CreditCardFraud.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CreditCardFraud",
      "version": "0.3.2",
      "provenance": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/varunsboy17/Credit-Card-Fraud/blob/master/CreditCardFraud.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KYgn0dtZqGu3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import gc,os,sys\n",
        "import re\n",
        "\n",
        "from sklearn import metrics, preprocessing\n",
        "from sklearn.model_selection import StratifiedKFold\n",
        "from sklearn.decomposition import PCA, KernelPCA\n",
        "from sklearn.cluster import KMeans\n",
        "from tqdm import tqdm\n",
        "\n",
        "pd.options.display.float_format = '{:.3f}'.format"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2OmTrYTEW_Ri",
        "colab_type": "code",
        "outputId": "67ed3ca7-845a-42ca-ccde-1e6070cd3ee8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 62,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ut5tC42jrZRU",
        "colab_type": "code",
        "outputId": "1ec56d4d-72df-4450-94c3-760a75bc0b3c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "%%time\n",
        "path_train_trans = \"/content/drive/My Drive/Credit Card Fraud Dataset/train_transaction.csv\"\n",
        "path_train_id = \"/content/drive/My Drive/Credit Card Fraud Dataset/train_identity.csv\"\n",
        "path_test_trans = \"/content/drive/My Drive/Credit Card Fraud Dataset/test_transaction.csv\"\n",
        "path_test_id = \"/content/drive/My Drive/Credit Card Fraud Dataset/test_identity.csv\"\n",
        "\n",
        "train_transaction = pd.read_csv(path_train_trans)\n",
        "train_identity = pd.read_csv(path_train_id)\n",
        "test_transaction = pd.read_csv(path_test_trans)\n",
        "test_identity = pd.read_csv(path_test_id)\n",
        "\n",
        "#train_transaction = pd.read_csv(b'D:\\\\Personal\\\\SR\\\\Credit card Fraud\\\\ieee-fraud-detection\\\\train_transaction.csv')\n",
        "#train_identity = pd.read_csv(\"D:\\\\Personal\\\\SR\\\\Credit card Fraud\\\\ieee-fraud-detection\\\\train_identity.csv\")\n",
        "#test_transaction = pd.read_csv(\"D:\\\\Personal\\\\SR\\\\Credit card Fraud\\\\ieee-fraud-detection\\\\test_transaction.csv\")\n",
        "#test_identity = pd.read_csv(\"D:\\\\Personal\\\\SR\\\\Credit card Fraud\\\\ieee-fraud-detection\\\\test_identity.csv\")"
      ],
      "execution_count": 63,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "CPU times: user 52.8 s, sys: 7.44 s, total: 1min\n",
            "Wall time: 1min 3s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rRBVCIKM5pJ_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def reduce_mem_usage(df):\n",
        "  start_mem = df.memory_usage().sum() /1024 **2\n",
        "  print(\"Memory usage of dataframe is {:.2f} MB\".format(start_mem))\n",
        "  for col in df.columns:\n",
        "    col_type = df[col].dtype\n",
        "    \n",
        "    if col_type != object:\n",
        "      cmin = df[col].min()\n",
        "      cmax = df[col].max()\n",
        "      if str(col_type)[:3] == 'int':\n",
        "        if cmin > np.iinfo(np.int8).min and cmax < np.iinfo(np.int8).max:\n",
        "          df[col] = df[col].astype(np.int8)\n",
        "        elif cmin > np.iinfo(np.int16).min and cmax < np.iinfo(np.int16).max:\n",
        "          df[col] = df[col].astype(np.int16)\n",
        "        elif cmin > np.iinfo(np.int32).min and cmax < np.iinfo(np.int32).max:\n",
        "          df[col] = df[col].astype(np.int32)\n",
        "        elif cmin > np.iinfo(np.int64).min and cmax < np.iinfo(np.int64).max:\n",
        "          df[col] = df[col].astype(np.int64)\n",
        "      else:\n",
        "        if cmin > np.finfo(np.float16).min and cmax > np.finfo(np.float16).max:\n",
        "          df[col] = df[col].astype(np.float16)\n",
        "        elif cmin > np.finfo(np.float32).min and cmax > np.finfo(np.float32).max:\n",
        "          df[col] = df[col].astype(np.float32)\n",
        "        else:\n",
        "          df[col] = df[col].astype(np.float64)\n",
        "    else:\n",
        "      df[col] = df[col].astype('category')\n",
        "  end_mem = df.memory_usage().sum() / 1024 ** 2\n",
        "  print(\"Memory usage after optimization is {:.2f} MB\".format(end_mem))\n",
        "  print(\"Decreased by {:.1f} %\".format(100 * (start_mem - end_mem) / start_mem))\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KYt8mHQIc4HT",
        "colab_type": "code",
        "outputId": "431cc8ad-ded6-4c5d-8367-480c21a850a1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 253
        }
      },
      "source": [
        "train_transaction.head()"
      ],
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>TransactionID</th>\n",
              "      <th>isFraud</th>\n",
              "      <th>TransactionDT</th>\n",
              "      <th>TransactionAmt</th>\n",
              "      <th>ProductCD</th>\n",
              "      <th>card1</th>\n",
              "      <th>card2</th>\n",
              "      <th>card3</th>\n",
              "      <th>card4</th>\n",
              "      <th>card5</th>\n",
              "      <th>card6</th>\n",
              "      <th>addr1</th>\n",
              "      <th>addr2</th>\n",
              "      <th>dist1</th>\n",
              "      <th>dist2</th>\n",
              "      <th>P_emaildomain</th>\n",
              "      <th>R_emaildomain</th>\n",
              "      <th>C1</th>\n",
              "      <th>C2</th>\n",
              "      <th>C3</th>\n",
              "      <th>C4</th>\n",
              "      <th>C5</th>\n",
              "      <th>C6</th>\n",
              "      <th>C7</th>\n",
              "      <th>C8</th>\n",
              "      <th>C9</th>\n",
              "      <th>C10</th>\n",
              "      <th>C11</th>\n",
              "      <th>C12</th>\n",
              "      <th>C13</th>\n",
              "      <th>C14</th>\n",
              "      <th>D1</th>\n",
              "      <th>D2</th>\n",
              "      <th>D3</th>\n",
              "      <th>D4</th>\n",
              "      <th>D5</th>\n",
              "      <th>D6</th>\n",
              "      <th>D7</th>\n",
              "      <th>D8</th>\n",
              "      <th>D9</th>\n",
              "      <th>...</th>\n",
              "      <th>V300</th>\n",
              "      <th>V301</th>\n",
              "      <th>V302</th>\n",
              "      <th>V303</th>\n",
              "      <th>V304</th>\n",
              "      <th>V305</th>\n",
              "      <th>V306</th>\n",
              "      <th>V307</th>\n",
              "      <th>V308</th>\n",
              "      <th>V309</th>\n",
              "      <th>V310</th>\n",
              "      <th>V311</th>\n",
              "      <th>V312</th>\n",
              "      <th>V313</th>\n",
              "      <th>V314</th>\n",
              "      <th>V315</th>\n",
              "      <th>V316</th>\n",
              "      <th>V317</th>\n",
              "      <th>V318</th>\n",
              "      <th>V319</th>\n",
              "      <th>V320</th>\n",
              "      <th>V321</th>\n",
              "      <th>V322</th>\n",
              "      <th>V323</th>\n",
              "      <th>V324</th>\n",
              "      <th>V325</th>\n",
              "      <th>V326</th>\n",
              "      <th>V327</th>\n",
              "      <th>V328</th>\n",
              "      <th>V329</th>\n",
              "      <th>V330</th>\n",
              "      <th>V331</th>\n",
              "      <th>V332</th>\n",
              "      <th>V333</th>\n",
              "      <th>V334</th>\n",
