504x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Check many transactions are labeled fraud\n",
        "print(df['isFraud'].value_counts())\n",
        "print(df['isFraud'].value_counts()/len(df))\n",
        "\n",
        "plt.figure(figsize = (7,5))\n",
        "graph = sns.countplot(df.isFraud)\n",
        "graph.set_xticklabels(graph.get_xticklabels())\n",
        "graph.set_title(\"Count Fraud\")\n",
        "\n",
        "for bar in graph.patches:\n",
        "  graph.annotate(format(bar.get_height(), '.2f'),\n",
        "                   (bar.get_x() + bar.get_width() / 2,\n",
        "                    bar.get_height()), ha='center', va='center',\n",
        "                   size=10, xytext=(0, 5),\n",
        "                   textcoords='offset points')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mtIAKAXCWmjb",
        "outputId": "af739ed9-5401-4fe8-bf1b-04acb31cce8b"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Lyft          25159\n",
              "Uber          25107\n",
              "gap.com       13760\n",
              "apple.com     13538\n",
              "target.com    13528\n",
              "Name: merchantName, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ],
      "source": [
        "# Check top merchants name transaction\n",
        "df['merchantName'].value_counts().head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "EsRqXRnaZg5I",
        "outputId": "d889f96b-f52f-4be5-f955-222fcf97a67f"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "         transactionAmount\n",
              "isFraud                   \n",
              "False           133.465217\n",
              "True            232.448969"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-f752ecae-53ce-435d-a5df-ca1c4e22aaf4\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>transactionAmount</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>isFraud</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>False</th>\n",
              "      <td>133.465217</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>True</th>\n",
              "      <td>232.448969</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f752ecae-53ce-435d-a5df-ca1c4e22aaf4')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-f752ecae-53ce-435d-a5df-ca1c4e22aaf4 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-f752ecae-53ce-435d-a5df-ca1c4e22aaf4');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ],
      "source": [
        "# Check average amount of transactions on Fraudulent Transactions and transactions on Legitimate Transactions?\n",
        "df.groupby('isFraud').mean()[['transactionAmount']]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 459
        },
        "id": "tbCc7dgMuKiY",
        "outputId": "b482a529-4418-40b8-e054-8abcd0ad41e1"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 648x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Check distribution fraud by transaction type\n",
        "plt.figure(figsize = (9,7))\n",
        "graph = sns.countplot(data=df, x='transactionType', hue='isFraud')\n",
        "graph.set_title(\"Transaction Type\")\n",
        "\n",
        "for bar in graph.patches:\n",
        "  graph.annotate(format(bar.get_height()),\n",
        "                   (bar.get_x() + bar.get_width() / 2,\n",
        "                    bar.get_height()), ha='center', va='center',\n",
        "                   size=10, xytext=(0, 7),\n",
        "                   textcoords='offset points')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NHo-mrt4bIlo",
        "outputId": "813659eb-ad68-4966-eeb3-a44e82e5bf0c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "       accountNumber customerId accountOpenDate transactionDateTime  isFraud\n",
            "305732     789836144  789836144      1985-12-25 2016-08-05 02:43:22    False\n",
            "       accountNumber customerId accountOpenDate transactionDateTime  isFraud\n",
            "452252     890796581  890796581      2015-12-31 2016-07-15 22:13:38    False\n"
          ]
        }
      ],
      "source": [
        "# Check the earliest account open date regarding the dataset?\n",
        "print(df.sort_values('accountOpenDate').head(1)[['accountNumber','customerId','accountOpenDate','transactionDateTime','isFraud']])\n",
        "\n",
        "# Check the most recent account open date regarding the dataset?\n",
        "print(df.sort_values('accountOpenDate',ascending=False,).head(1)[['accountNumber','customerId','accountOpenDate','transactionDateTime','isFraud']])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 370
        },
        "id": "mrHiUHyCjAOf",
        "outputId": "4824731f-76bf-4663-a75d-1a129f7f0a14"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/seaborn/distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
            "  warnings.warn(msg, FutureWarning)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Distribution Plot')"
            ]
          },
          "metadata": {},
          "execution_count": 27
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Check distribution transaction amount\n",
        "sns.distplot(df.transactionAmount).set_title(\"Distribution Plot\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "E0FPCZ4towXh",
        "outputId": "004ec655-c67f-4653-8d5c-430e798dcf24"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Boxplot Transaction Amount')"
            ]
          },
          "metadata": {},
          "execution_count": 28
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# check boxplot from transaction amount\n",
        "plt.boxplot(df['transactionAmount'])\n",
        "plt.title(\"Boxplot Transaction Amount\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 460
        },
        "id": "xPG1QqstiKwv",
        "outputId": "014ef223-ba54-45e2-9576-91ea015cd615"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f505ee8d390>"
            ]
          },
          "metadata": {},
          "execution_count": 29
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Check scatter plot from isfraud with transaction amount\n",
        "plt.figure(figsize=(6,7))\n",
        "sns.scatterplot(data=df, x='isFraud', y='transactionAmount')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 461
        },
        "id": "4yCPyU06KVo5",
        "outputId": "ca1badf2-1c87-45da-cd95-482edc012f9d"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f505ee4aa90>"
            ]
          },
          "metadata": {},
          "execution_count": 30
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Check scatter plot from transaction type with transaction amount\n",
        "plt.figure(figsize=(6,7))\n",
        "sns.scatterplot(data=df, x='transactionType', y='transactionAmount')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        },
        "id": "74gHXtxhjNTo",
        "outputId": "0e0a5da4-f7bd-470d-cb1a-3961af30171f"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Check distribution fraud by pos condition code\n",
        "plt.figure(figsize = (8,5))\n",
        "ax = sns.countplot(data = df, x='posConditionCode', hue = 'isFraud')\n",
        "for bar in ax.patches:\n",
        "  ax.annotate(format(bar.get_height()),\n",
        "                   (bar.get_x() + bar.get_width() / 2,\n",
        "                    bar.get_height()), ha='center', va='center',\n",
        "                   size=10, xytext=(0, 5),\n",
        "                   textcoords='offset points')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 388
        },
        "id": "ckEdcQuzjyYC",
