Coupon_redeem_project.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "MgbiwmZVTyt6"
   },
   "outputs": [],
   "source": [
    "#Import all packages\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from datetime import datetime\n",
    "#train test split\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "\n",
    "#function to get a mode of a dataset\n",
    "from statistics import mode\n",
    "#normalizing dataset\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "#Visualization\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "sns.set_style('darkgrid')\n",
    "\n",
    "#For resampling process to change the ratio of imbalanced ratio of the classes \n",
    "#testing various ratios\n",
    "from sklearn.metrics import roc_auc_score, roc_curve, auc\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "#resampling\n",
    "from imblearn.over_sampling import SMOTE\n",
    "from sklearn.model_selection import GridSearchCV, RandomizedSearchCV\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn import svm\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from xgboost import XGBClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn import svm\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from xgboost import XGBClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 502
    },
    "id": "iI0R-v4nUfFH",
    "outputId": "641953cd-5aff-445f-a0fe-93a50f00d564"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/IPython/core/interactiveshell.py:2882: DtypeWarning: Columns (26,27) have mixed types.Specify dtype option on import or set low_memory=False.\n",
      "  exec(code_obj, self.user_global_ns, self.user_ns)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-cca442b1-6c6e-48bb-a5ae-b5f9eee81d9b\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>redemption_id</th>\n",
       "      <th>c_id</th>\n",
       "      <th>pcode_id_x</th>\n",
       "      <th>cust_id</th>\n",
       "      <th>redeemed</th>\n",
       "      <th>date</th>\n",
       "      <th>prod_id</th>\n",
       "      <th>qty</th>\n",
       "      <th>price</th>\n",
       "      <th>dscnt_store</th>\n",
       "      <th>...</th>\n",
       "      <th>valid_date</th>\n",
       "      <th>manuf</th>\n",
       "      <th>brand_type</th>\n",
       "      <th>category</th>\n",
       "      <th>age_cohort</th>\n",
       "      <th>marital_status</th>\n",
       "      <th>home_is_rental</th>\n",
       "      <th>n_family</th>\n",
       "      <th>no_of_children</th>\n",
       "      <th>income_range</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>1017</td>\n",
       "      <td>1489</td>\n",
       "      <td>0</td>\n",
       "      <td>02/22/13</td>\n",
       "      <td>34047</td>\n",
       "      <td>1</td>\n",
       "      <td>91.19</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>02/22/13</td>\n",
       "      <td>56</td>\n",
       "      <td>Local</td>\n",
       "      <td>Grocery</td>\n",
       "      <td>46-55</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>1017</td>\n",
       "      <td>1489</td>\n",
       "      <td>0</td>\n",
       "      <td>02/22/13</td>\n",
       "      <td>35464</td>\n",
       "      <td>1</td>\n",
       "      <td>21.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>02/22/13</td>\n",
       "      <td>56</td>\n",
       "      <td>Local</td>\n",
       "      <td>Grocery</td>\n",
       "      <td>46-55</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>1017</td>\n",
       "      <td>1489</td>\n",
       "      <td>0</td>\n",
       "      <td>03/03/13</td>\n",
       "      <td>34047</td>\n",
       "      <td>1</td>\n",
       "      <td>91.19</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>03/03/13</td>\n",
       "      <td>56</td>\n",
       "      <td>Local</td>\n",
       "      <td>Grocery</td>\n",
       "      <td>46-55</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>1017</td>\n",
       "      <td>1489</td>\n",
       "      <td>0</td>\n",
       "      <td>03/03/13</td>\n",
       "      <td>33688</td>\n",
       "      <td>1</td>\n",
       "      <td>124.31</td>\n",
       "      <td>-10.69</td>\n",
       "      <td>...</td>\n",
       "      <td>03/03/13</td>\n",
       "      <td>939</td>\n",
       "      <td>Established</td>\n",
       "      <td>Grocery</td>\n",
       "      <td>46-55</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>1017</td>\n",
       "      <td>1489</td>\n",
       "      <td>0</td>\n",
       "      <td>03/05/13</td>\n",
       "      <td>34047</td>\n",
       "      <td>1</td>\n",
       "      <td>91.19</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>03/05/13</td>\n",
       "      <td>56</td>\n",
       "      <td>Local</td>\n",
       "      <td>Grocery</td>\n",
       "      <td>46-55</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1102837</th>\n",
       "      <td>125115</td>\n",
       "      <td>30</td>\n",
       "      <td>863</td>\n",
       "      <td>436</td>\n",
       "      <td>0</td>\n",
       "      <td>02/20/13</td>\n",
       "      <td>34336</td>\n",
       "      <td>1</td>\n",
       "      <td>213.36</td>\n",
       "      <td>-35.62</td>\n",
       "      <td>...</td>\n",
       "      <td>02/20/13</td>\n",
       "      <td>686</td>\n",
       "      <td>Established</td>\n",
       "      <td>Grocery</td>\n",
       "      <td>70+</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1102838</th>\n",
       "      <td>125115</td>\n",
       "      <td>30</td>\n",
       "      <td>863</td>\n",
       "      <td>436</td>\n",
       "      <td>0</td>\n",
       "      <td>03/11/13</td>\n",
       "      <td>69925</td>\n",
       "      <td>1</td>\n",
       "      <td>66.97</td>\n",
       "      <td>-15.67</td>\n",
       "      <td>...</td>\n",
       "      <td>03/11/13</td>\n",
       "      <td>524</td>\n",
       "      <td>Established</td>\n",
       "      <td>Grocery</td>\n",
       "      <td>70+</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1102839</th>\n",
       "      <td>125115</td>\n",
       "      <td>30</td>\n",
       "      <td>863</td>\n",
       "      <td>436</td>\n",
       "      <td>0</td>\n",
       "      <td>03/18/13</td>\n",
       "      <td>31922</td>\n",
       "      <td>1</td>\n",
       "      <td>66.97</td>\n",
       "      <td>-15.67</td>\n",
       "      <td>...</td>\n",
       "      <td>03/18/13</td>\n",
       "      <td>524</td>\n",
       "      <td>Established</td>\n",
       "      <td>Grocery</td>\n",
       "      <td>70+</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1102840</th>\n",
       "      <td>125115</td>\n",
       "      <td>30</td>\n",
       "      <td>863</td>\n",
       "      <td>436</td>\n",
       "      <td>0</td>\n",
       "      <td>03/26/13</td>\n",
       "      <td>26122</td>\n",
       "      <td>1</td>\n",
       "      <td>89.05</td>\n",
       "      <td>-17.45</td>\n",
       "      <td>...</td>\n",
       "      <td>03/26/13</td>\n",
       "      <td>56</td>\n",
       "      <td>Local</td>\n",
       "      <td>Packaged Meat</td>\n",
       "      <td>70+</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1102841</th>\n",
       "      <td>125115</td>\n",
       "      <td>30</td>\n",
       "      <td>863</td>\n",
       "      <td>436</td>\n",
       "      <td>0</td>\n",
       "      <td>03/29/13</td>\n",
       "      <td>7479</td>\n",
       "      <td>2</td>\n",
       "      <td>184.51</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>03/29/13</td>\n",
       "      <td>803</td>\n",
       "      <td>Established</td>\n",
       "      <td>Grocery</td>\n",
       "      <td>70+</td>\n",
       "      <td>Married</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1102842 rows × 29 columns</p>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-cca442b1-6c6e-48bb-a5ae-b5f9eee81d9b')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-cca442b1-6c6e-48bb-a5ae-b5f9eee81d9b button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-cca442b1-6c6e-48bb-a5ae-b5f9eee81d9b');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "         redemption_id  c_id  pcode_id_x  cust_id  redeemed      date  \\\n",
       "0                    9     8        1017     1489         0  02/22/13   \n",
       "1                    9     8        1017     1489         0  02/22/13   \n",
       "2                    9     8        1017     1489         0  03/03/13   \n",
       "3                    9     8        1017     1489         0  03/03/13   \n",
       "4                    9     8        1017     1489         0  03/05/13   \n",
       "...                ...   ...         ...      ...       ...       ...   \n",
       "1102837         125115    30         863      436         0  02/20/13   \n",
       "1102838         125115    30         863      436         0  03/11/13   \n",
       "1102839         125115    30         863      436         0  03/18/13   \n",
       "1102840         125115    30         863      436         0  03/26/13   \n",
       "1102841         125115    30         863      436         0  03/29/13   \n",
       "\n",
       "         prod_id  qty   price  dscnt_store  ...  valid_date  manuf  \\\n",
       "0          34047    1   91.19         0.00  ...    02/22/13     56   \n",
       "1          35464    1   21.02         0.00  ...    02/22/13     56   \n",
       "2          34047    1   91.19         0.00  ...    03/03/13     56   \n",
       "3          33688    1  124.31       -10.69  ...    03/03/13    939   \n",
       "4          34047    1   91.19         0.00  ...    03/05/13     56   \n",
       "...          ...  ...     ...          ...  ...         ...    ...   \n",
       "1102837    34336    1  213.36       -35.62  ...    02/20/13    686   \n",
       "1102838    69925    1   66.97       -15.67  ...    03/11/13    524   \n",
       "1102839    31922    1   66.97       -15.67  ...    03/18/13    524   \n",
       "1102840    26122    1   89.05       -17.45  ...    03/26/13     56   \n",
       "1102841     7479    2  184.51         0.00  ...    03/29/13    803   \n",
       "\n",
       "          brand_type       category age_cohort marital_status  home_is_rental  \\\n",
       "0              Local        Grocery      46-55        Married             0.0   \n",
       "1              Local        Grocery      46-55        Married             0.0   \n",
       "2              Local        Grocery      46-55        Married             0.0   \n",
       "3        Established        Grocery      46-55        Married             0.0   \n",
       "4              Local        Grocery      46-55        Married             0.0   \n",
       "...              ...            ...        ...            ...             ...   \n",
       "1102837  Established        Grocery        70+        Married             0.0   \n",
       "1102838  Established        Grocery        70+        Married             0.0   \n",
       "1102839  Established        Grocery        70+        Married             0.0   \n",
       "1102840        Local  Packaged Meat        70+        Married             0.0   \n",
       "1102841  Established        Grocery        70+        Married             0.0   \n",
       "\n",
       "         n_family no_of_children income_range  \n",
       "0             2.0            NaN          3.0  \n",
       "1             2.0            NaN          3.0  \n",
       "2             2.0            NaN          3.0  \n",
       "3             2.0            NaN          3.0  \n",
       "4             2.0            NaN          3.0  \n",
       "...           ...            ...          ...  \n",
       "1102837         2            NaN          8.0  \n",
       "1102838         2            NaN          8.0  \n",
       "1102839         2            NaN          8.0  \n",
       "1102840         2            NaN          8.0  \n",
       "1102841         2            NaN          8.0  \n",
       "\n",
       "[1102842 rows x 29 columns]"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "training = pd.read_csv('/content/drive/MyDrive/data/proj_training_data.csv')\n",
    "training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "F0Jl2QEzqZdE"
   },
   "outputs": [],
   "source": [
    "##Fill NAs\n",
    "def impute_missing_value(df):\n",
    "  df.fillna({'age_cohort':'unknown','marital_status':'unknown'}, inplace=True)\n",
    "  df.fillna({'home_is_rental':'0','n_family':'1','no_of_children':'0','income_range':df.income_range.mean()}, inplace=True)\n",
    "  return df "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "mX8Rt7-I7aaH"
   },
   "outputs": [],
   "source": [
    "##Function to convert date into datetime format\n",
    "def Change_to_datetime(date_col):\n",
    "    new_datetype = []\n",
    "    for i, date in enumerate(date_col):\n",
    "        date_format = \"%m/%d/%Y\" \n",
    "    \n",
    "        date= date[:-2]+'20'+date[-2:]#change the string format to the format convirtable to datetype object\n",
    "        date= datetime.strptime(date, date_format)\n",
    "        new_datetype.append(date)\n",
    "        \n",
    "    new_datetype = pd.Series(new_datetype)\n",
    "    return new_datetype"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "3iw1y0TR7FJ4"
   },
   "outputs": [],
   "source": [
    "## Function to do some date transformation\n",
    "def dataframe_transform(training):\n",