              "      <th>V335</th>\n",
              "      <th>V336</th>\n",
              "      <th>V337</th>\n",
              "      <th>V338</th>\n",
              "      <th>V339</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2987000</td>\n",
              "      <td>0</td>\n",
              "      <td>86400</td>\n",
              "      <td>68.500</td>\n",
              "      <td>W</td>\n",
              "      <td>13926</td>\n",
              "      <td>nan</td>\n",
              "      <td>150.000</td>\n",
              "      <td>discover</td>\n",
              "      <td>142.000</td>\n",
              "      <td>credit</td>\n",
              "      <td>315.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>19.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>2.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>14.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>13.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>117.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>117.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2987001</td>\n",
              "      <td>0</td>\n",
              "      <td>86401</td>\n",
              "      <td>29.000</td>\n",
              "      <td>W</td>\n",
              "      <td>2755</td>\n",
              "      <td>404.000</td>\n",
              "      <td>150.000</td>\n",
              "      <td>mastercard</td>\n",
              "      <td>102.000</td>\n",
              "      <td>credit</td>\n",
              "      <td>325.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>gmail.com</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2987002</td>\n",
              "      <td>0</td>\n",
              "      <td>86469</td>\n",
              "      <td>59.000</td>\n",
              "      <td>W</td>\n",
              "      <td>4663</td>\n",
              "      <td>490.000</td>\n",
              "      <td>150.000</td>\n",
              "      <td>visa</td>\n",
              "      <td>166.000</td>\n",
              "      <td>debit</td>\n",
              "      <td>330.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>287.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>outlook.com</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2987003</td>\n",
              "      <td>0</td>\n",
              "      <td>86499</td>\n",
              "      <td>50.000</td>\n",
              "      <td>W</td>\n",
              "      <td>18132</td>\n",
              "      <td>567.000</td>\n",
              "      <td>150.000</td>\n",
              "      <td>mastercard</td>\n",
              "      <td>117.000</td>\n",
              "      <td>debit</td>\n",
              "      <td>476.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>yahoo.com</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2.000</td>\n",
              "      <td>5.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>4.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>25.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>112.000</td>\n",
              "      <td>112.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>94.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>50.000</td>\n",
              "      <td>1758.000</td>\n",
              "      <td>925.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>354.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>135.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>50.000</td>\n",
              "      <td>1404.000</td>\n",
              "      <td>790.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2987004</td>\n",
              "      <td>0</td>\n",
              "      <td>86506</td>\n",
              "      <td>50.000</td>\n",
              "      <td>H</td>\n",
              "      <td>4497</td>\n",
              "      <td>514.000</td>\n",
              "      <td>150.000</td>\n",
              "      <td>mastercard</td>\n",
              "      <td>102.000</td>\n",
              "      <td>credit</td>\n",
              "      <td>420.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>gmail.com</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 394 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "   TransactionID  isFraud  TransactionDT  ...  V337  V338  V339\n",
              "0        2987000        0          86400  ...   nan   nan   nan\n",
              "1        2987001        0          86401  ...   nan   nan   nan\n",
              "2        2987002        0          86469  ...   nan   nan   nan\n",
              "3        2987003        0          86499  ...   nan   nan   nan\n",
              "4        2987004        0          86506  ... 0.000 0.000 0.000\n",
              "\n",
              "[5 rows x 394 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tL9nXIL132bX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#train_trns = reduce_mem_usage(train_transaction)\n",
        "#train_identity = reduce_mem_usage(train_identity)\n",
        "#test_transaction = reduce_mem_usage(test_transaction)\n",
        "#test_identity = reduce_mem_usage(test_identity)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iSbs7NKdHcI5",
        "colab_type": "code",
        "outputId": "a9ce1f48-26aa-486a-9121-4bc4ddcb8443",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "print(train_transaction.shape, test_transaction.shape)\n",
        "print(train_identity.shape, test_identity.shape)"
      ],
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(590540, 394) (506691, 393)\n",
            "(144233, 41) (141907, 41)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TbYwLzuYu2SV",
        "colab_type": "code",
        "outputId": "c9ea8bdb-c66c-4def-a23f-4821cccd74da",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 360
        }
      },
      "source": [
        "train_identity.head()"
      ],
      "execution_count": 68,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>TransactionID</th>\n",
              "      <th>id_01</th>\n",
              "      <th>id_02</th>\n",
              "      <th>id_03</th>\n",
              "      <th>id_04</th>\n",
              "      <th>id_05</th>\n",
              "      <th>id_06</th>\n",
              "      <th>id_07</th>\n",
              "      <th>id_08</th>\n",
              "      <th>id_09</th>\n",
              "      <th>id_10</th>\n",
              "      <th>id_11</th>\n",
              "      <th>id_12</th>\n",
              "      <th>id_13</th>\n",
              "      <th>id_14</th>\n",
              "      <th>id_15</th>\n",
              "      <th>id_16</th>\n",
              "      <th>id_17</th>\n",
              "      <th>id_18</th>\n",
              "      <th>id_19</th>\n",
              "      <th>id_20</th>\n",
              "      <th>id_21</th>\n",
              "      <th>id_22</th>\n",
              "      <th>id_23</th>\n",
              "      <th>id_24</th>\n",
              "      <th>id_25</th>\n",
              "      <th>id_26</th>\n",
              "      <th>id_27</th>\n",
              "      <th>id_28</th>\n",
              "      <th>id_29</th>\n",
              "      <th>id_30</th>\n",
              "      <th>id_31</th>\n",
              "      <th>id_32</th>\n",
              "      <th>id_33</th>\n",
              "      <th>id_34</th>\n",
              "      <th>id_35</th>\n",
              "      <th>id_36</th>\n",
              "      <th>id_37</th>\n",
              "      <th>id_38</th>\n",
              "      <th>DeviceType</th>\n",
              "      <th>DeviceInfo</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2987004</td>\n",
              "      <td>0.000</td>\n",
              "      <td>70787.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>100.000</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>nan</td>\n",