        "outputId": "e1d116a3-24ad-4320-c874-9db03471cae3"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Check distribution fraud by merchant country code\n",
        "plt.figure(figsize = (8,6))\n",
        "ax = sns.countplot(data = df, x='merchantCountryCode', hue = 'isFraud')\n",
        "for bar in ax.patches:\n",
        "  ax.annotate(format(bar.get_height()),\n",
        "                   (bar.get_x() + bar.get_width() / 2,\n",
        "                    bar.get_height()), ha='center', va='center',\n",
        "                   size=10, xytext=(0, 5),\n",
        "                   textcoords='offset points')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        },
        "id": "14euaqdlkWhe",
        "outputId": "e09c268a-7191-4994-889f-715961a898a8"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Check distribution fraud by card present\n",
        "plt.figure(figsize = (8,5))\n",
        "ax = sns.countplot(data = df, x='cardPresent', hue = 'isFraud')\n",
        "for bar in ax.patches:\n",
        "  ax.annotate(format(bar.get_height()),\n",
        "                   (bar.get_x() + bar.get_width() / 2,\n",
        "                    bar.get_height()), ha='center', va='center',\n",
        "                   size=10, xytext=(0, 5),\n",
        "                   textcoords='offset points')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-0P0NFCKutr-"
      },
      "source": [
        "# Data Pre-processing"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 270
        },
        "id": "Vp3nJZF77W6s",
        "outputId": "582145e9-5246-4e49-8be7-14780821efc6"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   creditLimit  availableMoney  transactionAmount acqCountry  \\\n",
              "0         5000         5000.00             111.33         US   \n",
              "1         5000         4888.67              24.75         US   \n",
              "2         5000         4863.92             187.40         US   \n",
              "3         5000         4676.52             227.34         US   \n",
              "4         5000         4449.18               0.00         US   \n",
              "\n",
              "  merchantCountryCode posEntryMode posConditionCode merchantCategoryCode  \\\n",
              "0                  US          5.0              1.0            rideshare   \n",
              "1                  US          9.0              1.0            rideshare   \n",
              "2                  US          5.0              1.0            rideshare   \n",
              "3                  US          2.0              1.0            rideshare   \n",
              "4                  US          2.0              1.0            rideshare   \n",
              "\n",
              "        transactionType  isFraud  currentBalance  cardPresent  \\\n",
              "0              PURCHASE     True            0.00        False   \n",
              "1              PURCHASE    False          111.33        False   \n",
              "2              PURCHASE    False          136.08        False   \n",
              "3              PURCHASE     True          323.48        False   \n",
              "4  ADDRESS_VERIFICATION    False          550.82        False   \n",
              "\n",
              "   expirationDateKeyInMatch  \n",
              "0                     False  \n",
              "1                     False  \n",
              "2                     False  \n",
              "3                     False  \n",
              "4                     False  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-aead2345-5201-483f-b8aa-720edbab4dc3\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>creditLimit</th>\n",
              "      <th>availableMoney</th>\n",
              "      <th>transactionAmount</th>\n",
              "      <th>acqCountry</th>\n",
              "      <th>merchantCountryCode</th>\n",
              "      <th>posEntryMode</th>\n",
              "      <th>posConditionCode</th>\n",
              "      <th>merchantCategoryCode</th>\n",
              "      <th>transactionType</th>\n",
              "      <th>isFraud</th>\n",
              "      <th>currentBalance</th>\n",
              "      <th>cardPresent</th>\n",
              "      <th>expirationDateKeyInMatch</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>5000</td>\n",
              "      <td>5000.00</td>\n",
              "      <td>111.33</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>True</td>\n",
              "      <td>0.00</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>5000</td>\n",
              "      <td>4888.67</td>\n",
              "      <td>24.75</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>9.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>False</td>\n",
              "      <td>111.33</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5000</td>\n",
              "      <td>4863.92</td>\n",
              "      <td>187.40</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>False</td>\n",
              "      <td>136.08</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>5000</td>\n",
              "      <td>4676.52</td>\n",
              "      <td>227.34</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>PURCHASE</td>\n",
              "      <td>True</td>\n",
              "      <td>323.48</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5000</td>\n",
              "      <td>4449.18</td>\n",
              "      <td>0.00</td>\n",
              "      <td>US</td>\n",
              "      <td>US</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>rideshare</td>\n",
              "      <td>ADDRESS_VERIFICATION</td>\n",
              "      <td>False</td>\n",
              "      <td>550.82</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-aead2345-5201-483f-b8aa-720edbab4dc3')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-aead2345-5201-483f-b8aa-720edbab4dc3 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-aead2345-5201-483f-b8aa-720edbab4dc3');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ],
      "source": [
        "#drop irrelevant column\n",
        "df = df.drop([\"accountNumber\", \"customerId\", \"transactionDateTime\", \"merchantName\",\n",
        "              \"currentExpDate\", \"accountOpenDate\", \"dateOfLastAddressChange\", \"cardCVV\",\n",
        "              \"enteredCVV\", \"cardLast4Digits\"], axis=1)\n",
        "df.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "LG5ugQ-p-j7j"
      },
      "outputs": [],
      "source": [
        "#Converting boolean values to binary\n",
        "for col in df.columns:\n",
        "    if(df[col].astype(str).str.contains('True').sum()>0 and df[col].astype(str).str.contains('False').sum()>0):\n",
        "        df.loc[df[col].astype(str).str.contains('True'), col] = 1\n",
        "        df.loc[df[col].astype(str).str.contains('False'), col] = 0"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "LR1v6wflxWpN"
      },
      "outputs": [],
      "source": [
        "#Converting distinct values using label encoding\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "obj_col = []\n",
        "for col in df.columns:\n",
        "    if(df[col].dtype == object):\n",
        "        obj_col.append(col)\n",
        "\n",
        "labelencoder = LabelEncoder()\n",
        "for i in obj_col:\n",
        "    df[i] = labelencoder.fit_transform(df[i].astype(str))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 270
        },
        "id": "dI5edBl2-P3Q",
        "outputId": "16e1bb95-50f2-43a7-b52b-c3deff81f35c"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   creditLimit  availableMoney  transactionAmount  acqCountry  \\\n",