    "  end_date =Change_to_datetime(training.end_date)\n",
    "  training['end_date'] = end_date\n",
    "  start_date =Change_to_datetime(training.start_date)\n",
    "  training['start_date'] = start_date\n",
    "  training['date'] = pd.to_datetime(training['date'])\n",
    "  func = lambda x: x.days\n",
    "  training['duration_days']=training.end_date - training.start_date\n",
    "  training['duration_days'] = training['duration_days'].apply(func)\n",
    "  training['coupon_used']= training.dscnt_pcode.apply(lambda x:1 if x !=0 else 0)\n",
    "  return training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ANeotPnyWD4Q"
   },
   "outputs": [],
   "source": [
    "## Clean dataframe, removing irelevant rows by logic \n",
    "def clean_dataframe(training):\n",
    "  training.sort_values(by='date') #sort df by date before split\n",
    "  mask = (training['date'] >= training['start_date']) & (training['date'] <= training['end_date'])\n",
    "  trans_valid_prom= training.loc[mask]\n",
    "  valid_list = training.loc[mask].index.to_list()\n",
    "  trans_expire_prom= training[~training.index.isin(valid_list)]\n",
    "  invalid_index = (trans_expire_prom[trans_expire_prom['redeemed'] == 1]).index.to_list()\n",
    "  training = training[~training.index.isin(invalid_index)]\n",
    "  problematic_data = training.query('redeemed ==1 and dscnt_store ==0 and dscnt_pcode ==0')\n",
    "  problematic_data_index= problematic_data.index.to_list()\n",
    "  training = training[~training.index.isin(problematic_data_index)]\n",
    "  return training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "irdKna4j12Wf"
   },
   "outputs": [],
   "source": [
    "##Aggregate dataframe by different indexes to reduce the total size\n",
    "def aggregate_data(training):\n",
    "  pivot_cust_redeemed = pd.pivot_table(training,index =['prod_id','cust_id','c_id','redeemed'],\n",
    "                                      values = ['manuf', 'brand_type', 'category',\n",
    "        'age_cohort', 'marital_status', 'home_is_rental', 'n_family',\n",
    "        'no_of_children', 'income_range', 'duration_days', 'coupon_used','qty','price','date','dscnt_store','dscnt_pcode','c_type']\n",
    "                                      ,aggfunc = {\n",
    "                          'qty':np.mean,\n",
    "                          'price':np.mean,\n",
    "                          'dscnt_store':np.mean,\n",
    "                          'dscnt_pcode':np.mean,\n",
    "                          'coupon_used': np.sum,\n",
    "                          'manuf':lambda x: x.value_counts().index[:1],\n",
    "                          'brand_type':lambda x: x.value_counts().index[:1], \n",
    "                          'category':lambda x: x.value_counts().index[:1],\n",
    "                          'age_cohort':lambda x: x.value_counts().index[:1],\n",
    "                          'marital_status':lambda x: x.value_counts().index[:1],\n",
    "                          'home_is_rental':lambda x: x.value_counts().index[:1],\n",
    "                          'n_family':lambda x: x.value_counts().index[:1],\n",
    "                          'no_of_children':lambda x: x.value_counts().index[:1],\n",
    "                          'income_range':lambda x: x.value_counts().index[:1],\n",
    "                          'duration_days':lambda x: x.value_counts().index[:1],\n",
    "                          'c_type':lambda x: x.value_counts().index[:1],\n",
    "                          'date' : lambda x:x.value_counts().index[:1]\n",
    "                          })\n",
    "  final_data = pivot_cust_redeemed.reset_index()\n",
    "  return final_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "PKG3Vhwd-i5j"
   },
   "outputs": [],
   "source": [
    "## Rename the column names\n",
    "def rename_cols(final_data):\n",
    "  new_cols = {'age_cohort':'mostfrequent_age_cohort',\n",
    "              'brand_type':'mostfrequent_brand_type',\n",
    "              'category':'mostfrequent_category',\n",
    "              'coupon_used':'total_coupoun_used',\n",
    "              'dscnt_pcode':'avg_dcnt_pcode',\n",
    "              'dscnt_store':'avg_dscnt_store',\n",
    "            'duration_days':'mostfrequent_duration_days',\n",
    "            'home_is_rental':'mostfrequent_home_is_rental',\n",
    "            'income_range':'mostfrequent_income_range',\n",
    "            'manuf':'mostfrequent_manuf',\n",
    "            'marital_status':'mostfrequent_marital_status',\n",
    "            'n_family':'mostfrequent_n_family',\n",
    "              'no_of_children':'mostfrequent_no_of_children',\n",
    "              'price':'avg_price',\n",
    "              'qty':'avg_qty',\n",
    "              'c_type':'mostfrequent_ctype'}\n",
    "  final_data.rename(columns=new_cols, inplace = True)\n",
    "  final_data_copy = final_data.copy()\n",
    "  final_data_copy['mostfrequent_n_family'] = final_data_copy['mostfrequent_n_family'].astype(str)\n",
    "  final_data_copy['mostfrequent_n_family'] = final_data_copy['mostfrequent_n_family'].replace({'1': '1.0', '2':'2.0', '3':'3.0', '4':'4.0'})\n",
    "  final_data_copy.date = pd.to_datetime(final_data_copy.date, yearfirst=True)\n",
    "  final_data_copy['date']=final_data_copy['date'].apply(lambda x: x.year)\n",
    "  return final_data_copy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "_1A2kAzKU_FF"
   },
   "outputs": [],
   "source": [
    "###Dummy columns\n",
    "def dummy_(final_data_copy):\n",
    "  final_data_copy= pd.get_dummies(data=final_data_copy, columns=['mostfrequent_age_cohort']\n",
    "        ,drop_first=True,dummy_na=False)\n",
    "  final_data_copy=pd.get_dummies(data=final_data_copy, columns=['mostfrequent_brand_type']\n",
    "        ,drop_first=True,dummy_na=False)\n",
    "  final_data_copy= pd.get_dummies(data=final_data_copy, columns=['mostfrequent_ctype']\n",
    "        ,drop_first=True,dummy_na=False)\n",
    "  final_data_copy=pd.get_dummies(data=final_data_copy, columns=['mostfrequent_category']\n",
    "        ,drop_first=True,dummy_na=False)\n",
    "  final_data_copy=pd.get_dummies(data=final_data_copy, columns=['mostfrequent_home_is_rental']\n",
    "        ,drop_first=True,dummy_na=False)\n",
    "  final_data_copy=pd.get_dummies(data=final_data_copy, columns=['mostfrequent_marital_status']\n",
    "        ,drop_first=True,dummy_na=False)\n",
    "  final_data_copy=pd.get_dummies(data=final_data_copy, columns=['mostfrequent_n_family']\n",
    "        ,drop_first=True,dummy_na=False)\n",
    "  final_data_copy=pd.get_dummies(data=final_data_copy, columns=['mostfrequent_no_of_children']\n",
    "        ,drop_first=True,dummy_na=False)\n",
    "  return final_data_copy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "5ENdA8H-0_xm"
   },
   "outputs": [],
   "source": [
    "###Split the data \n",
    "def split_data(final_data_copy):\n",
    "  x =final_data_copy.loc[:, final_data_copy.columns != 'redeemed']\n",
    "  x= x.drop(columns=['date'])\n",
    "  y=final_data_copy['redeemed']\n",
    "  X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.4,shuffle= False)\n",
    "  return X_train, X_test, y_train, y_test\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "X2YCvgpxhUn2"
   },
   "outputs": [],
   "source": [
    "## Split the test data into test1 and test 2, one for randomized search cv, one for final test\n",
    "def split_test(X_test,y_test):\n",
    "  X_test1, X_test2, y_test1, y_test2 = train_test_split(X_test, y_test, test_size=0.5,shuffle= False)\n",
    "  return X_test1, X_test2, y_test1, y_test2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "wPZpGNZH240Z"
   },
   "outputs": [],
   "source": [
    "## Smote to balance the dataset\n",
    "def smote_(X_train, X_test, y_train, y_test):\n",
    "# Now let's compare a few different ratios of minority class to majority class\n",
    "  ratios = [0.15, 0.25, 0.33, 0.5, 0.7, 1]\n",
    "  names = ['0.15', '0.25', '0.33','0.5','0.7','even']\n",
    "  colors = sns.color_palette('Set2')\n",
    "\n",
    "  plt.figure(figsize=(10, 8))\n",
    "\n",
    "  for n, ratio in enumerate(ratios):\n",
    "      # Fit a model\n",
    "      smote = SMOTE(sampling_strategy=ratio)\n",
    "      X_train_resampled, y_train_resampled = smote.fit_resample(X_train, y_train)\n",
    "      print(X_train_resampled)\n",
    "      logreg = LogisticRegression(fit_intercept=False, C=1e20, solver ='lbfgs')\n",
    "      model_log = logreg.fit(X_train_resampled, y_train_resampled)\n",
    "      print(model_log)\n",
    "\n",
    "      # Predict\n",
    "      y_hat_test = logreg.predict(X_test)\n",
    "\n",
    "      y_score = logreg.decision_function(X_test)\n",
    "\n",
    "      fpr, tpr, thresholds = roc_curve(y_test, y_score)\n",
    "      \n",
    "      print('AUC for {}: {}'.format(names[n], auc(fpr, tpr)))\n",
    "      print('-------------------------------------------------------------------------------------')\n",
    "      lw = 2\n",
    "      plt.plot(fpr, tpr, color=colors[n],\n",
    "              lw=lw, label='ROC curve {}, AUC:{}'.format(names[n], round(auc(fpr, tpr),3)))\n",
    "\n",
    "  plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n",
    "  plt.xlim([0.0, 1.0])\n",
    "  plt.ylim([0.0, 1.05])\n",
    "\n",
    "  plt.yticks([i/20.0 for i in range(21)])\n",
    "  plt.xticks([i/20.0 for i in range(21)])\n",
    "  plt.xlabel('False Positive Rate')\n",
    "  plt.ylabel('True Positive Rate')\n",
    "  plt.title('Receiver operating characteristic (ROC) Curve')\n",
    "  plt.legend(loc='lower right')\n",
    "  plt.show()\n",
    "  return X_train_resampled, y_train_resampled"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "65qnWi-n5Thu"
   },
   "outputs": [],
   "source": [
    "##Standardlize the values to reduce the error\n",
    "def Standardlize_(X_train_resampled,X_test1):\n",
    "  scaler = StandardScaler()\n",
    "  scaled_X_train_resampled = scaler.fit_transform(X_train_resampled)\n",
    "  scaled_X_test = scaler.transform(X_test1)\n",
    "  scaled_X_train_resampled = pd.DataFrame(scaled_X_train_resampled, columns = X_train_resampled.columns)\n",
    "  scaled_X_test = pd.DataFrame(scaled_X_test, columns = X_train_resampled.columns)\n",
    "  return scaled_X_train_resampled, scaled_X_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "H2BwO_BV-sOl"
   },
   "outputs": [],
   "source": [
    "def grid_search(scaled_X_train_resampled, y_train_resampled,scaled_X_test,y_test):\n",
    "  pipe_dt = Pipeline([('DT',DecisionTreeClassifier())])\n",
    "  pipe_lr = Pipeline([('LR', LogisticRegression(solver='lbfgs', max_iter=1000))])\n",
    "  pipe_svm = Pipeline([('SVM', svm.SVC())])\n",
    "  pipe_xgb = Pipeline([('xgboost',XGBClassifier())])\n",
    "  pipe_GaussianNB = Pipeline([('gussiannb',GaussianNB())])\n",
    "  pipe_randomforest = Pipeline([('randforesr',RandomForestClassifier())])\n",
    "\n",
    "  param_range = [1, 2, 3, 4, 5, 6]\n",
    "  param_range_fl = [1.0, 0.5, 0.1]\n",
    "  n_estimators = [50,100,150]\n",
    "  learning_rates = [.1,.2,.3]\n",
    "  lr_param_grid = [{\n",
    "                    'LR__C': param_range_fl}]\n",
    "  dt_param_grid = [{'DT__min_samples_leaf': [1,2,4,8,20]}]\n",
    "  svm_param_grid = [{'SVM__kernel': ['linear', 'rbf'], \n",
    "                      'SVM__C': param_range}]\n",
    "  xgb_boost_grid = { 'xgboost__max_depth': [3, 5, 6, 10, 15, 20],\n",
    "            'xgboost__learning_rate': [0.01, 0.1],\n",
    "            'xgboost__subsample': np.arange(0.5, 1.0, 0.1),\n",
    "            'xgboost__colsample_bytree': np.arange(0.4, 1.0, 0.2),\n",
    "            'xgboost__colsample_bylevel': np.arange(0.4, 1.0, 0.2),\n",
    "            'xgboost__n_estimators': [100,300, 500],\n",
    "            'xgboost__scale_pos_weight':[1,2,3,4,5,6,7,8]}\n",
    "          \n",
    "\n",
    "  random_forest_grid = {\n",
    "  'randforesr__max_depth': [10, 20, 30, 40],\n",
    "  'randforesr__max_features': ['auto', 'sqrt'],\n",
    "  'randforesr__min_samples_leaf': [1, 2, 4],\n",
    "  'randforesr__min_samples_split': [2, 5, 10],\n",
    "  'randforesr__n_estimators': [200, 400, 600]}\n",
    "\n",
    "  GaussianNB_param_grid ={'gussiannb__kernel': ['rbf','linear'],\n",
    "                          'gussiannb__C':[0.05,0.1,0.2],\n",
    "                            'gussiannb__gamma':[0.1,0.2,0.3]}\n",
    "  \n",
    "  \n",
    "\n",
    "  lr_grid_search = RandomizedSearchCV(estimator=pipe_lr,\n",
    "          param_distributions=lr_param_grid,\n",
    "          scoring='recall')\n",
    "  dt_grid_search = RandomizedSearchCV(estimator=pipe_dt,\n",
    "          param_distributions=dt_param_grid,\n",
    "          scoring='recall')\n",
    "\n",
    "  svm_grid_search = RandomizedSearchCV(estimator=pipe_svm,\n",
    "          param_distributions=svm_param_grid,\n",
    "          scoring='recall')\n",
    "  xgb_grid_search = RandomizedSearchCV(estimator = pipe_xgb,\n",
    "                                param_distributions = xgb_boost_grid,\n",
    "                                scoring='recall')\n",
    "  rf_grid_search = RandomizedSearchCV(estimator=pipe_randomforest,\n",
    "                                param_distributions=random_forest_grid,\n",
    "                                scoring='recall')\n",
    "\n",
    "\n",
    "\n",
    "  grids = [lr_grid_search, dt_grid_search, svm_grid_search,xgb_grid_search,rf_grid_search]\n",