              "      <td>-480.000</td>\n",
              "      <td>New</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>166.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>542.000</td>\n",
              "      <td>144.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>New</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>Android 7.0</td>\n",
              "      <td>samsung browser 6.2</td>\n",
              "      <td>32.000</td>\n",
              "      <td>2220x1080</td>\n",
              "      <td>match_status:2</td>\n",
              "      <td>T</td>\n",
              "      <td>F</td>\n",
              "      <td>T</td>\n",
              "      <td>T</td>\n",
              "      <td>mobile</td>\n",
              "      <td>SAMSUNG SM-G892A Build/NRD90M</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2987008</td>\n",
              "      <td>-5.000</td>\n",
              "      <td>98945.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>0.000</td>\n",
              "      <td>-5.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>100.000</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>49.000</td>\n",
              "      <td>-300.000</td>\n",
              "      <td>New</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>166.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>621.000</td>\n",
              "      <td>500.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>New</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>iOS 11.1.2</td>\n",
              "      <td>mobile safari 11.0</td>\n",
              "      <td>32.000</td>\n",
              "      <td>1334x750</td>\n",
              "      <td>match_status:1</td>\n",
              "      <td>T</td>\n",
              "      <td>F</td>\n",
              "      <td>F</td>\n",
              "      <td>T</td>\n",
              "      <td>mobile</td>\n",
              "      <td>iOS Device</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2987010</td>\n",
              "      <td>-5.000</td>\n",
              "      <td>191631.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>100.000</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>52.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>Found</td>\n",
              "      <td>Found</td>\n",
              "      <td>121.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>410.000</td>\n",
              "      <td>142.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Found</td>\n",
              "      <td>Found</td>\n",
              "      <td>NaN</td>\n",
              "      <td>chrome 62.0</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>F</td>\n",
              "      <td>F</td>\n",
              "      <td>T</td>\n",
              "      <td>T</td>\n",
              "      <td>desktop</td>\n",
              "      <td>Windows</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2987011</td>\n",
              "      <td>-5.000</td>\n",
              "      <td>221832.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>0.000</td>\n",
              "      <td>-6.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>100.000</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>52.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>New</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>225.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>176.000</td>\n",
              "      <td>507.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>New</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>NaN</td>\n",
              "      <td>chrome 62.0</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>F</td>\n",
              "      <td>F</td>\n",
              "      <td>T</td>\n",
              "      <td>T</td>\n",
              "      <td>desktop</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2987016</td>\n",
              "      <td>0.000</td>\n",
              "      <td>7460.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>100.000</td>\n",
              "      <td>NotFound</td>\n",
              "      <td>nan</td>\n",
              "      <td>-300.000</td>\n",
              "      <td>Found</td>\n",
              "      <td>Found</td>\n",
              "      <td>166.000</td>\n",
              "      <td>15.000</td>\n",
              "      <td>529.000</td>\n",
              "      <td>575.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Found</td>\n",
              "      <td>Found</td>\n",
              "      <td>Mac OS X 10_11_6</td>\n",
              "      <td>chrome 62.0</td>\n",
              "      <td>24.000</td>\n",
              "      <td>1280x800</td>\n",
              "      <td>match_status:2</td>\n",
              "      <td>T</td>\n",
              "      <td>F</td>\n",
              "      <td>T</td>\n",
              "      <td>T</td>\n",
              "      <td>desktop</td>\n",
              "      <td>MacOS</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   TransactionID  id_01  ...  DeviceType                     DeviceInfo\n",
              "0        2987004  0.000  ...      mobile  SAMSUNG SM-G892A Build/NRD90M\n",
              "1        2987008 -5.000  ...      mobile                     iOS Device\n",
              "2        2987010 -5.000  ...     desktop                        Windows\n",
              "3        2987011 -5.000  ...     desktop                            NaN\n",
              "4        2987016  0.000  ...     desktop                          MacOS\n",
              "\n",
              "[5 rows x 41 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BGVlwrTqu2sS",
        "colab_type": "code",
        "outputId": "0cfc5d6f-9e72-49df-b99f-f0dcb2bf91d5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 253
        }
      },
      "source": [
        "train_transaction.head()"
      ],
      "execution_count": 69,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>TransactionID</th>\n",
              "      <th>isFraud</th>\n",
              "      <th>TransactionDT</th>\n",
              "      <th>TransactionAmt</th>\n",
              "      <th>ProductCD</th>\n",
              "      <th>card1</th>\n",
              "      <th>card2</th>\n",
              "      <th>card3</th>\n",
              "      <th>card4</th>\n",
              "      <th>card5</th>\n",
              "      <th>card6</th>\n",
              "      <th>addr1</th>\n",
              "      <th>addr2</th>\n",
              "      <th>dist1</th>\n",
              "      <th>dist2</th>\n",
              "      <th>P_emaildomain</th>\n",
              "      <th>R_emaildomain</th>\n",
              "      <th>C1</th>\n",
              "      <th>C2</th>\n",
              "      <th>C3</th>\n",
              "      <th>C4</th>\n",
              "      <th>C5</th>\n",
              "      <th>C6</th>\n",
              "      <th>C7</th>\n",
              "      <th>C8</th>\n",
              "      <th>C9</th>\n",
              "      <th>C10</th>\n",
              "      <th>C11</th>\n",
              "      <th>C12</th>\n",
              "      <th>C13</th>\n",
              "      <th>C14</th>\n",
              "      <th>D1</th>\n",
              "      <th>D2</th>\n",
              "      <th>D3</th>\n",
              "      <th>D4</th>\n",
              "      <th>D5</th>\n",
              "      <th>D6</th>\n",
              "      <th>D7</th>\n",
              "      <th>D8</th>\n",
              "      <th>D9</th>\n",
              "      <th>...</th>\n",
              "      <th>V300</th>\n",
              "      <th>V301</th>\n",
              "      <th>V302</th>\n",
              "      <th>V303</th>\n",
              "      <th>V304</th>\n",
              "      <th>V305</th>\n",