              "0         5000         5000.00             111.33           3   \n",
              "1         5000         4888.67              24.75           3   \n",
              "2         5000         4863.92             187.40           3   \n",
              "3         5000         4676.52             227.34           3   \n",
              "4         5000         4449.18               0.00           3   \n",
              "\n",
              "   merchantCountryCode  posEntryMode  posConditionCode  merchantCategoryCode  \\\n",
              "0                    3             1                 0                    17   \n",
              "1                    3             3                 0                    17   \n",
              "2                    3             1                 0                    17   \n",
              "3                    3             0                 0                    17   \n",
              "4                    3             0                 0                    17   \n",
              "\n",
              "   transactionType  isFraud  currentBalance  cardPresent  \\\n",
              "0                1        1            0.00            0   \n",
              "1                1        0          111.33            0   \n",
              "2                1        0          136.08            0   \n",
              "3                1        1          323.48            0   \n",
              "4                0        0          550.82            0   \n",
              "\n",
              "   expirationDateKeyInMatch  \n",
              "0                         0  \n",
              "1                         0  \n",
              "2                         0  \n",
              "3                         0  \n",
              "4                         0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b72d1363-8ba1-483f-8bf9-aeace47bb8a0\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>creditLimit</th>\n",
              "      <th>availableMoney</th>\n",
              "      <th>transactionAmount</th>\n",
              "      <th>acqCountry</th>\n",
              "      <th>merchantCountryCode</th>\n",
              "      <th>posEntryMode</th>\n",
              "      <th>posConditionCode</th>\n",
              "      <th>merchantCategoryCode</th>\n",
              "      <th>transactionType</th>\n",
              "      <th>isFraud</th>\n",
              "      <th>currentBalance</th>\n",
              "      <th>cardPresent</th>\n",
              "      <th>expirationDateKeyInMatch</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>5000</td>\n",
              "      <td>5000.00</td>\n",
              "      <td>111.33</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>17</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>5000</td>\n",
              "      <td>4888.67</td>\n",
              "      <td>24.75</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>17</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>111.33</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5000</td>\n",
              "      <td>4863.92</td>\n",
              "      <td>187.40</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>17</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>136.08</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>5000</td>\n",
              "      <td>4676.52</td>\n",
              "      <td>227.34</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>17</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>323.48</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5000</td>\n",
              "      <td>4449.18</td>\n",
              "      <td>0.00</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>17</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>550.82</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b72d1363-8ba1-483f-8bf9-aeace47bb8a0')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-b72d1363-8ba1-483f-8bf9-aeace47bb8a0 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-b72d1363-8ba1-483f-8bf9-aeace47bb8a0');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ],
      "source": [
        "df.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 624
        },
        "id": "nmex13ifsih3",
        "outputId": "77a123df-d74d-4de8-ab10-7a104fc6791d"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f506419ac10>"
            ]
          },
          "metadata": {},
          "execution_count": 38
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# Chart correlation isFraud with other variabel\n",
        "dataC = abs(df.corr())['isFraud'].sort_values().reset_index()\n",
        "dataC = dataC[~dataC['index'].str.contains('isFraud')]\n",
        "\n",
        "plt.figure(figsize=(10,10))\n",
        "sns.barplot(data = dataC.reset_index(), y='index', x = 'isFraud')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4GqFuUCeGHct",
        "outputId": "399ecef7-ebe1-448e-be92-489a470f0707"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Correlation Values\n",
            "transactionAmount           0.087726\n",
            "posEntryMode                0.038323\n",
            "cardPresent                 0.014609\n",
            "transactionType             0.010996\n",
            "merchantCategoryCode        0.010092\n",
            "currentBalance              0.008464\n",
            "posConditionCode            0.005950\n",
            "merchantCountryCode         0.005394\n",
            "acqCountry                  0.004972\n",
            "creditLimit                 0.003248\n",
            "availableMoney              0.001422\n",
            "expirationDateKeyInMatch    0.001129\n",
            "Name: isFraud, dtype: float64\n"
          ]
        }
      ],
      "source": [
        "print(\"Correlation Values\")\n",
        "\n",
        "corr = abs(df.corr()).drop('isFraud')\n",
        "corr.sort_values(['isFraud'], ascending = False, inplace = True)\n",
        "print(corr.isFraud)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "v24kcmmQVU2O"
      },
      "outputs": [],
      "source": [
        "#Segrigating Features and Target variable\n",
        "X = df.drop(columns=['isFraud'])\n",
        "y = df['isFraud']\n",
        "feature = pd.DataFrame({'feature' : X.columns})"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w9xF8X2DWIeA",
        "outputId": "258b9f17-4886-44fd-bca5-874c3ce576a7"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(510628, 12) (510628,)\n",
            "(127657, 12) (127657,)\n"
          ]
        }
      ],
      "source": [
        "# Split the dataset into training data and test data\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 42)\n",
        "\n",
        "print(X_train.shape, y_train.shape)\n",
        "print(X_test.shape, y_test.shape)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Randomly under sample the majority class\n",
        "from collections import Counter\n",
        "from imblearn.under_sampling import RandomUnderSampler\n",
        "\n",
        "rus = RandomUnderSampler(random_state=42)\n",
        "X_train, y_train= rus.fit_resample(X_train, y_train)\n",
        "# Check the number of records after under sampling\n",
        "print(sorted(Counter(y_train).items()))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5PhdKpnVt7bU",
        "outputId": "3e786de6-2ae1-4b51-bf4a-1ed10e010c2e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[(0, 8818), (1, 8818)]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9gG5W9WrBFV6"
      },
      "source": [