    "  for pipe in grids:\n",
    "      pipe.fit(scaled_X_train_resampled, y_train_resampled)\n",
    "    \n",
    "  grid_dict = {0: 'Logistic Regression', 1: 'Decision Trees',  \n",
    "              2: 'Support Vector Machines',3:'XGBOOST',4:'randomforest'}\n",
    "  for i, model in enumerate(grids):\n",
    "      print('{} Recall is: {}'.format(grid_dict[i],\n",
    "      model.score(scaled_X_test,y_test)))\n",
    "      print('{} Best Params: {}'.format(grid_dict[i],model.best_params_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "zG-q-AgXHXYO"
   },
   "outputs": [],
   "source": [
    "training = impute_missing_value(training)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "cvGWPR_ZZUzc"
   },
   "outputs": [],
   "source": [
    "training= dataframe_transform(training)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "R0KPrHmIwlq0"
   },
   "outputs": [],
   "source": [
    "training= clean_dataframe(training)\n",
    "final_data = aggregate_data(training)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "WUlsxr9Mx7Pg"
   },
   "outputs": [],
   "source": [
    "final_data_copy = rename_cols(final_data)\n",
    "final_data_copy = dummy_(final_data_copy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "UF_g5MhQmDiD"
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test= split_data(final_data_copy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "mGCWBUS62G6G",
    "outputId": "8aba8773-4b0f-4ba0-b671-63eaa739fdd4"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['prod_id', 'cust_id', 'c_id', 'total_coupoun_used', 'avg_dcnt_pcode',\n",
       "       'avg_dscnt_store', 'mostfrequent_duration_days',\n",
       "       'mostfrequent_income_range', 'mostfrequent_manuf', 'avg_price',\n",
       "       'avg_qty', 'mostfrequent_age_cohort_26-35',\n",
       "       'mostfrequent_age_cohort_36-45', 'mostfrequent_age_cohort_46-55',\n",
       "       'mostfrequent_age_cohort_56-70', 'mostfrequent_age_cohort_70+',\n",
       "       'mostfrequent_age_cohort_unknown', 'mostfrequent_brand_type_Local',\n",
       "       'mostfrequent_ctype_Y', 'mostfrequent_category_Dairy, Juices & Snacks',\n",
       "       'mostfrequent_category_Flowers & Plants',\n",
       "       'mostfrequent_category_Garden', 'mostfrequent_category_Grocery',\n",
       "       'mostfrequent_category_Meat', 'mostfrequent_category_Miscellaneous',\n",
       "       'mostfrequent_category_Natural Products',\n",
       "       'mostfrequent_category_Packaged Meat',\n",
       "       'mostfrequent_category_Pharmaceutical',\n",
       "       'mostfrequent_category_Prepared Food', 'mostfrequent_category_Salads',\n",
       "       'mostfrequent_category_Seafood',\n",
       "       'mostfrequent_category_Skin & Hair Care',\n",
       "       'mostfrequent_category_Travel',\n",
       "       'mostfrequent_category_Vegetables (cut)',\n",
       "       'mostfrequent_home_is_rental_1.0', 'mostfrequent_home_is_rental_0',\n",
       "       'mostfrequent_marital_status_Single',\n",
       "       'mostfrequent_marital_status_unknown', 'mostfrequent_n_family_2.0',\n",
       "       'mostfrequent_n_family_3.0', 'mostfrequent_n_family_4.0',\n",
       "       'mostfrequent_n_family_5+', 'mostfrequent_no_of_children_2.0',\n",
       "       'mostfrequent_no_of_children_0', 'mostfrequent_no_of_children_1',\n",
       "       'mostfrequent_no_of_children_2', 'mostfrequent_no_of_children_3+'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "8g9rqKc1h8w-"
   },
   "outputs": [],
   "source": [
    "X_test_validation, X_test, y_test_validation, y_test = split_test(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "O2xps4CP-hr5",
    "outputId": "358dd1b4-8c96-4153-88df-3db798a50771"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       prod_id  cust_id  c_id  total_coupoun_used  avg_dcnt_pcode  \\\n",
      "0           22      637     8                   0        0.000000   \n",
      "1           22      637    30                   0        0.000000   \n",
      "2           27      637    30                   0        0.000000   \n",
      "3           31      637     8                   0        0.000000   \n",
      "4           31      637    30                   0        0.000000   \n",
      "...        ...      ...   ...                 ...             ...   \n",
      "26217    20808      445     8                   0        0.000000   \n",
      "26218     8403      747     8                   0        0.000000   \n",
      "26219    16811     1132     8                   0        0.000000   \n",
      "26220    29741     1527     8                   0      -17.897346   \n",
      "26221    15829     1350     8                   0        0.000000   \n",
      "\n",
      "       avg_dscnt_store  mostfrequent_duration_days  mostfrequent_income_range  \\\n",
      "0            -5.875000                          77                   6.000000   \n",
      "1            -5.875000                         133                   6.000000   \n",
      "2           -34.550000                         133                   6.000000   \n",
      "3             0.000000                          77                   6.000000   \n",
      "4            -4.272000                         133                   6.000000   \n",
      "...                ...                         ...                        ...   \n",
      "26217       -17.031242                          77                   3.821637   \n",
      "26218       -16.689110                          77                   4.866836   \n",
      "26219       -27.574093                          77                   4.528143   \n",
      "26220       -37.067618                          77                   4.971653   \n",
      "26221       -58.153380                          77                   4.901556   \n",
      "\n",
      "       mostfrequent_manuf   avg_price  ...  \\\n",
      "0                      56   29.745000  ...   \n",
      "1                      56   29.745000  ...   \n",
      "2                      56   66.970000  ...   \n",
      "3                      56   35.620000  ...   \n",
      "4                      56   38.472000  ...   \n",
      "...                   ...         ...  ...   \n",
      "26217                  56   38.560015  ...   \n",
      "26218                 723   85.469919  ...   \n",
      "26219                1716  121.662631  ...   \n",
      "26220                1351   83.921990  ...   \n",
      "26221                1032  182.851655  ...   \n",
      "\n",
      "       mostfrequent_marital_status_unknown  mostfrequent_n_family_2.0  \\\n",
      "0                                        1                          0   \n",
      "1                                        1                          0   \n",
      "2                                        1                          0   \n",
      "3                                        1                          0   \n",
      "4                                        1                          0   \n",
      "...                                    ...                        ...   \n",
      "26217                                    0                          1   \n",
      "26218                                    0                          0   \n",
      "26219                                    0                          0   \n",
      "26220                                    1                          0   \n",
      "26221                                    1                          0   \n",
      "\n",
      "       mostfrequent_n_family_3.0  mostfrequent_n_family_4.0  \\\n",
      "0                              0                          0   \n",
      "1                              0                          0   \n",
      "2                              0                          0   \n",
      "3                              0                          0   \n",
      "4                              0                          0   \n",
      "...                          ...                        ...   \n",
      "26217                          0                          0   \n",
      "26218                          0                          1   \n",
      "26219                          0                          0   \n",
      "26220                          0                          0   \n",
      "26221                          0                          0   \n",
      "\n",
      "       mostfrequent_n_family_5+  mostfrequent_no_of_children_2.0  \\\n",
      "0                             0                                0   \n",
      "1                             0                                0   \n",
      "2                             0                                0   \n",
      "3                             0                                0   \n",
      "4                             0                                0   \n",
      "...                         ...                              ...   \n",
      "26217                         0                                0   \n",
      "26218                         0                                0   \n",
      "26219                         0                                0   \n",
      "26220                         0                                0   \n",
      "26221                         0                                0   \n",
      "\n",
      "       mostfrequent_no_of_children_0  mostfrequent_no_of_children_1  \\\n",
      "0                                  1                              0   \n",
      "1                                  1                              0   \n",
      "2                                  1                              0   \n",
      "3                                  1                              0   \n",
      "4                                  1                              0   \n",
      "...                              ...                            ...   \n",
      "26217                              0                              0   \n",
      "26218                              0                              0   \n",
      "26219                              1                              0   \n",
      "26220                              1                              0   \n",
      "26221                              1                              0   \n",
      "\n",
      "       mostfrequent_no_of_children_2  mostfrequent_no_of_children_3+  \n",
      "0                                  0                               0  \n",
      "1                                  0                               0  \n",
      "2                                  0                               0  \n",
      "3                                  0                               0  \n",
      "4                                  0                               0  \n",
      "...                              ...                             ...  \n",
      "26217                              0                               0  \n",
      "26218                              0                               0  \n",
      "26219                              0                               0  \n",
      "26220                              0                               0  \n",
      "26221                              0                               0  \n",
      "\n",
      "[26222 rows x 47 columns]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/linear_model/_logistic.py:818: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=1e+20, fit_intercept=False)\n",
      "AUC for 0.15: 0.6133106242234908\n",
      "-------------------------------------------------------------------------------------\n",
      "       prod_id  cust_id  c_id  total_coupoun_used  avg_dcnt_pcode  \\\n",
      "0           22      637     8                   0             0.0   \n",
      "1           22      637    30                   0             0.0   \n",
      "2           27      637    30                   0             0.0   \n",
      "3           31      637     8                   0             0.0   \n",
      "4           31      637    30                   0             0.0   \n",
      "...        ...      ...   ...                 ...             ...   \n",
      "28497    29055      821     8                   0             0.0   \n",
      "28498    22326      594    25                   0             0.0   \n",
      "28499     7413     1317     9                   0             0.0   \n",
      "28500     8546      929     8                   0             0.0   \n",
      "28501     7389     1193     8                   0             0.0   \n",
      "\n",
      "       avg_dscnt_store  mostfrequent_duration_days  mostfrequent_income_range  \\\n",
      "0            -5.875000                          77                   6.000000   \n",
      "1            -5.875000                         133                   6.000000   \n",
      "2           -34.550000                         133                   6.000000   \n",
      "3             0.000000                          77                   6.000000   \n",
      "4            -4.272000                         133                   6.000000   \n",
      "...                ...                         ...                        ...   \n",
      "28497       -15.059613                          77                   4.571593   \n",