              "      <th>V306</th>\n",
              "      <th>V307</th>\n",
              "      <th>V308</th>\n",
              "      <th>V309</th>\n",
              "      <th>V310</th>\n",
              "      <th>V311</th>\n",
              "      <th>V312</th>\n",
              "      <th>V313</th>\n",
              "      <th>V314</th>\n",
              "      <th>V315</th>\n",
              "      <th>V316</th>\n",
              "      <th>V317</th>\n",
              "      <th>V318</th>\n",
              "      <th>V319</th>\n",
              "      <th>V320</th>\n",
              "      <th>V321</th>\n",
              "      <th>V322</th>\n",
              "      <th>V323</th>\n",
              "      <th>V324</th>\n",
              "      <th>V325</th>\n",
              "      <th>V326</th>\n",
              "      <th>V327</th>\n",
              "      <th>V328</th>\n",
              "      <th>V329</th>\n",
              "      <th>V330</th>\n",
              "      <th>V331</th>\n",
              "      <th>V332</th>\n",
              "      <th>V333</th>\n",
              "      <th>V334</th>\n",
              "      <th>V335</th>\n",
              "      <th>V336</th>\n",
              "      <th>V337</th>\n",
              "      <th>V338</th>\n",
              "      <th>V339</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2987000</td>\n",
              "      <td>0</td>\n",
              "      <td>86400</td>\n",
              "      <td>68.500</td>\n",
              "      <td>W</td>\n",
              "      <td>13926</td>\n",
              "      <td>nan</td>\n",
              "      <td>150.000</td>\n",
              "      <td>discover</td>\n",
              "      <td>142.000</td>\n",
              "      <td>credit</td>\n",
              "      <td>315.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>19.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>2.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>14.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>13.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>117.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>117.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2987001</td>\n",
              "      <td>0</td>\n",
              "      <td>86401</td>\n",
              "      <td>29.000</td>\n",
              "      <td>W</td>\n",
              "      <td>2755</td>\n",
              "      <td>404.000</td>\n",
              "      <td>150.000</td>\n",
              "      <td>mastercard</td>\n",
              "      <td>102.000</td>\n",
              "      <td>credit</td>\n",
              "      <td>325.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>gmail.com</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2987002</td>\n",
              "      <td>0</td>\n",
              "      <td>86469</td>\n",
              "      <td>59.000</td>\n",
              "      <td>W</td>\n",
              "      <td>4663</td>\n",
              "      <td>490.000</td>\n",
              "      <td>150.000</td>\n",
              "      <td>visa</td>\n",
              "      <td>166.000</td>\n",
              "      <td>debit</td>\n",
              "      <td>330.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>287.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>outlook.com</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2987003</td>\n",
              "      <td>0</td>\n",
              "      <td>86499</td>\n",
              "      <td>50.000</td>\n",
              "      <td>W</td>\n",
              "      <td>18132</td>\n",
              "      <td>567.000</td>\n",
              "      <td>150.000</td>\n",
              "      <td>mastercard</td>\n",
              "      <td>117.000</td>\n",
              "      <td>debit</td>\n",
              "      <td>476.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>yahoo.com</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2.000</td>\n",
              "      <td>5.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>4.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>25.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>112.000</td>\n",
              "      <td>112.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>94.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>50.000</td>\n",
              "      <td>1758.000</td>\n",
              "      <td>925.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>354.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>135.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>50.000</td>\n",
              "      <td>1404.000</td>\n",
              "      <td>790.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2987004</td>\n",
              "      <td>0</td>\n",
              "      <td>86506</td>\n",
              "      <td>50.000</td>\n",
              "      <td>H</td>\n",
              "      <td>4497</td>\n",
              "      <td>514.000</td>\n",
              "      <td>150.000</td>\n",
              "      <td>mastercard</td>\n",
              "      <td>102.000</td>\n",
              "      <td>credit</td>\n",
              "      <td>420.000</td>\n",
              "      <td>87.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>gmail.com</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>nan</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 394 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "   TransactionID  isFraud  TransactionDT  ...  V337  V338  V339\n",
              "0        2987000        0          86400  ...   nan   nan   nan\n",
              "1        2987001        0          86401  ...   nan   nan   nan\n",
              "2        2987002        0          86469  ...   nan   nan   nan\n",
              "3        2987003        0          86499  ...   nan   nan   nan\n",
              "4        2987004        0          86506  ... 0.000 0.000 0.000\n",
              "\n",
              "[5 rows x 394 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tYtNx2bgNZxk",
        "colab_type": "code",
        "outputId": "ced4050b-f2c8-424f-adf3-4d682008acde",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "[c for c in train_transaction.columns if c not in test_transaction.columns]"
      ],
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['isFraud']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TT-p3PvwORr1",
        "colab_type": "code",
        "outputId": "22db1dcd-a5b3-4e3a-8024-f5d5cd59148e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 328
        }
      },
      "source": [
        "fc = train_transaction[\"isFraud\"].value_counts(normalize=True).to_frame()\n",
        "fc.plot.bar()\n",
        "fc.T"
      ],
      "execution_count": 71,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>isFraud</th>\n",
              "      <td>0.965</td>\n",
              "      <td>0.035</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "            0     1\n",
              "isFraud 0.965 0.035"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 71
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD4CAYAAAAXUaZHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YClLSroBfIm_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#fig, ax = plt.subplots(2,1, figsize=(16,8))\n",
        "#train_transaction['_seq_day'] = train_transaction['TransactionDT']\n",
        "#train_transaction['_seq_week'] = train_transaction['_seq_day']\n",
        "#train_transaction.groupby('_seq_day')['isFraud'].mean().to_frame().plot.line(ax=ax[0])\n",
        "#train_transaction.groupby('_seq_week')['isFraud'].mean().to_frame().plot.line(ax=ax[1])\n",
        "#plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "URVvHArOoLfs",
        "colab_type": "code",