        "# DATA MODELLING"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Kvh33ruKBvem"
      },
      "source": [
        "### Logistic Regression"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "v3CcQ3pIB9WA",
        "outputId": "e7231da0-cf53-4692-b423-dcfe743c1d5f"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression()"
            ]
          },
          "metadata": {},
          "execution_count": 43
        }
      ],
      "source": [
        "# Fit a logistic regression model to our data\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "# define models\n",
        "model1 = LogisticRegression()\n",
        "model1.fit(X_train, y_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "IXcFRT4CGk_m"
      },
      "outputs": [],
      "source": [
        "# Obtain model predictions\n",
        "y1_pred = model1.predict(X_test)\n",
        "y1_pred_train = model1.predict(X_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "q8ep1G9JG71P",
        "outputId": "1bb7f266-40fe-4a3f-8139-203478ee649b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Classification report:\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.63      0.77    125442\n",
            "           1       0.03      0.60      0.05      2215\n",
            "\n",
            "    accuracy                           0.63    127657\n",
            "   macro avg       0.51      0.61      0.41    127657\n",
            "weighted avg       0.97      0.63      0.76    127657\n",
            "\n",
            "Confusion matrix:\n",
            " [[79143 46299]\n",
            " [  893  1322]]\n"
          ]
        }
      ],
      "source": [
        "# Print the classifcation report and confusion matrix\n",
        "from sklearn.metrics import classification_report, confusion_matrix\n",
        "print('Classification report:\\n', classification_report(y_test, y1_pred))\n",
        "conf_mat = confusion_matrix(y_test, y1_pred)\n",
        "print('Confusion matrix:\\n', conf_mat)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "cm = confusion_matrix(y_test, model1.predict(X_test))\n",
        "cm = pd.DataFrame(cm , index = ['0', '1'] , columns = ['0', '1'])\n",
        "plt.figure(figsize = (5,5))\n",
        "plt.title(\"Logistic Regression Confusion Matrix\")\n",
        "sns.heatmap(cm, cmap='Blues', annot = True, fmt='',xticklabels = [\"1\", \"0\"], yticklabels = [\"1\", \"0\"])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 354
        },
        "id": "VltPJwqZf_QY",
        "outputId": "48f5a5f8-1454-42b8-acce-dba18d2b3b7b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f5064a28a50>"
            ]
          },
          "metadata": {},
          "execution_count": 58
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x360 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 406
        },
        "id": "aozyApNTOlb_",
        "outputId": "2a1e5483-cbed-45dd-9857-06e26c28a444"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.6565326513349472\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x432 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUoAAAFzCAYAAABYXEBgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd5hU5fn/8fft0rsCNnoQhF06KIoFEEVAEROQYhT1S2I09vYTO0HsJaKQxIYtBokaIyqiQkDAjlQBURCkC4Ki9C33748ZhlnYMrvs7JnZ+byuay/mnDlnzn12l88+pz2PuTsiIpK/Q4IuQEQk0SkoRUQKoaAUESmEglJEpBAKShGRQigoRUQKUS7oAoqqTp063rhx46DLEJEy5ssvv/zR3evm9V7SBWXjxo2ZPXt20GWISBljZt/n954OvUVECqGgFBEphIJSRKQQCkoRkUIoKEVECqGgFBEphIJSRKQQCkoRkUIoKEVEChG3oDSzcWa20cy+yud9M7PHzWyZmS0wsw7xqkVE5GDEs0X5PNCrgPd7A83CX5cCf49jLSIixRa3oHT3GcCWAhbpB7zoIZ8CtczsqHjVIyKpYfO23UxbupEde7JK7DOD7BSjHrA6anpNeN76/Rc0s0sJtTpp2LBhqRQnIsllyfpf6D16ZmR62o3daFKnZCIuKS7muPtT7t7J3TvVrZtnL0gikqJ2ZWYzaeH6XCF5RfemHFWzUoltI8gW5VqgQdR0/fA8EZEC7cnK4e/Tl/P5ys18tGxzZH7/DvV5ZGDbEt9ekEE5EbjSzF4BOgNb3f2Aw24REYCcHOcfM5azYPVWJi/akOu9U5rV4eZeLWhVr2Zcth23oDSz8UA3oI6ZrQHuAsoDuPs/gElAH2AZsAO4JF61iEjyys5xzvjrh3y3aXuu+W0b1OLpoR2pW60iZhbXGuIWlO4+pJD3HbgiXtsXkeS2edtu7njzKyYt3Nd6HNCxPtee3oz6h1Yp1VqSbigIESm7dmVm85e3FjH+89UHvLdgRE9qVCofQFUKShFJANOXbuQvby1mxY+5D69v7tWCi7o0okqFYKNKQSkigdiTlcOLn6xk1DtLcs2/ontTLuvalOoBtR7zoqAUkVKVneP8ffoyHn7/m1zz/3FBR05veTjl0hLv9m4FpYjE1Z6sHB794BsWrdvKzG9/zPVe3eoVefOKkzi6VuWAqouNglJE4mJXZjZtRrzPnuycXPMrljuEAR3rc0PPYzmsaoWAqisaBaWIlKhdmdmcO/Yjvt7wa2Teqc3rcv/vWid8yzE/CkoRKRFzVv3Ehc98xvY92ZF5bevX5D9/Pom0Q+J7Q3i8KShF5KC8OW8t17wyL9e88zs3ZFS/VhyS5AG5l4JSRIrF3fnDC7OZ+vVGAOpUq8D1ZxzL+Z3LXleICkoRKbLHpnzDY1O+jUzf3S+DC09sHFxBcaagFJGYfbdpG6c98mFk+ogaFfnwpu5UKp8WYFXxp6AUkZjs3JOdKySn3tCVpnWrBVhR6VFQikiB3J0zH5vBNz9sA6D7sXV57pLjA66qdCkoRSRfG7bu4oT7pkam2zaoxbiLjwuwomAoKEXkAFnZObT5y/vsCN8TeYjBnDvOoFaV5HiSpqQpKEUklw+/2cRF4z6PTJ/XsT4PnVfy49AkEwWliADw5fc/0f/vH+eat+K+PnEfZiEZKChFUtyuzGxa3DE517zxfzyBE5vWDqiixKOgFElhz8z8LlfHuY8Pac/ZrY8qM48elhQFpUiKGjdrRSQkT2lWh3EXH0f5BOw0NxEoKEVSzMZfdnH8vftu+bnvd60ZcnzZez67JCkoRVLIqs07OPWhaZHpVy49gRN+o3ORhVFQiqSI6Nt+yqcZ397TJ+CKkodOSIikgG9++DUSkmekH8E3o3oHXFFyUYtSpIyb9e2PXPDsZwCc3vJwnh7aKeCKko+CUqQMG/nWYsZ9tAIIXdl+5qLUe067JCgoRcqoN+auiYTk8N4tuKxr04ArSl4KSpEyaOmGX7luwnwAHj6vLQM61g+4ouSmizkiZcyW7Xs487EZAJzb7miFZAlQi1KkDLn2lbn8d946AM5qcxSPDW4fcEVlg4JSpIxoPPydyOu+bY/m8cHtAqymbFFQiiS5z1ds4f+e/yIyPeX6rhxzeGqMZVNaFJQiSezKf83h7QXrI9OLR55JlQr6b13S9B0VSUKffbeZW99YyPJN2wG47vTmXN3jGHWyGycKSpEkM2LiIp7/eGVkWrf/xJ+CUiRJrPhxO90fnh6ZHtE3nYtPahJcQSlEQSmS4JZt3MbItxcz45tNkXlTrj+VYw6vHmBVqUVBKZLAHpz8NX+bvjwyfXGXxow4JyPAilKTglIkQc1b/XMkJM9qcxSPD25PmsayCYSCUiQBzfx2Exc+G+o/8okh7enb9uiAK0ptetZbJMFk53gkJDOOrqGQTAAKSpEE88T/vgWgS9PavHP1KQFXI6CgFEkov+zK5LEpoaB8dKCe1U4UCkqRBPHxsh9pM+J9AE4+pg5H1qwUcEWyl4JSJAF888OvnP/MZ5Hpl4YdH2A1sj9d9RYJ2PuLNnDpS18CcN/vWjPk+IYBVyT7U1CKBOTh95YyZtqyyHSdahUVkglKQSkSgMlfrY+EZPVK5fjrwHacnn5EwFVJfhSUIqXs0hdn8/7iHwDdTJ4s4noxx8x6mdlSM1tmZsPzeL+hmU0zs7lmtsDM+sSzHpEgzVn1E42HvxMJyZvOPFYhmSTi1qI0szRgLHAGsAb4wswmuvviqMVuB/7t7n83s3RgEtA4XjWJBMHd6frQdFZt2RGZN+eOMzisaoUAq5KiiOeh9/HAMnf/DsDMXgH6AdFB6UCN8OuawLo41iNS6i589jNmfvtjZPrufhlceGLj4AqSYolnUNYDVkdNrwE677fMCOB9M7sKqAqcHsd6RErNlu176HD3B5HpWlXK89mtPahYLi3AqqS4gr6YMwR43t0fMbMTgZfMrJW750QvZGaXApcCNGyo2yckcbk7N766gNfnrAHg0Crl+Xh4DypXUEAms3gG5VqgQdR0/fC8aMOAXgDu/omZVQLqABujF3L3p4CnADp16uTxKljkYPzwyy463zs1Mn1ex/o8OKCNBvwqA+IZlF8AzcysCaGAHAycv98yq4AewPNm1hKoBGxCJMm89uUabnx1fmT6y9tPp3a1igFWJCUpbkHp7llmdiXwHpAGjHP3RWY2Epjt7hOBG4Cnzew6Qhd2LnZ3tRglqXy/eXskJM9tdzSPDW4fcEVS0uJ6jtLdJxG65Sd63p1RrxcDJ8WzBpF42f+CzQUnNGTUua0DrEjiJeiLOSJJKbojC4BrejTjmh7NAqxI4klBKVJEQ576lE++2xyZXnFfH12wKeMUlCIxenbWCu5+e9/zEmPOb8/ZbfQIYipQUIoUYvWWHZzy4LRc8z67tQdH1FAP5KlCQSmSj89XbGHI05+SnbPvRoy3rzqZVvVqBliVBEFBKbKfTb/u5rh7puSad+3pzbj29OYBVSRBU1CK7Cc6JO/5bSt+37lRgNVIIlBQioTl5Dj9xn4UmV55/1kBViOJREEpAmTnOE1v3fdsxORrTwmwGkk0CkpJee5Ot4f3XdVeOKIn1SuVD7AiSTQKSkl5fR6fxeotOwFYdk9vyqVpuHvJTb8RktIGPfkJS9b/AsCCET0VkpIntSglJW3bnUWX+6byy64sAP45rDM1dLgt+VBQSsr512eruPWNhZHpGTd1p2HtKgFWJIlOQSkpY1dmNul3TmbvgzaHV6/Ix8NP0+G2FEpBKSlhV2Y2Le6YHJm+/ayW/OGU3wRYkSQTBaWUeft3aqFu0aSodMwhZdryTdsUknLQ1KKUMuvByV/zt+nLAWjboBZvXqFRR6R41KKUMmnyVxsiIXnhCY0UknJQ1KKUMucPL3zBlCWhoeFf/kNnTjqmTsAVSbJTi1LKlHGzVkRC8s/dmiokpUSoRSllxrbdWYwMj2kz9YauNK1bLeCKpKxQi1LKhJ+276HVXe8BcPIxdRSSUqIUlJL0Js5fR/u7PwCgWsVy/PMPnQOuSMoaHXpLUms38n1+3pEJQPMjqvH+dV0DrkjKIgWlJCV356rxcyMhOf6PJ3Bi09oBVyVllYJSktKgpz7l8xVbAHjjz11o3/DQgCuSskxBKUnnb9OXRUJy5v/rToPD1EWaxJeCUpLKpS/O5v3FPwDw9NBOCkkpFQpKSRqrt+yIhOSzF3WiR8sjAq5IUoVuD5Kk4O6RXoDu+11rhaSUKgWlJIULnv0s8nrI8Q0DrERSkQ69JaG5O3966Us+WrYZgK/v7hVwRZKKFJSSsHJynN/cOiky/dzFx1GpfFqAFUmqUlBKwooOyfl39aRmZQ0nK8FQUEpC+nnHnsjrb+/pTXmNlCgB0m+fJJzvN2+n3chQJxejzm2lkJTA6TdQEsrE+evo+tD0yPTvO+sKtwRPQSkJ4/mPVnD1+LkADD6uASvvP0sjJkpC0DlKSQjPf7SCEW+Feif/yzkZXNSlcbAFiUSJuUVpZnqoVuLiuaiQvKZHM4WkJJxCg9LMupjZYuDr8HRbM/tb3CuTlDB03Of8JRySw3u34LozmgdckciBYjn0/itwJjARwN3nm9mpca1KUkK/MbOYv2YrAP/6Y2e6NNWIiZKYYjpH6e6