      "28498       -62.849802                         122                   4.000000   \n",
      "28499       -25.291989                          79                   1.000000   \n",
      "28500       -34.935802                          77                   8.000000   \n",
      "28501        -3.384500                          77                   5.220483   \n",
      "\n",
      "       mostfrequent_manuf   avg_price  ...  \\\n",
      "0                      56   29.745000  ...   \n",
      "1                      56   29.745000  ...   \n",
      "2                      56   66.970000  ...   \n",
      "3                      56   35.620000  ...   \n",
      "4                      56   38.472000  ...   \n",
      "...                   ...         ...  ...   \n",
      "28497                 699   72.359284  ...   \n",
      "28498                 250  128.498937  ...   \n",
      "28499                  59   74.496839  ...   \n",
      "28500                1487  170.222358  ...   \n",
      "28501                  56   42.557704  ...   \n",
      "\n",
      "       mostfrequent_marital_status_unknown  mostfrequent_n_family_2.0  \\\n",
      "0                                        1                          0   \n",
      "1                                        1                          0   \n",
      "2                                        1                          0   \n",
      "3                                        1                          0   \n",
      "4                                        1                          0   \n",
      "...                                    ...                        ...   \n",
      "28497                                    0                          0   \n",
      "28498                                    0                          0   \n",
      "28499                                    0                          0   \n",
      "28500                                    0                          0   \n",
      "28501                                    0                          1   \n",
      "\n",
      "       mostfrequent_n_family_3.0  mostfrequent_n_family_4.0  \\\n",
      "0                              0                          0   \n",
      "1                              0                          0   \n",
      "2                              0                          0   \n",
      "3                              0                          0   \n",
      "4                              0                          0   \n",
      "...                          ...                        ...   \n",
      "28497                          0                          0   \n",
      "28498                          0                          0   \n",
      "28499                          0                          0   \n",
      "28500                          0                          0   \n",
      "28501                          0                          0   \n",
      "\n",
      "       mostfrequent_n_family_5+  mostfrequent_no_of_children_2.0  \\\n",
      "0                             0                                0   \n",
      "1                             0                                0   \n",
      "2                             0                                0   \n",
      "3                             0                                0   \n",
      "4                             0                                0   \n",
      "...                         ...                              ...   \n",
      "28497                         0                                0   \n",
      "28498                         0                                0   \n",
      "28499                         0                                0   \n",
      "28500                         0                                0   \n",
      "28501                         0                                0   \n",
      "\n",
      "       mostfrequent_no_of_children_0  mostfrequent_no_of_children_1  \\\n",
      "0                                  1                              0   \n",
      "1                                  1                              0   \n",
      "2                                  1                              0   \n",
      "3                                  1                              0   \n",
      "4                                  1                              0   \n",
      "...                              ...                            ...   \n",
      "28497                              0                              0   \n",
      "28498                              0                              0   \n",
      "28499                              0                              0   \n",
      "28500                              0                              0   \n",
      "28501                              1                              0   \n",
      "\n",
      "       mostfrequent_no_of_children_2  mostfrequent_no_of_children_3+  \n",
      "0                                  0                               0  \n",
      "1                                  0                               0  \n",
      "2                                  0                               0  \n",
      "3                                  0                               0  \n",
      "4                                  0                               0  \n",
      "...                              ...                             ...  \n",
      "28497                              0                               0  \n",
      "28498                              0                               0  \n",
      "28499                              0                               0  \n",
      "28500                              0                               0  \n",
      "28501                              0                               0  \n",
      "\n",
      "[28502 rows x 47 columns]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/linear_model/_logistic.py:818: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=1e+20, fit_intercept=False)\n",
      "AUC for 0.25: 0.6351891031497745\n",
      "-------------------------------------------------------------------------------------\n",
      "       prod_id  cust_id  c_id  total_coupoun_used  avg_dcnt_pcode  \\\n",
      "0           22      637     8                   0        0.000000   \n",
      "1           22      637    30                   0        0.000000   \n",
      "2           27      637    30                   0        0.000000   \n",
      "3           31      637     8                   0        0.000000   \n",
      "4           31      637    30                   0        0.000000   \n",
      "...        ...      ...   ...                 ...             ...   \n",
      "30321    23613      441     8                   0        0.000000   \n",
      "30322    13022      824     8                   0        0.000000   \n",
      "30323     7877      903     8                   0        0.000000   \n",
      "30324    16811     1219     8                   2       -2.860429   \n",
      "30325    16317       51    22                   0      -17.127891   \n",
      "\n",
      "       avg_dscnt_store  mostfrequent_duration_days  mostfrequent_income_range  \\\n",
      "0            -5.875000                          77                   6.000000   \n",
      "1            -5.875000                         133                   6.000000   \n",
      "2           -34.550000                         133                   6.000000   \n",
      "3             0.000000                          77                   6.000000   \n",
      "4            -4.272000                         133                   6.000000   \n",
      "...                ...                         ...                        ...   \n",
      "30321        -7.120000                          77                   3.415410   \n",
      "30322        -6.853082                          77                   8.000000   \n",
      "30323       -40.582334                          77                   7.641810   \n",
      "30324       -17.531005                          77                   5.518176   \n",
      "30325        -6.334934                         148                   4.901556   \n",
      "\n",
      "       mostfrequent_manuf  avg_price  ...  \\\n",
      "0                      56  29.745000  ...   \n",
      "1                      56  29.745000  ...   \n",
      "2                      56  66.970000  ...   \n",
      "3                      56  35.620000  ...   \n",
      "4                      56  38.472000  ...   \n",
      "...                   ...        ...  ...   \n",
      "30321                1101  63.760000  ...   \n",
      "30322                  56  39.839029  ...   \n",
      "30323                 128  69.644860  ...   \n",
      "30324                 794  60.478995  ...   \n",
      "30325                  56  97.390146  ...   \n",
      "\n",
      "       mostfrequent_marital_status_unknown  mostfrequent_n_family_2.0  \\\n",
      "0                                        1                          0   \n",
      "1                                        1                          0   \n",
      "2                                        1                          0   \n",
      "3                                        1                          0   \n",
      "4                                        1                          0   \n",
      "...                                    ...                        ...   \n",
      "30321                                    0                          0   \n",
      "30322                                    0                          0   \n",
      "30323                                    0                          0   \n",
      "30324                                    0                          0   \n",
      "30325                                    1                          0   \n",
      "\n",
      "       mostfrequent_n_family_3.0  mostfrequent_n_family_4.0  \\\n",
      "0                              0                          0   \n",
      "1                              0                          0   \n",
      "2                              0                          0   \n",
      "3                              0                          0   \n",
      "4                              0                          0   \n",
      "...                          ...                        ...   \n",
      "30321                          0                          0   \n",
      "30322                          0                          0   \n",
      "30323                          0                          0   \n",
      "30324                          0                          0   \n",
      "30325                          0                          0   \n",
      "\n",
      "       mostfrequent_n_family_5+  mostfrequent_no_of_children_2.0  \\\n",
      "0                             0                                0   \n",
      "1                             0                                0   \n",
      "2                             0                                0   \n",
      "3                             0                                0   \n",
      "4                             0                                0   \n",
      "...                         ...                              ...   \n",
      "30321                         0                                0   \n",
      "30322                         0                                0   \n",
      "30323                         0                                0   \n",
      "30324                         0                                0   \n",
      "30325                         0                                0   \n",
      "\n",
      "       mostfrequent_no_of_children_0  mostfrequent_no_of_children_1  \\\n",
      "0                                  1                              0   \n",
      "1                                  1                              0   \n",
      "2                                  1                              0   \n",
      "3                                  1                              0   \n",
      "4                                  1                              0   \n",
      "...                              ...                            ...   \n",
      "30321                              0                              0   \n",
      "30322                              0                              0   \n",
      "30323                              0                              0   \n",
      "30324                              1                              0   \n",
      "30325                              1                              0   \n",
      "\n",
      "       mostfrequent_no_of_children_2  mostfrequent_no_of_children_3+  \n",
      "0                                  0                               0  \n",
      "1                                  0                               0  \n",
      "2                                  0                               0  \n",
      "3                                  0                               0  \n",
      "4                                  0                               0  \n",
      "...                              ...                             ...  \n",