        "outputId": "66586ad5-675e-4840-c51d-38091439b4ea",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 985
        }
      },
      "source": [
        "# Fraud transaction rate by weekday, hour, month-day, and year-month\n",
        "import datetime\n",
        "START_DATE = '2017-12-01'\n",
        "start_date = datetime.datetime.strptime(START_DATE, '%Y-%m-%d')\n",
        "train_transaction['Date'] = train_transaction['TransactionDT'].apply(lambda x:start_date + datetime.timedelta(seconds=x))\n",
        "train_transaction['_year_Month'] = train_transaction['Date'].dt.year.astype(str) + '/' + train_transaction['Date'].dt.month.astype(str)\n",
        "train_transaction['_weekday'] = train_transaction['Date'].dt.dayofweek\n",
        "train_transaction['_hour'] = train_transaction['Date'].dt.hour\n",
        "train_transaction['_day'] = train_transaction['Date'].dt.day\n",
        "\n",
        "fig, ax = plt.subplots(4, 1, figsize=(16,16))\n",
        "train_transaction.groupby('_weekday')['isFraud'].mean().to_frame().plot.bar(ax=ax[0])\n",
        "train_transaction.groupby('_hour')['isFraud'].mean().to_frame().plot.bar(ax=ax[1])\n",
        "train_transaction.groupby('_day')['isFraud'].mean().to_frame().plot.bar(ax=ax[2])\n",
        "train_transaction.groupby('_year_Month')['isFraud'].mean().to_frame().plot.bar(ax=ax[3])"
      ],
      "execution_count": 73,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f758086d1d0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 73
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA7IAAAO3CAYAAAAeTXhHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 1152x1152 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LnJMHkXY000Q",
        "colab_type": "code",
        "outputId": "46601cfe-5ab4-4754-f3ac-c5e42a1974da",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Fraud transaction rate by weekday hour\n",
        "train_transaction['_weekday_hour'] = train_transaction['_weekday'].astype(str) + '_' + train_transaction['_hour'].astype(str)\n",
        "train_transaction.groupby('_weekday_hour')['isFraud'].mean().to_frame().plot.line(figsize=(16,3))"
      ],
      "execution_count": 74,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f757ff54828>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 74
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6sAAADVCAYAAACxImapAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nVVEURVEURVEURTE6/w++UvvOuGJOIwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1152x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PF7YZffp5AkS",
        "colab_type": "code",
        "outputId": "61aa84cd-d363-4553-88db-baab0857b388",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        }
      },
      "source": [
        "df = train_transaction.groupby('_weekday')['isFraud'].mean().to_frame()\n",
        "df.sort_values(by='isFraud', ascending=False)"
      ],
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>isFraud</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_weekday</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.037</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>0.037</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>0.036</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.036</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.035</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.033</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.031</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          isFraud\n",
              "_weekday         \n",
              "4           0.037\n",
              "6           0.037\n",
              "5           0.036\n",
              "0           0.036\n",
              "3           0.035\n",
              "2           0.033\n",
              "1           0.031"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 75
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nIFFlYiE5hne",
        "colab_type": "code",
        "outputId": "2c2f3aa9-872c-4900-b6df-e0112aa68829",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 390
        }
      },
      "source": [
        "df = train_transaction.groupby('_hour')['isFraud'].mean().to_frame()\n",
        "df.sort_values(by='isFraud', ascending=False).head(10)"
      ],
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>isFraud</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_hour</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>0.106</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>0.093</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>0.090</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>0.078</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>0.070</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>0.053</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.052</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>0.039</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.038</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.037</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       isFraud\n",
              "_hour         \n",
              "7        0.106\n",
              "8        0.093\n",
              "9        0.090\n",
              "6        0.078\n",
              "5        0.070\n",
              "10       0.053\n",
              "4        0.052\n",
              "11       0.039\n",
              "3        0.038\n",
              "2        0.037"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 76
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9tV9o5lx6Wbj",
        "colab_type": "code",
        "outputId": "378c7450-1689-4d64-b81a-396a64009f71",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 390
        }
      },
      "source": [
        "df = train_transaction.groupby('_weekday_hour')['isFraud'].mean().to_frame()\n",
        "df.sort_values(by='isFraud', ascending=False).head(10)"
      ],
      "execution_count": 77,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>isFraud</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_weekday_hour</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>6_7</th>\n",
              "      <td>0.143</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0_7</th>\n",
              "      <td>0.142</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4_8</th>\n",
              "      <td>0.136</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5_8</th>\n",
              "      <td>0.135</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5_9</th>\n",
              "      <td>0.117</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0_8</th>\n",
              "      <td>0.110</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6_10</th>\n",
              "      <td>0.109</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5_7</th>\n",
              "      <td>0.106</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1_7</th>\n",
              "      <td>0.103</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0_11</th>\n",
              "      <td>0.102</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "               isFraud\n",
              "_weekday_hour         \n",
              "6_7              0.143\n",
              "0_7              0.142\n",
              "4_8              0.136\n",
              "5_8              0.135\n",
              "5_9              0.117\n",
              "0_8              0.110\n",
              "6_10             0.109\n",
              "5_7              0.106\n",
              "1_7              0.103\n",
              "0_11             0.102"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 77
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AqC4XvHl9V4-",
        "colab_type": "code",