r9zupnh2fciRVzF/9cyQkPxp+GvVqVQ64IpH8xRKUq82sC+BmVh64BlgS37KkLMvOcfqN/QiAe3/bWiEpCS+WizmXAVcA9YC1QDvgz/EsSsquuat+omn40cSjalbifN0nKUkglhblse7+++gZZnYS8FF8SpKy6r9z13LthHmR6Y9uPi3AakRiF0uL8okY54nka9a3P0ZC8pHz2rLy/rM45BDdTC7JId8WpZmdCHQB6prZ9VFv1QDU15XEzN0jHe9eclJj+nesH3BFIkVT0KF3BaBaeJnqUfN/AQbEsygpO9ydJrfs6y7trr4ZAVYjUjz5BqW7fwh8aGbPu/v3xflwM+sFjCbUAn3G3e/PY5mBwAjAgfnufn5xtiWJZ/+Q/Pae3gFWI1J8sVzM2WFmDwEZQKW9M929wDPxZpYGjAXOANYAX5jZRHdfHLVMM+AW4CR3/8nMDi/GPkgCysrO4Zjb3o1ML7+3D2k6JylJKpaLOS8TenyxCfAXYCXwRQzrHQ8sc/fv3H0P8ArQb79l/giMdfefANx9Y4x1S4I7NTxiIsDSUb0UkpLUYgnK2u7+LJDp7h+6+/8BsdzXUQ9YHTW9JjwvWnOguZl9ZGafhg/VD2Bml5rZbDObvWnTphg2LUEaMXER67buAuC7e/tQsZyu/UlyiyUoM8P/rjezs8ysPXBYCW2/HNAM6AYMAZ42s1r7L+TuT7l7J3fvVLdu3RLatMTDkx8u5/mPVwLw5hUn6RYgKRNiOUc5ysxqAjcQun+yBnBtDOutBRpETdcPz4u2BvjM3TOBFWb2DaHgjOXQXhLM4nW/cN+7XwPw/CXH0bbBAX/zRJJSoS1Kd3/b3be6+1fu3t3dOwJbYvjsL4BmZtbEzCoAgwn3QBTlv4Rak5hZHUKH4t8VZQckMezKzKbP4zMB6NfuaLodq+tyUnbkG5RmlmZmQ8zsRjNrFZ53tpl9DIwp7IPdPQu4EniPUCca/3b3RWY20szOCS/2HrA53N/lNOAmd998kPskAWhxx+TI69GD2wdYiUjJK+jQ+1lCh86fA4+b2TqgEzDc3f8by4e7+yRg0n7z7ox67cD14S9JUrf8Z2Hk9fJ7+wRYiUh8FBSUnYA27p5jZpWADUBTtfgk2ucrtjD+81UAfHHb6boNSMqkgs5R7nH3HAB33wV8p5CUaJt+3c3AJz8BQh1d1K1eMeCKROKjoBZlCzNbEH5tQNPwtBE6am4T9+okYW3etpvj7pkCQKt6NdTRhZRpBQVly1KrQpLKgjU/c86YUHek1SqW4+2rTgm4IpH4KqhTjGJ1hCFlm7tHQvLYI6oz+VqFpJR9MY/rLZKZnUOru94DoEvT2rx33ansN+icSJkU0yiMIjk5TrOo3oD+9vsOAVYjUrpialGaWWUzOzbexUjiOvdv+4ZI+mZUb2pVqRBgNSKlq9CgNLO+wDxgcni6nZnt/yiilGGfr9jCgvAY3N+M6k2FcjpjI6kllt/4EYT6lvwZwN3nEeqbUlKAu0fulXztshMVkpKSYupmzd237jfP41GMJJas7JxcQzl0bHRogNWIBCeWizmLzOx8IC08dMPVwMfxLUuCNnbaMh56b2lk+uu7e+kKt6SsWILyKuA2YDfwL0I9/oyKZ1ESnBU/bqf7w9Mj092PrcszFx2nZ7glpcUSlC3c/TZCYSll2Iatu3KF5LvXnELLo2oEV5BIgoglKB8xsyOB14AJ7v5VnGuSAMxf/TP9xoZuATqu8aG8elmXgCsSSRyx9HDeHegObAKeNLOFZnZ73CuTUrNs47ZISJ6RfoRCUmQ/Md3r4e4b3P1x4DJC91TeWcgqkkSumzAPgP87qQlPD+0UcDUiiSeWG85bmtkIM1tIaHCxjwkNFCZlwLhZK1i4NnT315190wOuRiQxxXKOchwwATjT3dfFuR4pRcs2/srItxcD8NwlxwVcjUjiKjQo3f3E0ihEStf23Vmc/ugMAIb3bkF3jZookq98g9LM/u3uA8OH3NFP4qiH8zKg1YhQd2nHHF6Ny7o2DbgakcRWUIvymvC/Z5dGIVJ6Hpj8NR7+0/fO1ScHW4xIEsj3Yo67rw+//LO7fx/9Bfy5dMqTkubu/H36cgCm39iNiuXSAq5IJPHFcnvQGXnM613ShUj8RfcE9Lv29Whcp2rAFYkkh4LOUV5OqOX4m6jRGAGqAx/lvZYkqj1ZOTS/fV8P5cP7tAiwGpHkUtA5yn8B7wL3AcOj5v/q7lviWpWUuN6jZ0ReLxjRkxqVygdYjUhyKSgo3d1XmtkV+79hZocpLJPHf+euZfmm7YB6KBcpjsJalGcDXxK6PSi6ny0HfhPHuqSE/LhtN9eGH1F8bFA7haRIMRQ0rvfZ4X817EMS6zRqCgAXd2nMue3rBVyNSHKK5Vnvk8ysavj1BWb2qJk1jH9pcrAeeu/ryOsR52QEWIlIcovlOOzvwA4zawvcACwHXoprVVIixk4L3S/510FtA65EJLnFEpRZ7u5AP2CMu48ldIuQJLAr/jUHgEa1q/Db9ursSeRgxNJ70K9mdgtwIXCKmR0C6N6SBPfOgtCDVf8c1jngSkSSXywtykGEBhb7P3ffQKgvyofiWpUclNe+XAPAaS0Op8FhVQKuRiT5xTIUxAbgZaCmmZ0N7HL3F+NemRTbja/OB+AudcQrUiJiueo9EPgcOA8YCHxmZgPiXZgUz9QlPwBQvWI5GtXWs9wiJSGWc5S3Ace5+0YAM6sLTCE0KqMkkKzsHIa9MBtQj+UiJSmWc5SH7A3JsM0xrielbMA/Qj0DnfCbw+jU+LCAqxEpO2JpUU42s/eA8eHpQcCk+JUkxfHFyi3MW/0zAC//4YSAqxEpW2IZM+cmM/sdsLcr7Kfc/Y34liVFsTsrm/PCrclb+7Qg7RArZA0RKYqC+qNsBjwMNAUWAje6+9rSKkxi9/jUbyOvLz1V49+IlLSCzjWOA94G+hPqQeiJUqlIimRXZnbkUcWlo3oFXI1I2VTQoXd1d386/Hqpmc0pjYIkdu5OizsmA1CrSnmNfyMSJwUFZSUza8++figrR0+7u4IzYMfdMzXyes7teQ1tJCIloaCgXA88GjW9IWragdPiVZQUbuGarfy4bTcAX9/di0N0AUckbgrquLd7aRYiscvMzqHvmFlAqNfySuV1yC0ST7pxPAmd/XgoJKtXKqdey0VKQSw3nEsCaX7bu+zJzgHgi9tOD7gakdSgFmUSeXvBukhIfnLLaTrkFiklsfQeZOGxcu4MTzc0s+PjX5pEy8rO4cp/zQXgfzd05aialQOuSCR1xNKi/BtwIjAkPP0rMDaWDzezXma21MyWmdnwApbrb2ZuZp1i+dxUs2NPFsfc9i4AHRsdym/qVgu4IpHUEktQdnb3K4BdAO7+E1ChsJXMLI1QoPYG0oEhZnZAT7JmVh24BvisCHWnlFMfnB55PeFSdXghUtpiCcrMcOg5RPqjzIlhveOBZe7+nbvvAV4hNEDZ/u4GHiAcxJLb1h2ZkfslV9zXh3JpOq0sUtpi+V/3OPAGcLiZ3QPMAu6NYb16wOqo6TXheRFm1gFo4O7vFPRBZnapmc02s9mbNm2KYdNlx/X/ngfANT2aYaabykWCEEs3ay+b2ZdAD0KPL57r7ksOdsPh0RwfBS6OoYangKcAOnXq5Ae77WSxKzObqV+H+ky+9vRmAVcjkroKDUozawjsAN6KnufuqwpZdS3QIGq6fnjeXtWBVsD0cEvpSGCimZ3j7rNjK79s+9u0ZQAMOb6hWpMiAYrlhvN3CJ2fNKAS0ARYCmQUst4XQDMza0IoIAcD5+990923AnX2TpvZdEJ9Xiokw54IB6VGUxQJViyH3q2jp8PnFf8cw3pZZnYl8B6QBoxz90VmNhKY7e4Ti1lzStiyfQ/uUKdaRd1YLhKwIj/C6O5zzKxzjMtOYr/xddz9znyW7VbUWsqyK14O9WKn1qRI8GI5R3l91OQhQAdgXdwqEiZ8sYpPvtsMwNltjgq4GhGJpUVZPep1FqFzlq/HpxzZlZnNza8vBOCOs9N1EUckARQYlOEbzau7+42lVE/K23vf5KnN6zLs5CYBVyMiUMAN52ZWzt2zgZNKsZ6UtnTDr0xauAGApy7sGHA1IrJXQS3Kzwmdj5xnZhOBV4Hte9909//EubaUc+ZjMwC44ISGutItkkBiOUdZCdhMaIycvfdTOqCgLEF3vvkVAIdWKc+oc1sXsrSIlKaCgvLw8BXvr9gXkHulzGOEpSEnx3nxk+8BmHZjt2CLEZEDFBSUaUA1cgfkXgrKEnTVK6EOeU9pVodaVQrtwU5ESlmBw9W6+8hSqyRF5eQ47yxYD8CjA9sFXI2I5KWgbtZ0A18peGzqtwBc0b0pdatXDLgaEclLQUHZo9SqSFGL1/3C4+GgPK9jg0KWFpGg5BuU7r6lNAtJRWOmhUJy9OB2NK5TNeBqRCQ/GlcgIN9v3h65ubxfu3qFLC0iQVJQBqTrQ9MBuLL7McEWIiKFUlAG4I8v7uub+MYzjw2wEhGJhYKylC3b+CsfLP4BgI+HnxZwNSISCwVlKev12EwAHh/SnqNrVQ64GhGJhYKyFG3dmUlWTuihpr7qkFckaSgoS9Fxo6YA8GD/NuqQVySJKChLybKN29iTnQNAv/ZHB1yNiBSFgrKUjA4/gXPvb1tTsZz6mhRJJgrKUrAnK4e35ofGYxt0nB5VFEk2Cso4c3ea3/4uAG3q1yTtEJ2bFEk2Cso4e3bWisjrN6/Q8EMiyUhBGUe7MrMZ9c4SAD655TRd6RZJUgrKOLr8n18CcFTNShxVUzeXiyQrBWUcTVu6CYAp13cNuBIRORgKyjiZEn6e+9TmdalaMZbBLkUkUSko4+Sq8aEBwx7or6FnRZKdgjIOcnKcnZnZADo3KVIGKCjj4KmZ3wFwRvoRAVciIiVBQVnCdmVmc/+7XwOhxxVFJPkpKEvY3pBs26CWhp8VKSMUlCXs+Y9XAvDKH08IthARKTEKyhJ02xsLgdAz3ZUrqIcgkbJCQVlCsnOclz9bBcA//9A54GpEpCQpKEvImP8tA6BXxpHUqFQ+4GpEpCQpKEvIhC9CrclHB7UNuBIRKWkKyhKQlZ3Duq27qFerMlUq6HFFkbJGQXmQZnyziWNuC3XM2zNDN5iLlEUKyoOwfXcWQ8d9DkDVCmnceXZ6wBWJSDwoKA/C4/8LDRjWpWltFo3spY55RcooBeVBePLD0DPdf/99x4ArEZF4UlAW0/qtOwFofkQ1albR7UAiZZmCsphu+Pd8AG7u1SLgSkQk3hSUxZCVncPHyzcD0KOlrnSLlHUKymJ4YHKoh6AhxzcIuBIRKQ0KyiJyd56eGRqre3jvlgFXIyKlQUFZRI9NCd0S1KpeDWpW1kUckVSgoCyi0VNDQfnyMPU3KZIq4hqUZtbLzJaa2TIzG57H+9eb2WIzW2BmU82sUTzrOVhbd2ZGXuuWIJHUEbegNLM0YCzQG0gHhpjZ/s/4zQU6uXsb4DXgwXjVUxL6jZkFwF199aiiSCqJZ4vyeGCZu3/n7nuAV4B+0Qu4+zR33xGe/BSoH8d6Dsr6rTtZuTlU6oUnJHTDV0RKWDyDsh6wOmp6TXhefoYB78axnoPS/eHpANx+VkvKpenUrkgqSYjOE83sAqAT0DWf9y8FLgVo2LBhKVYWsm13FrsycwD4wym/KfXti0iw4tk0WgtE35FdPzwvFzM7HbgNOMfdd+f1Qe7+lLt3cvdOdevWjUuxBXl34XoA7u6XUerbFpHgxTMovwCamVkTM6sADAYmRi9gZu2BJwmF5MY41lJsOTnOTa8tAKBnxpEBVyMiQYhbULp7FnAl8B6wBPi3uy8ys5Fmdk54sYeAasCrZjbPzCbm83GBueqVuZHXR9SoFGAlIhKUuJ6jdPdJwKT95t0Z9fr0eG7/YG3fncU7C0KH3d+M6h1wNSISFF2+LcAT4SFoz0g/ggrl9K0SSVX631+AcR+FOr/466B2AVciIkFSUObj6w2/sCcrh7b1a1KtYkLcRSUiAVFQ5uOZcFdqv++sp3BEUp2CMh+ffhfqwfy8Tgn7VKWIlBIFZR6ysnNY89NOGh5WRUPQioiCMi+TvtoAQP8Oak2KiIIyTw/uHROns8bEEREF5QF27MlizU+hMbsPr64ncUREQXmAP788B4CBuogjImEKyv1MX7oJgAf6twm4EhFJFArKKL/uCo2J06nRobraLSIRCsooSzf8CsC57QvqiF1EUo2CMsp7i0K3BWUcXSPgSkQkkSgoo+ztUq1N/VoBVyIiiURBGfb95u2s27oLgLRDdH5SRPZRUIb97+vQSBRXn3ZMwJWISKJRUIbdNyn0NM6fuysoRSQ3BSWwc082e7JzKJ9mVCqfFnQ5IpJgFJTAfe8uATRmt4jkTUEJfL5iCwDXnd484OmqEcoAABXFSURBVEpEJBEpKIHVW3bQuHYVDSAmInlK+WTIznG278nmmMOrB12KiCSolA/Kdxbuvcm8ZsCViEiiSvmgfC/cm/mg49RJr4jkLaWDcvvurEiL8oga6qRXRPKW0kH579mrAXWCISIFS+mg3NsJxiuXnhBwJSKSyFI2KLOyc5j9/U8AVK9UPuBqRCSRpWxQ7j03+bsO6qRXRAqWskH53EcrARjeu0WwhYhIwkvZoJy3+mdAQ9KKSOFSMii37gwNInZ6yyMCrkREkkFKBuWzM78D4MwMBaWIFC4lg/L1OWsB+K1GWxSRGKRkUK79eSfVK5WjXFpK7r6IFFHKJcWCNaGLOOd3bhhwJSKSLFIuKN+ctw6APq2OCrgSEUkWKReU+8buVrdqIhKblArKnXuy2fDLLuofWhkzjd0tIrFJqaB87ctQb0FDjtf5SRGJXUoF5bSlmwAFpYgUTUoF5f++3gjAYVUrBFyJiCSTlAnKL1aGhqRNP0qd9IpI0aRMUC5e9wsAI87JCLgSEUk2KROUY6ctAyBdwz6ISBGVC7qA0pCZncPGX3dTrWI5qlVMiV2Ou8zMTNasWcOuXbuCLkWkSCpVqkT9+vUpXz72kQ1SIjVe/OR7AP5wSpOAKyk71qxZQ/Xq1WncuLHuSZWk4e5s3ryZNWvW0KRJ7HmQEofeUxb/AMAlJykoS8quXbuoXbu2QlKSiplRu3btIh8JpURQfvLdZg6vXpGalTWIWElSSEoyKs7vbZkPyh9+Cf3laFy7asCVSDz897//xcz4+uuvAZg+fTpnn312rmUuvvhiXnvtNSB0bnX48OE0a9aMDh06cOKJJ/Luu+/GtK3du3czaNAgjjnmGDp37szKlSvzXO7nn39mwIABtGjRgpYtW/LJJ58AMGLECOrVq0e7du1o164dkyZNiqyzYMECTjzxRDIyMmjdunWkxdOtWzeOPfbYyDobN4buBf7HP/5B69atadeuHSeffDKLFy8G4IMPPqBjx460bt2ajh078r///S+yjT179nDppZfSvHlzWrRoweuvvw7A888/T926dSPbeOaZZyLr9OrVi1q1ah3wPR02bBht27alTZs2DBgwgG3bthVY116rVq2iWrVqPPzww5F5o0ePplWrVmRkZPDYY48d8P185JFHMDN+/PFHAB566KFIra1atSItLY0tW7awevVqunfvTnp6OhkZGYwePTq/H2XRuXvcvoBewFJgGTA8j/crAhPC738GNC7sMzt27OhF8