      "30321                              0                               0  \n",
      "30322                              0                               0  \n",
      "30323                              0                               0  \n",
      "30324                              0                               0  \n",
      "30325                              0                               0  \n",
      "\n",
      "[30326 rows x 47 columns]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/linear_model/_logistic.py:818: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=1e+20, fit_intercept=False)\n",
      "AUC for 0.33: 0.643375151999495\n",
      "-------------------------------------------------------------------------------------\n",
      "       prod_id  cust_id  c_id  total_coupoun_used  avg_dcnt_pcode  \\\n",
      "0           22      637     8                   0         0.00000   \n",
      "1           22      637    30                   0         0.00000   \n",
      "2           27      637    30                   0         0.00000   \n",
      "3           31      637     8                   0         0.00000   \n",
      "4           31      637    30                   0         0.00000   \n",
      "...        ...      ...   ...                 ...             ...   \n",
      "34198     5288      860     8                   0         0.00000   \n",
      "34199     8646      802     8                   0         0.00000   \n",
      "34200    23226     1038     8                   1       -14.89528   \n",
      "34201    23124     1477     8                   0         0.00000   \n",
      "34202    18089      173     8                   0         0.00000   \n",
      "\n",
      "       avg_dscnt_store  mostfrequent_duration_days  mostfrequent_income_range  \\\n",
      "0            -5.875000                          77                   6.000000   \n",
      "1            -5.875000                         133                   6.000000   \n",
      "2           -34.550000                         133                   6.000000   \n",
      "3             0.000000                          77                   6.000000   \n",
      "4            -4.272000                         133                   6.000000   \n",
      "...                ...                         ...                        ...   \n",
      "34198       -38.619903                          77                   7.427052   \n",
      "34199       -61.930540                          77                   2.416976   \n",
      "34200        -9.383199                          77                   4.901556   \n",
      "34201        -3.469999                          77                   5.000000   \n",
      "34202       -56.154693                          77                   7.060882   \n",
      "\n",
      "       mostfrequent_manuf   avg_price  ...  \\\n",
      "0                      56   29.745000  ...   \n",
      "1                      56   29.745000  ...   \n",
      "2                      56   66.970000  ...   \n",
      "3                      56   35.620000  ...   \n",
      "4                      56   38.472000  ...   \n",
      "...                   ...         ...  ...   \n",
      "34198                4615   57.092315  ...   \n",
      "34199                1950  196.776880  ...   \n",
      "34200                1807  131.590736  ...   \n",
      "34201                1101   17.810000  ...   \n",
      "34202                1473   79.666539  ...   \n",
      "\n",
      "       mostfrequent_marital_status_unknown  mostfrequent_n_family_2.0  \\\n",
      "0                                        1                          0   \n",
      "1                                        1                          0   \n",
      "2                                        1                          0   \n",
      "3                                        1                          0   \n",
      "4                                        1                          0   \n",
      "...                                    ...                        ...   \n",
      "34198                                    0                          0   \n",
      "34199                                    1                          0   \n",
      "34200                                    1                          0   \n",
      "34201                                    1                          0   \n",
      "34202                                    1                          0   \n",
      "\n",
      "       mostfrequent_n_family_3.0  mostfrequent_n_family_4.0  \\\n",
      "0                              0                          0   \n",
      "1                              0                          0   \n",
      "2                              0                          0   \n",
      "3                              0                          0   \n",
      "4                              0                          0   \n",
      "...                          ...                        ...   \n",
      "34198                          0                          0   \n",
      "34199                          0                          0   \n",
      "34200                          0                          0   \n",
      "34201                          0                          0   \n",
      "34202                          0                          0   \n",
      "\n",
      "       mostfrequent_n_family_5+  mostfrequent_no_of_children_2.0  \\\n",
      "0                             0                                0   \n",
      "1                             0                                0   \n",
      "2                             0                                0   \n",
      "3                             0                                0   \n",
      "4                             0                                0   \n",
      "...                         ...                              ...   \n",
      "34198                         0                                0   \n",
      "34199                         0                                0   \n",
      "34200                         0                                0   \n",
      "34201                         0                                0   \n",
      "34202                         0                                0   \n",
      "\n",
      "       mostfrequent_no_of_children_0  mostfrequent_no_of_children_1  \\\n",
      "0                                  1                              0   \n",
      "1                                  1                              0   \n",
      "2                                  1                              0   \n",
      "3                                  1                              0   \n",
      "4                                  1                              0   \n",
      "...                              ...                            ...   \n",
      "34198                              0                              0   \n",
      "34199                              0                              0   \n",
      "34200                              1                              0   \n",
      "34201                              1                              0   \n",
      "34202                              1                              0   \n",
      "\n",
      "       mostfrequent_no_of_children_2  mostfrequent_no_of_children_3+  \n",
      "0                                  0                               0  \n",
      "1                                  0                               0  \n",
      "2                                  0                               0  \n",
      "3                                  0                               0  \n",
      "4                                  0                               0  \n",
      "...                              ...                             ...  \n",
      "34198                              0                               0  \n",
      "34199                              0                               0  \n",
      "34200                              0                               0  \n",
      "34201                              0                               0  \n",
      "34202                              0                               0  \n",
      "\n",
      "[34203 rows x 47 columns]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/linear_model/_logistic.py:818: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=1e+20, fit_intercept=False)\n",
      "AUC for 0.5: 0.6612091784578646\n",
      "-------------------------------------------------------------------------------------\n",
      "       prod_id  cust_id  c_id  total_coupoun_used  avg_dcnt_pcode  \\\n",
      "0           22      637     8                   0        0.000000   \n",
      "1           22      637    30                   0        0.000000   \n",
      "2           27      637    30                   0        0.000000   \n",
      "3           31      637     8                   0        0.000000   \n",
      "4           31      637    30                   0        0.000000   \n",
      "...        ...      ...   ...                 ...             ...   \n",
      "38758    13540      948     8                   0        0.000000   \n",
      "38759    23898     1405     8                   0        0.000000   \n",
      "38760    16754     1391     8                   0       -7.176759   \n",
      "38761    24410      880     8                   0        0.000000   \n",
      "38762    11083      536     8                   0        0.000000   \n",
      "\n",
      "       avg_dscnt_store  mostfrequent_duration_days  mostfrequent_income_range  \\\n",
      "0            -5.875000                          77                   6.000000   \n",
      "1            -5.875000                         133                   6.000000   \n",
      "2           -34.550000                         133                   6.000000   \n",
      "3             0.000000                          77                   6.000000   \n",
      "4            -4.272000                         133                   6.000000   \n",
      "...                ...                         ...                        ...   \n",
      "38758       -12.202621                          77                   8.000000   \n",
      "38759       -38.409873                          77                   2.574498   \n",
      "38760       -61.276472                          77                   7.388151   \n",
      "38761       -22.071568                          77                   3.925398   \n",
      "38762      -112.356949                          77                   4.917225   \n",
      "\n",
      "       mostfrequent_manuf   avg_price  ...  \\\n",
      "0                      56   29.745000  ...   \n",
      "1                      56   29.745000  ...   \n",
      "2                      56   66.970000  ...   \n",
      "3                      56   35.620000  ...   \n",
      "4                      56   38.472000  ...   \n",
      "...                   ...         ...  ...   \n",
      "38758                 116   91.556723  ...   \n",
      "38759                  56   49.023549  ...   \n",
      "38760                1072  112.494637  ...   \n",
      "38761                 482  134.851112  ...   \n",
      "38762                1255  105.903004  ...   \n",
      "\n",
      "       mostfrequent_marital_status_unknown  mostfrequent_n_family_2.0  \\\n",
      "0                                        1                          0   \n",
      "1                                        1                          0   \n",
      "2                                        1                          0   \n",
      "3                                        1                          0   \n",
      "4                                        1                          0   \n",
      "...                                    ...                        ...   \n",
      "38758                                    0                          0   \n",
      "38759                                    0                          0   \n",
      "38760                                    0                          0   \n",
      "38761                                    0                          0   \n",
      "38762                                    0                          0   \n",
      "\n",
      "       mostfrequent_n_family_3.0  mostfrequent_n_family_4.0  \\\n",
      "0                              0                          0   \n",
      "1                              0                          0   \n",
      "2                              0                          0   \n",
      "3                              0                          0   \n",
      "4                              0                          0   \n",
      "...                          ...                        ...   \n",
      "38758                          0                          0   \n",
      "38759                          0                          0   \n",
      "38760                          0                          0   \n",
      "38761                          0                          0   \n",
      "38762                          0                          0   \n",
      "\n",
      "       mostfrequent_n_family_5+  mostfrequent_no_of_children_2.0  \\\n",