        "outputId": "3ba9de64-bc3f-469d-a3cd-4038e91cb3c2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 390
        }
      },
      "source": [
        "train_transaction['_amount_qcut'] = pd.qcut(train_transaction['TransactionAmt'], 10)\n",
        "df = train_transaction.groupby('_amount_qcut')['isFraud'].mean().to_frame()\n",
        "df.sort_values(by='isFraud', ascending=False)"
      ],
      "execution_count": 78,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>isFraud</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>_amount_qcut</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>(0.25, 25.95]</th>\n",
              "      <td>0.056</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(275.293, 31937.391]</th>\n",
              "      <td>0.051</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(117.0, 159.95]</th>\n",
              "      <td>0.043</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(159.95, 275.293]</th>\n",
              "      <td>0.038</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(68.769, 100.0]</th>\n",
              "      <td>0.036</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(35.95, 49.0]</th>\n",
              "      <td>0.032</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(25.95, 35.95]</th>\n",
              "      <td>0.032</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(57.95, 68.769]</th>\n",
              "      <td>0.029</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(100.0, 117.0]</th>\n",
              "      <td>0.020</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(49.0, 57.95]</th>\n",
              "      <td>0.019</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                      isFraud\n",
              "_amount_qcut                 \n",
              "(0.25, 25.95]           0.056\n",
              "(275.293, 31937.391]    0.051\n",
              "(117.0, 159.95]         0.043\n",
              "(159.95, 275.293]       0.038\n",
              "(68.769, 100.0]         0.036\n",
              "(35.95, 49.0]           0.032\n",
              "(25.95, 35.95]          0.032\n",
              "(57.95, 68.769]         0.029\n",
              "(100.0, 117.0]          0.020\n",
              "(49.0, 57.95]           0.019"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 78
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sBkpcTo-sTFP",
        "colab_type": "code",
        "outputId": "884bf125-011a-429a-8102-69c48f52d058",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "fraud_id = train_transaction[train_transaction['isFraud'] == 1]['TransactionID']\n",
        "fraud_id_in_identity = [i for i in fraud_id if i in train_identity['TransactionID'].values]\n",
        "print(f'Fraud id in transaction:{len(fraud_id)} and in identity:{len(fraud_id_in_identity)}')"
      ],
      "execution_count": 79,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Fraud id in transaction:20663 and in identity:11318\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Iuq1Z4ewEDLP",
        "colab_type": "code",
        "outputId": "24b6822d-1047-46c6-85b3-f9f47014dee0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "train_id_trns = pd.merge(train_identity, train_transaction[['isFraud', 'TransactionID']])\n",
        "train_id_trns_f0 = train_id_trns[train_id_trns['isFraud'] == 0]\n",
        "train_id_trns_f1 = train_id_trns[train_id_trns['isFraud'] == 1]\n",
        "del train_id_trns\n",
        "\n",
        "print(train_id_trns_f0.shape)\n",
        "print(train_id_trns_f1.shape)\n",
        "\n",
        "def plotHistByFraud(col, bins=20, figsize=(8,3)):\n",
        "  with np.errstate(invalid='ignore'):\n",
        "    plt.figure(figsize=figsize)\n",
        "    plt.hist([train_id_trns_f0[col], train_id_trns_f1[col]], bins=bins, density=True, color=['royalblue', 'orange'])\n",
        "    \n",
        "def plotCategoryRateBar(col, topN=np.nan, figsize=(8,3)):\n",
        "  a, b = train_id_trans_f0, train_id_trns_f1\n",
        "  if topN==topN:\n",
        "    vals = b[col].value_counts(normalize=True).to_frame().iloc[:topN,0]\n",
        "    subA = a.loc[a[col].isin(vals.index.values), col]\n",
        "    df = pd.DataFrame({'normal': subA.value_counts(normalize=True), 'fraud':vals})\n",
        "  else:\n",
        "    df = pd.DataFrame({'normal':a[col].value_counts(normalize=True), 'fraud':b[col].value_counts(normalize=True)})\n",
        "  df.sort_values('fraud', ascending=False).plot.bar(figsize=figsize)"
      ],
      "execution_count": 80,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(132915, 42)\n",
            "(11318, 42)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mjjYc9y6x_5S",
        "colab_type": "code",
        "outputId": "a4a1411f-71f4-4c33-b7a5-cc0fff32ca1b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 214
        }
      },
      "source": [
        "plotHistByFraud('id_01')"
      ],
      "execution_count": 81,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAADFCAYAAAB5PKoUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYWhLktQRhrYkSR3xf+kxtIMQ+ZHKAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 576x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lbxqoVmV8EvW",
        "colab_type": "code",
        "outputId": "c7b1339e-e2a5-4337-c9f1-3f5bdea23ec2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 214
        }
      },
      "source": [
        "plotHistByFraud('id_02')"
      ],
      "execution_count": 82,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgYAAADFCAYAAAAv3DqTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngSRJ6pgYSJKkjomBJEnqmBhIkqTO/wexejIs+aXF/AAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 576x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wVy7Hny_8QM1",
        "colab_type": "code",
        "outputId": "447d4906-346b-4b64-cda1-9603c8a491f8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 214
        }
      },
      "source": [
        "plotHistByFraud('id_07')"
      ],
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAADFCAYAAAB5PKoUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ndG5QrhMAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 576x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3-BgOHpQAEBm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_transaction = pd.read_csv(path_train_trans)\n",
        "train_identity = pd.read_csv(path_train_id)\n",
        "test_transaction = pd.read_csv(path_test_trans)\n",
        "test_identity = pd.read_csv(path_test_id)\n",
        "\n",
        "id_cols = list(train_identity.columns.values)\n",
        "trn_cols = list(train_transaction.drop('isFraud', axis=1).columns.values)\n",
        "\n",
        "\n",
        "X_train = pd.merge(train_transaction[trn_cols + ['isFraud']], train_identity[id_cols], how='left')\n",
        "X_test = pd.merge(test_transaction[trn_cols], test_identity[id_cols], how='left')\n",
        "\n",
        "X_train_id = X_train.pop('TransactionID')\n",
        "X_test_id = X_test.pop('TransactionID')\n",
        "\n",
        "\n",
        "del train_transaction, train_identity, test_transaction, test_identity\n",
        "\n",
        "all_data = X_train.append(X_test, sort=False).reset_index(drop=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XsFBz-0EBHsh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "all_data['_nulls'] = all_data.isnull().sum(axis=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KlsIdXoOv9tY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "vcols = [f'V{i}' for i in range(1,340)]\n",
        "sc = preprocessing.MinMaxScaler()\n",
        "pca = PCA(n_components=2)\n",
        "vcol_pca = pca.fit_transform(sc.fit_transform(all_data[vcols].fillna(-1)))\n",
        "\n",
        "all_data['_vcol_pca0'] = vcol_pca[:,0]\n",