dgH33ijm9/2179cXaT1pGCLFy8OugR3dx84cKCffPLJfuedd7q7+7Rp0/yss87KtcxFF13kr776qru733zzzT506FDftWuXu7tv2LDBJ0yYENO2xo4d63/605/c3X38+PE+cODAPJcbOnSoP/300+7uvnv3bv/pp5/c3f2uu+7yhx566IDlMzMzvXXr1j5v3jx3d//xxx89KyvL3d27du3qX3zxxQHrbN26NfL6zTff9DPPPNPd3efMmeNr1651d/eFCxf60UcfHVnuzjvv9Ntuu83d3bOzs33Tpk3u7v7cc8/5FVdckee+TJkyxSdOnHjA9zR6+9ddd53fd999Bda1V//+/X3AgAGR78PChQs9IyPDt2/f7pmZmd6jRw//9ttvI8uvWrXKe/bs6Q0bNozUG23ixInevXt3d3dft26df/nll+7u/ssvv3izZs180aJFee5XXr+/wGzPJ3fi1qI0szRgLNAbSAeGmFn6fosNA35y92OAvwIPlHQdr89ZA8CZGUeW9EdLwLZt28asWbN49tlneeWVVwpdfseOHTz99NM88cQTVKxYEYAjjjiCgQMHxrS9N998k4suugiAAQMGMHXq1L1/8CO2bt3KjBkzGDZsGAAVKlSgVq1aBX7u+++/T5s2bWjbti0AtWvXJi0trcB1atTYd5vb9u3bI4eT7du35+ijjwYgIyODnTt3snv3bgDGjRvHLbfcAsAhhxxCnTp1Ct3nHj16UL169Xy37+7s3Lkzsv386oJQ679JkyZkZOy7l3nJkiV07tyZKlWqUK5cObp27cp//vOfyPvXXXcdDz74YL6Hy+PHj2fIkCEAHHXUUXTo0AGA6tWr07JlS9auXVvoPsYinle9jweWuft3AGb2CtAPiG6L9wNGhF+/BowxM/P9f/uKyd1ZtWUHh1evSFXdFhQ3f3lrUeSG/pKSfnQN7upb8MMBb775Jr169aJ58+bUrl2bL7/8ssDlly1bRsOGDXP9Z442aNAgli5desD866+/nqFDh7J27VoaNGgAQLly5ahZsyabN2/OFTgrVqygbt26XHLJJcyfP5+OHTsyevRoqlYNnfoZM2YML774Ip06deKRRx7h0EMP5ZtvvsHMOPPMM9m0aRODBw/m//2//xf5zEsuuYS0tDT69+/P7bffHgmNsWPH8uijj7Jnz55ch9h7vf7663To0IGKFSvy88+hDqvvuOMOpk+fTtOmTRkzZgxHHHFEZNkZM2bQvHlz/vrXv0b2syCXXHIJkyZNIj09nUceeSQyP6+6tm3bxgMPPMAHH3yQ67C7VatW3HbbbWzevJnKlSszadIkOnXqBIR+vvXq1Yv8Adnfjh07mDx5MmPGjDngvZUrVzJ37lw6d+5c6H7EIp7nKOsBq6Om14Tn5bmMu2cBW4Ha+3+QmV1qZrPNbPamTZtiLmB3Vg7djq3Ln7o2LWrtkgTGjx/P4MGDARg8eDDjx4/Pt+URywn8CRMmMG/evAO+hg4dGnNNWVlZzJkzh8svv5y5c+dStWpV7r//fgAuv/xyli9fzrx58zjqqKO44YYbIuvMmjWLl19+mVmzZvHGG28wdepUAF5++WUWLlzIzJkzmTlzJi+99FJkW1dccQXLly/ngQceYNSoUbnqWLRoETfffDNPPvlkZBtr1qyhS5cuzJkzhxNPPJEbb7wRgL59+7Jy5UoWLFjAGWecEWk1F+a5555j3bp1tGzZkgkTJhRY14gRI7juuuuoVq1ars9o2bIlN998Mz179qRXr160a9eOtLQ0duzYwb333svIkSPz3f5bb73FSSedxGGHHZZr/rZt2+jfvz+PPfZYvn8Uiyy/Y/KD/QIGAM9ETV8IjNlvma+A+lHTy4E6BX1uUc9RSnwEfY5y8+bNXrlyZW/YsKE3atTI69ev7w0aNPAFCxZ4ly5dci3bt29fnz59um/fvt0PO+ywXOfRog0cONDbtm17wNcLL7zg7u49e/b0jz/+2N1D5xVr167tOTk5uT5j/fr13qhRo8j0jBkzvE+fPgdsa8WKFZ6RkeHuofOdQ4cOjbw3cuRIf/DBBw9YJ79zidnZ2V6jRo3I9OrVq71Zs2Y+a9asyLycnByvUqWKZ2dnu3vo3F96evoBn5WVlZXrs9zzPu8b7cMPP8zz/ei6Tj75ZG/UqJE3atTIa9as6Yceeqg/8cQTB6xzyy23+NixY33BggVet27dyDppaWneoEEDX79+fWTZc889119++eVc6+/Zs8d79uzpjzzySL71uifQOUpgLRDdfq8fnpfnMmZWDqgJbI5jTVJGvPbaa1x44YV8//33rFy5ktWrV9OkSRO2bNnCunXrWLJkCQDff/898+fPp127dlSpUoVhw4ZxzTXXsGfPHgA2bdrEq6++ChTeojznnHN44YUXIts/7bTTDmipHnnkkTRo0CByCD916lTS00On5tevXx9Z7o033qBVq1YAnHnmmSxcuJAdO3aQlZXFhx9+SHp6OllZWZErvZmZmbz99tuRdb799tvIZ73zzjs0a9YMCF1xP+uss7j//vs56aSTIsuYGX379mX69OkF1jVx4kRatmxZ4Pfe3Vm2bFnk9cSJE2nRokWBdc2cOZOVK1eycuVKrr32Wm699VauvPJKgMiV/FWrVvGf//yH888/n9atW7Nx48bIOvXr12fOnDkceWToWsPWrVv58MMP6devX666hg0bRsuWLbn++usL3Iciyy9BD/aL0PnP74AmQAVgPpCx3zJXAP8Ivx4M/Luwz1WLMjEE3aLs1q2bv/vuu7nmjR492i+77DKfNWuWd+7c2du2beudOnXy999/P7LM7t27/aabbvKmTZt6RkaGH3/88T558uSYtrlz504fMGCAN23a1I877jhfvny5u7uvXbvWe/fuHVlu7ty53rFjR2/durX369fPt2zZ4u7uF1xwgbdq1cpbt27tffv29XXr1kXWeemllzw9Pd0zMjL8pptucnf3bdu2eYcOHbx169aenp7uV199deRq+NVXX+3p6enetm1b79atm3/11Vfu7n733Xd7lSpVcrWIf/jhB3d3X7lypZ9yyineunVrP+200/z77793d/fhw4d7enq6t2nTxrt16+ZLliyJ1HXyySd7nTp1vFKlSl6vXj2fPHmyZ2dne5cuXbxVq1aekZHh559/fqSVnl9d0fa/+n/yySd7y5YtvU2bNj5lypQ8v/eNGjXKddX7ueee80GDBuVaZubMmQ5469atI/v+zjvv5Pl5RW1RmpfMdZM8mVkf4DEgDRjn7veY2chwQRPNrBLwEtAe2AIM9vDFn/x06tTJZ8+eHbeaJTZLliwptOUhkqjy+v01sy/dvVNey8f1UrC7TwIm7TfvzqjXu4Dz4lmDiMjBKvNP5oiIHCwFpYhIIRSUUmzxPL8tEi/F+b1VUEqxVKpUic2bNyssJam4h/qjrFSpUpHW03N9Uiz169dnzZo1FOVJKZFEsLeH86JQUEqxlC9fvkg9RIskMx16i4gUQkEpIlIIBaWISCHi+ghjPJjZJuD7Iq5WB/gxDuWUtrKyH6B9SVRlZV+Ksx+N3L1uXm8kXVAWh5nNzu8ZzmRSVvYDtC+JqqzsS0nvhw69RUQKoaAUESlEqgTlU0EXUELKyn6A9iVRlZV9KdH9SIlzlCIiByNVWpQiIsVWpoLSzHqZ2VIzW2Zmw/N4v6KZTQi//5mZNS79KgsXw35cb2aLzWyBmU01s0ZB1BmLwvYlarn+ZuZmlrBXXGPZFzMbGP7ZLDKzf5V2jbGI4feroZlNM7O54d+xPkHUGQszG2dmG83sq3zeNzN7PLyvC8ysQ7E2lN8YEcn2RWi4ieXAb9g3Rk/6fsv8mdxj9EwIuu5i7kd3oEr49eWJuB+x7kt4uerADOBToFPQdR/Ez6UZMBc4NDx9eNB1F3M/ngIuD79OB1YGXXcB+3Mq0AH4Kp/3+wDvAgacAHxWnO2UpRbl8cAyd//O3fcArwD99lumH/BC+PVrQA+LZcDn0lXofrj7NHffEZ78lNAIl4kolp8JwN3AA8Cu0iyuiGLZlz8CY939JwB331jKNcYilv1wYO+A2DWBdaVYX5G4+wxC423lpx/wood8CtQys6OKup2yFJT1gNVR02vC8/Jcxt2zgK1A7VKpLnax7Ee0YYT+YiaiQvclfCjUwN3fKc3CiiGWn0tzoLmZfWRmn5pZr1KrLnax7McI4AIzW0NozKurSqe0uCjq/6c8qZu1JGZmFwCdgK5B11IcZnYI8ChwccCllJRyhA6/uxFq5c8ws9bu/nOgVRXdEOB5d3/EzE4EXjKzVu6eE3RhQSlLLcq1QIOo6frheXkuY2blCB1WbC6V6mIXy35gZqcDtwHnuPvuUqqtqArbl+pAK2C6ma0kdA5pYoJe0Inl57IGmOjume6+AviGUHAmklj2YxjwbwB3/wSoROjZ6WQU0/+nwpSloPwCaGZmTcysAqGLNRP3W2YicFH49QDgfx4+45tACt0PM2sPPEkoJBPxPNheBe6Lu2919zru3tjdGxM633qOuyfiwO2x/H79l1BrEjOrQ+hQvMBx6gMQy36sAnoAmFlLQkGZrF3ZTwSGhq9+nwBsdff1Rf6UoK9alfAVsD6E/oovB24LzxtJ6D8fhH7grwLLgM+B3wRdczH3YwrwAzAv/DUx6JqLuy/7LTudBL3qHePPxQidSlgMLAQGB11zMfcjHfiI0BXxeUDPoGsuYF/GA+uBTEIt+mHAZcBlUT+TseF9XVjc3y89mSMiUoiydOgtIhIXCkoRkUIoKEVECqGgFBEphIJSRKQQCkqJiZllm9m8qK/GBSy7rQS297yZrQhva074CZGifsYzZpYefn3rfu99fLA1hj9n7/flKzN7y8xqFbJ8u0TujUfyptuDJCZmts3dq5X0sgV8xvPA2+7+mpn1BB529zYH8XkHXVNhn2tmLwDfuPs9BSx/MaF7+a4s6VokftSilGIxs2rhvjDnmNlCMzugVyAzO8rMZkS1uE4Jz+9pZp+E133VzAoLsBnAMeF1rw9/1ldmdm14XlUze8fM5ofnDwrPn25mnczsfqByuI6Xw+9tC//7ipmdFVXz82Y2wMzSzOwhM/si3I/hn2L4tnxCuMMFMzs+vI9zzexjMzs2/CTMSGBQuJZB4drHmdnn4WXz6l1Jghb0nfX6So4vIJt9TwK9QagDiBrh9+oQetpp7xHKtvC/N7DvyY80Qs921yEUfFXD828G7sxje88DA8KvzwM+AzoSerqiKlANWAS0B/oDT0etWzP873TCT2LsrSlqmb01/hZ4Ify6AqGeZioDlwK3h+dXBGYDTfKoc1vU/r0K9ApP1wDKhV+fDrwefn0xMCZq/XuBC8KvaxF6YqZq0D9vfeX+Uu9BEqud7t5u74SZlQfuNbNTgRxCLakjgA1R63wBjAsv+193n2dmXQk/IhfuCrQCoZZYXh4ys9sJPWc8jNDzx2+4+/ZwDf8BTgEmA4+Y2QOEDtdnFmG/3gVGm1lFoBcww913hg/325jZgPByNQl1cLFiv/Urm9m88P4vAT6IWv4FM2tGqH/H8vlsvydwjpndGJ6uBDQMf5YkCAWlFNfvgbpAR3fPDPf+Uyl6AXefEQ7Ss4DnzexR4CfgA3cfEsM2bnL31/ZOmFmPvBZy928s1K9lH2CUmU1195Gx7IS77zKz6cCZwCBCHdlC6Bnhq9z9vUI+Yqe7tzOzKsB7wBXA44Q6I57m7r8NX/ians/6BvR396Wx1CvB0DlKKa6awMZwSHYHDhi3x0Jj+fzg7k8DzxDqsv9T4CQz23vOsaqZNY9xmzOBc82siplVJXTYPNPMjgZ2uPs/gYfC29lfZrhlm5cJwCXsa51CKPQu37uOmTUPbzNPHupx/mrgBtvXhd/e7rwujlr0V0KnIPZ6D7jKws1rC/UMJQlGQSnF9TLQycwWAkOBr/NYphsw38zmEmqtjXb3TYSCY7yZLSB02N0ilg26+xxC5y4/J3TO8hl3nwu0Bj4PHwLfBYzKY/WngAV7L+bs531CnR9P8dDwCBAK9sXAHAsNXPUkhRyBhWtZQKjj2weB+8L7Hr3eNCB978UcQi3P8uHaFoWnJcHo9iARkUKoRSkiUggFpYhIIRSUIiKFUFCKiBRCQSkiUggFpYhIIRSUIiKFUFCKiBTi/wNYyF48nnx2WQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "#define metrics\n",
        "from sklearn.metrics import roc_auc_score, roc_curve\n",
        "y1_pred_proba = model1.predict_proba(X_test)[::,1]\n",
        "fpr, tpr, _ = roc_curve(y_test,  y1_pred_proba)\n",
        "aucLR = roc_auc_score(y_test, y1_pred_proba)\n",
        "print(aucLR)\n",
        "\n",
        "#create ROC curve\n",
        "plt.figure(figsize = (5,6))\n",
        "plt.plot(fpr,tpr,label=\"AUC=\"+str(aucLR))\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.legend(loc=4)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "importance_lg = abs(model1.coef_[0])\n",
        "feature['score'] = importance_lg\n",
        "feature\n",
        "# plot feature importance\n",
        "sns.barplot(data = feature, y = 'feature', x = 'score', color='steelblue')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        },
        "id": "c-3VxlMmbYJx",
        "outputId": "33aa41a3-abf9-49f2-beb9-8b8b3fb01b21"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f5062fa1a50>"
            ]
          },
          "metadata": {},
          "execution_count": 50
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Dengan menggunakan model Logistic Regression, feature yang paling berpengaruh\n",
        "terhadap isFraud adalah merchantCategoryCode\n",
        "\n",
        "\n",
        "---"
      ],
      "metadata": {
        "id": "_nduZbZtJdNa"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6dfQwH39VgdZ"