      "0                             0                                0   \n",
      "1                             0                                0   \n",
      "2                             0                                0   \n",
      "3                             0                                0   \n",
      "4                             0                                0   \n",
      "...                         ...                              ...   \n",
      "38758                         1                                0   \n",
      "38759                         0                                0   \n",
      "38760                         0                                0   \n",
      "38761                         0                                0   \n",
      "38762                         0                                0   \n",
      "\n",
      "       mostfrequent_no_of_children_0  mostfrequent_no_of_children_1  \\\n",
      "0                                  1                              0   \n",
      "1                                  1                              0   \n",
      "2                                  1                              0   \n",
      "3                                  1                              0   \n",
      "4                                  1                              0   \n",
      "...                              ...                            ...   \n",
      "38758                              0                              0   \n",
      "38759                              0                              0   \n",
      "38760                              1                              0   \n",
      "38761                              0                              0   \n",
      "38762                              1                              0   \n",
      "\n",
      "       mostfrequent_no_of_children_2  mostfrequent_no_of_children_3+  \n",
      "0                                  0                               0  \n",
      "1                                  0                               0  \n",
      "2                                  0                               0  \n",
      "3                                  0                               0  \n",
      "4                                  0                               0  \n",
      "...                              ...                             ...  \n",
      "38758                              0                               1  \n",
      "38759                              0                               0  \n",
      "38760                              0                               0  \n",
      "38761                              0                               0  \n",
      "38762                              0                               0  \n",
      "\n",
      "[38763 rows x 47 columns]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/linear_model/_logistic.py:818: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=1e+20, fit_intercept=False)\n",
      "AUC for 0.7: 0.6714798311255339\n",
      "-------------------------------------------------------------------------------------\n",
      "       prod_id  cust_id  c_id  total_coupoun_used  avg_dcnt_pcode  \\\n",
      "0           22      637     8                   0        0.000000   \n",
      "1           22      637    30                   0        0.000000   \n",
      "2           27      637    30                   0        0.000000   \n",
      "3           31      637     8                   0        0.000000   \n",
      "4           31      637    30                   0        0.000000   \n",
      "...        ...      ...   ...                 ...             ...   \n",
      "45599    11461      561    30                   0        0.000000   \n",
      "45600    14091      607     8                   0      -15.388540   \n",
      "45601     9852     1542     8                   3       -7.724697   \n",
      "45602    24698     1546    30                   0      -35.453883   \n",
      "45603    12763     1263     8                   0        0.000000   \n",
      "\n",
      "       avg_dscnt_store  mostfrequent_duration_days  mostfrequent_income_range  \\\n",
      "0            -5.875000                          77                   6.000000   \n",
      "1            -5.875000                         133                   6.000000   \n",
      "2           -34.550000                         133                   6.000000   \n",
      "3             0.000000                          77                   6.000000   \n",
      "4            -4.272000                         133                   6.000000   \n",
      "...                ...                         ...                        ...   \n",
      "45599       -18.103565                         133                   4.000000   \n",
      "45600       -19.476640                          77                   6.728402   \n",
      "45601       -14.685001                          77                   7.843175   \n",
      "45602      -213.221649                         133                   4.995336   \n",
      "45603       -86.992383                          77                   6.550235   \n",
      "\n",
      "       mostfrequent_manuf   avg_price  ...  \\\n",
      "0                      56   29.745000  ...   \n",
      "1                      56   29.745000  ...   \n",
      "2                      56   66.970000  ...   \n",
      "3                      56   35.620000  ...   \n",
      "4                      56   38.472000  ...   \n",
      "...                   ...         ...  ...   \n",
      "45599                 278   92.398475  ...   \n",
      "45600                 489  110.390005  ...   \n",
      "45601                1337   51.482958  ...   \n",
      "45602                 241  212.513282  ...   \n",
      "45603                4374  105.639806  ...   \n",
      "\n",
      "       mostfrequent_marital_status_unknown  mostfrequent_n_family_2.0  \\\n",
      "0                                        1                          0   \n",
      "1                                        1                          0   \n",
      "2                                        1                          0   \n",
      "3                                        1                          0   \n",
      "4                                        1                          0   \n",
      "...                                    ...                        ...   \n",
      "45599                                    0                          1   \n",
      "45600                                    0                          0   \n",
      "45601                                    0                          0   \n",
      "45602                                    0                          0   \n",
      "45603                                    1                          0   \n",
      "\n",
      "       mostfrequent_n_family_3.0  mostfrequent_n_family_4.0  \\\n",
      "0                              0                          0   \n",
      "1                              0                          0   \n",
      "2                              0                          0   \n",
      "3                              0                          0   \n",
      "4                              0                          0   \n",
      "...                          ...                        ...   \n",
      "45599                          0                          0   \n",
      "45600                          0                          0   \n",
      "45601                          0                          0   \n",
      "45602                          0                          0   \n",
      "45603                          0                          0   \n",
      "\n",
      "       mostfrequent_n_family_5+  mostfrequent_no_of_children_2.0  \\\n",
      "0                             0                                0   \n",
      "1                             0                                0   \n",
      "2                             0                                0   \n",
      "3                             0                                0   \n",
      "4                             0                                0   \n",
      "...                         ...                              ...   \n",
      "45599                         0                                0   \n",
      "45600                         0                                0   \n",
      "45601                         0                                0   \n",
      "45602                         0                                0   \n",
      "45603                         0                                0   \n",
      "\n",
      "       mostfrequent_no_of_children_0  mostfrequent_no_of_children_1  \\\n",
      "0                                  1                              0   \n",
      "1                                  1                              0   \n",
      "2                                  1                              0   \n",
      "3                                  1                              0   \n",
      "4                                  1                              0   \n",
      "...                              ...                            ...   \n",
      "45599                              1                              0   \n",
      "45600                              0                              0   \n",
      "45601                              0                              0   \n",
      "45602                              0                              0   \n",
      "45603                              0                              0   \n",
      "\n",
      "       mostfrequent_no_of_children_2  mostfrequent_no_of_children_3+  \n",
      "0                                  0                               0  \n",
      "1                                  0                               0  \n",
      "2                                  0                               0  \n",
      "3                                  0                               0  \n",
      "4                                  0                               0  \n",
      "...                              ...                             ...  \n",
      "45599                              0                               0  \n",
      "45600                              0                               0  \n",
      "45601                              0                               0  \n",
      "45602                              0                               0  \n",
      "45603                              0                               0  \n",
      "\n",
      "[45604 rows x 47 columns]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/linear_model/_logistic.py:818: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=1e+20, fit_intercept=False)\n",
      "AUC for even: 0.6639063668433053\n",
      "-------------------------------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train_resampled, y_train_resampled = smote_(X_train, X_test_validation, y_train, y_test_validation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "7NHTR9VkE0zz",
    "outputId": "570e5e06-1a7b-4658-a924-0ee0816e5cab"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original class distribution: \n",
      "\n",
      "0    37939\n",
      "1     3801\n",
      "Name: redeemed, dtype: int64\n",
      "-----------------------------------------\n",
      "Synthetic sample class distribution: \n",
      "\n",
      "0    22802\n",
      "1    15961\n",
      "Name: redeemed, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# # Previous original class distribution\n",
    "print('Original class distribution: \\n')\n",
    "y=final_data_copy['redeemed']\n",
    "print(y.value_counts())\n",
    "SMOTE_ratio=0.7\n",
    "smote = SMOTE(sampling_strategy=SMOTE_ratio)\n",
    "X_train_resampled, y_train_resampled = smote.fit_resample(X_train, y_train) \n",
    "# Preview synthetic sample class distribution\n",
    "print('-----------------------------------------')\n",
    "print('Synthetic sample class distribution: \\n')\n",
    "print(pd.Series(y_train_resampled).value_counts()) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "oKKVOFkvFuLW"
   },
   "outputs": [],
   "source": [
    "scaled_X_train_resampled, scaled_X_test = Standardlize_(X_train_resampled,X_test_validation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "_mdcwLwm01wg",
    "outputId": "f025be87-5fbb-40ff-bbe7-e247c8cdf7f2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['avg_dscnt_store', 'mostfrequent_marital_status_unknown', 'cust_id',\n",
      "       'mostfrequent_n_family_2.0', 'mostfrequent_income_range',\n",
      "       'avg_dcnt_pcode', 'mostfrequent_marital_status_Single',\n",
      "       'mostfrequent_n_family_3.0', 'total_coupoun_used',\n",
      "       'mostfrequent_n_family_4.0', 'mostfrequent_n_family_5+',\n",
      "       'mostfrequent_age_cohort_unknown', 'mostfrequent_age_cohort_36-45',\n",
      "       'mostfrequent_age_cohort_26-35', 'mostfrequent_age_cohort_46-55',\n",
      "       'prod_id', 'mostfrequent_duration_days', 'c_id',\n",
      "       'mostfrequent_age_cohort_56-70', 'avg_price',\n",
      "       'mostfrequent_age_cohort_70+', 'mostfrequent_manuf',\n",