        "all_data['_vcol_pca1'] = vcol_pca[:,1]\n",
        "all_data['_vcol_nulls'] = all_data[vcols].isnull().sum(axis=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JR_Tlg1vFUGG",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "all_data.drop(vcols, axis=1, inplace=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4nKU928zGlLc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "all_data['card4'].fillna('unknown', inplace=True)\n",
        "all_data['card6'].fillna('unknown', inplace=True)\n",
        "\n",
        "all_data['P_emaildomain'].fillna('unknown', inplace=True)\n",
        "all_data['R_emaildomain'].fillna('unknown', inplace=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vS5Cl0uVHCk2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "START_DATE = '2017-12-01'\n",
        "start_date = datetime.datetime.strptime(START_DATE, '%Y-%m-%d')\n",
        "all_data['Date'] = all_data['TransactionDT'].apply(lambda x: start_date + datetime.timedelta(seconds=x))\n",
        "all_data['_weekday'] = all_data['Date'].dt.weekday\n",
        "all_data['_hour'] = all_data['Date'].dt.hour\n",
        "\n",
        "all_data['_weekday'] = all_data['_weekday'].astype(str)\n",
        "all_data['_hour'] = all_data['_hour'].astype(str)\n",
        "all_data['_weekday__hour'] = all_data['_weekday'] + all_data['_hour']\n",
        "\n",
        "all_data.drop(['TransactionDT','Date'], axis=1, inplace=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QwpxAWkMJ8xD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "all_data['id_31_ua'] = all_data['id_31'].apply(lambda x : x.split()[0] if x == x else 'unknown')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MxVV5UnmTSdv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "all_data['_P_emaildomain__addr1'] = all_data['P_emaildomain'] + '__' + all_data['addr1'].astype(str)\n",
        "all_data['_card1__card2'] = all_data['card1'].astype(str) + '__' + all_data['card2'].astype(str)\n",
        "all_data['_card1__addr1'] = all_data['card1'].astype(str) + '__' + all_data['addr1'].astype(str)\n",
        "all_data['_card2__addr1'] = all_data['card2'].astype(str) + '__' + all_data['addr1'].astype(str)\n",
        "all_data['_card12__addr1'] = all_data['_card1__card2'] + '__' + all_data['addr1'].astype(str)\n",
        "all_data['_card_all__addr1'] = all_data['_card12__addr1'] + '__' + all_data['addr1'].astype(str)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kYaRIonqVkXW",
        "colab_type": "code",
        "outputId": "523f85e7-1abc-4d55-e2bd-7d2f975987fe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 359
        }
      },
      "source": [
        "all_data['_amount_decimal'] = ((all_data['TransactionAmt'] - all_data['TransactionAmt'].astype(int)) * 1000).astype(int)\n",
        "all_data['_amount_decimal_len'] = all_data['TransactionAmt'].apply(lambda x : len(re.sub('0+$','',str(x)).split('.')[1]))\n",
        "all_data['_amount_fraction'] = all_data['TransactionAmt'].apply(lambda x : float('0.' + re.sub('^[0-9]|\\.|0+$','',str(x))))\n",
        "all_data[['TransactionAmt','_amount_decimal','_amount_decimal_len','_amount_fraction']].head(10)"
      ],
      "execution_count": 92,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>TransactionAmt</th>\n",
              "      <th>_amount_decimal</th>\n",
              "      <th>_amount_decimal_len</th>\n",
              "      <th>_amount_fraction</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>68.500</td>\n",
              "      <td>500</td>\n",
              "      <td>1</td>\n",
              "      <td>0.850</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>29.000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.900</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>59.000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.900</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>50.000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>50.000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>49.000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.900</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>159.000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.590</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>422.500</td>\n",
              "      <td>500</td>\n",
              "      <td>1</td>\n",
              "      <td>0.225</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>15.000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.500</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>117.000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.170</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   TransactionAmt  _amount_decimal  _amount_decimal_len  _amount_fraction\n",
              "0          68.500              500                    1             0.850\n",
              "1          29.000                0                    0             0.900\n",
              "2          59.000                0                    0             0.900\n",
              "3          50.000                0                    0             0.000\n",
              "4          50.000                0                    0             0.000\n",
              "5          49.000                0                    0             0.900\n",
              "6         159.000                0                    0             0.590\n",
              "7         422.500              500                    1             0.225\n",
              "8          15.000                0                    0             0.500\n",
              "9         117.000                0                    0             0.170"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 92
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "58fKLYIQcmMc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "cols = ['ProductCD','card1','card2','card5','card6','P_emaildomain','_card_all__addr1']\n",
        "\n",
        "for f in cols:\n",
        "  all_data[f'_amount_mean_{f}'] = all_data['TransactionAmt'] / all_data.groupby([f])['TransactionAmt'].transform('mean')\n",
        "  all_data[f'_amount_std_{f}'] = all_data['TransactionAmt'] / all_data.groupby([f])['TransactionAmt'].transform('std')\n",
        "  all_data[f'_amount_pct_{f}'] = (all_data['TransactionAmt'] - all_data[f'_amount_mean_{f}']) / all_data[f'_amount_std_{f}']\n",
        "  \n",
        "# Frequency encoding\n",
        "for f in cols:\n",
        "  vc = all_data[f].value_counts(dropna=False)\n",
        "  all_data[f'_count_{f}'] = all_data[f].map(vc)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "282MR0l7kV0l",
        "colab_type": "code",
        "outputId": "78eb97eb-7261-4448-b31c-54c9c3a9c632",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print(\"Features:\", all_data.shape)"
      ],
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Features: (1097231, 138)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bvfHnTpZyOry",
        "colab_type": "code",
        "outputId": "6838e766-d61f-4184-9694-187115fde51d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        }
      },
      "source": [
        "cat_cols = [f'id_{f}' for i in range(12,39)]\n",
        "for i in cat_cols:\n",
        "  if i in all_data.columns:\n",
        "    all_data[i] = all_data[i].astype(str)\n",
        "    all_data[i].fillna('unknown', inplace=True)\n",
        "    \n",
        "enc_cols = []\n",
        "\n",
        "for i,t in all_data.loc[:,all_data.columns != 'isFraud'].dtypes.iteritems():\n",
        "  if t == object:\n",
        "    enc_cols.append(i)\n",
        "    all_data[i] = pd.factorize(all_data[i])[0]\n",
        "    \n",
        "print(enc_cols)"
      ],