      },
      "source": [
        "### Random Forest"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Yi1ACK4HYm37"
      },
      "outputs": [],
      "source": [
        "# Define the model as the random forest\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "\n",
        "# Hyper Parameter Tuning\n",
        "def tuning(model,parameter,n_cv):\n",
        "    random = GridSearchCV(model, parameter, cv = n_cv)\n",
        "    random.fit(X_train,y_train)\n",
        "    best_param = random.best_params_\n",
        "    return best_param\n",
        "param = {\n",
        "    'max_features': ['log2', 'sqrt','auto'],\n",
        "    'criterion':['entropy','gini'],\n",
        "    'max_depth': [2, 5, 10, 15, 30, None],\n",
        "    \"min_samples_split\":[5,8,10,20,30],\n",
        "    \"min_samples_leaf\":[3,5,10,15,20]\n",
        "}\n",
        "param_best = tuning(RandomForestClassifier(),param,5)\n",
        "\n",
        "# Fit the model to our training set\n",
        "rf_model = RandomForestClassifier(**param_best)\n",
        "rf_model.fit(X_train,y_train)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the model as the random forest\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "\n",
        "# Fit the model to our training set\n",
        "rf_model = RandomForestClassifier(max_depth=15, min_samples_leaf=5, min_samples_split=8)\n",
        "rf_model.fit(X_train,y_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uTKa7onUeSWQ",
        "outputId": "a5f95789-5e06-4c77-d5b8-a73b393aa080"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomForestClassifier(max_depth=15, min_samples_leaf=5, min_samples_split=8)"
            ]
          },
          "metadata": {},
          "execution_count": 52
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-mPuzc9Jazsu"
      },
      "outputs": [],
      "source": [
        "# Obtain predictions from the test data\n",
        "y2_pred = rf_model.predict(X_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "guO2qWlzbuW7",
        "outputId": "1568549f-f443-430c-866d-e0e99248ad25"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.67      0.80    125442\n",
            "           1       0.04      0.69      0.07      2215\n",
            "\n",
            "    accuracy                           0.67    127657\n",
            "   macro avg       0.51      0.68      0.44    127657\n",
            "weighted avg       0.98      0.67      0.79    127657\n",
            "\n",
            "\n",
            "Confusion Matrix:\n",
            "[[84556 40886]\n",
            " [  678  1537]]\n"
          ]
        }
      ],
      "source": [
        "# Print the ROC curve, classification report and confusion matrix\n",
        "print('\\nClassification Report:')\n",
        "print(classification_report(y_test, y2_pred))\n",
        "print('\\nConfusion Matrix:')\n",
        "print(confusion_matrix(y_test, y2_pred))"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "cm = confusion_matrix(y_test, rf_model.predict(X_test))\n",
        "cm = pd.DataFrame(cm , index = ['0', '1'] , columns = ['0', '1'])\n",
        "plt.figure(figsize = (5,5))\n",
        "plt.title(\"Random Forest Confusion Matrix\")\n",
        "sns.heatmap(cm, cmap='Blues', annot = True, fmt='',xticklabels = [\"0\", \"1\"], yticklabels = [\"0\", \"1\"])\n",
        "plt.xlabel('Predicted')\n",
        "plt.ylabel('Actual')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 368
        },
        "id": "c8Kn_Fgw7_Ug",
        "outputId": "512e3098-f2e5-41b6-b3a8-dced2c7e27b8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(24.0, 0.5, 'Actual')"
            ]
          },
          "metadata": {},
          "execution_count": 80
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x360 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 422
        },
        "id": "MOsp1EmhdNPC",
        "outputId": "9b9e40ba-278c-4bde-cfbc-2809eca3e0ba"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.7465162967044243\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "#define metrics\n",
        "y2_pred_proba = rf_model.predict_proba(X_test)[::,1]\n",
        "fpr, tpr, _ = roc_curve(y_test,  y2_pred_proba)\n",
        "aucRF = roc_auc_score(y_test, y2_pred_proba)\n",
        "print(aucRF)\n",
        "\n",
        "#create ROC curve\n",
        "plt.figure(figsize=(5,6))\n",
        "plt.plot(fpr,tpr,label=\"AUC=\"+str(aucRF))\n",
        "plt.title('ROC AUC Score Rnadom Forest')\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.legend(loc=4)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "importances_rf = rf_model.feature_importances_"
      ],
      "metadata": {
        "id": "0LnRFz9yVZVO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "indices_rf = np.argsort(importances_rf)\n",
        "\n",
        "fig, ax = plt.subplots()\n",
        "ax.barh(range(len(importances_rf)), importances_rf[indices_rf], color='steelblue')\n",
        "ax.set_yticks(range(len(importances_rf)))\n",
        "_ = ax.set_yticklabels(np.array(X_train.columns)[indices_rf])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "K0QUX7YYZbza",
        "outputId": "89aad472-4483-455c-f6de-659573fa0584"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Dengan menggunakan model Random Forest, feature yang paling berpengaruh terhadap isFraud adalah transactionAmount\n",
        "\n",
        "---"
      ],
      "metadata": {
        "id": "de1NuuJoHsVn"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v27O06-SnS5p"
      },
      "source": [
        "### Decision Tree"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mefSJ-p3l9JM",
        "outputId": "2b8ce82c-d9ca-44b0-cf4e-a7bc051412a4"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeClassifier(criterion='entropy', max_depth=8, max_features='log2',\n",
              "                       min_samples_leaf=15, min_samples_split=5)"
            ]
          },
          "metadata": {},
          "execution_count": 66
        }
      ],
      "source": [
        "# Define the model as the Decision Tree\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "\n",
        "# Hyper Parameter Tuning\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "def tuning(model,parameter,n_cv):\n",
        "    random = GridSearchCV(model, parameter, cv = n_cv)\n",
        "    random.fit(X_train,y_train)\n",
        "    best_param = random.best_params_\n",
        "    return best_param\n",
        "param = {\n",
        "    'max_features': ['log2', 'sqrt','auto'],\n",
        "    'criterion':['entropy','gini'],\n",
        "    'max_depth': [5, 8, 10, 15, 30, None],\n",
        "    \"min_samples_split\":[5,10,15,30],\n",
        "    \"min_samples_leaf\":[5,8,10,15,30]\n",
        "}\n",
        "param_best = tuning(DecisionTreeClassifier(),param,5)\n",
        "\n",
        "# Fit the model to our training set\n",
        "dct_tune = DecisionTreeClassifier(**param_best)\n",
        "dct_tune.fit(X_train,y_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FlIZv6S3nEdZ",
        "outputId": "a3612b66-9fce-42fc-a0ab-7a5d1a201298"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.69      0.81    125442\n",
            "           1       0.03      0.62      0.07      2215\n",
            "\n",
            "    accuracy                           0.69    127657\n",
            "   macro avg       0.51      0.66      0.44    127657\n",
            "weighted avg       0.97      0.69      0.80    127657\n",
            "\n",
            "\n",
            "Confusion Matrix:\n",
            "[[86726 38716]\n",
            " [  831  1384]]\n"
          ]
        }
      ],