      "       'mostfrequent_no_of_children_2.0', 'avg_qty',\n",
      "       'mostfrequent_home_is_rental_1.0', 'mostfrequent_no_of_children_0',\n",
      "       'mostfrequent_category_Grocery', 'mostfrequent_ctype_Y',\n",
      "       'mostfrequent_category_Packaged Meat', 'mostfrequent_no_of_children_2',\n",
      "       'mostfrequent_no_of_children_1', 'mostfrequent_brand_type_Local',\n",
      "       'mostfrequent_category_Natural Products',\n",
      "       'mostfrequent_no_of_children_3+', 'mostfrequent_category_Meat',\n",
      "       'mostfrequent_category_Seafood', 'mostfrequent_category_Pharmaceutical',\n",
      "       'mostfrequent_category_Dairy, Juices & Snacks',\n",
      "       'mostfrequent_category_Vegetables (cut)',\n",
      "       'mostfrequent_home_is_rental_0', 'mostfrequent_category_Travel',\n",
      "       'mostfrequent_category_Skin & Hair Care',\n",
      "       'mostfrequent_category_Salads', 'mostfrequent_category_Prepared Food',\n",
      "       'mostfrequent_category_Miscellaneous', 'mostfrequent_category_Garden',\n",
      "       'mostfrequent_category_Flowers & Plants'],\n",
      "      dtype='object')\n"
     ]
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "model = DecisionTreeClassifier()\n",
    "model.fit(scaled_X_train_resampled,y_train_resampled)\n",
    "model.feature_importances_\n",
    "importances = pd.Series(model.feature_importances_,index=scaled_X_train_resampled.columns)\n",
    "importances = importances.sort_values(ascending=False)\n",
    "print(importances.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ftIQr3U0PEvT"
   },
   "outputs": [],
   "source": [
    "scaled_X_train_resampled = scaled_X_train_resampled.drop(columns =['mostfrequent_age_cohort_56-70', 'mostfrequent_age_cohort_70+',\n",
    "       'mostfrequent_no_of_children_0', 'mostfrequent_category_Grocery',\n",
    "       'mostfrequent_no_of_children_2.0', 'mostfrequent_brand_type_Local',\n",
    "       'mostfrequent_category_Packaged Meat', 'mostfrequent_no_of_children_1',\n",
    "       'mostfrequent_ctype_Y', 'mostfrequent_home_is_rental_1.0',\n",
    "       'mostfrequent_category_Natural Products',\n",
    "       'mostfrequent_no_of_children_3+',\n",
    "       'mostfrequent_category_Pharmaceutical', 'mostfrequent_no_of_children_2',\n",
    "       'mostfrequent_category_Meat',\n",
    "       'mostfrequent_category_Dairy, Juices & Snacks',\n",
    "       'mostfrequent_home_is_rental_0', 'mostfrequent_category_Seafood',\n",
    "       'mostfrequent_category_Vegetables (cut)',\n",
    "       'mostfrequent_category_Travel',\n",
    "       'mostfrequent_category_Skin & Hair Care',\n",
    "       'mostfrequent_category_Salads', 'mostfrequent_category_Prepared Food',\n",
    "       'mostfrequent_category_Miscellaneous', 'mostfrequent_category_Garden',\n",
    "       'mostfrequent_category_Flowers & Plants','prod_id','cust_id','c_id'],inplace=True)\n",
    "scaled_X_test = scaled_X_test.drop(columns=['mostfrequent_age_cohort_56-70', 'mostfrequent_age_cohort_70+',\n",
    "       'mostfrequent_no_of_children_0', 'mostfrequent_category_Grocery',\n",
    "       'mostfrequent_no_of_children_2.0', 'mostfrequent_brand_type_Local',\n",
    "       'mostfrequent_category_Packaged Meat', 'mostfrequent_no_of_children_1',\n",
    "       'mostfrequent_ctype_Y', 'mostfrequent_home_is_rental_1.0',\n",
    "       'mostfrequent_category_Natural Products',\n",
    "       'mostfrequent_no_of_children_3+',\n",
    "       'mostfrequent_category_Pharmaceutical', 'mostfrequent_no_of_children_2',\n",
    "       'mostfrequent_category_Meat',\n",
    "       'mostfrequent_category_Dairy, Juices & Snacks',\n",
    "       'mostfrequent_home_is_rental_0', 'mostfrequent_category_Seafood',\n",
    "       'mostfrequent_category_Vegetables (cut)',\n",
    "       'mostfrequent_category_Travel',\n",
    "       'mostfrequent_category_Skin & Hair Care',\n",
    "       'mostfrequent_category_Salads', 'mostfrequent_category_Prepared Food',\n",
    "       'mostfrequent_category_Miscellaneous', 'mostfrequent_category_Garden',\n",
    "       'mostfrequent_category_Flowers & Plants','prod_id','cust_id','c_id'],inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "vu_0UUK1NgD-"
   },
   "outputs": [],
   "source": [
    "#dummy classifier 2 with 1 variables to check if a particular product has an effect on coupon redemption\n",
    "clf = LogisticRegression(C = 1.0)\n",
    "clf = clf.fit(scaled_X_train_resampled[[\"prod_id\"]], y_train_resampled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "NM-4qb1S0UEL",
    "outputId": "3f2bb21b-c021-4d95-cf25-3082c3d30b90"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_search.py:296: UserWarning: The total space of parameters 3 is smaller than n_iter=10. Running 3 iterations. For exhaustive searches, use GridSearchCV.\n",
      "  UserWarning,\n",
      "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_search.py:296: UserWarning: The total space of parameters 5 is smaller than n_iter=10. Running 5 iterations. For exhaustive searches, use GridSearchCV.\n",
      "  UserWarning,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression Recall is: 0.16526610644257703\n",
      "Logistic Regression Best Params: {'LR__C': 0.5}\n",
      "Decision Trees Recall is: 0.45938375350140054\n",
      "Decision Trees Best Params: {'DT__min_samples_leaf': 8}\n",
      "Support Vector Machines Recall is: 0.22549019607843138\n",
      "Support Vector Machines Best Params: {'SVM__kernel': 'rbf', 'SVM__C': 6}\n",
      "XGBOOST Recall is: 0.8907563025210085\n",
      "XGBOOST Best Params: {'xgboost__subsample': 0.6, 'xgboost__scale_pos_weight': 5, 'xgboost__n_estimators': 100, 'xgboost__max_depth': 5, 'xgboost__learning_rate': 0.01, 'xgboost__colsample_bytree': 0.4, 'xgboost__colsample_bylevel': 0.6000000000000001}\n",
      "randomforest Recall is: 0.28291316526610644\n",
      "randomforest Best Params: {'randforesr__n_estimators': 200, 'randforesr__min_samples_split': 2, 'randforesr__min_samples_leaf': 1, 'randforesr__max_features': 'auto', 'randforesr__max_depth': 30}\n"
     ]
    }
   ],
   "source": [
    "# grid_search(scaled_X_train_resampled, y_train_resampled,scaled_X_test,y_test_validation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "A5zlyVFPNhzl"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "xR1CDL7iOLzY",
    "outputId": "5370272b-9f7b-4a58-b155-bdf7d48908e6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.93      0.97      0.95      4690\n",
      "           1       0.56      0.32      0.41       528\n",
      "\n",
      "    accuracy                           0.91      5218\n",
      "   macro avg       0.74      0.65      0.68      5218\n",
      "weighted avg       0.89      0.91      0.89      5218\n",
      "\n",
      "DecisionTreeClassifier()accuracy = 0.9057110003832887\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning: Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.92      0.93      0.92      4690\n",
      "           1       0.29      0.27      0.28       528\n",
      "\n",
      "    accuracy                           0.86      5218\n",
      "   macro avg       0.60      0.60      0.60      5218\n",
      "weighted avg       0.85      0.86      0.86      5218\n",
      "\n",
      "LogisticRegression()accuracy = 0.8602912993484093\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning: Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import confusion_matrix,accuracy_score\n",
    "from sklearn.dummy import DummyClassifier\n",
    "\n",
    "example_model =[DecisionTreeClassifier(min_samples_leaf = 8),\n",
    "                LogisticRegression(C = 0.5),\n",
    "                svm.SVC(kernel= 'linear', C=5,probability=True),\n",
    "                XGBClassifier(subsample=0.6,scale_pos_weight=5,n_estimators=100,max_depth=5,learning_rate=0.01,colsample_bytree=0.4,colsample_bylevel=0.6),\n",
    "                RandomForestClassifier(n_estimators=200,min_samples_split=2,min_samples_leaf=1,max_features='auto',max_depth=30),\n",
    "                DummyClassifier(strategy='stratified')] #naive classifier 1 with strategy of selecting a random class\n",
    "              \n",
    "for models in example_model:\n",
    "  # Remember to use the training data here!! \n",
    "  models.fit(scaled_X_train_resampled, y_train_resampled)\n",
    "  # And to use the test data here!!\n",
    "  pred = models.predict(scaled_X_test)\n",
    "  print(classification_report(y_true = y_test_validation, y_pred = pred))\n",
    "  metrics.plot_confusion_matrix(models, X = scaled_X_test, y_true = y_test_validation, cmap='Greys')\n",
    "  accuracy=accuracy_score(y_test_validation,pred)\n",
    "  print(str(models) +'accuracy =',accuracy)\n",
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "WmDeaG66MU5J"
   },
   "outputs": [],
   "source": [
    "## Test on test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "8D4llYMI3CJw"
   },
   "outputs": [],
   "source": [
    "scaled_X_train_resampled, scaled_X_test2 = Standardlize_(X_train_resampled,X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "Br681jZh-3b2",
    "outputId": "023816d2-2298-4650-f8e7-ee0897deb4e1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.94      0.96      0.95      7503\n",
      "           1       0.54      0.45      0.49       845\n",
      "\n",
      "    accuracy                           0.91      8348\n",
      "   macro avg       0.74      0.70      0.72      8348\n",
      "weighted avg       0.90      0.91      0.90      8348\n",
      "\n",
      "DecisionTreeClassifier(min_samples_leaf=8)accuracy = 0.9053665548634403\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning: Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.91      0.97      0.94      7503\n",
      "           1       0.37      0.17      0.23       845\n",
      "\n",
      "    accuracy                           0.89      8348\n",
      "   macro avg       0.64      0.57      0.58      8348\n",
      "weighted avg       0.86      0.89      0.87      8348\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning: Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=0.5)accuracy = 0.8870388116914231\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.58      0.73      7503\n",
      "           1       0.19      0.87      0.31       845\n",
      "\n",
      "    accuracy                           0.61      8348\n",
      "   macro avg       0.58      0.73      0.52      8348\n",
      "weighted avg       0.90      0.61      0.69      8348\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning: Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBClassifier(colsample_bylevel=0.6, colsample_bytree=0.4, learning_rate=0.01,\n",
      "              max_depth=5, scale_pos_weight=5, subsample=0.6)accuracy = 0.6106851940584571\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.92      0.99      0.96      7503\n",
      "           1       0.72      0.29      0.41       845\n",
      "\n",
      "    accuracy                           0.92      8348\n",
      "   macro avg       0.82      0.64      0.68      8348\n",
      "weighted avg       0.90      0.92      0.90      8348\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning: Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForestClassifier(max_depth=30, n_estimators=200)accuracy = 0.9166267369429804\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.90      0.59      0.71      7503\n",
      "           1       0.10      0.43      0.17       845\n",
      "\n",
      "    accuracy                           0.57      8348\n",
      "   macro avg       0.50      0.51      0.44      8348\n",
      "weighted avg       0.82      0.57      0.66      8348\n",
      "\n",