      "execution_count": 95,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['ProductCD', 'card4', 'card6', 'P_emaildomain', 'R_emaildomain', 'M1', 'M2', 'M3', 'M4', 'M5', 'M6', 'M7', 'M8', 'M9', 'id_12', 'id_15', 'id_16', 'id_23', 'id_27', 'id_28', 'id_29', 'id_30', 'id_31', 'id_33', 'id_34', 'id_35', 'id_36', 'id_37', 'id_38', 'DeviceType', 'DeviceInfo', '_weekday', '_hour', '_weekday__hour', 'id_31_ua', '_P_emaildomain__addr1', '_card1__card2', '_card1__addr1', '_card2__addr1', '_card12__addr1', '_card_all__addr1']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wMHFYGe8lAem",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_train = all_data[all_data['isFraud'].notnull()]\n",
        "X_test = all_data[all_data['isFraud'].isnull()].drop('isFraud', axis=1)\n",
        "Y_train = X_train.pop('isFraud')\n",
        "del all_data"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tNrlyxk2nxQK",
        "colab_type": "text"
      },
      "source": [
        "Prediction"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HHaRGvVpnycy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import lightgbm as lgb"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5RWhdu_WopVR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "params = {'learning_rate':0.01,\n",
        "         'objective':'binary',\n",
        "         'metric':'auc',\n",
        "         'num_threads':-1,\n",
        "         'num_leaves':256,\n",
        "         'verbose':1,\n",
        "         'random_state':42,\n",
        "         'bagging_fraction':1,\n",
        "         'feature_fraction':0.85}\n",
        "\n",
        "oof_preds = np.zeros(X_train.shape[0])\n",
        "sub_preds = np.zeros(X_test.shape[0])\n",
        "\n",
        "clf = lgb.LGBMClassifier(**params, n_estimators=3000)\n",
        "clf.fit(X_train, Y_train)\n",
        "\n",
        "oof_preds = clf.predict_proba(X_train, num_iteration=clf.best_iteration_)[:,1]\n",
        "sub_preds = clf.predict_proba(X_test, num_iteration=clf.best_iteration_)[:,1]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dArGmCe6d-62",
        "colab_type": "code",
        "outputId": "9aac1436-77b9-43d7-f919-11cf8e995b97",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "fpr, tpr, thresholds = metrics.roc_curve(Y_train, oof_preds)\n",
        "auc = metrics.auc(fpr, tpr)\n",
        "\n",
        "plt.plot(fpr, tpr, label='ROC curve (ares = %.3f)' %auc)\n",
        "plt.legend()\n",
        "plt.title('ROC curve')\n",
        "plt.xlabel('False Positive Range')\n",
        "plt.xlabel('True Positive Range')\n",
        "plt.grid(True)"
      ],
      "execution_count": 99,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEWCAYAAACdaNcBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRCSGFO4iIjH0/wHM6j3XcvSUFwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EHpXeY6BgijI",
        "colab_type": "code",
        "outputId": "8f39a551-c61f-4149-d346-4cbdbb3e8497",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 730
        }
      },
      "source": [
        "feature_importance = clf.feature_importances_\n",
        "feature_importance = 100 * (feature_importance/feature_importance.max())\n",
        "sorted_idx = np.argsort(feature_importance)\n",
        "sorted_idx = sorted_idx[len(feature_importance) - 50:]\n",
        "pos = np.arange(sorted_idx.shape[0]) + 0.5\n",
        "\n",
        "plt.figure(figsize = (10,12))\n",
        "plt.barh(pos, feature_importance[sorted_idx], align='center')\n",
        "plt.yticks(pos, X_train.columns[sorted_idx])\n",
        "plt.xlabel('Relative Importance')\n",
        "plt.title('Variable Importance')\n",
        "plt.show()"
      ],
      "execution_count": 100,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAusAAALJCAYAAAANnW0AAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 720x864 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bZtI1YuSgtlv",
        "colab_type": "code",
        "outputId": "9bad84de-4a94-43c1-c431-90b08f57bdd4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        }
      },
      "source": [
        "X_train.columns[np.argsort(-feature_importance)].values"
      ],
      "execution_count": 101,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['_amount_pct__card_all__addr1', 'card1', '_P_emaildomain__addr1',\n",
              "       '_card2__addr1', '_weekday__hour', 'addr1',\n",
              "       '_count__card_all__addr1', '_card1__card2', '_amount_pct_card1',\n",
              "       '_count_card1', 'card2', '_amount_fraction', '_card1__addr1',\n",
              "       'D15', '_amount_std__card_all__addr1', '_count_card2',\n",
              "       '_amount_pct_card2', '_vcol_pca0', '_hour', 'dist1', '_vcol_pca1',\n",
              "       '_amount_decimal', 'C13', '_amount_std_card2', '_amount_std_card1',\n",
              "       '_card12__addr1', 'D4', 'id_20', 'D2',\n",
              "       '_amount_mean__card_all__addr1', '_amount_mean_card2', 'D10',\n",
              "       '_amount_mean_card5', 'C1', '_amount_std_card5', 'id_19', 'id_02',\n",
              "       'D1', '_amount_mean_P_emaildomain', '_amount_mean_card1', '_nulls',\n",
              "       'D8', 'id_31', 'C2', '_amount_std_P_emaildomain', 'D11',\n",
              "       '_amount_std_ProductCD', 'DeviceInfo', 'P_emaildomain', 'dist2',\n",
              "       '_amount_mean_ProductCD', '_amount_pct_P_emaildomain', 'D3',\n",
              "       '_amount_pct_card5', 'C6', '_count_P_emaildomain', 'id_13',\n",
              "       'card5', 'C11', 'C14', 'id_06', '_amount_mean_card6', 'D5',\n",
              "       'TransactionAmt', 'id_05', 'C9', 'D14', 'R_emaildomain',\n",
              "       '_amount_std_card6', '_count_card5', 'id_01', 'M4', 'D9', 'M6',\n",
              "       'M5', '_weekday', 'id_33', 'C5', 'id_31_ua', 'C8', 'C10', 'id_30',\n",
              "       'D13', 'id_18', 'D6', 'card4', 'D12', 'C12', 'id_14',\n",
              "       '_amount_decimal_len', '_card_all__addr1', 'card3', 'C4',\n",
              "       '_vcol_nulls', 'M7', 'M3', 'D7', 'M8', 'M9', 'ProductCD', 'id_38',\n",
              "       'card6', 'id_03', 'id_17', 'id_09', 'DeviceType', 'M2', 'id_11',\n",
              "       'id_15', 'id_37', '_amount_pct_card6', 'C7', 'id_32', 'id_07',\n",
              "       'id_12', '_count_ProductCD', 'id_04', 'id_08', 'C3', 'id_21',\n",
              "       '_amount_pct_ProductCD', 'id_36', 'id_26', 'id_34', 'addr2',\n",
              "       'id_25', 'id_16', 'id_10', '_count_card6', 'id_24', 'M1', 'id_28',\n",
              "       'id_35', 'id_22', 'id_23', 'id_29', 'id_27'], dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 101
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5zf2H21tlJ_w",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "submission = pd.DataFrame()\n",
        "submission['TransactionID'] = X_test_id\n",
        "submission['isFraud'] = sub_preds\n",
        "submission.to_csv('submission.csv', index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1AqbFSL6lOLf",
        "colab_type": "code",
        "outputId": "5f0501ea-66a6-44d2-be74-6de8d4c5d555",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "np.mean(sub_preds)"
      ],
      "execution_count": 103,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.020789868164036312"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 103
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sc2uoV1IlqTU",
        "colab_type": "code",
        "outputId": "7bb17322-732f-451e-e876-e0950fc2883f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "len(sub_preds[sub_preds > 0.99]), len(sub_preds)"
      ],
      "execution_count": 104,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(2205, 506691)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 104
        }
      ]
    }
  ]
}