      "source": [
        "# Obtain model predictions\n",
        "y3_pred = dct_tune.predict(X_test)\n",
        "\n",
        "# Print the ROC curve, classification report and confusion matrix\n",
        "print('\\nClassification Report:')\n",
        "print(classification_report(y_test, y3_pred))\n",
        "print('\\nConfusion Matrix:')\n",
        "print(confusion_matrix(y_test, y3_pred))"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "cm = confusion_matrix(y_test, dct_tune.predict(X_test))\n",
        "cm = pd.DataFrame(cm , index = ['0', '1'] , columns = ['0', '1'])\n",
        "plt.figure(figsize = (5,5))\n",
        "plt.title(\"Decision Tree Confusion Matrix\")\n",
        "sns.heatmap(cm, cmap='Blues', annot = True, fmt='',xticklabels = [\"0\", \"1\"], yticklabels = [\"0\", \"1\"])\n",
        "plt.xlabel('Predicted')\n",
        "plt.ylabel('Actual')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 368
        },
        "id": "33mpTLQ7O6It",
        "outputId": "3d854680-8cf5-44cd-a84a-b5dcf051375d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(24.0, 0.5, 'Actual')"
            ]
          },
          "metadata": {},
          "execution_count": 82
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x360 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 406
        },
        "id": "j0Zn_O2Knfb9",
        "outputId": "ea46c49f-d293-494a-b3fa-2a974ec30ffe"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.7063397370914504\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "#define metrics\n",
        "y3_pred_proba = dct_tune.predict_proba(X_test)[::,1]\n",
        "fpr, tpr, _ = roc_curve(y_test,  y3_pred_proba)\n",
        "aucDT = roc_auc_score(y_test, y3_pred_proba)\n",
        "print(aucDT)\n",
        "\n",
        "#create ROC curve\n",
        "plt.figure(figsize=(5,6))\n",
        "plt.plot(fpr,tpr,label=\"AUC=\"+str(aucDT))\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.legend(loc=4)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "importances_dt = dct_tune.feature_importances_"
      ],
      "metadata": {
        "id": "v1DNN9rOZsPW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "indices_dt = np.argsort(importances_dt)\n",
        "\n",
        "fig, ax = plt.subplots()\n",
        "ax.barh(range(len(importances_dt)), importances_dt[indices_dt], color='steelblue')\n",
        "ax.set_yticks(range(len(importances_dt)))\n",
        "_ = ax.set_yticklabels(np.array(X_train.columns)[indices_dt])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "tdQy38jJZ0qz",
        "outputId": "74194740-2557-4547-d108-d495b7039ead"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Dengan menggunakan model Decision Tree, feature yang paling berpengaruh terhadap isFraud adalah transactionAmount\n",
        "\n",
        "\n",
        "---"
      ],
      "metadata": {
        "id": "-509LmdVH8aI"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Compare Accuracy Model"
      ],
      "metadata": {
        "id": "Wddyz2TfAvlx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score\n",
        "\n",
        "logreg_acc = accuracy_score(y_test, model1.predict(X_test))\n",
        "logreg_recall = recall_score(y_test, model1.predict(X_test))\n",
        "logreg_precision = precision_score(y_test, model1.predict(X_test))\n",
        "logreg_f1 = f1_score(y_test, model1.predict(X_test))\n",
        "logreg_auc = aucLR\n",
        "\n",
        "rf_acc = accuracy_score(y_test, rf_model.predict(X_test))\n",
        "rf_recall = recall_score(y_test, rf_model.predict(X_test))\n",
        "rf_precision = precision_score(y_test, rf_model.predict(X_test))\n",
        "rf_f1 = f1_score(y_test, rf_model.predict(X_test))\n",
        "rf_auc = aucRF\n",
        "\n",
        "dt_acc = accuracy_score(y_test, dct_tune.predict(X_test))\n",
        "dt_recall = recall_score(y_test, dct_tune.predict(X_test))\n",
        "dt_precision = precision_score(y_test, dct_tune.predict(X_test))\n",
        "dt_f1 = f1_score(y_test, dct_tune.predict(X_test))\n",
        "dt_auc = aucDT"
      ],
      "metadata": {
        "id": "JQXxxhXsrIuA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_performances = pd.DataFrame({\n",
        "    \"Model\": [\"Logistic Regression\", \"Random Forest\", \"Decision Tree\"],\n",
        "    \"Accuracy\": [logreg_acc, rf_acc, dt_acc],\n",
        "    \"Recall\": [logreg_recall, rf_recall, dt_recall],\n",
        "    \"Precision\": [logreg_precision, rf_precision, dt_precision],\n",
        "    \"F1-Score\": [logreg_f1, rf_f1, dt_f1],\n",
        "    \"AUC Score\": [logreg_auc, rf_auc, dt_auc]\n",
        "})\n",
        "\n",
        "model_performances"
      ],
      "metadata": {
        "id": "egNyhv-qr1S-",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "outputId": "84ba9ecb-f138-4095-cef7-ca5d0431c2d1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                 Model  Accuracy    Recall  Precision  F1-Score  AUC Score\n",
              "0  Logistic Regression  0.630322  0.596840   0.027761  0.053054   0.656533\n",
              "1        Random Forest  0.674409  0.693905   0.036230  0.068865   0.746516\n",
              "2        Decision Tree  0.690209  0.624831   0.034514  0.065414   0.706340"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-849a3d61-2869-4f1f-bfce-bb931fb7de3a\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Recall</th>\n",
              "      <th>Precision</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC Score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Logistic Regression</td>\n",
              "      <td>0.630322</td>\n",
              "      <td>0.596840</td>\n",
              "      <td>0.027761</td>\n",
              "      <td>0.053054</td>\n",
              "      <td>0.656533</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Random Forest</td>\n",
              "      <td>0.674409</td>\n",
              "      <td>0.693905</td>\n",
              "      <td>0.036230</td>\n",
              "      <td>0.068865</td>\n",
              "      <td>0.746516</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Decision Tree</td>\n",
              "      <td>0.690209</td>\n",
              "      <td>0.624831</td>\n",
              "      <td>0.034514</td>\n",
              "      <td>0.065414</td>\n",
              "      <td>0.706340</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-849a3d61-2869-4f1f-bfce-bb931fb7de3a')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-849a3d61-2869-4f1f-bfce-bb931fb7de3a button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-849a3d61-2869-4f1f-bfce-bb931fb7de3a');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 77
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Dari perbandingan diatas dapat dilihat model yang bagus untuk memprediksi suatu transaksi fraud atau legitimate adalah Random Forest"
      ],
      "metadata": {
        "id": "zY3wiQn6J5qt"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Kesimpulan"
      ],
      "metadata": {
        "id": "cOuijyoxKhtf"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Feature yang paling berpengaruh terhadap suatu transaksi fraud atau legitimate berdasarkan model yang dipakai adalah sebagai berikut,\n",
        "1. Logistic Regression: merchantCategoryCode\n",
        "2. Random Forest: transactionAmount\n",
        "3. Decision Tree: transactionAmount\n",
        "\n",
        "Serta model yang terbaik untuk memprediksi suatu transaksi fraud atau legitimate adalah Random Forest."
      ],
      "metadata": {
        "id": "xCo8UoimKm68"
      }
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}