      "DummyClassifier(strategy='stratified')accuracy = 0.5709151892668902\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning: Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAT8AAAEGCAYAAAAT05LOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dfVjUdb7/8ecAoqjciMKgLlF4s3aUZC/1qIu30zWgIIIIh6y1lZOXad6s2tEwyxtM0l1NN909xWXHtHXTNIGSChU3kMy7FGk9Wkc3FG8YDBAUjYHx+/uDn99LMpgZnRGYeT+ui+tivvO9+XzMXn4+38/3+/loFEVREEIIJ+PS3AUQQojmIOEnhHBKEn5CCKck4SeEcEoSfkIIp+TW3AW417Vr17hw4UJzF8PmgoKCHLJeAL6+vs1dBLvw8/Pj2rVrzV0Mm/Pz88PT0/OhzvHFF1/QpUsXi/YdOHDgQ13LnlpU+F24cIFBgwY1dzFs7tixYw5ZL4DnnnuuuYtgFykpKSxZsqS5i2FzKSkpDx1+Xbp0sfjvc0t+kq5FhZ8QonXQaDTNXYSHJuEnhLCai0vrHy6Q8BNCWEWj0Uj4CSGck3R7hRBOScJPCOGUJPyEEE5Ho9FI+AkhnJOrq2tzF+GhSfgJIawmLT8hhNORbq8QwmlJ+AkhnJKEnxDC6Wg0GhnwEEI4J0do+bX+F/SEEI/c3UEPcz+WMJlMxMbG8uKLLwJQXFxMQkICer2euXPnYjQaATAajcydOxe9Xk9CQgKXLl1Sz/Huu++i1+uJiIjg4MGDFl1Xwk8IYTVbht/WrVvp0aOH+nnNmjVMmTKFffv24eXlxa5duwDYuXMnXl5e7Nu3jylTprBmzRoAzp07R1ZWFllZWWzatInly5djMpnMXlfCTwhhFUuDz5LwKykp4csvvyQ+Ph6on/z08OHDREREADBhwgRycnIAOHDgABMmTAAgIiKCr7/+GkVRyMnJISoqCnd3dwIDAwkKCqKwsNDsteWenxDCapa26srLy5k6dar6OTExkcTERPVzamoqCxYsoLq6GoCKigq8vLxwc6uPpoCAAAwGAwAGg4GuXbsC4ObmhqenJxUVFRgMBvr376+eU6vVqsc0RcJPCGE1S0d7fX192b179y9+949//ANfX1/69evHkSNHbFk8i0j4CSGsYqs3PE6cOMGBAwfIy8ujpqaGmzdvsnLlSqqqqqirq8PNzY2SkhK0Wi1Q36K7evUqAQEB1NXVcePGDTp16oRWq6WkpEQ9r8FgUI9pitzzE0JYzRb3/F5++WXy8vI4cOAAb731FkOGDGHt2rUMHjyY7OxsANLT09HpdADodDrS09MByM7OZsiQIWg0GnQ6HVlZWRiNRoqLiykqKuKpp54yWwdp+QkhrGbP5/wWLFjAvHnzWL9+PU8++SQJCQkAxMfHs2DBAvR6Pd7e3qxbtw6AXr16MXbsWCIjI3F1dWXJkiUWdcsl/IQQVrN1+A0ePJjBgwcDEBgYqD7ecq+2bdvy9ttv/+LxM2bMYMaMGVZdU8JPCGEVeb1NCOG0HOH1Ngk/IYTVJPyEEE5Jwk8I4XRk0XIhhNOSlp8QwunIaK8QwmlJy08I4ZTknp8QwilJy08I4XRk3V4hhNOSbq8QwunIaK8QwmlJt1cI4ZSk2yuEcEqO0PJr/fEthHikbLV0ZU1NDfHx8YwfP56oqCh1otLk5GR0Oh0xMTHExMRw5swZoH5ZyzfeeAO9Xk90dDSnT59Wz5Wenk54eDjh4eHqVPfmSMtPCGE1Wwx4uLu7s2XLFjp06EBtbS3PPvssI0aMAGDhwoWMGTOmwf55eXkUFRWxd+9eTp06xbJly9i5cyfXr19n48aNfPzxx2g0GuLi4tDpdHh7ezd5fWn5CSGscndWF0t+zJ2nQ4cOANTV1VFXV9dkazEnJ4fY2Fg0Gg2hoaFUVVVRWlpKfn4+YWFh+Pj44O3tTVhYGAcPHjRbDwk/IYTVLO32lpeXExcXp/7s2LGjwXlMJhMxMTH89re/5be//a26+Pi6deuIjo4mNTUVo9EI1C9JGRAQoB57d0Hzn2+XRcuFEHZj6YBHU4uWQ333OTMzk6qqKmbOnMn333/P/Pnz8fPzo7a2ltdff520tDRmzZplq6KrpOUnhLCaLbq99/Ly8mLw4MEcPHgQf39/NBoN7u7uxMXF8e233wLctzj53QXNZdFyIcQjYavR3vLycqqqqgD46aefOHToEMHBwZSWlgL1o7v79++nV69eQP2i5RkZGSiKQkFBAZ6envj7+zNs2DDy8/OprKyksrKS/Px8hg0bZrYe0u0VQljNFqO9paWlJCcnYzKZUBSFMWPGMHr0aJ5//nkqKipQFIU+ffqwfPlyAEaOHElubi56vR4PDw9SU1MB8PHx4aWXXiI+Ph6AmTNn4uPjY/b6En5CCKvZ4iHnPn36kJGRcd/2rVu3NnrNpUuX/uJ38fHxavhZSsJPCGEVWcBI4OLiwvHjx7l8+TLR0dFs2rSJgQMHotFo+P7775kyZQrV1dUAJCQksGzZMhRF4dSpUzz33HMAfP755wwZMoT8/Hyio6ObszoOrU2bNrz22mu4ubnh6urK0aNH2b17N3q9njFjxqDVapk+fTo3b95scFxwcDBLly5l48aNHDt2jMcee4ykpCQ8PDy4c+cOmZmZHDlypJlq1Xwc4fU2u4ZfXl4eK1eu5M6dOyQkJDBt2jR7Xu6R+8Mf/sCZM2fw8vICYN68edy4cQOAtWvXMmvWLFavXk3btm1ZtGgRYWFhXL9+HT8/P/Ucf/rTn2jfvj0vvvhis9TBWdTW1pKamkpNTQ2urq68/vrrnDp1iu+//56TJ0+yePHi+47RaDQkJiaqo40ARqORd955B4PBgI+PD2+88Qbffvstt27depTVaXaO0PKzWw1MJhMpKSls2rSJrKws9uzZw7lz5+x1uUeue/fuREVFsWnTJnXb3eAD8PDwQFEUAPz8/PjLX/7C9evXAbh27Zq634EDBxocJ+ynpqYGqL9Z7+ZW/+/+hQsX+PHHH39x//DwcI4dO6aOSEL94xV3H6C9fv06lZWVeHp62rnkLYutRnubm93Cr7CwkKCgIAIDA3F3dycqKoqcnBx7Xe6RW79+PQsXLuTOnTsNtv/P//wPJSUl9OnThw0bNgDQtm1bevfuTX5+Pl9//TURERHNUWSnp9FoWLlyJX/961/59ttvOX/+fKP7urq6MnDgwCb/zgYHB+Pm5qY+muFMXF1dLfppyTTK3eaJjX3xxRccPHiQlStXApCRkUFhYSFLlixp9Jhr165x4cIFexTHpry9vfH29ubixYt4enqi1Wrva9U+9thjVFdXU1ZWRmhoKDdu3OBf//oXbdq0oU+fPpw+fRqTyQTQ6DlaA19f3+YugtVcXFzQarX8+OOP1NbWAhAYGMjly5fVf8yCgoIoKSmhpqYGPz8/bt26pd6/hfr/+bt27cq1a9fUFmVrERwc/FDHnzt3jldeecWifT/++OOHupY9tagBjwsXLjBo0KDmLoZZqampTJ48mS5dutCuXTvc3d05e/YskydPVvcZPnw4CxcuJDo6GoPBwCuvvML7778PwP79+0lOTub48eNA/fNL//Vf/9UqBzzuDty0NrGxsRiNRj777DOg/l3SVatWqQMemzdvxt3dHXd3d9q0aUO7du34+9//zjfffIOHhweLFy/mv//7vzl27FhzVsNqKSkpD32O1tCltYTdur0P+spJa/Dqq68SGBjIE088wTPPPMOBAweYPHkyPXr0UPcZP348Z8+eBervDY0aNQqAzp0707t3b/71r381R9GdlqenJ+3btwfqR35DQkK4cuVKo/sXFxczb9485s2bx9GjR3n//ff55ptvcHV1Ze7cuRw8eLDVBZ8tOcI9P7u1/EJCQigqKqK4uBitVktWVhZr16611+WanUajYcuWLXh5eaHRaDh16hQzZswAoKqqirKyMrWru2DBAsrLy4H6EfE+ffrQsWNHiouLeeGFF9i7d29zVsUh+fj48OKLL+Li4oJGo+HIkSMUFBQQHh7OuHHj8Pb25s033+TUqVMNBrF+bsiQIfz617+mY8eO6txz7777LhcvXnxUVWkRWnqwWcJu9/wAcnNzSU1NxWQyMXHiRDUMGnP8+PFW0e211rFjxxyyXtB6u73mpKSkNHl/urVKSUl56Ht+58+f/8VHg37J9u3bH+pa9mTXe34jR45k5MiR9ryEEOIRaw1dWku0qAEPIUTr4AgPOUv4CSGsJi0/IYRTkvATQjgduecnhHBacs9PCOGUHKHl1/rjWwjxSNlqVpeamhri4+MZP348UVFRvP3220D92zUJCQno9Xrmzp2rLl1pNBqZO3cuer2ehIQELl26pJ7r3XffRa/XExERYdGavSDhJ4R4ALYIP3d3d7Zs2cInn3xCRkYGBw8epKCggDVr1jBlyhT27duHl5cXu3btAmDnzp14eXmxb98+pkyZwpo1a4D6iRaysrLIyspi06ZNLF++XJ00pCkSfkIIq9ki/DQaDR06dACgrq6Ouro6NBoNhw8fVqd9mzBhgjqt2IEDB5gwYQIAERERfP311yiKQk5ODlFRUbi7uxMYGEhQUBCFhYVm6yD3/IQQVrP0nl95eTlTp05VPycmJpKYmKh+NplMxMXFcfHiRZ599lkCAwPx8vJSJ5sNCAhQJ481GAx07doVADc3Nzw9PamoqMBgMNC/f3/1nFqtVj2mKRJ+QgirWLOAka+vL7t37270e1dXVzIzM6mqqmLmzJmPdLYj6fYKIaxm6ymtvLy8GDx4MAUFBVRVVVFXVwfULxtwdyo8rVbL1atXgfpu8o0bN+jUqdMDT58n4SeEsJotwq+8vFxdH+Wnn37i0KFD9OjRg8GDB5OdnQ1Aeno6Op0OAJ1OR3p6OgDZ2dkMGTIEjUaDTqcjKysLo9FIcXExRUVFPPXUU2brIN1eIYTVbPGcX2lpKcnJyZhMJhRFYcyYMYwePZqePXsyb9481q9fz5NPPklCQgJQvzD5ggUL0Ov1eHt7s27dOgB69erF2LFjiYyMxNXVlSVLlli0foiEnxDCarYIvz59+pCRkXHf9sDAQPXxlnu1bdtWfRbw52bMmGF2vtCfk/ATQljFmgGPlkzCTwhhNUd4vU3CTwhhNQk/IYRTkvATQjgdmc9PCOG0JPyEEE5JRnuFEE5HHnURQjgt6fYKIZyShJ8Qwik5dPitWLGiyQq+9tprdimQEKJlc/h7fv369XuU5RBCtCIO3fK7O1f+Xbdv38bDw8PuBRJCtHyOEH5m264nT54kMjKSsWPHAnD27FmWLVtm73IJIVowW8/k3BzMhl9qairvvfcePj4+QP0cXMePH7d7wYQQLZdThB+grpikHuQANzuFEA/GVouWX716lcmTJxMZGUlUVBRbtmwBYMOGDQwfPpyYmBhiYmLIzc1Vj2lscfK8vDwiIiLQ6/WkpaVZVA+zj7p07dqVEydOoNFoqK2tZevWrfTo0cOikwshHJMl08Rbco7k5GT69u3LzZs3mThxImFhYQBMmTKFF154ocH+9y5ObjAYSEpKUtf6SElJYfPmzWi1WuLj49HpdPTs2bPJ65ttwi1btoxt27ZhMBgYPnw4Z86cYcmSJQ9aXyGEA7BFy8/f35++ffsC0LFjR4KDg5tcb7exxckLCwsJCgoiMDAQd3d3oqKi1IXOm2K25efr68vatWvNnkgI4RzscT/v0qVLnDlzhv79+3PixAm2bdtGRkYG/fr1Izk5GW9v7yYXJw8ICGiwvbCw0Ow1zbb8iouLmT59OkOGDGHo0KHMmDGD4uLiB6mfEMJBWNryKy8vJy4uTv3ZsWPHfeeqrq5mzpw5vPrqq3Ts2JFJkyaxb98+MjMz8ff3Z9WqVXapg9mW38svv8yzzz7Lxo0bAcjKymL+/Pns3LnTLgUSQrR8lrb8fH192b17d6Pf19bWMmfOHKKjowkPDwegS5cu6vcJCQlMnz4doMnFye2yaPnt27eJjY3Fzc0NNzc3YmJiqKmpMXtiIYTjssU9P0VRWLx4McHBwSQlJanbS0tL1d/3799Pr169ABpdnDwkJISioiKKi4sxGo1kZWWpC503pdGW3/Xr1wEYMWIEaWlpREZGotFo+Oyzzxg5cqTZEwshHJOt3u395ptvyMzMpHfv3sTExAAwf/589uzZw9mzZwHo3r07KSkpQNOLky9ZsoSpU6diMpmYOHGiGphNaTT84uLi0Gg0KIoCwPbt29XvNBoNL7/88gNWWQjR2tliwGPgwIF89913921vqnHV2OLkI0eOtLpR1mj4HThwwKoTCSGchyO86GDRfH7ff/89586dw2g0qttiY2PtVighRMvW0l9ds4TZ8Nu4cSNHjhzh/PnzjBw5kry8PAYMGCDhJ4STag3v7VrCbNs1OzubLVu20KVLF958800yMzO5cePGoyibEKKFcnFxseinJTPb8mvbti0uLi64ublx8+ZNOnfuzNWrVx9F2YQQLZQjtPzMhl+/fv2oqqoiISGBuLg42rdvz29+85tHUTYhRAvlFOF3d+LSSZMmMXz4cG7evEmfPn3sXS4hRAvlKPf8Gg2/06dPN3rQ6dOn1dkYhBDOx6HDr6mXiTUaDVu3brV5YQYOHKg+VO1oHLVeZWVlzV0Eu/D29ubPf/5zcxfD5ry9vW1yHocOvw8++OBRlkMI0UpoNBqbTGba3GTRciGE1Ry65SeEEI2R8BNCOKWW/gCzJczWQFEUMjMz1clMr1y5YtEU0UIIx2Sr1duam0ULGBUUFJCVlQVAhw4dWL58ud0LJoRouZwi/AoLC1m6dClt27YF6ofKa2tr7V4wIUTL5RTv9rq5uWEymdQULy8vb/GVEkLYj61mcm5uZmswefJkZs6cSVlZGevWrWPSpEm8+OKLj6JsQogWyhbd3qtXrzJ58mQiIyOJiopiy5YtQP0SGklJSYSHh5OUlERlZSVQP/7wxhtvoNfriY6ObvAWWnp6OuHh4YSHh5Oenm5RHcy2/MaPH0/fvn05fPgwiqLw17/+lR49elh0ciGEY7LF/TxXV1eSk5Pp27cvN2/eZOLEiYSFhbF7926GDh3KtGnTSEtLIy0tjQULFpCXl0dRURF79+7l1KlTLFu2jJ07d3L9+nU2btzIxx9/jEajIS4uDp1OZ/ZtFrMtvytXruDh4cHo0aPR6XR4eHhw5cqVh664EKL1skXLz9/fX50joGPHjgQHB2MwGMjJyVEnS46NjWX//v0A6naNRkNoaChVVVWUlpaSn59PWFgYPj4+eHt7ExYWxsGDB83WwWzL794ubk1NDZcuXeKJJ55QR3+FEM7FmtfbysvLmTp1qvo5MTGRxMTE+/a7dOkSZ86coX///pSVleHv7w+An5+f+v64wWAgICBAPSYgIACDwXDfdq1Wi8FgMFs2s+H36aefNvh8+vRp/v73v5s9sRDCcdlq0XKA6upq5syZw6uvvkrHjh3vu469Hpmxesimb9++8pCzEE7OVs/51dbWMmfOHKKjowkPDwegc+fO6sLlpaWl+Pr6AvUtupKSEvXYkpIStFrtfdsNBgNardbstc22/DZv3qz+fufOHf73f/9XbZIKIZyTLVpjiqKwePFigoODSUpKUrfrdDoyMjKYNm0aGRkZPP300+r2v/3tb0RFRXHq1Ck8PT3x9/dn2LBhvPXWW+qocH5+PvPnzzd7fbPhV11drf7u6urKyJEjiYiIsLqiQgjHYKvn/L755hsyMzPp3bs3MTExAMyfP59p06Yxd+5cdu3aRbdu3Vi/fj1QvzB5bm4uer0eDw8PUlNTAfDx8eGll14iPj4egJkzZ+Lj42O+HkoTs2yaTCbWrFnDK6+88tAVFY7JkSczvduScCTe3t64uT3cfCYVFRV8+eWXFu07YcKEh7qWPTX6p1BXV4ebmxsnTpx4lOURQrRwjvKGR6Phl5CQQHp6On369GH69OmMGTOG9u3bq9/fvTkphHA+Dh1+dxmNRjp16sSRI0cabJfwE8J5tfQZWyzRaPiVlZWxefNmevXqhUajabAAjyNUXAjx4BwhAxoNvzt37jQY6RVCCHCCdXv9/PyYNWvWoyyLEKKVcOjwc9R1ZoUQD8+hl658//33H2ExhBCthcN3ey15QloI4ZwcOvyEEKIxTvGcnxBC/Jy0/IQQTsfh7/kJIURjJPyEEE5Jwk8I4ZQk/IQQTskRwq/1j1cLIR4pS9fvsCQgFy1axNChQxk3bpy6bcOGDQwfPpyYmBhiYmLIzc1Vv3v33XfR6/VEREQ0WJ4yLy+PiIgI9Ho9aWlpFtVDWn5CCKvZquUXFxfH7373u/tmi58yZQovvPBCg23nzp0jKyuLrKwsDAYDSUlJZGdnA5CSksLmzZvRarXEx8ej0+no2bNnk9eW8BNCWM1W4Tdo0CAuXbpk0b45OTlERUXh7u5OYGAgQUFB6kqSQUFBBAYGAhAVFUVOTo7Z8JNurxDCarbq9jZm27ZtREdHs2jRInUtlcYWJ3/QRcsl/IQQVrM0/MrLy4mLi1N/duzYYfbckyZNYt++fWRmZuLv78+qVavsUgfp9gohrGJNq87X15fdu3dbdf4uXbqovyckJDB9+nTg/kXL712c/EEWLZeWnxDCai4uLhb9PIjS0lL19/3799OrVy+gftHyrKwsjEYjxcXFFBUV8dRTTxESEkJRURHFxcUYjUaysrLQ6XRmryMtPyGE1Ww14DF//nyOHj1KRUUFI0aMYPbs2Rw9epSzZ88C0L17d1JSUgDo1asXY8eOJTIyEldXV5YsWaJOqrpkyRKmTp2KyWRi4sSJamA2WYemFi0XwhxZtLx1scWi5dXV1Wo4mTNgwICHupY9SctPCGE1R3jDQ8JPCGEVmdJKCOG0JPyEEE5JprEXQjglafkJIZyO3PMTQjgtRwi/1t9xbwH+8z//E39/f/r166duKy8vR6/X06tXL/R6PRUVFQBUVlYSHR1N//796du3L5s3b26uYjut//u//2PUqFHqz+OPP84777zDm2++yYgRIxg1ahRjx47l6tWrQP38cnf3HTZsGP7+/up/T9F62e0h50WLFvHll1/SuXNn9uzZY49LtBh5eXl07NiR559/nn/+858ALFy4EF9fX5KTk1m1ahUVFRWsXr2a1NRUKisrWb16NdeuXePXv/41JSUluLu7N3MtHkxrf8jZZDIREhJCdnY2Pj4+eHp6ArB161ZOnTrF2rVrG+z/xRdf8M4775CRkdEcxX1otnjI+fbt2/zwww8W7ftv//ZvD3Ute7Jbyy8uLo5NmzbZ6/QtyogRI/D19W2wLTMzk9///vcA/P73v1f/Z9FoNNy4cQNFUbh58ya+vr4P/ZdRPLi8vDwef/xxAgMD1eADuHXr1i927Xbv3k1cXNyjLGKLZM93ex8Vu/1fZ80khY7IYDDQtWtXAAICAtT5xWbNmsX48ePp1q0bN27cYMeOHS3+L4kjS09PbxBmK1euZMeOHXTq1ImPP/64wb63bt3iwIEDrF69+lEXs0WRAQ9hsXv/smRnZxMaGsqBAwc4f/48er2e4cOH4+Xl1cylfDDe3t7NXYQHZjQayc7O5o9//KNajz/+8Y/qz9/+9jeWLl2q7p+dnU1YWBiPP/54M5W45ZDwE43SarVcvXqVrl27cvXqVfz9/QHYvHkzycnJaDQaevbsyRNPPMHZs2f593//92Yu8YNpzS//f/bZZ4SEhNCuXbv76vHMM88QFRXF3Llz1W3btm1j/PjxrbrOtvrHyhHCT/pbdjJ+/Hi2bNkCwJYtW4iJiQHgscceIycnB6jvGn/33XcEBwc3Wzmd2c/v350/f179/ZNPPmkwLVJVVRWHDh1i7Nixj7SMwn6k5WcDkyZN4ssvv+THH3/kV7/6FcuXLyc5OZn/+I//4L333iMoKIiPPvoIgNdff50pU6YQEhKCoiisXr26wcy14tGorq4mNzeXt956S922YsUKzp07h4uLC0888USD6dOzsrIYNWoUHTp0aI7itjiOcJ/abo+63DtJYefOnZk9ezYJCQn2uJRoRq39UZfGyHx+jfvpp5+4fPmyRfv26NHjoa5lT3Zr+d37L6oQwnE4ymhv62+7CiEeOVstXblo0SKGDh3KuHHj1G3Xr18nKSmJ8PBwkpKS1Ba4oii88cYb6PV6oqOjOX36tHpMeno64eHhhIeHk56eblEdJPyEEFazVfj90ssQaWlpDB06lL179zJ06FDS0tKA+gfSi4qK2Lt3LytWrGDZsmVAfVhu3LiRjz76iJ07d7Jx40aLbllI+Akhms2gQYPue/wmJyeH2NhYAGJjY9m/f3+D7RqNhtDQUKqqqigtLSU/P5+wsDB8fHzw9vYmLCyMgwcPmr22jPYKIaxm6WhveXk5U6dOVT8nJiaSmJjY5DFlZWXqc7F+fn7qoJrBYCAgIEDd7+6bUz/frtVq1TeqmiLhJ4SwmwdZtPxe9hxckW6vEMIqlt7ve9DQ6ty5s7pweWlpqTppiFarpaSkRN2vpKQErVZ733aDwYBWqzV7HQk/IYTV7Bl+Op1OnQUpIyODp59+usF2RVEoKCjA09MTf39/hg0bRn5+PpWVlVRWVpKfn8+wYcPMXke6vUIIq9mqK3rvyxAjRoxg9uzZTJs2jblz57Jr1y66devG+vXrARg5ciS5ubno9Xo8PDxITU0FwMfHh5deeon4+HgAZs6ciY+Pj/k62OsND+Ec5A2P1sUWb3gYjUauXbtm0b7du3d/qGvZk7T8hBBWc4R3eyX8hBBWkdfbhBCiFZOWnxDCao7Q8pPwE0JYTcJPCOGUJPyEEE5Jwk8I4XRktFcIIVoxafkJIazmCC0/CT8hhNUk/IQQTknCTwjhlBwh/GTAQwjhlKTlJ4SwijzqIoQQrZi0/IQQVnOElp+EnxDCarYKP51OR4cOHXBxccHV1ZXdu3dz/fp15s2bx+XLl+nevTvr16/H29sbRVFYuXIlubm5tGvXjlWrVtG3b98HvrZ0e4UQzWrLli1kZmaqS1ympaUxdOhQ9u7dy9ChQ0lLSwMgLy+PoqIi9nyzuW4AAAfMSURBVO7dy4oVK1i2bNlDXVfCTwhhNXuu3paTk0NsbCwAsbGx7N+/v8F2jUZDaGgoVVVV6hKXD0K6vUIIuykvL2fq1Knq58TERBITExvs88ILL6DRaNTvysrK8Pf3B8DPz09dJMtgMBAQEKAeFxAQgMFgUPe1loSfEMJqlrbqfH191e7sL/nwww/RarWUlZWRlJREcHDwfdex1+CKdHuFEFaxtMtrSWhptVoAOnfujF6vp7CwkM6dO6vd2dLSUnx9fdV9S0pK1GNLSkrU4x+EhJ8QolncunWLmzdvqr9/9dVX9OrVC51OR0ZGBgAZGRk8/fTTAOp2RVEoKCjA09Pzgbu8IN1eIcQDsEVXtKysjJkzZwJgMpkYN24cI0aMICQkhLlz57Jr1y66devG+vXrARg5ciS5ubno9Xo8PDxITU19uDooiqI8dC2E07p7M9rReHt7U1lZ2dzFsDlvb2/c3B6uzXPnzh2MRqNF+7Zr1+6hrmVP0u0VQjgl6fYKIawmr7cJIZySI4SfdHuFEE5JWn5CCKtJy08IIVopafkJIawiMzkLIUQrJi0/IYTVHKHlJ+EnhLCaI4SfdHuFEE5JWn5CCKtJy08IIVopafkJIaziCK0+kCmthBBOSrq9QginJOEnhHBKEn5CCKck4SeEcEoSfkIIpyThJ4RwShJ+QginJOFnR3l5eURERKDX60lLS2vu4ggLLFq0iKFDhzJu3LjmLoqwMwk/OzGZTKSkpLBp0yaysrLYs2cP586da+5iCTPi4uLYtGlTcxdDPAISfnZSWFhIUFAQgYGBuLu7ExUVRU5OTnMXS5gxaNAgvL29m7sY4hGQ8LMTg8FAQECA+lmr1WIwGJqxREKIe0n4CSGckoSfnWi1WkpKStTPBoMBrVbbjCUSQtxLws9OQkJCKCoqori4GKPRSFZWFjqdrrmLJYT4/2RKKzvKzc0lNTUVk8nExIkTmTFjRnMXSZgxf/58jh49SkVFBZ07d2b27NkkJCQ0d7GEHUj4CSGcknR7hRBOScJPCOGUJPyEEE5Jwk8I4ZQk/IQQTknCrxV58skniYmJYdy4ccyZM4fbt28/8LmSk5P54osvAFi8eHGTky4cOXKEEydOWH0NnU5HeXm5xdvv9Zvf/Maqa23YsIH33nvPqmOEc5Pwa0XatWtHZmYme/bsoU2bNmzfvr3B93V1dQ903pUrV9KzZ89Gvz969CgnT558oHML0VLJouWt1MCBA/nuu+84cuQIf/7zn/Hy8uKHH37gs88+Y82aNRw9ehSj0chzzz3HM888g6IorFixgq+++oquXbvSpk0b9VyTJ09m4cKFhISEkJeXx7p16zCZTHTq1ImVK1eyfft2XFxc+OSTT3j99dcJDg5m6dKlXLlyBYBXX32VAQMGUFFRwcsvv4zBYCA0NBRLHiF96aWXKCkpoaamhueff57ExET1u9TUVL766iu6dOnCunXr8PX15eLFiyxfvpyKigratWvHihUr6NGjh+3/gIXjU0SrERoaqiiKotTW1irTp09Xtm3bphw+fFjp37+/cvHiRUVRFGX79u3KX/7yF0VRFKWmpkaZMGGCcvHiRSU7O1uZMmWKUldXp5SUlCgDBgxQPv/8c0VRFOV3v/udUlhYqJSVlSkjRoxQz1VRUaEoiqK8/fbbyqZNm9RyzJ8/Xzl27JiiKIpy+fJlZcyYMYqiKMqKFSuUDRs2KIqiKP/4xz+U3r17K2VlZffVY/To0er2u9e4ffu2EhUVpZSXlyuKoii9e/dWMjMzFUVRlA0bNijLly9XFEVRnn/+eeWHH35QFEVRCgoKlMmTJ/9iGYUwR1p+rchPP/1ETEwMUN/yi4+P5+TJk4SEhBAYGAjAV199xXfffUd2djYAN27c4MKFCxw7doyoqChcXV3RarUMGTLkvvMXFBQwcOBA9Vw+Pj6/WI5Dhw41uEd48+ZNqqurOXbsGBs3bgRg1KhRFs2L98EHH7Bv3z4Arl69yoULF+jUqRMuLi5ERkYCEBMTw6xZs6iurubkyZP84Q9/UI83Go1mryHEL5Hwa0Xu3vP7ufbt26u/K4rCa6+9xvDhwxvsk5uba7Ny3Llzh48++oi2bds+1HmOHDnCoUOH2LFjBx4eHkyePJmamppf3Fej0aAoCl5eXr/4ZyCEtWTAw8EMGzaMDz/8kNraWgB++OEHbt26xaBBg/j8888xmUyUlpZy5MiR+44NDQ3l+PHjFBcXA3D9+nUAOnToQHV1dYNrfPDBB+rnM2fOAPWzIH/66adAfdhWVlY2WdYbN27g7e2Nh4cH58+fp6CgQP3uzp07auv1008/ZcCAAXTs2JFf/epXfP7550B90J89e9a6PyAh/j8JPweTkJBAz549iYuLY9y4cSxZsgSTyYRerycoKIjIyEheeeUVQkND7zvW19eXlJQUZs+ezfjx45k3bx4Ao0ePZt++fcTExHD8+HEWL17MP//5T6Kjo4mMjOTDDz8EYObMmRw/fpyoqCj27dtHt27dmizriBEjqKurY+zYsaxdu7ZBmdq3b09hYSHjxo3j8OHDzJw5E4A//elP7Nq1i/HjxxMVFcX+/ftt9UcnnIzM6iKEcErS8hNCOCUJPyGEU5LwE0I4JQk/IYRTkvATQjglCT8hhFOS8BNCOKX/BxRscfSppmQWAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import confusion_matrix,accuracy_score\n",
    "from sklearn.dummy import DummyClassifier\n",
    "\n",
    "\n",
    "example_model =[DecisionTreeClassifier(min_samples_leaf = 8),\n",
    "                LogisticRegression(C = 0.5),\n",
    "                XGBClassifier(subsample=0.6,scale_pos_weight=5,n_estimators=100,max_depth=5,learning_rate=0.01,colsample_bytree=0.4,colsample_bylevel=0.6),\n",
    "                RandomForestClassifier(n_estimators=200,min_samples_split=2,min_samples_leaf=1,max_features='auto',max_depth=30),\n",
    "                DummyClassifier(strategy='stratified')] #naive classifier 1 with strategy of selecting a random class\n",
    "for models in example_model:\n",
    "  # Remember to use the training data here!! \n",
    "  models.fit(scaled_X_train_resampled, y_train_resampled)\n",
    "  # And to use the test data here!!\n",
    "  pred = models.predict(scaled_X_test2)\n",
    "  print(classification_report(y_true = y_test, y_pred = pred))\n",
    "  metrics.plot_confusion_matrix(models, X = scaled_X_test2, y_true = y_test, cmap='Greys')\n",
    "  accuracy=accuracy_score(y_test,pred)\n",
    "  print(str(models) +'accuracy =',accuracy)"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "Copy of Copy of “finalllllllllll2.0",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}